Vol. 5 N.1 - Journal of Aerospace Technology and Management by Journal of Aerospace Technology and Management

Journal of aerospace technology and management

JOURNAL OF AEROSPACE TECHNOLOGY AND MANAGEMENT Vol. 5 N. 1 Jan./Mar. 2013 ISSN 1984-9648 ISSN 2175-9146 (online)

www.jatm.com.br

V.5, n. 1, Jan./mar., 2013

Journal of Aerospace Technology and Management

General Information Journal of Aerospace Technology and Management (JATM) is a technoscientific publication serialized, published by Departamento de Ciência e Tecnologia Aeroespacial (DCTA) and aims to serve the international aerospace community. It contains articles that have been selected by an Editorial Committee composed of researchers and technologists from the scientific community. The journal is quarterly published, and its main objective is to provide an archival form of presenting scientific and technological research results related to the aerospace field, as well as promote an additional source of diffusion and interaction, providing public access to all of its contents, following the principle of making free access to research and generate a greater global exchange of knowledge. JATM is added/indexed in the following databases; SCOPUS - Elsevier; CAS - Chemical Abstracts Service; DOAJ - Directory of Open Access Journals; J-GATE - The e-journal gateway from global literature; LIVRE - Portal to Free Access Journals; GOOGLE SCHOLAR; SUMÁRIOS.ORG - Summaries of Brazilian Journals; EZB - Electronic Journals Library; ULRICHSWEB - Ulrich´s Periodicals Directory; SOCOL@R - China Educational Publications; LATINDEX - Regional Cooperative Online Information System for Scholarly Journals from Latin America, the Caribbean, Spain and Portugal; REDALYC - Red de Revistas Científicas de América Latinay el Caribe, España y Portugal; and PERIÓDICOS CAPES. In WEB QUALIS System, JATM is classified as B4 in the Geosciences and Engineering III areas. JATM is affiliated to ABEC - Brazilian Association of Scientific Editors and all published articles contain DOI numbers attributed by CROSSREF.

Correspondence All correspondence should be sent to: Journal of Aerospace Technology and Management Instituto de Aeronáutica e Espaço Praça Mal. Eduardo Gomes, 50 - Vila das Acácias CEP 12228-901 São José dos Campos/ São Paulo/Brazil Contact Phone: (55) 12-3947- 6493/5122 E-mail: editor@jatm.com.br Web: http://www.jatm.com.br Published by: Departamento de Ciência e Tecnologia Aeroespacial Distributed by: Instituto de Aeronáutica e Espaço Editing, proofreading and standardization: Zeppelini Editorial Printing: Type Brasil Edition: 750 São José dos Campos/SP, Brazil ISSN 1984-9648

JATM is supported by:

Journal of Aerospace Technology and Management Vol. 5, n.1 (Jan./Mar. 2013) – São José dos Campos: Zeppelini Editorial, 2013 Quartely issued Aerospace sciences Technologies Aerospace engineering CDU: 629.73

Historical Note: JATM was created in 2009 after the iniciative of the diretor of Instituto de Aeronáutica e Espaço (IAE), Brigadeiro Engenheiro Francisco Carlos Melo Pantoja. In order to reach the goal of becoming a journal that could represent knowledge in science and aerospace technology, JATM searched for partnerships with others institutions in the same field from the beginning. From September 2011, it has been edited by the Departamento de Ciência e Tecnologia Aeroespacial (DCTA), and it also started to be financially supported by Fundação Conrado Wessel. The copyright on all published material belongs to Departamento de Ciência e Tecnologia Aeroespacial (DCTA)

ISSN 1984-9648 ISSN 2175-9146 (online)

Journal of Aerospace Technology and Management Vol. 5 No. 1 - Jan./Mar. 2013 Editor in Chief

Executive Editor

Ana Cristina Avelar Instituto de Aeronáutica e Espaço São José dos Campos/SP – Brazil editor@jatm.com.br

Ana Marlene F. Morais Instituto de Aeronáutica e Espaço São José dos Campos/SP – Brazil secretary@jatm.com.br

SCIENTIFIC COUNCIL Angelo Passaro Instituto de Estudos Avançados São José dos Campos/SP – Brazil

Eduardo Morgado Belo Escola de Engenharia de São Carlos São Carlos/SP – Brazil

Marco A. Sala Minucci Vale Soluções em Energia São José dos Campos/SP – Brazil

Antonio Pascoal Del’Arco Jr Instituto de Aeronáutica e Espaço São José dos Campos/SP – Brazil

Francisco Carlos M. Pantoja Diretoria de Engenharia da Aeronáutica Rio de Janeiro/RJ – Brazil

Mischel Carmen N. Belderrain Instituto Tecnológico de Aeronáutica São José dos Campos/SP – Brazil

Carlos Antônio M. Kasemodel Instituto de Aeronáutica e Espaço São José dos Campos/SP – Brazil

Francisco Cristovão L. Melo Instituto de Aeronáutica e Espaço São José dos Campos/SP – Brazil

Paulo Tadeu de Melo Lourenção EMBRAER São José dos Campos/SP – Brazil

Carlos de Moura Neto Instituto Tecnológico de Aeronáutica São José dos Campos/SP – Brazil

João Marcos T. Romano Universidade Estadual de Campinas Campinas/SP – Brazil

Rita de Cássia L. Dutra Instituto de Aeronáutica e Espaço São José dos Campos/SP – Brazil

Acoustics

Applied computation

Ceramic Materials

Marcello Faraco de Medeiros Escola de Engenharia de São Carlos São Carlos/SP – Brazil

José Márcio Machado Instituto de Biociências, Letras e Ciências Exatas São José do Rio Preto/SP – Brazil

José Maria Fonte Ferreira Universidade de Aveiro Aveiro – Portugal

Aerodynamics

Romis R. F. Attux Universidade Estadual de Campinas Campinas/SP – Brasil

Circuitry

ASSOCIATE EDITORS Bert Pluymers Katholieke Universiteit Leuven Leuven – Belgium

Acir Mércio Loredo Souza Universidade Federal do Rio Grande do Sul Porto Alegre/RS – Brazil João Luiz F. Azevedo Instituto de Aeronáutica e Espaço São José dos Campos/SP – Brazil

Aerospace Meteorology Gilberto Fisch Instituto de Aeronáutica e Espaço São José dos Campos/SP – Brazil Willian W. Vaughan University of Alabama Huntsville/AL – USA

Carlos Henrique Netto Lahoz Instituto de Aeronáutica e Espaço São José dos Campos/SP – Brazil

Astrodynamics

Antonio Sergio Bezerra Sombra Universidade Federal do Ceará Fortaleza/CE – Brazil

Altamiro Susin Universidade Federal do Rio Grande do Sul Porto Alegre/RS – Brazil

Antonio F. Bertachini Instituto Nacional de Pesquisas Espaciais São José dos Campos/SP – Brazil

Raimundo Freire Universidade Federal de Campina Grande Campina Grande/PB – Brazil

Othon Cabo Winter Faculdade de Engenharia de Guaratinguetá Guaratinguetá/SP – Brazil

Composites

Edson Cocchieri Botelho Faculdade de Engenharia de Guaratinguetá Guaratinguetá/SP – Brazil Flamínio Levy Neto Universidade de Brasília Brasília/DF – Brazil

Computational fluid dynamics

Joern Sesterhenn Technische Universität Berlin Berlin – Germany John Cater University of Auckland Auckland – New Zealand Paulo Celso Greco Escola de Engenharia de São Carlos São Carlos/SP – Brazil

Defense Systems

Adam S. Cumming Defence Science and Technology Laboratory Salisbury/Wiltshire – England Wim P. C. de Klerk Netherlands Organisation for Applied Scientific Research Rijswijk/SH – Netherlands

Eletromagnetic Compatibility

Alain Azoulay École Supérieure d’Electricité Gif–Sur–Yvette – France Cynthia Junqueira Instituto de Aeronáutica e Espaço São José dos Campos/SP – Brazil

Energetic Materials Elizabeth da Costa Mattos Instituto de Aeronáutica e Espaço São José dos Campos/SP – Brazil

Guidance, Navigation and Control

Radars and Tracking Systems

David Murray–Smith University of Glasgow Glasgow – Scotland

Marc Lesturgie Office National d’Etudes et de Recherches Aérospatiales Palaiseau – France

Daniel Alazard Institut Supérieur de l’Aéronautique et de l’Espace Toulouse – France

Waldemar de Castro Leite Filho Instituto de Aeronáutica e Espaço São José dos Campos/SP – Brazil

Management Systems

André Fenili Universidade Federal do ABC Santo André/SP – Brazil

Sadek Crisostomo Absi Alfaro Universidade de Brasília Brasília/DF – Brazil

Antonio Henriques de Araújo Jr Centro Universitário de Volta Redonda Volta Redonda/RJ – Brazil

Structures

Metallic Materials

José Rubens G. Carneiro Pontifícia Universidade Católica de Minas Gerais Belo Horizonte – Brazil

Photonics

Álvaro Damião Instituto de Estudos Avançados São José dos Campos/SP – Brazil

Polimeric Materials Cristina Tristão de Andrade Instituto de Macromoléculas Rio de Janeiro/RJ – Brazil

Mirabel Cerqueira Rezende Instituto de Aeronáutica e Espaço São José dos Campos/SP – Brazil

Fluid Dynamics and Turbulence

Processing of Aerospace Materials

Vassilis Theofilis Universidad Politécnica de Madrid Madrid – Spain

Robotics and Automation

Adiel Teixeira de Almeida Universidade Federal de Pernambuco Recife/PE – Brazil

José Leandro Andrade Campos Universidade de Coimbra Coimbra – Portugal

Marcos Pinotti Barbosa Universidade Federal de Minas Gerais Belo Horizonte/MG – Brazil

Hugo H. Figueroa Universidade Estadual de Campinas Campinas/SP – Brazil

Alexandre Queiroz Bracarense Universidade Federal de Minas Gerais Belo Horizonte/MG – Brazil

Propulsion and Combustion

Fernando de Souza Costa Instituto Nacional de Pesquisa Espacial São José dos Campos/SP, Brazil

Sérgio Frascino M. Almeida Instituto Tecnológico de Aeronáutica São José dos Campos/SP – Brazil

Synthesis and Characterization of Aerospace Materials

Gilson da Silva Instituto Nacional da Propriedade Industrial Rio de Janeiro/RJ – Brazil Roberto Costa Lima Instituto de Pesquisas da Marinha Rio de Janeiro/RJ – Brazil

Thermal Sciences

Márcia B. H. Mantelli Universidade Federal de Santa Catarina Florianópolis/SC – Brazil Renato Machado Cotta Universidade Federal do Rio de Janeiro Rio de Janeiro/RJ – Brazil

Vibration and Structural Dynamics Carlos Cesnik University of Michigan Ann Arbor/MI – USA

Valder Steffen Junior Universidade Federal de Uberlândia Uberlândia/MG – Brazil

Carlos Henrique Marchi Universidade Federal do Paraná Curitiba/PR – Brazil

Editorial Production Glauco da Silva Instituto de Aeronáutica e Espaço São José dos Campos/SP – Brazil

Lucia Helena de Oliveira Depart. Ciência e Tecnologia Aeroespacial São José dos Campos/SP – Brazil

Helena Prado A.Silva Instituto de Aeronáutica e Espaço São José dos Campos/SP – Brazil

Mônica E. Rocha de Oliveira Instituto Nacional de Pesquisas Espaciais São José dos Campos/SP – Brazil

Rosilene Maria M. Costa Instituto de Estudos Avançados São José dos Campos/SP – Brazil

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, 2013

ISSN 1984-9648 | ISSN 2175-9146 (online)

CONTENTS Editorial 5

Critical Technologies for Aerospace and Defense Applications: The Pursuit of Autonomy Vilson Rosa de Almeida

REVIEW ARTICLE 7

Aerospace Meteorology: An Overview of Some Key Environmental Elements William W. Vaughan, Dale L. Johnson

ORIGINAL PAPERS 15

A Sensitivity Investigation on the Aeroelastic Dynamic Stability of Slender Spinning Sounding Rockets Roberto Gil Annes da Silva, José Guido Damilano, João Luiz F. Azevedo

Unsteady Blade Element-Momentum Method Including Returning Wake Effects Cláudio Tavares Silva, Maurício Vicente Donadon

Calculation of The Vehicle Drag and Heating Reduction at Hypervelocities with Laser-Induced Air Spike Israel da Silveira Rêgo, Paulo Gilberto de Paula Toro, Marco Antonio Sala Minucci, José Brosler Chanes Júnior, Felipe Jean da Costa, Antonio Carlos de Oliveira

Thermal Protection System and Trajectory Optimization for Orbital Plane Change Aeroassisted Maneuver Antonio Mazzaracchio

Study by FT-IR Technique and Adhesive Properties of Vulcanized EPDM Modified with Plasma Renata Patrícia dos Santos,, Mauro Santos de Oliveira Junior, Elizabeth da Costa Mattos, Milton Faria Diniz, Rita de Cássia Lazzarini Dutra

Maximization of Fundamental Frequency of Laminated Composite Cylindrical Shells by Ant Colony Algorithm Rubem Matimoto Koide, Marco Antonio Luersen

Architecture for ES Receiver Systems Targeted at Commercial Wireless Communications Warren Paul du Plessis

An Asynchronous Interface with Robust Control for Globally-Asynchronous Locally-Synchronous Systems Duarte Lopes de Oliveira, Eduardo Lussari, Sandro Shoiti Sato, Lester de Abreu Faria

103 Fuzzy Modeling of Annoyance Caused by Aircraft Noise Using Laeq and Laeq Metrics d n Tarcilene Heleno, Jules Ghislain Slama 111 IMFLAR: An Intuitive Method for Logical Avionics Reliability Nilson Silva, Luís Gonzaga Trabasso THESIS ABSTRACTS 127 A Contribution for the Pre-galactic Universe Study Eduardo dos Santos Pereira 127 Tack Study in Prepregs of Epoxy Resin Reinforced with Carbon Fiber Eduardo Gouveia Martins Romão

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, 2013

ISSN 1984-9648 | ISSN 2175-9146 (online)

128 Global Instability Analysis of Compressible Flow Elmer Mateus Gennaro 128 The Influence of Erosion and Wear on the Accretion and Adhesion of Ice for Nano Reinforced Polymeric Composites Used in Aeronautics Omid Gohardani 129 INSTRUCTIONS TO AUTHORS

CORRECTIONS J. Aerosp. Technol. Manag. Vol.1, No 1, pp. 49, Jan.-Jul., 2009 Title, instead of: Iosipesco shear resistance in composites of carbon and glass fiber with epoxi resin Insert: Iosipescu shear resistance in composites of carbon and glass fiber with epoxi resin

Editorial Critical Technologies for Aerospace and Defense Applications: The Pursuit of Autonomy Vilson Rosa de Almeida1

he Brazilian National Strategy of Defense defines three strategically important sectors: space, cybernetics and nuclear, which are reckoned to be essential and decisive for the national defense. According to this document, partnerships and the purchase of products and services abroad should be harmonized with the purpose of ensuring a wide range of capacities and technologies in Brazil. Moreover, it states that our national independence has to be achieved by an autonomous technological capacity building, and warns that whoever does not master critical technologies is neither independent for defense nor for development. However, achieving autonomy on critical technologies for aerospace and defense applications depend upon a multilayer set of actions and institutional structures. It is well known that Aerospace and Defense systems are intrinsically complex structures, which depend on the interdisciplinary knowledge and the mastering of several areas of science and technology. Any large aerospace or defense system comprises a large number of subsystems, whose performances depend upon the quality grade of several components, sensors and materials. Harnessing of critical and restricted access technologies for Aerospace and Defense Applications is not an easy enterprise, which usually demands several distinct technological task levels

to be accomplished. Examples of critical and restricted access technologies are: liquid rocket engines, inertial navigation/ stabilization systems, special composite materials and alloys, hypersonics, compact nuclear power reactors, laser isotope separation, amongst several others, not to mention all the sensing and control technologies intrinsically associated with those. The Aerospace Science and Technology Department – DCTA – is composed of several institutes; each of them is primarily devoted to a specific technological task level on the process of seeking solutions for the associated critical technologies, including: academic education, basic and advanced research, development, certification and flight testing. Amongst these institutes, the Institute for Advanced Studies (IEAv) focus mostly on the research level pursuit of critical technologies, although it also contributes with PostGraduate education and the development of some systems in areas of expertise that are exclusively or mainly available at IEAv. Over its 30 years of existence, IEAv has strongly contributed to the national technological independence by mastering science and technology in several areas of strategic interest, such as: laser isotope separation processes, Fiber Optic Gyroscope (FOG), hypersonics, nuclear reactors for space applications, infrared sensors, geospatial intelligence, just to mention a few.

Instituto de Estudos Avançados – São José dos Campos/SP – Brazil 1.Graduated in Aeronautical Sciences by Academia da Força Aérea – AFA (1987) and in Electronics Engineering by Instituto Tecnológico de Aeronáutica – ITA (1997), Master of Science in Electronics and Computer Engineering by ITA (1998), and PhD in Electrical and Computer Engineering by Cornell University (2004). Master of Business Administration in Advanced Development of Executives – Area of Process Management, by Universidade Federal Fluminense (2008). He has published several technical publications in the field of Photonics, with more than 2500 citations. Main career and academic positions: Fighter Aviation Pilot and Instructor at 1º/4º GAv (1989-1992), Post-Graduate Professor at ITA since 2006, Head of Photonics Division at IEAv (2010), Vice-Director of IEAv (2011). He is currently the Director of IEAv. E-mail: vilson@ieav.cta.br

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.5-6, Jan.-Mar., 2013

Almeida, V.R.

Similarly, the other institutes in DCTA also perform some extent of research activities in specific areas of expertise, besides endeavoring in other technological task levels, depending mostly upon the managerial and technical boundary conditions imposed by each critical technology, rather than relying on more formal analyses or benchmarking, e.g., by applying the concept of Technology Readiness Level. Nevertheless, keeping synergic relationship among all the institutes of DCTA is mandatory for the successful achievement of technological independence in the aerospace and defense realm. This relationship extends beyond the borders of DCTA, bringing to the game board other governmental institutions, universities, funding and regulatory agencies, industries and research groups that are involved with science and technology for aerospace or defense. The result of decades of investment on aerospace and defense science and technology has rewarded Brazil with strategic solutions and spin-offs in several technological areas, more than paying off all the investments made along its history; for instance, it is worth to mention the biofuel engines for terrestrial and aerial vehicles, the birth of EMBRAER, amongst many others.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.5-6, Jan.-Mar., 2013

The balance between education, research and development is also a key aspect to overcome future asymmetry or scarcity in some of these task levels, which are of foremost relevance for the survival of technological independence in the aerospace and defense fields. Any misbalanced initiative that is intended for shortterm results may only lead to short-lived technological independence, due to the lack of mid/long-term updating on technological evolutionary trend, usually by loss of expertise or by premature obsolescence of the adopted technologies. Additionally, the reduced level of replacement of expertise with the desired and specific skills demanded for the search of solutions for critical aerospace and defense technologies has posed a great challenge for a smooth and efficient transfer of knowledge and experience from retiring researchers and technicians to the new generations of aerospace and defense personnel. Future is unpredictable; however, science and technology history has proved that a solid groundwork on education, research and development is indispensable for any selfsustainable enterprise in strategic areas. Furthermore, keeping the know-how and expertise in already harnessed strategic technologies is as important as attempting to unveil new ones.

doi: 10.5028/jatm.v5i1.188

Aerospace Meteorology: An Overview of Some Key Environmental Elements William W. Vaughan1, Dale L. Johnson2

Abstract: The natural terrestrial environment plays a significant role in the design and operation of aerospace vehicles (space vehicles and rockets) and in the associated integrity of aerospace systems and elements. Addressed herein are some of the key vehicle and environment areas of concern plus “lessons learned” that have been identified over a number of years. Many of these aerospace meteorology related events occurred during the development and interpretation of natural environment inputs, especially those of terrestrial environment, used in the design and development of the Saturn Apollo and Space Shuttle vehicles plus associated mission operations. Background is given regarding the actions needed to avoid having to re-learn these lessons. Keywords: Aerospace meteorology, Launch vehicle development, Mission operations.

INTRODUCTION This paper is primarily based on and reflects the content of the recent revision conducted by the authors and their colleagues of the report “Terrestrial Environment (Climatic) Criteria Guidelines for Use in Aerospace Vehicle Development” (Johnson, 2008). The Earth’s terrestrial environment is typically defined as the altitude region up to approximately 90 km. Aerospace vehicle design guidelines are provided in the report for the following: winds; atmospheric models and thermodynamic properties; thermal radiation; U.S. and world surface extremes; humidity; precipitation, fog, and icing; cloud phenomena and cloud cover models; atmospheric electricity; atmospheric constituents; aerospace vehicle exhaust and toxic chemical release; tornadoes and hurricanes; geologic hazards; and sea state. Also included is information on mission analysis, prelaunch monitoring, flight evaluation, physical constants, and metric/English unit conversion factors. The first version of this report was published in 1962 and has subsequently been updated periodically by the NASA Marshall Space Flight Center. The terrestrial environment guidelines provided in the report are intended to be used in the development of specific terrestrial environment design requirements based on an aerospace vehicle’s mission requirements and on the program’s engineering design philosophy. These terrestrial environment criteria guidelines were formulated based on discussions with and requests from engineers who were involved in aerospace vehicle design, development and operations. Therefore, they represent responses to actual engineering problems and are not just a general compilation of environmental data. NASA Centers, various other government agencies, and their associated

1.University of Alabama – Huntsville/AL – USA 2.NASA Retired – Huntsville/AL – USA Author for correspondence: William W. Vaughan | 301 Sparkman Drive Northwest – 35899 – Huntsville/AL – USA | E-mail: vaughan@nsstc.uah.edu Received: 06/11/12 | Accepted: 14/12/12

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.7-14, Jan.-Mar., 2013

Vaughan, W.W. and Johnson, D.L.

contractors responsible for the design, mission planning, and operational studies have used these guidelines, and associated design requirements developed from them, extensively. Another key document related to the scope of this paper is the “Guide to Reference and Standard Atmosphere Models” (Vaughan, 2010). This document contains the description of over 75 reference and standard atmosphere models, Earth and planetary, prepared by national and international organizations. It provides information on the scope, data bases, uncertainty, sources for codes, and applicable references. It was prepared based on the contributions of numerous authors. The objective of the guide is to enable the reader to more readily ascertain the applicability of a model for their intended use. Another element of aerospace meteorology, weather support, is reviewed in relation to the Space Shuttle operations by Bellue et al. (2006). The mission’s success and safety of aerospace vehicles present unique weather support challenges. Weather support requirements to ensure the safe processing, launch, and landing of these vehicles have been continuously improved since the first successful missile launches. Weather equipment for operational support has also significantly improved since these first launches. The effective use of weather information is translated both into significant annual cost savings through the timely management decisions, and into paramount contributions to safety. Ideally, aerospace vehicle design should accommodate all expected operational terrestrial environment conditions. However, this is neither economically nor technically feasible. For this reason, consideration must be given to the protection of the vehicle from some extremes by use of support equipment, special facilities, and specialized forecast personnel to advise on the expected occurrence of critical terrestrial environment conditions. Services of specialized weather forecast personnel have proved very economical in comparison to a more extensive vehicle design that would be necessary to cope with all terrestrial environment possibilities. In general, terrestrial environment design requirement documents do not specify how the designer should use the data. Such specifications may be established only through analysis and study of a particular design problem. ENGINEERING RELIVANCE All aerospace vehicles (launch vehicle and spacecraft) have one thing in common. To one degree or another, they J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.7-14, Jan.-Mar., 2013

must be designed to accommodate the expected operational mission requirements relative to the terrestrial environment. Thus, terrestrial environment phenomena and elements play a significant role in how aerospace vehicles are designed and operated and in protecting the integrity of their systems and components (Vaughan et al., 1997). It is important to recognize the need to define the terrestrial environment design requirements very early in the design and development cycle of any new or to be modified aerospace vehicle in order for the vehicles to meet their mission operational requirements. Table 1 provides an overview of the key engineering systems and terrestrial environment interactions relative to vehicle engineering systems and mission phase. Specific definitions of the terrestrial environment can be provided which, if the aerospace vehicle is designed to accommodate, will enable the desired operational capability within the defined design risk level. It is very important that those responsible for the terrestrial environment definitions for use in the design of an aerospace vehicle have a close working relationship with program management and design engineers. This will ensure that the desired operational capabilities are reflected in the terrestrial environment requirements specified for design and development of the vehicle and, accordingly, their interpretation relative to engineering applications. An aerospace vehicle’s response to terrestrial environment design criteria must be carefully evaluated in order to provide an acceptable design relative to the desired operational requirements. The choice of criteria depends on the specific launch and landing location(s), vehicle configuration, and expected mission plan(s). Vehicle design, operation, and flight procedures can be separated into particular categories for proper assessment of environmental influences and impact on the life history of each vehicle and all associated systems. These categories include: (1) purpose and concept of the vehicle, (2) preliminary engineering design, (3) structural design, (4) control system design, (5) flight mechanics, orbital mechanics, and performance (trajectory shaping), (6) optimization of design limits regarding the various natural environmental factors, and (7) final assessment of the terrestrial environmental capability for launch and flight operations. Another important matter that must be recognized is the necessity to have a coordinated and consistent set of terrestrial environment requirements to be used in a new aerospace vehicle’s design and development or modification

Aerospace Meteorology: An Overview of Some Key Environmental Elements

Table 1. Key Terrestrial Environment Parameters Needed versus Engineering Systems (X) and Mission Phase (P).

X P

X P X P X P P X P P P P P P X P

Mission Operations

X P

P P P

X P

X P X P P X P P X P X P X P

X X P X X P P X P

X X P P

P X P

X P

X P P X P X X P P P X P P P X P

Atmospheric Constituents

X X P

of an existing vehicle. This is particularly important when diverse groups are involved in the development, and is of utmost importance for any international endeavor. A “central control point” holding responsibility for the definition and interpretation of the natural environment inputs is critical to the successful design and operation of any new or modified aerospace vehicle. Without this control, different terrestrial environment values or models may be used with costly results in terms of money, time, and vehicle performance. This central control point should also include responsibility for terrestrial environment inputs related to mission analysis, test support requirements, flight evaluation, launch constraints, and operational support. The close association between the design and test engineering groups and those responsible (central control point) for the terrestrial environment inputs is key to the success of the vehicle’s development process. This procedure has been followed in many NASA aerospace vehicle developments and is of particular importance for any

X X P X P

X X P P P P P X P

X X X X X

P P P P P P X P

X P P P

X P X X P

X P X P X P X P P P X P P X P

X P P

P P P X P

Mission Phase

X P

Geologic Hazards

X P

Severe Weather

X P

Sea State

X P

X P X P P

X P X X P

X P

Precipitation or Hail

X P P

Humidity

X P X P X P

Clouds & Fog

X P X X P

Atmospheric Electricity

Atmospheric Thermo-dynamics

System Propulsion Engine Sizing Structures / Airframe Performance / Trajectory/ G&N Aerodynamics Thermal Loads / Aerodynamic Heat Control Loads Avionics Materials Electric Power Optics Thermal Control Telemetry / Tracking / Comms

Solar / Thermal Radiation

Winds & Gusts

Terrestrial Environment Parameters

Launch Vehicle Systems (Sub-)

Mission Analysis Manufacturing Testing Transportation / Ground Handling Roll Out / On Pad Pre-Launch / DOL Count Liftoff / Ascent Stages Recovery Flight Orbital Descent Landing Post Landing Ferry / Transport Facility / Special Eq Refurbishment Storage

new aerospace vehicle. Feedback is critical for the vehicle development process relative to terrestrial environment requirements and thus the ability to produce a viable vehicle design and operational capability. Finally, although often not considered to be significant, it is very important that all aerospace vehicle design review meetings include a representative from the natural environment group (central control point) assigned to support the program. This will ensure the good understanding of design requirements and timely opportunity to incorporate terrestrial environment inputs and interpretations, which are tailored to the desired operational objectives into the design process. It is also necessary that any proposed deviations from the specified terrestrial environment design requirements, including those used in preliminary design tradeoff studies, be approved by the responsible terrestrial environment central control point. This will ensure that all program elements are using the same baseline terrestrial environment inputs. Included is the need for the maintenance of a centralized and J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.7-14, Jan.-Mar., 2013

Vaughan, W.W. and Johnson, D.L.

controlled terrestrial environment definition document. This is also important in order to make sure that the program manager understands the operational impact of any change in terrestrial environment requirements before they are implemented into the design. Otherwise, gross errors and deficiencies in design, and thus operational capability, may result from using different inputs selected from various sources. One must remember that the flight profile of any aerospace vehicle includes travel through the terrestrial environment. Terrestrial environment definitions are usually limited to information below ≈90 km (≈295,000 ft). Thus, aerospace vehicle operations will always be influenced to some degree by the terrestrial environment with which they interact. As a result, the definition of the terrestrial environment and its interpretation is one of the most important aerospace vehicle design and development inputs. The definition plays a significant role in the areas of structures, control systems, trajectory shaping (performance), aerodynamic heating, and takeoff/landing capabilities. The aerospace vehicle’s capabilities which result from the design, in turn, determine the constraints and flight opportunities for tests and mission operations. SOME TERRESTRIAL ENVIRONMENT CONSIDERATIONS Experience gained and lessons learned from developing terrestrial environment design criteria for previous aerospace vehicle programs have proven that, in order to be most effective, the terrestrial environment design criteria for a new vehicle should: (1) Be developed once the mission is defined, in order to ensure the desired operational performance of the aerospace vehicle. (2) Be issued under the signature of the program manager and be part of the controlled program definition and requirement documentations. (3) Specify terrestrial environment requirements for all phases of activity including prelaunch, launch, ascent, on-orbit, descent, and landing. For extremes in the terrestrial environments, there generally is no known physical upper or lower bound. However, wind speed does have a strict physical lower bound of zero. Essentially, all observed extreme conditions have a finite probability of being exceeded. Consequently, terrestrial environment extremes used to develop design requirements must be accepted with the knowledge that there is some J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.7-14, Jan.-Mar., 2013

risk of the values being exceeded. The measurement of many environmental parameters may not be as accurate as desired. Also, in some cases theoretical model estimates are believed to be of more value for design than those indicated by empirical distributions from short periods of record. Therefore, theoretical values are given considerable weight when selecting extreme values for some parameters; e.g., peak surface winds. Criteria guidelines are normally presented for various percentiles based on the available data. There should be caution in the interpretation of these percentiles in aerospace vehicle studies to ensure consistency with physical reality and the specific design and operational problems of concern. Aerospace vehicles are not normally designed for launch and flight in severe weather conditions such as hurricanes, thunderstorms, ice storms, and squalls. Environmental elements associated with severe weather that may be hazardous to aerospace vehicles and associated ground support equipment include strong ground and in-flight winds, strong wind shears and gusts, turbulence, icing conditions, and electrical activity. Terrestrial environment guidelines should provide information concerning those severe weather characteristics that should be included in vehicle and associated facilities of design necessities and specifications, if required to meet the program’s mission operational requirements. Knowledge of the terrestrial environment is also necessary to establish test requirements for aerospace vehicles and in the design of associated support equipment and facilities. Such data are required to define fabrication, storage, transportation, test, and preflight design conditions, and should be considered for both the whole vehicle system and the components which comprise the system. The group with the central control point responsibility and authority to define and interpret terrestrial environment design requirements must also be in a position to pursue environment input-related applied research studies and engineering assessments and updates. This is necessary to ensure accurate and timely terrestrial environment inputs tailored to the program’s needs. Design engineers and program managers should not assume they can simply draw on the vast statistical databases and numerous models of the terrestrial environment currently available in literature without interpretation and tailoring to specific vehicle design needs. Otherwise, this can prove to be a major deterrent to the successful development and operation of an aerospace vehicle. Although vehicle design ideally should accommodate all expected operational environment conditions, it is neither

Aerospace Meteorology: An Overview of Some Key Environmental Elements

economically nor technically feasible to design an aerospace vehicle to withstand all terrestrial environment extremes. For this reason, consideration should be given to protecting vehicles from some situations. This can be achieved by using support equipment and specialized weather forecast personnel to provide advice on the expected occurrence of critical terrestrial environment conditions. The services of specialized forecast personnel may be very economical compared to the more expensive vehicle designs that would be required to cope with all terrestrial environment possibilities. Good engineering judgment must be exercised to apply terrestrial environment requirements to the aerospace vehicle’s design analysis. Consideration must be given to the overall vehicle mission and system performance requirements. Knowledge is still lacking on relationships between some of the terrestrial environment parameters that are required as inputs for the design of aerospace vehicles. Also, interrelationships between vehicle parameters and terrestrial environment variables cannot always be clearly defined. Therefore, a close working relationship and team philosophy must exist between the design and operational engineer and the respective organization’s terrestrial environment central control point specialists. In many cases, it is impossible to clearly define the limiting extreme values for a particular terrestrial environment element that may occur during the desired operational lifetime of the vehicle. However, a lower value may be defined so that the probability is small that the lower value will be exceeded during the desired operational lifetime of the vehicle. Risks of launch delay may also be acceptable versus the expense of additional design considerations. Because of these and other considerations, a value that is lower than the extreme may be a more appropriate design requirement. The terrestrial environment specialist is responsible for providing the program manager and chief engineer with pertinent information so they can determine the highest risk value that is feasible for the program. Therefore, it is very important that the aerospace vehicle program manager and the chief engineer have a good understanding of the operational risks due to the selected design terrestrial environment. Terrestrial environment elements may significantly affect multiple areas of an aerospace vehicle’s design, and thus operational capabilities, including structure, control, trajectory shaping (performance), heating, takeoff and landing capabilities.

SOME LESSONS LEARNED CONCERNING THE TERRESTRIAL ENVIRONMENT AND AEROSPACE VEHICLE DEVELOPMENT INTERACTIONS The NASA Marshall Space Flight Center’s Natural Environments Branch and its predecessor organizations have over 50 years of experience in the development and interpretation of terrestrial environment requirements to be used in the design and operation of aerospace vehicles. During this period, a large number of “lessons learned” have formed the basis for the contents of the report “Terrestrial Environment (Climatic) Criteria Guidelines for Use in Aerospace Vehicle Development” (Johnson, 2008) and a conference presentation by the authors (Vaughan and Johnson, 2010). Some of these lessons learned are summarized in the following list. Wind Vectors Versus Engineering Vector Conventions • Background – Flight mechanics use of wind vectors versus conventional meteorological usage. In the case of flight mechanics, the vector is stated in relation to the direction in which force is being applied. However, in meteorology, the wind vector is stated relative to the direction from which the wind force is coming. • Lesson – The proper interpretation and application of wind vectors is important to avoid a 180 deg error in structural loads and control system response calculations. Design Requirements, Not Climatology • Background – While based on climatology and models, both physical and statistical, terrestrial environment requirements are part of the overall vehicle design effort that is necessary to ensure that mission operational requirements are met. Thus, they must be selected and defined on this basis. Simply making reference to climatology or databases will not produce the desired vehicle design. • Lesson – Members of the natural environment groups assigned as the control point for inputs to a program must also be part of the vehicle design team and participate in all reviews etc., to ensure proper interpretation and application of natural environment definitions/requirements concerning the overall vehicle design needs.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.7-14, Jan.-Mar., 2013

Vaughan, W.W. and Johnson, D.L.

Early Input of Natural Environment Requirements Based on Interpretation of Mission Purpose and Operational Expectations • Background – One needs to develop the natural environment definitions and requirements for a program as soon as possible after knowing the level required for the program’s mission. Thus, all of those concerned with the development will have a common base with associated control on changes made to natural environment definitions/requirements and associated vehicle operational impacts. • Lesson – The definition of natural environment needs for a vehicle that are necessary to meet the mission requirements is important for all of those concerned with the program. This provides visibility to all, especially the program manager and systems engineers, in relation to the impact on the operation of the vehicle and to the natural environment design requirements on the program’s mission. Consistent Input for all Users More Important for Tradeoff and Design Studies Than Different Inputs on Natural Environment Topic • Background – The natural environment is one of the key drivers for many of the design efforts on an aerospace vehicle’s thermal, structural, and material control. Differences in natural environment inputs used by various design groups can mask critical engineering design inputs if not avoided by consistent and coordinated natural environmental inputs and interpretations for engineering applications. • Lesson – The need for a focused natural environment group which provides coordinated and consistent environment definitions/requirements/interpretations is key to have all of those involved direct their efforts toward the same inputs, thus contributing to engineering applications that can readily be interpreted from a common base. Ability to Test New or Changes in Natural Environment Requirements Versus Results Important Before Implementing Them as Formal Requirements • Background – Preliminary assessment of natural environment definitions and requirements must first be accomplished in collaboration with a responsible engineering group in order to identify design drives versus mission requirements. Based on this information, J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.7-14, Jan.-Mar., 2013

the appropriate natural environment definitions and requirements can be implemented and controlled accordingly. • Lesson – To avoid problems with the engineering interpretation of natural environment definitions and requirements, the natural environment groups in charge must first interact directly with an appropriate engineering group to ensure proper use and interpretation when formally implemented as part of the overall program requirements. Maintain Natural Environment Requirements for Design and Operation of Vehicle as Base From Which Other Requirements are Related • Background – Taking this action provides a viable and robust operational vehicle capability that will meet the vehicle’s mission requirements. Otherwise, a vehicle will be produced with lower operational capability based on natural environment conditions. Natural environment operational requirements can be monitored, and decisions can be made regarding launch operations, etc., or, in case monitoring is not practical or in an emergency, the vehicle will then be functional concerning probable natural environment conditions established on the basis of past records and mission requirements. • Lesson – Do not design an aerospace vehicle with the natural environment design requirements incorporated as one of the non-nominal inputs and RSS (root sum squared) in the final vehicle design decision. Natural Environment Elements That Cannot be Monitored Prior to Operational Decision Must be Minimum Risk Level Possible Consistent With Mission Capability Requirements • Background – For an aerospace vehicle launch, most terrestrial environment elements can be monitored and thus taken into account before making a launch decision. The same is true for some on-orbit and deep-space spacecraft natural environment operational requirements. In such cases, lower probability of occurrence environments may be considered, consistent with mission requirements, along with subsequent cost savings on design. Vehicle ascent winds through max Q versus re-entry winds is an example of lower probability (higher risk of occurrence) versus higher probability (lower risk of occurrence) natural environment design requirements

Aerospace Meteorology: An Overview of Some Key Environmental Elements

for a vehicle. However, minimum risk of natural environment requirements must be used for design to ensure operational capability when natural environments cannot be measured or monitored. • Lesson – It is necessary to carefully analyze the mission requirements relative to vehicle operations and provide the natural environment definitions and requirements accordingly in collaboration with the vehicle program manager to enable the understanding of the implications of environments provided for design. Maintain Natural Environment Requirements for Design as a Separate Document but Integral to Overall Mission Requirements for Vehicle • Background – The natural environment definitions and requirements for the Space Shuttle and Space Station were provided so they could be controlled and available in separate program documents, as part of the overall design requirement documentation. This not only provided direct access for all of those concerned with use of natural environment inputs into design and mission planning but also provided an easy control of inputs. Changes, where required, were readily enabled with the change of one document that had application for all natural environment design requirement inputs to the program. • Lesson – Each vehicle development program should have only one natural environment definition and requirement document. It should be an integral part of the overall mission requirements for the vehicle design, development, and operations, and be controlled accordingly. Atmospheric and Space Parameter Analysis Model • Background – The ability for a program manager to easily access information on the operational capability impact of a vehicle design change according to the natural environment is an important tool for decision making. In addition, such a tool provides additional insight into mission planning activities, including launch and landing delay probabilities. • Lesson – Knowledge by mission managers, chief engineers, mission planners, etc. on the availability of an Atmospheric and Space Parameter Analysis Model is a valuable decisionmaking tool and should be used to make the tradeoff decision when the desired operational natural environment is a factor.

Reference Period for Design Statements of Natural Environment Definitions and Requirements Relative to Launch and On-Orbit Operations • Background – For launch statements on natural environment definitions and requirements, the worst reference month should be used. This provides an operational capability regarding the natural environment that ensures that for any given month, the desired operational capability will be met. Thus, for the worst month reference period, the minimum risk of launch delay due to the natural environment will occur, and all the other months will have less probability of launch delay. The same situation exists for natural environments associated with on-orbit operational capability and deep-space operations. In other words, for these cases the anticipated lifetime in these operational conditions must be taken into account along with the acceptable risk for comprising the mission concerning the natural environment conditions exceeding the design requirements. • Lesson – All launch terrestrial environment definitions and requirements for the design of a vehicle must be made with respect to the worst month reference period. For natural environments associated with on-orbit and deep-space operations, the anticipated lifetime in these operational conditions must be taken into account along with acceptable risks for operations. Life-Cycle Cost Estimates and Natural Environment Operational Constraints of Vehicle • Background – Once a vehicle has been developed, the constraints relative to operations in the natural environment should be assessed based on the resulting capability of the vehicle. This is the case for launch, on-orbit, and deep-space aspects of the mission. An Atmospheric and Space Environment Parameter Analysis Model can be especially helpful in this regard. The resulting information should be incorporated into the development of the full life-cycle cost estimates and model for the vehicle program. • Lesson – Consideration needs to be given to the effect of natural environmental constraints on launch vehicle and spacecraft operation when developing full life-cycle cost estimates and models.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.7-14, Jan.-Mar., 2013

Vaughan, W.W. and Johnson, D.L.

CONCLUSION The natural terrestrial environment plays a significant role in the design and operation of aerospace vehicles and in the associated integrity of aerospace systems and elements. This paper has addressed some of the key vehicle

and terrestrial environment areas of concern plus some of the “lessons learned” that have been identified and documented over a number of years. Additional lessons learned from an engineering perspective may be found in publications by Blair et al. (2001), Pearson et al. (1996), and Ryan et al. (1996).

REFERENCES Bellue, D.G., Boyd, B.F., Vaughan, W.W., Garner, T., Weems, J.W., Madura, J.T. and Herring, H.C. 2006, “Weather Support to the Space Shuttle An Overview: paper Number AIAA-2006-0684”, 44th AIAA Aerospace Sciences Meeting, Reno, NV. Blair, J.C., Ryan, R.S., Schutzenhofer, L.A. and Humphries, W.R. 2001, “Launch Vehicle Design Process: Characterization, Technical Integration and Lessons Learned” NASA/TP—2001–210992, NASA Marshall Space Flight Center, AL. Johnson, D.L., 2008, “Terrestrial Environment (Climatic) Criteria Guidelines for Use in Aerospace Vehicle Development”, NASA/ TM-2008-215633, NASA Marshall Space Flight Center, AL, from http://ntrs.nasa.gov/search.jsp?R=20090022159 Pearson, S.D., Vaughan, W.W., Batts, G.W. and Jasper, G.L., 1996, “Importance of the Natural Terrestrial Environment With Regard to Advanced Launch Vehicle Design and Development”, NASA TM 108511, NASA Marshall Space Flight Center, AL.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.7-14, Jan.-Mar., 2013

Ryan, R., Blair, J., Townsend, J. and Verderaime, V., 1996, “Working on the Boundaries: Philosophies and Practices of the Design Process”, NASA TP-3642, NASA Marshall Space Flight Center, AL. Vaughan, W.W., 2010, “Guide to Reference and Standard Atmosphere Models”, AIAA-G-003C-2010, American Institute of Aeronautics and Astronautics, Reston, VA, from http://www.aiaa.org. Vaughan, W.W. and Johnson, D.L., 2010, “Aerospace Meteorology: Some Lessons Learned”, Proceeding of the 14th Conference on Aviation, Range and Aerospace Meteorology (ARAM), American Meteorological Society, Boston, MA. Vaughan, W.W., Johnson, D.L., Pearson, S.D. and Batts G.W., 1997, “The Role of Aerospace Meteorology in the Design, Development and Operation of New Advance Launch Vehicles”, Proceedings of the Seventh Conference on Aviation, Range and Aerospace Meteorology, American Meteorological Society, Boston, MA.

doi: 10.5028/jatm.v5i1.192

A Sensitivity Investigation on the Aeroelastic Dynamic Stability of Slender Spinning Sounding Rockets Roberto Gil Annes da Silva1, José Guido Damilano1, João Luiz F. Azevedo1

Abstract: The present work addresses a sensitivity analysis investigation of the aeroelastic stability margins for the VSB-30 sounding rocket during the atmospheric flight phase. Parametric stability analyses are performed considering variations of the inertia properties of the modular payload. Such variations can be caused by different type and/or number of experiments (payload modules). The aerodynamic model is based on a supersonic unsteady potential aerodynamic method. Freestream conditions depend on the flight speed and atmosphere. An equivalent structural dynamic model of the rocket is represented by a beam-like structure. The objective of this investigation is to establish an aeroelastic model for aeroelastic stability and response analyses, as well as a procedure for the identification of stability margins for rockets. The resulting aeroelastic model should be further used in MDO processes for the improvement of the vehicle flight performance. The results of the present effort indicate that the flutter behavior of the VSB-30 sounding rocket is sufficiently robust inside the operational envelope, even considering the environmental and loading conditions. The spinning effect, in this case, does not play a significant role, because the flutter margins remain almost unaltered with and without VSB-30 body spin. Keywords: Sounding rockets, Aeroelastic analysis, Flutter margins, Sensitivity studies.

INTRODUCTION European research on microgravity required a new sounding rocket with performance similar to the one delivered by the English rocket Skylark-7, whose production had been discontinued. Instituto de Aeronáutica e Espaço (IAE), in Brazil, took on the task of the rocket development and complete integration in a joint effort with the Deutsches Zentrum für Luft und Raumfahrt (DLR), responsible for the payload of microgravity experiments. Several modifications were performed on the previously developed VS-30 sounding rocket, in order to satisfy the required specifications for new scientific and technological experiments in the microgravity environment. The resulting modified vehicle, named VSB-30, is a two-stage spinning-stabilized slender sounding rocket, with two sets of three fins on each stage, whose engines use a solid propellant (Duarte et al., 2005). The VSB-30 rocket flight operation is divided into two phases: the first stage flight (FSF) and the second stage flight (SSF). The engines accelerate the vehicle to a ballistic flight path towards the desired microgravity condition. A more recent application of the baseline VSB-30 sounding rocket configuration regards in-flight experiments on the aerodynamic behavior and thermal problems of an unconventional asymmetric shape for reentry vehicles comprising multifaceted surfaces with sharp edges (SHEFEX) (Turner et al., 2006). The aim of the experiment was to correlate numerical analysis results with actual flight data regarding the aerodynamic effects and structural concept for the thermal protection system. Hence, the development of an aeroelastic model of the VSB-30 vehicle should be very useful

1.Instituto de Aeronáutica e Espaço – São José dos Campos/SP – Brazil Author for correspondence: João Luiz F. Azevedo | Praça Marechal Eduardo Gomes, 50 – Vila das Acácias | CEP 12228-903 São José dos Campos/SP – Brazil | E-mail: joaoluiz.azevedo@gmail.com Received: 14/11/12 | Accepted: 06/02/13

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.15-26, Jan.-Mar., 2013

Silva, R.G.A., Damilano, J.G. and Azevedo, J.L.F.

for the aeroelastic stability and response investigation for any type of proposed configurations, including SHEFEX. Typically, slender-body finned vehicles can present two types of flutter mechanisms, i.e., fin bending-torsion flutter and body pitch-bending flutter (Martin, 1958). Examples of rocket flutter analysis have been well documented in several reports and articles (Azevedo, 1988; Bae et al., 2004; Francesco Capri et al., 2006; Garcia-Fogeda and Liu, 1988; McNamara and Friedmann, 2006; Paek et al., 2002). Some of them concern fin flutter, for example, when the rocket presents deployable fins. Nonlinear phenomena, such as free-play, also lead to nonlinear aeroelastic behavior (Bae et al., 2004). In the literature, several efforts have been well documented (Chae, 2004; Chae and Hodges, 2003; Haddadpour, 2006; Heddadj and Cayzac, 2000; Livshits et al., 1996; Meyers, 1973; Murphy and Mermagen, 2001; Murphy and Mermagen, 2005; Platus, 1992), regarding the investigation of the flight dynamic/aeroelastic response behavior of free flight spinning (or not) rockets and missiles. Moreover, in some studies the coupling between flight mechanics, aeroelasticity, and control system (aeroservoelastic coupling) has been investigated (Haddadpour, 2006). Motivations for these studies are usually associated to the fact that flight vehicles with slender configurations might experience aeroelastic instability during flight due to coupling between a short period rigid-body mode and a body-bending mode. Such a coupling might affect the planned trajectory of the vehicle and, in the case of military rockets, the weapon aiming system could be compromised. Most of the previously cited investigations employ slender body theory for the aerodynamic model of the elastic body, combined with a quasi-steady aerodynamic model representing the body global aerodynamic coefficients. However, for the best assessment of aeroelastic instabilities, a more accurate prediction method to compute the unsteady aerodynamics was presented by Garcia-Fogeda and Liu (1988), the so-called Harmonic Potential Panel (HPP) method. Further development on more accurate methods, based on lifting surface theory, for the prediction of unsteady airloads aeroelastic to analyze supersonic wingbody configurations were presented by Chen and Liu (1990), and Liu et al. (1997), which is, in fact, a unified hypersonicsupersonic lifting surface method. Nevertheless, lifting surface methods are limited to linear computation of the unsteady airloads. For this reason,

computational fluid dynamics for unsteady aerodynamic modeling of transonic airloads has also been used for transonic aeroelastic analysis of body alone rockets (Azevedo, 1988) and wing-body-like vehicles (Francesco Capri et al., 2006). A review of the state-of-the-art on advanced methods for computational fluid dynamics and heat transfer applications of hypersonic aeroelasticity and aerothermoelasticity is presented by MacNamara and Friedmann (2006). The scope of the present investigation is the study of the aeroelastic stability (flutter) of a free flight spinning rocket. The strategy for the parametric investigation regarding a sensitivity analysis should be based on the identification of the flutter margins as a function of the variation of the inertial properties of the payload, solid propellant consumption, as well as the flight environmental conditions. A set of adequate steps for an aeroelastic investigation of sounding rockets is proposed. The procedure for the analysis, as well as the assumed hypotheses, should be mainly based on the flight dynamic behavior of the vehicle during its operation. It is important to note that differently from conventional aircraft, the rocket atmospheric flight does not present a steady behavior as in a cruise condition. The vehicle should be subjected to strong environmental variations associated to its flight level, flow speed and its inertial characteristics. MATHEMATICAL MODELS Structural Dynamic Model The VSB-30 airframe is represented by a finite element structural dynamic reduced order model, composed by beam and plate elements connected by nodes to represent the body, the fin spars and the fin surfaces, respectively. Such an approach is sufficient for the aeroelastic stability analysis of the vehicle. In the vehicle body nodes, the distribution of inertial characteristics is approximated by lumped masses representing the inertia properties. A sketch of the finite element model, including the distributed masses, will be presented in the forthcoming paragraphs. Since the rocket engines burn the solid propellant, a considerable rate of change in mass should be considered in the modeling of the vehicle structure. Thus, mass properties represented in the structural dynamic model, for each flight condition, change as a function of time. Furthermore, depending on the payload configuration, there is also a significant change in mass to be considered for the structural dynamic modeling. The structural

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.15-26, Jan.-Mar., 2013

A Sensitivity Investigation on the Aeroelastic Dynamic Stability of Slender Spinning Sounding Rockets

model was developed with the MSC/NASTRAN® software (MSC/NASTRAN, 1995) to compute the natural frequencies and mode shapes for the analysis, assuming a piecewise linear model, because the mass variation should be assumed as constant at a given instant. The stability of the VSB-30 rocket is guaranteed by the fins and the induced spin after liftoff. Therefore, it would be desirable to include spinning effects in the structural dynamic modeling. Spinning effects can lead to changes on mode shapes and natural frequencies of structures. Paek et al. (2002) presented results on the investigation of wraparound finned projectiles flutter. Results of their aeroelastic analysis indicate that flutter speeds depend on the roll rate. Furthermore, in that special case, there was a significant deformation of the wrapped fin with the increase of the roll rate. For their analysis the order of the roll rate was about three times the first natural frequency of the body and ten times smaller than the first natural frequency of the fin (1st fin bending). So, it was possible to conclude that the spinning effects had to be taken into account for the identification of the rocket mode shapes and natural frequencies. Figure 1 presents the roll rate for the VSB-30 vehicle until 40 seconds of flight. It is observed that the rolling frequencies are moderately low. Therefore, such an effect will not be considered in the VSB-30 rocket analysis because the spinning rates are between four to six times smaller than the first mode shape of the vehicle (body 1st bending), thus not inducing significant changes in stiffness nor in gyroscopic effects. On the other hand, even in this case, an investigation on the spinning effects was performed, supposing that this

Roll rate (Hz)

2.5 2 1.5 1 0.5 0

20 Time (s)

Figure 1. Roll rates of the VSB-30 vehicle after liftoff.

aeroelastic analysis methodology could be applicable to other classes of sounding rockets. From the dynamics rotating system, it is desirable to consider spinning effects, which might include the Coriolis and centrifugal effects by introducing the gyroscopic damping matrix into the dynamic equations, as well as the differential stiffness due to nonlinear effects (Heddadj and Cayzac, 2000). These effects are significant when the spinning roll rates are larger than the current ones (Fig. 1). The inclusion of such effects may lead to structural dynamic changes that might influence the flutter mechanisms of the airframe. A MSC/NASTRAN DMAP option was used to compute the modal characteristics of the rotating vehicle, and the corresponding theory is well documented in MSC/NASTRAN V70 (1995). Unsteady Aerodynamic Model In the present investigation, only aeroelastic analyses at supersonic flow conditions will be performed. The aerodynamic modeling method of unsteady linear potential flows is based on the discrete kernel function approach. The development of discrete element kernel function methods is based on integral solutions of the small disturbances linearized potential flow equation. In this work, the ZONA7U method (ZAERO, 2003), implemented in the ZAERO® software system, has been used to model the unsteady airloads for flutter analysis (Liu et al., 1997). Chen and Liu (1990) present the mathematical formulation of the ZONA7 method for supersonic aerodynamic modeling of wing body configurations. Its extension, which accounts for the lifting surface thickness effects, is named ZONA7U method (Liu et al., 1997), and is also implemented in the ZAERO® software package. The ZONA7U method is different from ZONA7 due to the inclusion, in the former, of the thickness effect correction. Thickness corrections are an important issue to be considered, leading to the improvement of the flutter speed prediction (Liu et al., 1997). In most cases, the flutter speed decreases when the fin thickness is considered (Liu et al., 1997). Special care should be taken during the meshing process. For example, depending on the Mach number, the lifting surface elements (panels) should have an adequate aspect ratio for the computation of supersonic flows. Another important issue to be considered is the need for a mesh convergence study based on the resulting flutter speed dispersion.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.15-26, Jan.-Mar., 2013

Silva, R.G.A., Damilano, J.G. and Azevedo, J.L.F.

Such study might be seen as a tool to identify the best costbenefit relation regarding the size of the aerodynamic mesh for modeling the body and lifting surfaces unsteady aerodynamics. AEROELASTIC MODELING The scope of the aeroelastic analysis to be presented here is the flutter investigation of a spinning sounding rocket based on the unsteady aerodynamic model of the full vehicle. It is usual to investigate rocket fin flutter by just modeling the isolated fin (Bae et al., 2004; Paek et al., 2002). However, possible interference effects could influence the unsteady aerodynamic behavior of the vehicle, and, for this reason, the full vehicle should be modeled. The structural dynamic behavior is also assumed as linear, and the spinning effect is neglected for the aforementioned reasons. It is supposed, as a first approach, that the present aeroelastic model is linear regarding the aerodynamic and structural models. The scope here is the investigation of the vehicle flutter mechanisms. Small aeroelastic displacements are assumed and the development of this baseline aeroelastic model will be important for further studies on Multidisciplinary Design and Optimization (MDO) applied to the vehicle design and upgrade processes. Also, this baseline linear model should be important for future development on computational aeroelasticity applied to transonic aeroelastic stability and response, assuming nonlinear aerodynamic models for correction of the linear aerodynamic methods (ZAERO, 2003). Description of the Vehicle The VSB-30 sounding rocket is composed of two tandem rocket engines. Its total length is 12.699 m and it has a total mass of about 2,200 kg at liftoff. Evidently, this may vary depending

on the payload configuration. In the present study, only the aeroelastic analysis of the VSB-30 sounding rocket assembled with modular cylindrical payload modules is investigated. These modular payloads are assembled in tandem as a set of two, three or four modules. They are mass balanced in relation to the vehicles longitudinal axis (assumed here as the “x” axis). Figure 2 presents the vehicle finite element structural model and the corresponding CAD drawing. Basic Hypotheses and Assumptions The VSB-30 vehicle is a ballistic sounding rocket whose flight altitude range from sea level is 250 km. Its flight path is designed in order to achieve sufficient time in microgravity conditions for scientific and technological experiments. The vehicle reentry flight phase is not included in the scope of the present investigation. Figure 3 shows the typical altitude, Mach number, dynamic pressure and angle of attack for this class of vehicles as a function of time. The data shown represent the case for which the vehicle comprises four payload modules for scientific experiments. Information presented in Fig. 3 allows the formulation of hypotheses and conditions to set the aeroelastic analysis procedure. One should observe that the Mach number varies up to the hypersonic flow condition, remaining constant afterwards. However, while the vehicle flies within the atmosphere, the maximum Mach number is supersonic (3.313) at the maximum dynamic pressure, which can be obtained from the corresponding chart in Fig. 3. In fact, one can observe that there are two dynamic pressure peaks. Furthermore, each of these peaks is associated with the vehicle configuration in each stage: in the first stage, when the vehicle remains with two tandem rocket engines

Payload 2nd Stage fins S31 Booster z

S30 Motor

Booster adapter 1st Stage fins

Figure 2. VSB-30 rocket: the finite element beam-plate model and the CAD drawing.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.15-26, Jan.-Mar., 2013

Motor-payload adapter

Mach number

200000 150000 100000 50000 0

60 Time (s)

100

120

Altitude (km)

200 160 120 80 40 0

60 80 Time (s)

100

120

Angle of Attack (deg)

Dynamic pressure (pa)

A Sensitivity Investigation on the Aeroelastic Dynamic Stability of Slender Spinning Sounding Rockets

8 7 6 5 4 3 2 1 0

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

20 30 Time (s)

Figure 3. Typical time history curves describing the performance of the VSB-30 vehicle.

and two set of fins, and, in the second stage, in which the vehicle is composed by a single rocket engine and a single set of fins. This is the reason why the aeroelastic analysis procedure should be divided into first and second stage flight analyses. Since the stages separate, it will be necessary to consider two distinct dynamic models of the airframe, one before separation and another after separation. In the case of a non matched-point flutter analysis, parametric variations around the maximum dynamic pressure condition are assumed. These parametric variations should be set as freestream speed and Mach number, flight altitude and mass variations regarding the solid propellant consumption and number of payload modules. This first approach for the aeroelastic analysis provides the sensitivity of the aeroelastic stability due to payload and solid propellant variations. The spinning behavior of the vehicle is associated with the body roll rate. The spin may promote Coriolis and Magnus effects. Such effects could have an important influence on the flight dynamic behavior of the vehicle. The idea of inducing a spin during flight is related to improving the flight dynamic longitudinal stability margins of the vehicle. The reader should keep in mind that a sounding rocket is essentially a ballistic body, that is, there are no controls acting during its flight path. Coriolis effects are associated to gyroscopic forces which result from the spin. For this reason, the structural dynamic behavior of the vehicle should be altered by the appearance of a gyroscopic damping matrix and a differential stiffness due

to centrifugal forces (Francesco Capri et al., 2006). However, this type of effect can be negligible when the vehicle roll rates are small. For the present investigation, it is assumed that roll rates of 1.42 Hz and 0.91 Hz at the first and second stages, respectively, are sufficiently small and can be neglected, since they are not able to promote significant changes in the structural dynamic behavior of the vehicle. Another effect associated to a spinning body is the Magnus effect. This type of effect is represented by a lift force induced by the rotating body in angle of attack. This effect changes the mean flow around the vehicle, and it is significant when the angle of attack is high. Thus, the small disturbance flow should not be altered by the loadingchangesaroundameansteadyflowcondition.Furthermore, from the corresponding chart in Fig. 3, it is possible to observe that the angle of attack is sufficiently small throughout the flight path, resulting in negligible Magnus effect. Structural Dynamic Model The VSB-30 rocket is represented by a beam-like equivalent structure, sufficiently accurate to capture the global mode shapes of the airframe. Moreover, the fittings are represented by a set of constraints which might represent the physical coupling between the parts of the vehicle. The fins are represented by a combination of beams and plate elements, as in the real structure. The second stage engine fin internal structure resembles a “spider web” beam structure, as observed in Fig. 4.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.15-26, Jan.-Mar., 2013

Silva, R.G.A., Damilano, J.G. and Azevedo, J.L.F.

Figure 4. Finite element model of the “spider web” like structure of the second stage fins.

The mass of the solid propellant is distributed in several points along the vehicle longitudinal axis, and its consumption is represented by subtracting their corresponding value, as the propellant is burned, for each instant of analysis. Aerodynamic Model The aerodynamic model is developed using the ZAERO software (Chen, 1999). A body of revolution subdivided into panels was assumed to represent the aerodynamic shape of the vehicle. For the fins, a flat plate aerodynamic representation is assumed, including the thickness correction (ZAERO, 2003). The correction for thickness effects is important for Mach numbers higher than 1.2. The reader should remember that the aeroelastic stability behavior is less conservative when this correction is not considered (ZAERO, 2003). Figure 5 shows the aerodynamic meshes used for the present investigation. Each payload configuration is represented by a proper aerodynamic mesh. Furthermore, for the first stage flight phase, the aerodynamic model represents the whole

vehicle including the two sets of fins. The chosen Mach numbers for the stability investigation are the corresponding values at maximum dynamic pressure conditions. Table 1 shows the main flight parameters for each configuration and flight phases. The maximum dynamic pressure identified during the flight of the first stage occurs when the flight altitude is around 3,400 m, at Mach number 1.6, that is, 12.5 seconds after liftoff, for the case of the vehicle configured with two payload modules. These conditions are taken as reference to perform the aeroelastic stability analyses. They are different from other configurations of the vehicle, since they also depend on the number of payload modules. It is assumed that the Mach number and flight altitude, for the reference conditions, are constant, but the freestream speed varies around the specified Mach number. Table 1 shows the flight parameters for each of the configurations analyzed, as well as the fight phases represented as “first” and “second” stages. The number of panels, used in the aerodynamic mesh for the body, depends on the configuration to be

Table 1. Reference conditions for the aeroelastic stability analyses. Configuration flight phase

Time (s) 1st stage

Dyn. Pressure (Pa)

2nd stage

1st stage

2nd stage

Mach number 1st stage

2nd stage

Altitude (m) 1st stage

2nd stage

2 modules

12.5

24.2

117420.0

179525.5

1.599

3.375

3518

11050

3 modules

12.5

24.5

112684.4

175899.4

1.558

3.337

3434

11037

4 modules

12.5

24.8

108724.7

172895.0

1.523

3.313

3363

11062

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.15-26, Jan.-Mar., 2013

A Sensitivity Investigation on the Aeroelastic Dynamic Stability of Slender Spinning Sounding Rockets

analyzed. However, the number of panels, used to mesh the fins, remains constant since these structures do not change according to the configurations to be analyzed. The remaining task to proceed in the flutter analysis is a mesh convergence study to evaluate the robustness of the aerodynamic model for stability investigations. The idea is to identify which are the flutter modes and to proceed with the mesh refinement only in the parts of the vehicle in which modal displacements contribute for the flutter mode. Such investigations will be discussed further, as soon as the flutter mechanisms are identified. FLuTTER ANALySIS RESuLTS The aeroelastic analysis procedure is based on the investigation of the aeroelastic stability within the vehicle flight envelope. The strategy adopted here is to observe the dynamic pressure-time history, demonstrated in Fig. 6. In this figure, the red line indicates that the points below it may not be included in the aeroelastic analysis procedure. This is because this decrement in the dynamic pressure does not offer any risk in terms of flutter, since larger dynamic pressures are reached before, as shown in Fig. 6. The non-matched point flutter analysis will be based on the variation of the freestream speed at each of the maximum dynamic pressure peaks, each one associated to the first and second stage flight phases. This is not the best approach for flutter substantiation since only a single eigenvalue will

Dyn. Press (Pa)

figure 5. aerodynamic meshes for the vsb-30 rocket for two stages configuration and four modules (right), and the single second stage for 2 to 4 modules (left).

200000 180000 160000 140000 120000 100000 80000 60000 40000 20000 0

15 20 25 Time (s)

figure 6. Typical dynamic pressure-time history.

represent the true aerodynamic damping (Chen, 1999) and the associated aeroelastic frequency. Nevertheless, this approach is fast enough to qualitatively identify the flutter mechanism to be explored on a subsequent matchedpoint flutter analysis. The reference Mach number and density are the same values identified at the maximum dynamic pressure condition. The resulting flutter mechanism identified by the nonmatched point flutter analysis is represented by the coupling of the 7th and 9th anti-symmetric modes. These modes are the second stage fin bending and torsion modes, which are illustrated in Fig. 7. It is important to note that a possible introduction of the spinning effect should be responsible for altering the

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.15-26, Jan.-Mar., 2013

Silva, R.G.A., Damilano, J.G. and Azevedo, J.L.F.

structural dynamic behavior. The flutter mechanism could be anticipated or delayed, depending on the coalescent modes. For this reason, further investigation is recommended in order to introduce the spinning effect in the vehicle dynamic model, due to the influence of the gyroscopic effects produced by the spin. The results of the sensitivity investigation based on the number of payload modules are summarized in Table 2.

7th mode (37.2 hz)

The reader may observe that the computed flutter speeds associated with the number of modules are related to the same flutter mechanism, and the flutter speed reduction is less that 4%. This fact is a good indication that the vehicle is sufficiently robust in terms of the variations of payload mass regarding the aeroelastic stability. Moreover, this result makes sense because the flutter mode results from a coalescence of two anti-symmetric fin mode shapes. That is, the dynamic contributions of the vehicle body modes do not play an important role in the flutter mechanism. From the discussion above, it is possible to conclude that this change in flutter speed should be mostly related to the vehicle environmental conditions, since the dynamic pressure for each of the configurations is different. Figure 8 shows a comparison between the dynamic pressures at the same time frame for two, three and four payload modules. The most critical configurations occur when the vehicle flies with two payload modules. Such configuration, during the atmospheric flight phase, subjects the vehicle to dynamic pressures higher than the others do. The next step in the study was an analysis still based on a nonmatched point flutter solution. However, unlike the previous procedure, this one was repeated for different Mach numbers, each of which associated to a different flight condition. In this case for the unsteady aerodynamic model used to compute the flutter condition, these reference Mach numbers are used to

Dyn. Press (Pa)

9th mode (66.9 hz)

200000 180000 160000 140000 120000 100000 80000 60000 40000 20000 0

2 modules 3 modules 4 modules

flutter mode (39.7 hz)

figure 7. flutter mechanism in the first stage flight phase.

20 30 Time (s)

figure 8. comparison of the dynamic pressure time history for two, three and four payload modules.

table 2. flutter speeds and dynamic pressure for the vsb-30 vehicle in the first stage of flight â&#x20AC;&#x201C; non-matched point flutter analysis. Number of modules

7th and 9th

Flutter speed (m/s)

556.4935

550.0458

536.4167

Flutter frequency (Hz)

40.0591

39.8965

39.6766

Dynamic pressure (Pa)

1.355E+05

1.324E+05

1.259E+05

Flutter mode

J. Aerosp. Technol. Manag., SĂŁo JosĂŠ dos Campos, Vol.5, No 1, pp.15-26, Jan.-Mar., 2013

A Sensitivity Investigation on the Aeroelastic Dynamic Stability of Slender Spinning Sounding Rockets

generate the aerodynamic influence coefficients. Figure 9 shows the computed flutter dynamic pressures, when the vehicle is configured with two payload modules. The results represented in Fig. 9 indicate that flutter dynamic pressures are above the dynamic pressure imposed on the vehicle during actual flight. It is possible to conclude, based on the studied trajectories, that increasing the number of payload modules reduces the chance for flutter, because the higher vehicle masses will decrease the flight speeds in lower altitudes. In the second stage flight, environmental conditions are more favorable because the higher altitudes lead to lower densities and dynamic pressures. On the other hand, flight speeds are higher, approaching two times the Mach number of the first stage flight. For this analysis, the mass ratio concept is introduced to help understand why at higher dynamic pressures the same airframe remains stable in terms of flutter, as demonstrated in Table 3. A lower mass ratio leads to a lower amount of energy needed to promote the coalescence of aeroelastic modes which may lead to flutter. On the other hand, there is a need for high kinetic energy flow from which energy would be extracted to 160000 140000

Dyn. Press (Pa)

120000 100000 80000 60000 40000

VSB-30 Dyn. Press Flutter Dyn. Press

20000 0

1.1

1.2

1.3

1.4

1.5

1.6

Mach Number figure 9. flutter dynamic pressures for vsb-30 rocket, 1st stage flight, two payload modules.

cause flutter. Looking at Fig. 10, it is possible to understand why, at higher altitudes and Mach numbers, even though the dynamic pressure is high, the mass ratio is also high enough, thus retarding the coalescence of the modes for flutter. This explains why it should be necessary to concentrate efforts on analyzing flutter stability nearby transonic Mach numbers. The matched-point flutter analysis considering the pair density (altitude)/Mach number is the subsequent step in the whole process. Thus, the true aeroelastic damping curves, as functions of the freestream flow conditions, are computed. This kind of analysis is more expensive, computationally speaking, since it is necessary to calculate an aerodynamic influence coefficient matrix for each Mach number included in the profile to be investigated. Once these matrices are generated, it is possible to repeat the flutter computation for a set of Mach-altitude (density) pairs. The results of the matched-point flutter analysis indicate that the vehicle is free from flutter mechanisms inside its flight envelope. However, the critical configuration, in terms of flutter, is the one in which the vehicle is configured with two payload modules. Figure 11 shows that in the typical matched-point flutter analysis it is likely that the 7th mode is unstable, even though its corresponding natural frequency remains practically constant. The inclusion of spinning effects considers the vehicle rotating between 0.5 Hz to 10 Hz. In the latter, the roll rates are beyond those at the maximum dynamic pressure condition during first stage flight. Here, the same conditions described above, i.e., vehicle with two payload modules, are investigated regarding the sensitivity of the flutter dynamic pressures due to a projected increase in roll rates. Figure 12 shows the sensitivity of the flutter dynamic pressure as a function of the roll rate. This result indicates that increasing the roll rate leads to a decrease in the flutter dynamic pressures, in the case of the coupling of the 7th and 9th modes. The reader should remember that these results refer to a non-matched flutter point analysis. Looking at Fig. 13, this behavior is clearly understood. Damping and frequency evolution, as the roll rates increase,

table 3. flutter speeds and dynamic pressure for the vsb-30 vehicle during the second stage flight. Number of modules

5th

No flutter

Flutter speed (m/s)

1950.773

No flutter

Flutter frequency (Hz)

54.1746

No flutter

Dynamic pressure (Pa)

7.376E+05

No flutter

Flutter mode

J. Aerosp. Technol. Manag., SĂŁo JosĂŠ dos Campos, Vol.5, No 1, pp.15-26, Jan.-Mar., 2013

Silva, R.G.A., Damilano, J.G. and Azevedo, J.L.F.

160 140

Mass Ratio

120 100 80 65 40 20 0 0

10 11

Altitude (km) 4

can be seen in the right side of Fig. 13. The coupling mechanism is altered in such a way that the flutter speed is lower with the increase of the roll rate. Such behavior occurs because the increase of the natural frequencies of the first fin antisymmetric bending mode is mostly subjected to differential stiffness effects in relation to the first fin anti-symmetric torsion mode. That is, the apparent increase in stiffness of the fin leads to an increase in the natural frequency. Therefore, there is a contribution for the coupling between the bending and torsion modes, while the frequency remains almost unaltered. Another interesting behavior to be noted concerns a second and third coupling between the 6th and 8th, and 5th and 10th modes, respectively. The coalescence of these modes leads to higher flutter speeds, which are, therefore, not taken into account in the present analysis. However, the reader should observe that the spinning effects for such couplings are more significant than

3.5

140 139

2.5

With roll rate

138 Dyn. Press (KPa)

Mach Number

2 1.5 1 0.5

W/O roll rate

137 136 135 134 133 132

0 0

131

10 11

130

Altitude (km)

Roll rate (Hz)

0 -0.01 -0.02 -0.03 G

-0.04 -0.05 -0.06 -0.07 -0.08 -0.09

1.2

1.3

Mach Number 1.4 1.5

1.6

figure 12. dynamic pressure as a function of the vehicle roll rate.

MODE-1 MODE-2 MODE-3 MODE-4 MODE-5 MODE-6 MODE-7 MODE-8 MODE-9 MODE-10 MODE-11 MODE-12 MODE-13 MODE-14 MODE-15 MODE-16 MODE-17 MODE-18 MODE-19 MODE-20

120 100 Frequency (Hz)

figure 10. Mass ratio and Mach number as functions of the flight altitude.

80 60 40 20 0

1.2

1.3

1.4 1.5 Mach Number

figure 11. Matched point flutter analysis results for the two-module payload configuration. J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.15-26, Jan.-Mar., 2013

1.6

A Sensitivity Investigation on the Aeroelastic Dynamic Stability of Slender Spinning Sounding Rockets

WHZ,MODE-5

WHZ,MODE-5 - 0,5 Hz

WHZ,MODE-6

WHZ,MODE-6 - 0,5 Hz

WHZ,MODE-7

WHZ,MODE-7 - 0,5 Hz

WHZ,MODE-8

WHZ,MODE-9

WHZ,MODE-10 WHZ,MODE-5 - 1 Hz

WHZ,MODE-6 - 1 Hz

WHZ,MODE-7 - 1 Hz

Frequency (Hz)

30 300

WHZ,MODE-9 - 1 Hz

30 300

G,MODE-5 G,MODE-6 G,MODE-7 G,MODE-8 G,MODE-9 G,MODE-10 G,MODE-5 - 1 Hz G,MODE-6 - 1 Hz G,MODE-7 - 1 Hz G,MODE-8 - 1 Hz G,MODE-9 - 1 Hz

Freestream Speed (m/s)

WHZ,MODE-5 - 10 Hz WHZ,MODE-6 - 10 Hz

40 35

1 0.8 0.6 0.4 0.2 0 -0.2 -0.4 -0.6 -0.8 -1 -1.2 -1.4

WHZ,MODE-10 - 0,5 Hz

WHZ,MODE-8 - 1 Hz

WHZ,MODE-10 - 1 Hz

WHZ,MODE-9 - 0,5 Hz

35 600 900 1200 1500 Freestream Speed (m/s)

WHZ,MODE-8 - 0,5 Hz

1 0.8 0.6 0.4 0.2 0 -0.2 -0.4 -0.6 -0.8 -1 -1.2 -1.4

WHZ,MODE-7 - 10 Hz WHZ,MODE-8 - 10 Hz WHZ,MODE-9 - 10 Hz

600 900 1200 1500 Freestream Speed (m/s)

G,MODE-10 - 1 Hz

WHZ,MODE-10 - 10 Hz

G,MODE-5 - 0,5 Hz G,MODE-6 - 0,5 Hz G,MODE-7 - 0,5 Hz G,MODE-8 - 0,5 Hz G,MODE-9 - 0,5 Hz

Frequency (Hz)

G,MODE-10 - 0,5 Hz G,MODE-5 - 10 Hz G,MODE-6 - 10 Hz G,MODE-7 - 10 Hz G,MODE-8 - 10 Hz G,MODE-9 - 10 Hz

Freestream Speed (m/s)

G,MODE-10 - 10 Hz

Figure 13. Evolution of the damping and frequencies as a function of the freestream speed.

the effects observed in the 7th–9th mode coupling. The reason for the augmented sensitivity due to increasing roll rates is associated to the types of mode shapes involved in the coupling. The modes, which primarily compose the associated flutter mechanisms, have an important contribution from the body displacement terms. Since body modes contribute to the coupling, the spin “softening effect” may also contribute for the early coalescence of the modes involved in the flutter mechanism.

CONCLUSIONS AND RECOMMENDATIONS Results from the present investigation show the aeroelastic behavior of the VSB-30 sounding rocket regarding flutter. The present effort is the beginning of an investigation whose objectives are beyond the scope the aeroelastic analysis presented herein. Unlike conventional aeroelastic analysis of aircraft, aeroelastic stability of rockets depends mostly on the environmental conditions and operational aspects, such as the flight phase. The present study serves as a guideline for future enhancement on the aeroelastic analysis process and for the design of flight instrumentation for test-analysis correlations. As previously

discussed, the vehicle is free from flutter throughout its flight envelope, considering reasonable flutter margins. However, the spinning effect is an influence to be considered in further investigations. In the present analyses, such effect was neglected due to vehicle low roll rates. As expected, it was found that the smallest flutter margins occur at lower mass ratio conditions, when the flow is transonic. Therefore, further investigation of the VSB-30 transonic aeroelastic stability margins is strongly recommended, preferably using a higher fidelity aerodynamic formulation. Moreover, aeroelastic dynamic responses should also be explored, besides aeroelastic stability, in order to quantify the vibration levels of the vehicle and to correlate them with flight vibration data. Actually, as a continuation of the present investigation, a study of the correlation of the aeroelastic response with flight data would be extremely helpful. At transonic flow conditions, there are severe shock induced vibration characteristics which can compromise the operation of the scientific instrumentation at the payload modules. Future refinements of the analyses here presented, in an attempt to further enhance the methodology for aeroelastic clearance of spinning sounding rockets, must include the effects of the gyroscopic stiffness and damping matrices. Aerodynamic model development for transonic flow

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.15-26, Jan.-Mar., 2013

Silva, R.G.A., Damilano, J.G. and Azevedo, J.L.F.

conditions must also be included. The aerodynamic models for transonic flow should be based on field panel methods or linear lifting surface methods corrected for transonic flow conditions. These approaches are the most adequate when the aeroelastic analysis is included in a MDO process. Furthermore, computational aeroelasticity analyses, based on the solution of the non-linear fluid dynamic equations, could be considered for validation of the transonic correction methods previously mentioned. Finally, since the VSB-30 rocket could be used for reentry aerodynamics, aerothermoelasticity at hypersonic flow conditions may also be included in the design process. The linear aeroelastic model, or the one based on non-linear unsteady

aerodynamics, could be employed for the aeroelastic stability and response analyses at reentry conditions.

ACKNOWLEDGMENTS The authors would like to acknowledge Mr. Sidney Servulo Cunha Yamanaka, for providing the flight dynamic data of the VSB-30 sounding rocket. Partial support for this work was provided by Conselho Nacional de Desenvolvimento Científico e Tecnológico, CNPq, under Research Grants No. 312064/2006-3 and No. 471592/2011-0.

REFERENCES Azevedo, J.L.F., 1988, “Transonic Aeroelastic Analysis of Launch Vehicle Configurations,” NASA CR – 4186.

Theory to a Unified Hypersonic-Supersonic Lifting Surface Method,” Journal of Aircraft, Vol. 34, No. 3, pp. 304-312.

Bae, J.S., Kim, D.K., Shih, W.H., Lee, I. and Kim, S.H., “Nonlinear Aeroelastic Analysis of a Deployable Missile Control Fin,” Journal of Spacecraft and Rockets, Vol. 41, No. 2, pp. 264-271.

Livshits, D.S., Yaniv, S. and Karpel, M., 1996, “Dynamic Stability of Free Flight Rockets,” 37th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference and Exhibit, Salt Lake City, UT.

Chae, S., 2004, “Effect of Follower Forces on Aeroelastic Stability of Flexible Structures,” Ph.D. Thesis, School of Aerospace Engineering, Georgia Institute of Technology, Atlanta, GA.

Martin, D.J., 1958, “Summary of Flutter Experiences as a Guide to the Preliminary Design of Lifting Surfaces on Missiles,” NACA TN – 4197.

Chae, S. and Hodges, D.H., 2003, “Dynamics and Aeroelastic Analysis of Missiles,” 44th AIAA/ASME/ASCE/AHS Structures, Structural Dynamics, and Materials Conference, Norfolk, VA. Chen, P.C., 1999, “A Damping Perturbation Method for Flutter Solution: The g-Method,” 2nd AIAA/CEAS International Forum on Aeroelasticity and Structural Dynamics, Williamsburg, VA. Chen, P.C. and Liu, D.D., 1990, “Unsteady Supersonic Computation for Arbitrary Wing-Body Configuration Including External Stores,” Journal of Aircraft, Vol. 27, No. 2, pp. 108-116. Duarte, J.A.A., Damilano J.G., Almeida, F.E. and Odaguiri, P.K., 2005, “Flight Data Used on the Evaluation of the Acceptance Testing Specifications for the VSB-30 Sounding Rocket,” Proceedings of the 17th ESA Symposium, Sandefjord, Norway. Francesco Capri, F., Mastroddi, F. and Pizzicaroli, A., 2006, “Linearized Aeroelastic Analysis for a Launch Vehicle in Transonic Flight Conditions,” Journal of Spacecraft and Rockets, Vol. 43, No. 1, pp. 92-104. Garcia-Fogeda, P. and Liu, D.D., 1988, “Supersonic Aeroelastic Applications of Harmonic Potential Panel Method to Oscillating Flexible Bodies,” Journal of Spacecraft and Rockets, Vol. 25, No. 4, pp. 271-277. Haddadpour, H., 2006, “Aeroservoelastic Stability of Supersonic SlenderBody Flight Vehicles,” Journal of Guidance, Control and Dynamics, Vol. 29, No. 6, pp. 1423-1427. Heddadj, S. and Cayzac, R., 2000, “Aeroelasticity of High L/D Supersonic Bodies: Theoretical and Numerical Approach,” 38th AIAA Aerospace Sciences Meeting & Exhibit, Reno, NV. Liu, D.D., Yao, Z.X., Sarhaddi, D. and Chavez, F., 1997, “From Piston

McNamara, J.J. and Friedmann, P.P., 2006, “Aeroelastic and Aerothermoelastic Analysis of Hypersonic Vehicles: Current Status and Future Trends,” AIAA Paper No. 2006-8058, 14th AIAA/AHI Space Planes and Hypersonic Systems and Technologies Conference, Canberra, Australia. Meyers, S.C., 1973, “Aeroelastic Analysis of Sounding Rocket Vehicles,” AIAA Paper No. 73-284. MSC/NASTRAN, 1995, Software Package User´s Guide, V. 68.2, MacNeal-Schwendler Corporation, Los Angeles, CA. MSC/NASTRAN V70, 1995, Advanced Dynamics User´s Guide, MacNealSchwendler Corporation, Los Angeles, CA. Murphy, C.H. and Mermagen, W.H, 2001, “Flight Mechanics of an Elastic Symmetric Missile,” Journal of Guidance, Control and Dynamics, Vol. 24, No. 6, pp. 1125-1132. Murphy, C.H. and Mermagen, W.H., 2005, “Aero-Elastic Motion of a SpinStabilized Projectile,” ARL-TR-3453, Aberdeen Proving Ground, MD. Paek, S.K., Bae, J.S. and Lee, I., 2002, “Flutter Analysis of a Wraparound Fin Projectile Considering Rolling Motion,” Journal of Spacecraft and Rockets, Vol. 39, No. 1, pp. 66-72. Platus, D.H., 1992, “Aeroelastic Stability of Slender, Spinning Missiles,” Journal of Guidance, Control and Dynamics, Vol. 15, No.1, pp. 144-151 . Turner, J., Hörschgen, M., Jung, W., Stamminger, A. and Turner, P., 2006, “SHEFEX–Hypersonic Re-entry Flight Experiment Vehicle and Subsystem Design, Flight Performance and Prospects,” AIAA Paper No. 2006-8115, 14th AIAA/AHI Space Planes and Hypersonic Systems and Technologies Conference, Canberra, Australia. ZAERO, 2003, Software Package, Ver. 6.5, Zona Technology, Scottsdale, AZ.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.15-26, Jan.-Mar., 2013

doi: 105028/jatm.v5i1.163

Unsteady Blade Element-Momentum Method Including Returning Wake Effects Cláudio Tavares Silva1,2, Maurício Vicente Donadon1

Abstract: The wind energy research has grown substantially in the past few years, considerably fostered by the pursuit for a clean and sustainable energy source. Improvements on the design methods are increasingly needed. The purpose of this research is to investigate the use of the Loewy´s lift deficiency function (LDF), also named Returning Wake Model, coupled with a non-stationary Blade Element-Momentum Method (BEM). The LDF simulates the influence of the wake behind the wind turbine on its capacity to generate power. It is expected that this model reduce the dependency of the several empirical parameters necessary in other wake models which are currently used. Aiming to validate the results obtained in this new approach they are compared with those provided by commercial computational software and they have proven to be very consistent. It is concluded that the method is feasible to be used as an efficient design and optimization tool of upwind horizontal axis wind turbine blades. Keywords: Blade element method, Loewy´s lift deficiency function, Non stationary BEM, Wind turbine, Wind turbine blade design, Returning wake model.

INTRODUCTION Wind turbines operate in a hostile environment in which strong flow fluctuations, mainly due to the nature of the wind, can produce high and variable loads on its components. These loads, combined with the elastic behavior of the turbine structural components, compromise the overall energy generation efficiency and cannot be neglected during the design phase. The need for experimental and computational approaches to investigate the behavior of unsteady loads produced on a wind turbine blade has grown in proportion to the growing nominal power and size of the actual horizontal axis wind turbines. The objective of this work is to evaluate the mathematical model of the Loewy´s lift deficiency function (LDF) to be included in a computational package enabling the optimal design of horizontal axis wind turbine blades. This model allows the inclusion of non-stationarities as well as an approximation of the effects of the downwind wake dynamics on turbine aerodynamic performance. This inclusion of non-stationarities and the effects of the downwind wake allow more realistic results for the aerodynamic loads than the simple Blade Element-Momentum (BEM) method, and consequently leads to better designs of structural components. The overwhelming majority of computer packages and procedures actually use semi-empirical models to represent the influence of the wake in the overall aerodynamic performance of a wind turbine blade. The LDF (Loewy, 1957) is an analytical solution based on the classical Theodorsen theory (Theodorsen, 1935). The use of an analytical model reduces the dependence on experimental tests for the adjustment of empirical parameters needed to calibrate

1.Instituto Tecnológico de Aeronáutica – São José dos Campos/SP – Brazil 2.Universidade Tecnológica Federal do Paraná – Curitiba/PR – Brazil Author for correspondence: Cláudio Tavares Silva | Praça Marechal Eduardo Gomes, 50 – Vila das Acácias | CEP 12228-900 São José dos Campos/SP – Brazil | E-mail: ctavares@ita.br Received: 19/09/12 | Accepted: 07/02/13

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.27-42, Jan.-Mar., 2013

Silva, C.T. and Donadon, M.V.

these semi-empirical engineering models. Furthermore, an analytical resolution results in a more elegant mathematical solution, computationally feasible and sufficiently fast to provide blade design optimization. The induced effects generated by the shed wake have been modeled using two general approaches: dynamic inflow methods and vortex wake methods. The principles of the dynamic inflow approach are attributed to Carpenter and Freidovich (1953). The idea is to consider the unsteady aerodynamic lag of the inflow development over the rotor disk in response to changes in blade pitch inputs of changes in rotor thrust. They are written in the form of ordinary differential equations, with a time constant (or constants) representing the dynamic lag in the build-up of the inflow. One of its less satisfying aspects is that time constants must be obtained by experimental calibrations. Vortex wake models are based on the assumption of an incompressible potential flow, with all vorticity being assumed concentrated within vortex filaments (which in the case of rotors require a coupling to the blade lift distribution). The induced velocity field can be determined through the application of the Biot-Savart low. Different approaches are encompassed ranging from prescribed to free vortex techniques. The prescribed wake models are strictly applicable when the operating conditions are nominally steady-state, i.e., in a steady wind. The free vortex methods have fewer potential limitations. They have been widely developed to be used in helicopters rotor analyses. Free vortex methods are based on discretized, finite-difference representation of the governing equations for the wake, and when solved, they track the evolution of discrete vortex elements through the flow. The number of discrete elements per vortex filaments can be very large, making the tracking process memory intensive and computationally demanding. Leishman (2002) presents more details about dynamic inflow models and vortex wake models.

using different settings on wind speed, rotational speed and pitch angle. The method couples the momentum theory with local events taking place at the actual blades. The blade is analyzed as a number of independent stream tubes. In each one, the induced velocity is calculated by performing the conservation of momentum, and the aerodynamic forces are found with the 2D aerodynamic theory and airfoil data. The stream tubes are discretized into N annular elements. The lateral boundary of the elements does not admit any flow across them. Some assumptions are made for the annular elements: no radial dependence, that is, one element cannot be affected by the others; the forces from the blade on the flow are constant in each annular element, corresponding to a rotor with a number of blades. A correction known as Prandtl´s tip loss factor (Glauert, 1935) is introduced to correct this latter assumption in order to compute a rotor with a finite number or blades. A relative velocity Vrel seen by a blade section is a combination of axial velocity V0(1–a), in which a is the axial induction factor, and the tangential velocity (1–a') ϖr, where a' is the radial induction factor at the rotor plane (Fig. 1). The angle θ is the local pitch angle of the blade element, i.e., the local angle between the chord and the plane of rotation. It is a combination of the pitch angle, measured between the tip chord, the rotor plane and the twist of the blade, relative to the tip chord. ϕ is the flow angle, measured between the plane of rotation and the relative velocity. The local angle of attack α is obviously found.

(1 + a ) Rotor plane

THEORY AND METHODS The classical Blade Element Momentum Method The BEM method presented by Glauert (1935) enables to calculate the steady loads and also the thrust and power

V0 (1 a)

Vrel

Figure 1. Velocities at rotor plane.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.27-42, Jan.-Mar., 2013

Unsteady Blade Element-Momentum Method Including Returning Wake Effects

The algorithm for the BEM model can be summarized as the sequence of steps that follows. Since different control volumes are assumed to be independent, each blade element can be treated separately and the solution at one radius can be computed before solving another radius. The following algorithm is applied for each control volume. Step (1): Initialize a' (axial induction factor) and á (radial induction factor), typically a = a' = 0. Step (2): Compute the flow angle ϕ. Step (3): Compute the local angle of attack α. Step (4): Read off the lift Cl(α) and drag Cd(α) coefficients from a table. Step (5): Compute Cn and Ct, respectively normal and tangential aerodynamic force coefficients. Step (6): Recalculate a and a'. Step (7): If a and a' have changed more than a tolerable amount, repeat step (2), or else finish. Step (8): Compute local loads on the blade element. All equations necessary to perform the above algorithm can are described by Burton (2001) and Hansen (2007). Unsteady BEM Model In order to obtain good estimates of the annual energy production of a wind turbine, a steady BEM method is adequate to compute the steady power curve. But in reality the rotor of a wind turbine feels the inherent unsteadiness of the wind caused by atmospheric turbulence, wind shear and the presence of the tower. It is necessary to use an unsteady BEM method to compute realistically this variable behavior of the wind. One simple model and additional coordinate systems can by placed at the wind turbine and its blades so it is possible to know the relative position of any blade element at any time. This simple model is depicted at Fig. 2. An inertial system of coordinates is placed at tower base and named System 1. System 2 is non-rotating and fixed in the nacelle. System 3 is solidary to the shaft turbine and rotates with it, and system 4 is aligned with one of the blades. The tilt and cone angles are shown and the azimuthal position of the blade is set by de wing angle θwing, not depicted. The undisturbed wind velocity seen by the blade is found by a simple coordinate transformation clearly detailed by Hansen (2007).

P x4 r

tilt

cone

Figure 2. Coordinate systems.

The essence of the BEM method is to determine the induced velocity and thus the local angle of attack. This is achieved by a summation of vectors, Vrel = V0+ Vrot+W, all written at the element blade coordinate system (System 4, Fig. 3), in which the induction velocity is the term W. With the induced velocity known, the flow angle and angle of attack are found.

tanϕ =

Vrel ,z -Vrel ,y

α =ϕ −( β +θ p )

(1)

Bramwell et al. (1976) states that Glauert’s relation between thrust and this induced velocity for a gyrocopter in

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.27-42, Jan.-Mar., 2013

Silva, C.T. and Donadon, M.V.

Vrot +

Rotor plane

Vrel

Figure 3. Relative and induced wind speeds in System 4.

forward flight (similar to the result of the lifting line for an elliptically loaded circular wing) is:

Wn = n ⋅ W =

T 2 ρ A V0 + n (n ⋅W )

(2)

in which n is the unit vector in the direction of the thrust, which in System 3 has the coordinates n=[0,0,1]T. It is assumed that only the lift contributes to the induced velocity, and that the induced velocity acts in the opposite direction to the lift. The force from this blade at radial position is assumed to affect the air in the area dA = 2πrdr/B, so that all B blades cover the entire annulus of the rotor disc at radius r. The following expression can be derived for one blade, according to Hansen (2007), Wn = Wz

− Lcos ϕ dr − BLcos ϕ = 2π rdr 2ρ F V0 + n (n ⋅W ) 4πρ rF V0 + n (n ⋅W ) (3) B

For the tangential component a similar expression is postulated,

Wt = Wy =

−BL sin ϕ 4πρ rF V0 + n (n ⋅W )

(4)

in which F is Prandtl’s tip loss factor. If the rotor is yawed (and/or tilted), there will be an azimuthal variation of the induced velocity, so that it is lower when the blade is pointing upstream in relation to when the same blade, half a revolution later, is pointing downstream. The physical explanation for it is that a

blade pointing downstream is deeper into the wake than a blade pointing upstream. This means that an upstream blade sees a higher wind speed and thus produces higher loads than the downstream blade, which produces a beneficial yawing moment that will try to turn the rotor more into the wind, thus enhancing yaw stability. The yaw model describes the distribution of the induced velocity. If a yaw model is not included, the BEM method will not be able to predict the restoring yaw moment, according to Hansen (2007):

⎛ ⎞ r ⎛χ⎞ W = W0 ⎜1 + tan⎜ ⎟ cos (θ wing − θ0 ) ⎟ 2 R ⎝ ⎠ ⎝ ⎠

(5)

in which the wake skew angle, X, is defined as the angle between the wind velocity in the wake and the rotational axis of the rotor. θ0 is the angle in which the blade is deepest into the wake. The skew angle can be found as: cos X =

n ⋅ ( V0 + W ) n V0 + W

(6)

The skew angle is assumed to be constant with the radius and can be computed at a radial position close to r/R = 0.7. The induced velocity is now known at the new azimuthal position at time t+∆t, θwing(t+ ∆t)= θwing(t)+ ω∆t. The angle of attack can thus be evaluated from the equation and the lift and drag coefficients can be looked up from a table. The normal, pz, and tangential, py, loads can be determined from:

pz = L cosφ + Dsin φ

p y = Lsin φ − Dcosφ

(7)

in which L = 1 ρVrel 2 cC l 2

D = 1 ρ Vrel 2cC d 2

(8)

The algorithm can be resumed like the following: Step (1): Initialize all necessary data (geometry and run parameters); Step (2): Initialize the position and velocity of blades; Step (3): Discretize the blades into N elements; Step (4): Initialize the induced velocity; - for n=1 to max time step (t=n∆t) - for each blade - for each element 1 to N

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.27-42, Jan.-Mar., 2013

Unsteady Blade Element-Momentum Method Including Returning Wake Effects

Step (5): Compute relative velocity to the blade element using old values for induced velocity; Step (6): Calculate flow angle and thus the angle of attack (Eq. 1); Step (7): Determine static drag and lift coefficients from tables; Step (8): Compute lift and drag for each blade element (Eq. 8); Step (9): Compute loads (Eq. 7); Step (10): Compute new equilibrium values for induced velocities (Eqs. 3 and 4); Step (11) Calculate the azimuthal variation from Eq. 5 and compute the induced velocity for each blade; Step (12): Compute momentum, thrust and power; Step (13): Increment time step and repeat from step (5). The equations of the BEM method must be solved iteratively. The flow angle and thus the angle of attack depend on induced velocity. But the described algorithm is unsteady, therefore time is used as relaxation. After blades have moved in one time step an azimuthal angle of ∆θwing=ω∆t (for small ∆t´s), values from the previous time step are used on the right hand side of equations for W when updating new values for induced velocity. This can be taken into account since induced velocity changes relatively slowly in time. This eliminates the need of calculating the induction factors and use of tolerance. Deterministic wind model The exponential model used to simulate wind shear. It controls the wind speed according to the altitude, and the shear parameter used is 0.2. Detailed information about this wind shear model was described by Hansen (2007). The wind is also influenced by the presence of the tower. The simple model used to simulate the tower shadow assumes potential flow. All details about this simple model were also described by Hansen (2007). This model is a bad approximation for a downwind machine, in which each blade passes the tower wake once every revolution. However, for an upwind machine, the object of this study, the model provides good estimation. Also, the turbulent part of the real atmospheric wind should be added for a realistic time simulation for a wind turbine. For this initial investigation no atmospheric turbulence is added to the simulation.

Loewy´s lift deficiency function The problem of calculating the aerodynamic loading on an oscillating profile was first approached by Glauert (1929), but it was only properly solved by Theodorsen (1935). Theodorsen’s approach gives the solution for unsteady aerodynamic loading on a 2D oscillating airfoil in an inviscid and incompressible flow, and subject to the assumption of small disturbances. Theodorsen’s problem is to obtain the solution for loading on the surface of the airfoil under the condition of forced harmonic oscillations. For a simple harmonic motion of the airfoil the solution given by Theodorsen in a way that represents a transfer function relating the forcing input (angle of attack) and the aerodynamic response (pressure distribution, lift, and pitching moment). The approach is summarized by Bisplinghoff et al. (1955). See also Bramwell et al. (1976) and Johnson (1980) for a detailed exposition of the theory. Theodorsen’s theory is not suitable for studies involving rotors. In these types of problems sections of blades can find wake vorticity due to other rotor blades, as well as the returning wake from the blade in question. This fact was recognized by Loewy (1957) and Jones (1958). They built a two-dimensional model of a 2-D blade section with a returning shed wake, as shown in Fig. 4. As in Theodorsen’s model, the shed wake is modeled with 2-D flat surfaces of vortices, but now with a series of surfaces below the airfoil section with a vertical separation h, which depends on the speed induced by the rotor disc V and the number of blades Nb. Loewy shows that, in this case, the lift on the blade can be expressed by replacing the function Theodorsen by the named Loewy’s function.

(

Cʹ k,ω

)

,h =

H1( 2 ) ( k ) + 2 J 1 (k )W

2 1

( k ) + iH0(2) ( k ) + 2 ( J1 (k ) +iJ 0 (k ))W

(9)

in which it is known for Loewy’s function of Loewy’s lift deficiency function. For a rotor with Nb blades the complex function W is written as:

(

W kh , ω , Δψ , Nb b Ω

)

i 2π

i (Δψ ) ω Ω Nb Ω e

−1

(10)

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.27-42, Jan.-Mar., 2013

Silva, C.T. and Donadon, M.V.

γb

iω t

α=θ

∞

γb = 0

γw

∆ψ C

C C

n=0, m=0

x,ζ

m=1

∞ m=2 n=1, m=0 m=1 m=2 n=2, m=0

h 4λ b =σ

m=1 m=2

figure 4. loewy’s returning wake problem (leishman, 2000).

The wake spacing ratio h/b can be determined from the spacing of vortex sheets that are laid down below the rotor. If an average induced velocity v i=λΩR is assumed, then during a single rotor revolution the shed wake generated by a single blade will be at a distance h=(2π/Ω)v i below the rotor. For a multiple blade rotor the spacing is (2π)v i/ ΩNb, i.e. h λΩ R 2π 4λ = = Ω Nb b σ b

(11)

in which σ is the rotor solidity. Representative results from Loewy’s theory show that the main consequence of including shed vorticity below the blade is that it serves to amplify or attenuate the unsteady lift response, depending on the reduced frequency, wake spacing and wake phase. The most important effects are for lower reduced frequencies, with oscillations at the harmonics of the rotor rotational frequency (Leishman, 2000). Several aspects related to the non-stationarity of the operating environment of a wind turbine need to be addressed during its project. Among them there are the main variations in wind speed (gust and wind shear),

dynamic inflow, yaw and tower shadow, turbulence, wake dynamics and interactions blade/wake, and the dynamic stall. The adoption of a non-stationary BEM method using a dynamic inflow model is a solution that tends to enhance the results, proving to be a good option to introduce the study of non-stationarity in the design of a wind turbine blade. However the same problems related to the physics of the method remain. Unsatisfactory aspects of the inflow theory are the so-called dynamic time constants employed in the methods. They are developed using the concept of apparent mass and inertia of the fluid surrounding the rotor (noncirculatory effect) as opposed to the delay of the dynamic evolution of wake vortices (circulatory effects). The concept of apparent mass applied to the rotor also assumes equivalence between the apparent force of the rotor disk accelerating in a stopped fluid and the force in a fluid accelerating through a permeable actuator disc, which certainly is not an accurate analogy. Loewy proposes a solution to the problem of aerodynamics of rotors affected by non-stationarity generated by shed wake. It is based on Theodorsen’s solution applying a suitable physical model for

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.27-42, Jan.-Mar., 2013

Unsteady Blade Element-Momentum Method Including Returning Wake Effects

rotorcraft aerodynamics. The results obtained in the solution of problems related to helicopter hovers, and duly validated with experimental results, confirm the efficiency of the method. The aerodynamics of a helicopter hover resembles in many aspects the aerodynamics of the blades of a wind generator. This similarity, coupled with the need for a mathematical model for the wind generator blade design that considers the conditions of non-stationarity of the phenomena involved in its aerodynamics, are the main motivators for this work. The corrections using Loewy’s model are applied for the Equation 2.8 in order to provide the lift deficiency described above. The corrected equations can be rewritten. No corrections are applied to the drag force. L = 1 ρVrel 2 cCl C 2

(12)

Verification against an industrystandard software The commercial package used for result validation is the GH Bladed V4.1, distributed by GL Garrad Hassan (2012), an independent renewable energy consultancy. GH Bladed is an industry-standard integrated software package for the design and certification of onshore and offshore turbines. It provides user with a design tool that has been extensively validated against measured data from a wide range of turbines and enables the conduction of the full range of performance and loading calculations (Garrad Hassan & Partners, 2011). Its manual postulates that GH Bladed uses the same methods employed at the present work. That is, combined blade element and momentum theory, wake rotation with radial induction, tip and hub loss models, which is suppressed for the present comparison, dynamic wake model and dynamic stall, also suppressed. An educational version of de GH Bladed GH Bladed 4.1, with a limitation of 10 blade elements, is used. The object of study is a 2MW wind turbine. The main used parameters and data are shown in Tables 1 and 2. All necessary data, like angle of attack, lift, drag and moment coefficients are described by McGee and Beasley (1976). Just for visualization, BEM NE has a graphical interface that shows a simplified illustration of the wind turbine and his main geometrical parameter. This is shown at Fig. 5.

Table 1. General characteristics of rotor and turbine.

Rotor diameter

Number of blades

Hub height

61.5

Tower height

Tilt angle of rotor to horizontal

deg

Cone angle of rotor

deg

Blade set angle

deg

Rotor overhang

3.7

Rotational sense of rotor, viewed from upwind

Clockwise

Position of rotor relative to tower

Upwind

Aerodynamic control surfaces

Pitch

Radial position of root station

1.25

Cut in windspeed

m/s

Cut out windspeed

m/s

Table 2. Blade geometry. Distance along blade (m)

Chord (m)

Aerodynamic twist (deg)

Aerofoil section

2.07

cylinder

1.15

2.07

cylinder

3.44

2.76

cylinder

5.74

3.44

NASA LS(1)-0421

9.19

3.44

NASA LS(1)-0421

16.07

2.76

7.8

NASA LS(1)-0421

26.41

1.84

3.3

NASA LS(1)-0417

35.59

1.15

0.3

NASA LS(1)-0413

38.23

0.69

2.75

NASA LS(1)-0413

38.75

0.03

NASA LS(1)-0413

The developed computational routine uses these same data to be compared with the results of both packages. It also uses de same model for wind shear, i.e., the exponential

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.27-42, Jan.-Mar., 2013

Silva, C.T. and Donadon, M.V.

vertical shear model with wind shear coefficient 0.2 and the potential flow model for the tower shadow. Details about the models for wind shear end tower shadow were described by Hansen (2007).

Figures 6 and 7 illustrate the wind condition for one blade element. Figure 6 is an illustration of the magnitude of the wind for a blade element. Figure 7 shows the wind speed distribution influenced by the presence of the tower (tower shadow).

Wind spedd vectors for θyaw=0 degress

Wind turbine 3D view - yaw=0º, tilt=-4, wing=0º, cone=0º 100

60 50

x[m]

50 45

z[m]

-5 30

-20 -10

-30 Radial position (r=30 m) Blade element (theta=0 degree)

y[m]

Tower (height=60 m)

Nacele (overhang=3.7 m) Pás (lenght=40 m)

wind speed Figure 5. 3D view of the wind turbine.

-5 25

-5

-25 -20 -15 -10

lateral distance from tower axis

Figure 6. Vectors representing wind.

Z component for θyaw=0 degress

ground height

γ component for θyaw=0 degress

0.4

0.3

Wind speed (component Y) [m/s]

Wind speed (component Z) [m/s]

12.5

11.5

0.2

0.1

-0.1

-0.2

10.5

-0.3

Blade azimuth, θwing[rad] Figure 7. Wind speed components. J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.27-42, Jan.-Mar., 2013

-0.4 0

Blade azimuth, θwing[rad]

Unsteady Blade Element-Momentum Method Including Returning Wake Effects

RESULTS AND DISCUSSION

The most important aerodynamic and performance results are shown and compared here. It is important to mention that the GH Bladed package has a pitch control module that changes the pitch angle of the blades in each time step according to the wind speed, rotor speed and output power. The developed computational routine, named here BEM NE, uses this same pitch angle control map in order to achieve the correct level of the shaft power, once it does not have a control module. This control map is depicted in Fig. 8b as also the shaft power (Fig. 8a) against different wind speeds. The blades of the wind turbine used in this study are designed to produce 2MW power between wind speeds of 12 and 25 m/s, at the rotor speed 18 rpm. Results at 12 and 18 m/s wind speed are compared. At first the results obtained for 12 m/s wind speed is shown. Figures 9 and 10 compare the geometrical results of flow angle and angle of attack, respectively. The results shown are obtained from the blade element located 26.407 m from the rotor center. This blade element is the most representative one. It is located in 70% length position of the blade. During the 12 seconds of simulation, deviations on the results are minimal, as demonstrated in Figs. 9 and 10. Mean relative error at inflow angle is -1.47% and mean relative error at angle of attack is -2.71%. Figures 11 and 12 show the comparison of the relative wind speed and the corresponding lift coefficient of the investigated

blade element. For the BEM NE results for the lift coefficient, Loewy’s lift deficiency function is applied to correct its value. Again, the mean relative errors are small: respectively +0.28% e +7.68% for relative wind speed and lift coefficient. The most important result, the output shaft power, is shown in Fig. 13. The relative mean error is now -0.66%. Results obtained until this point show a good correlation between the GH Bladed and the BEM NE for the wind speed 12 m/s. Now the same comparison of results for the wind speed 18 m/s is shown in the middle of the operation wind speed range. Figures 14 until 18 show these comparison. For wind speed 18 m/s the correspondent pitch angle is 14.9 degrees, as shown in Fig. 8. Relative mean errors are compatible with the ones obtained for wind speed 12 m/s. The most important parameter, the shaft power, has a relative error of -4.1%. It is observed that the power curve obtained from BEM NE has more spikes then the one from GH Bladed. The discontinuities coincide with the passage of the blade through the tower shadow. Even though both packages use the same model for the tower shadow, these discrepancies can be justified by the absence of any kind of pitch control on the BEM NE. At the current stage of the present study, the model is simplified because its primary objective is to investigate the viability of LDF as a simulation for the influence of the wake shed behind the rotor. Therefore the pitch angle is maintained constant during all time steps simulation. On the contrary, Bladed has a variation of the pitch angle during the simulation, as shown in Fig. 19.

(a)

(b)

1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0

Hub wind speed [m/s]

Pitch angle [deg]

Shaft power [MW]

1.8

2.2 20 15 10 5 0 -5

Hub wind speed [m/s]

Bladed Educational - Licensed to: Instituto Tecnológico de Aeronáutica

2.0

Bladed Educational - Licensed to: Instituto Tecnológico de Aeronáutica

2.2

Figure 8. Shaft Power (a) and Pitch angle (b) control map.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.27-42, Jan.-Mar., 2013

Silva, C.T. and Donadon, M.V.

Inflow angle for blade 1 at wind speed 12 m/s - Distance along blade 26.407 m 11

GH Bladed mean (GH Bladed) BEM NE

10.5

mean (BEM NE)

Inflow angle [deg]

9.5

8.5

7.5

Time [s]

Figure 9. Inflow angle (wind speed 12 m/s).

Angle of attack for blade 1 at wind speed 12 m/s - Distance along blade 26.407 m 7

GH Bladed mean (GH Bladed) BEM NE

6.5

mean (BEM NE)

Angle of attack [deg]

5.5

4.5

3.5

Time [s]

Figure 10. Angle of attack (wind speed 12 m/s).

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.27-42, Jan.-Mar., 2013

Unsteady Blade Element-Momentum Method Including Returning Wake Effects

Relative wind speed for blade 1 at wind speed 12 m/s - Distance along blade 26.407 m

GH Bladed mean (GH Bladed) BEM NE

Relative wind speed [m/s]

54.5

mean (BEM NE)

53.5

52.5

Time [s]

Figure 11. Relative wind speed (12 m/s wind speed).

Angle of attack for blade 1 at wind speed 12 m/s - Distance along blade 26.407 m 7

GH Bladed mean (GH Bladed) BEM NE

6.5

mean (BEM NE)

Angle of attack [deg]

5.5

4.5

3.5

Time [s]

Figure 12. Lift coefficient (wind speed 12 m/s).

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.27-42, Jan.-Mar., 2013

Silva, C.T. and Donadon, M.V.

Angle of attack for blade 1 at wind speed 12 m/s - Distance along blade 26.407 m 7

GH Bladed mean (GH Bladed) BEM NE

6.5

mean (BEM NE)

Angle of attack [deg]

5.5

4.5

3.5

Time [s]

Figure 13. Measured shaft power (wind speed 12 m/s).

Inflow angle for blade 1 at wind speed 18 m/s - Distance along blade 26.407 m

GH Bladed mean (GH Bladed) BEM NE

mean (BEM NE)

Inflow angle [deg]

Time [s]

Figure 14. Inflow angle (wind speed 18 m/s).

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.27-42, Jan.-Mar., 2013

Unsteady Blade Element-Momentum Method Including Returning Wake Effects

Angle of attack for blade 1 at wind speed 18 m/s - Distance along blade 26.407 m GH Bladed

mean (GH Bladed) BEM NE mean (BEM NE)

Angle of attack [deg]

-1

-2

-3

-4

Time [s]

Figure 15. Angle of attack (wind speed 18 m/s).

Relative wind speed for blade 1 at nominal wind speed 18 m/s - Distance along blade 26.407 m

GH Bladed mean (GH Bladed) BEM NE mean (BEM NE)

Realative wind speed [m/s]

Time [s]

Figure 16. Relative wind speed (wind speed 18 m/s).

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.27-42, Jan.-Mar., 2013

Silva, C.T. and Donadon, M.V.

Lift coefficient for blade 1 at wind speed 18 m/s - Distance along blade 26.407 m

0.8

GH Bladed mean (GH Bladed) BEM NE

0.7

mean (BEM NE)

0.6

Lift coefficient

0.5

0.4

0.3

0.2

0.1

Time [s]

Figure 17. Lift coefficient (wind speed 18 m/s).

Measured shaft power at wind speed 18 m/s

2.2

GH Bladed mean (GH Bladed) BEM NE

Measured shaft power [MW]

2.15

mean (BEM NE)

2.1

2.05

1.95

1.9

1.85

Time [s]

Figure 18. Measured shaft power (wind speed 18 m/s).

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.27-42, Jan.-Mar., 2013

Unsteady Blade Element-Momentum Method Including Returning Wake Effects

(a)

(b)

0.80

15.02

0.74 0.72 0.70 0.68 0.66 0

15.00

Nominal pitch angle [deg]

Blade 1 pitch angle [deg]

0.76

14.98 14.96 14.94 14.92 14.90 14.88 14.86 14.84 14.82

Bladed Educational - Licensed to: Instituto Tecnológico de Aeronáutica

0.78

0.64

Time [s]

Figure 19. Pitch angle for wind speed (a) 12 m/s and (b) 18 m/s from Bladed.

CONCLUSION The work presented an alternative approach to predict the performance of upwind horizontal-axis wind turbine design using the unsteady BEM theory and an analytical model for the wake shed behind the rotor. This alternative approach is employed in order to verify the viability of using an analytical model for the returning wake effects, which does not have any empirical parameters or the need for experimental calibration. The numerical approach is very stable and fast, even being written in an interactive computation environment. The computational processing time necessary for any case is less than 10 seconds, and there were not numerical crashes. Based on the good approximation of results, when compared with others provided by commercial and established computational package, the mathematical model presented in this paper introduces an alternative tool for the wind turbine design, especially for upwind rotors.

The comparisons show that the model has good performance in terms of computational speed and the differences between its results and those provided by the commercial software used as validation parameter are very small, being compatible with the optimization design method.

ACKNOWLEDGEMENTS The authors acknowledge the financial support from Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), by Programa Institucional de Qualificação Docente para a Rede Federal de Educação Profissional e Tecnológica (PIQDTec), of Universidade Tecnológica Federal do Paraná (UTFPR). The authors acknowledge the financial support received for this work from Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), contract number 303287/2009-8.

REFERENCES Bisplinghoff, R. L., Ashley, H. and Halfman, R. L, 1955, “Aeroelasticity”, Addison-Wesley Publishing Co., Reading, MA.

Burton,T., 2001, “Wind Energy Handbook”, Ed. John Wiley and Sons Ltd., Chichester, New York, 624 p.

Bramwell, A. R. S., Done, G. and Balmford, D., 1976, “Helicopter Dynamics”, Edward Arnold, Great Britain.

Carpenter, P.J. and Fridovich, B., 1953, “Effect of A Rapid Blade-Pitch Increase on the Thrust and Induced-Velocity Response of a Full-Scale Helicopter Rotor”, NACA TN 3044.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.27-42, Jan.-Mar., 2013

Silva, C.T. and Donadon, M.V.

Garrad Hassan & Partners, 2011, “Bladed Theory Manual Version 4.1 Multibody Dynamics”, St. Vincent’s Works, Silverthome Lane, Bristol BS2 0QD, England.

Leishman, J.G., 2002, “Challenges in Modeling the Unsteady Aerodynamics of Wind Turbines”, 21st ASME Wind Energy Symposium and 40th AIAA Aerospace Sciences Meeting Reno, NV.

GL Garrad Hassan, 2012, http://www.gl-garradhassan.com.

Leishman, J.G., 2000, “Principles of Helicopter Aerodynamics”, Cambridge University Press, The Edinburgh Building, Cambridge CB2 2RU, UK.

Glauert, H., 1935, “Airplane Propellers”, Aerodynamic Theory (W.F. Durand, ed.), Div. L, Chapter XI. Berlin:Springer Verlag. Glauert, H., 1929, “The Force and Moment on an Oscillating Airfoil”, Rep. Mem. Aeronaut., Res. Comm., Great Britain, No. 1561. Hansen, M.O.L., 2007, “Aerodynamics of wind turbines”, Earthscan, Camden High Street London, NW1 0JH, UK, 2nd ed, pp 8-12. Johnson,W., 1980, “Helicopter Theory”, Princeton University Press. Jones, J.P., 1958, “The Influence of the Wake on the Flutter and Vibration of Rotor Blades”, the Aeronaut. Quart., Vol. 9, No 3, pp 258-286.

Loewy, R.G., 1957, “A Two-dimensional Approximation to the Unsteady Aerodynamics of Rotary Wings”, J. Aeronaut. Sci. Vol. 24, No 2, pp 81-92. McGee, R.J. and Beasley, W.D., 1976, “The Aerodynamic Characteristics of An Initial Low-speed Family of Airfoils for General Aviation Applications”, NASA TM X-72843, NASA Langley Research Center, Hampton, VA, USA. Theodorsen, T., 1935, “General Theory of Aerodynamic Instability and the Mechanism of Flutter”, NACA report 496, pp 413-33.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.27-42, Jan.-Mar., 2013

doi: 10.5028/jatm.v5i1.180

Calculation of The Vehicle Drag and Heating Reduction at Hypervelocities with LaserInduced Air Spike Israel da Silveira Rêgo1, Paulo Gilberto de Paula Toro2, Marco Antonio Sala Minucci3, José Brosler Chanes Júnior2, Felipe Jean da Costa4, Antonio Carlos de Oliveira2

Abstract: Scientists at the laboratory of Força Aérea Brasileira (FAB) have demonstrated experimentally that laser-induced “air spikes” (DEAS) may reduce effectively both total vehicle drag and heating at hypervelocities. Now, we apply the Rayleigh flow to directly determine the degree of reduction in vehicle drag and convective heat flux into the airframe of a hypersonic blunt-body when laser energy is added upstream of the flight path. Our numerical findings are in accordance with the physical trends observed in our previous hypersonic laser-induced DEAS experiments. Keywords: Laser, Flow control, Hypersonic, Rayleigh flow.

INTRODUCTION Hypersonic technologies now being developed by a few countries, including Brazil, could within two or three decades yield limitless possibilities for air and space travel. In particular, hypersonic flight poses numerous engineering challenges involving many more structural difficulties due to severe heating and dynamic loads, new materials and structures for airframe, predictive models for hypersonic flow, advanced control techniques for hypervelocity flights, new types of airbreathing propulsion systems and proper aerodynamic integration of both airframe and propulsion systems (Heiser and Pratt, 1994). The Directed-Energy Air Spike (DEAS) is a promising technique for hypersonic flight control, according to which an air spike is induced upstream in relation to the flight path serving to reduce the vehicle drag and to lower heat transfer into the airframe (Myrabo and Raizer, 1994). Air spike production is obtained through many means, including electric arcs (Toro, 1998; Minucci et al., 2000), microwaves (Myrabo and Lewis, 2009) and laser beams (Minucci et al., 2005; Oliveira et al., 2008; Oliveira, 2008; Salvador et al., 2006). A series of experiments on the concept of hypersonic laser-induced DEAS was performed at the Aerothermodynamics and Hypersonic Laboratory Henry T. Nagamatsu, at Instituto de Estudos Avançados (IEAv), in São José dos Campos, Brazil, conclusively proving that air spikes do work for enhancing flight performance (Minucci et al., 2005; Oliveira et al., 2008; Oliveira, 2008; Salvador et al., 2006). Figure 1 shows the historic hypersonic laser-induced DEAS experiments at the IEAv.

1.Universidade Federal do ABC – Santo André/SP – Brazil 2.Instituto de Estudos Avançados – São José dos Campos/SP – Brazil 3.Vale Soluções em Energia – São José dos Campos/SP – Brazil 4.Instituto Tecnológico de Aeronáutica – São José dos Campos/SP – Brazil Author for correspondence: Israel da Silveira Rêgo | Rua Santa Adélia, 166 – Bangu | CEP 09210-170 Santo André/SP – Brazil | E-mail: israel.rego@ufabc.edu.br Received: 03/09/12 | Accepted: 27/11/12

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.43-48, Jan.-Mar., 2013

Rêgo, I.S., Toro, P.G.P., Minucci, M.A.S., Chanes Júnior, J.B.,Costa, F.J. and Oliveira, A.O.

Bow shock

ρ0SL

Model

Laser Beam

ρ2

ρ1

M1 p01

p2 Shock layer

Freestream

Figure 2. Rectangular control volume for one-dimensional airflow with laser energy addition.

However, the LSD wave is not taken into account for sake of simplicity. Thus, the changes in the onedimensional flow occurring inside the control volume in Fig. 2 is caused only by the heat added to the hypersonic airflow by the intense beam of laser radiation without the presence of any shock wave. The governing equations of the hypersonic airflow are continuity, momentum and energy, written as follows (Anderson, 2003): ρ1V1 = ρ2V2 Figure 1. IEAv experiments on hypersonic laser-induced DEAS concept.

METHODOLOGY Control volume and considerations Consider Fig. 2, which illustrates a control volume for onedimensional flow. Inside this control volume, a laser-supported detonation (LSD) wave and the heating itself cause the hypersonic airflow properties in region 2 to be different than in region 1.

p1 + ρ1V1 = p2 + ρ2V2 2

Now, we will use the Rayleigh flow to investigate the laserinduced DEAS in hypersonic airflow in a manner that the degree of reduction in vehicle drag and convective heat flux can be easily determined. The methodology is validated by comparison with our previous experimental findings (Minucci et al., 2005; Oliveira et al., 2008; Oliveira, 2008; Salvador et al., 2006).

(1) (2) 2

h1 + V1 + qL = h2 + V2 2 2

(3)

They say that mass is conserved (Eq. 1), force equals time rate of change of momentum (Eq. 2), and energy is also conserved (Eq. 3). Note that Eq. 2 neglects body forces and viscous stresses, and Eq. 3 does not include shaft work and work done by viscous stresses, but takes into account the amount of laser energy added by unit of mass, qL. Since conditions in region 1 are known, for a given qL these equations, along with the appropriate equations of state and the specific case of a calorically perfect gas, can be analytically solved for conditions in region 2. Although the airflow is hypersonic, we assume that the air is calorically perfect, that is, air with constant specific heats and unchangeable composition. Note that conditions in regions 1 and 2 correspond to freestream conditions of the hypersonic airflow before and after adding the laser energy.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.43-48, Jan.-Mar., 2013

Calculation of The Vehicle Drag and Heating Reduction at Hypervelocities with Laser-Induced Air Spike

Rayleigh flow The analytical relations for the one-dimensional hypersonic flow with laser energy addition can be given as follows (Anderson, 2003): qL = cp (T02 − T01)

(4)

p2 = 1 + γM1

(5)

p1 = 1 + γM2

T2 = M2

ρ2 = M1

1+ γM2 2

1+ γM2

T01

(7)

ρ1 = M2

T02

(6)

T1 = M1

1+ γM1

1+ (γ-1)M2 / 2

1+ (γ-1)M1 / 2

= 2 (8)

p02 P01

1+ γM2

1+ (γ-1) M2 / 2

1+ γM1

1+ (γ-1) M1 / 2

S2 − S1 = cp ln

T2 T1

− R ln

P2 P1

γ γ-1

(9)

M2<M1 u2<u1 T2>T1 p2>p1 p02<p01 T02>T01 s2–s1>0

Mach number Velocity Temperature Pressure Total pressure Total temperature Entropy

1+γ

(11)

1 + γM

T = M2 1 + γ T* 1 + γM2 ρ 1 1 + γM2 = ρ* M2 1 + γ

Table 1. Physical trends for Rayleigh flow. Quantity Variation for qL>0 and M1>1

2 2

1+ γM1

(12)

(13)

2 = (1 + γ) M2 2 2 +( γ-1) M2 T*0 (1 + γM )

p0 P*0

1+ γ

2 + (γ-1)M 2

1+ γM

1+ γ

(14) γ γ-1

(15)

(10)

They are the basic relations, which are valid for any flow with heat addition or removal, known as Rayleigh flow relations. Table 1 shows the variation in hypersonic flow quantities produced by laser energy addition. Sonic flow condition and Rayleigh table A methodology for the solution of Rayleigh flow relations is now outlined as follows. All conditions in region 1 are experimentally known. Also, the amount of laser energy added by unit of mass is experimentally estimated a priori. Therefore, given qL, T02 can be obtained from Eq. 4. Now, instead of finding the solution of Eq. 8 for M2 by trial and error, for convenience of calculation we use sonic flow as a reference condition and Rayleigh table for air, which we will explain briefly here. Let M1=1. Then, the corresponding flow properties are denoted by p1= p* , T1= T *, r1= r *, p01= p*01, and T01= T*01 . The flow properties at any other value of M are then obtained by inserting M1= 1 and M2= M into Eqs. 5 to 9, given (Anderson, 2003):

Equations 11 to 15 are tabulated as a function of M for air in appendix A.3 of Anderson’s book (Anderson, 2003). Here, the reference sonic conditions p*, T*, ρ*, p0*, and T0*, are those in the one-dimensional flow that would exist if enough laser energy is added to achieve Mach 1. The reference sonic conditions achieved when enough laser energy is added to bring the flow to Mach 1 are exactly the same, no matter whether the laser energy is added as q*L1, downstream of region 1 or as q*L2 downstream of region 2. This is why Eqs. 11 to 15 are simply reference quantities that are fixed for a given airflow with heat addition. With this concept, we can easily obtain M2. Consequently, any condition in region 2 is calculated directly with Eqs. 5 to 10. Reduction in drag and heating For the aerodynamic application addressed here, both drag and convective heat flux to the vehicle moving through the hypersonic airflow, before and after adding laser energy upstream of the flight path, should be compared separately. . The total drag D and the convective heat flux q are proportional to ρV2 and ρV3, respectively (Anderson, 2007). Hence:

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.43-48, Jan.-Mar., 2013

Rêgo, I.S., Toro, P.G.P., Minucci, M.A.S., Chanes Júnior, J.B.,Costa, F.J. and Oliveira, A.O.

D2 D1

ρ2 ρ1

T2 T1

M2 M1 3/2

•

(16)

q2 T2 M2 ρ = 2 (17) • ρ1 T1 M1 q 1

We have used Eqs. 16 and 17 along with the sonic flow condition and Rayleigh tables to discuss theoretically the aerodynamic applicability of laser-induced DEAS to control vehicle drag and heating at hypervelocities. The main advantage of this methodology is simplicity, which allows us to easily compare and contrast the numerical findings with the facts observed in our hypersonic laser-induced DEAS experiments (Minucci et al., 2005; Oliveira et al., 2008; Oliveira, 2008; Salvador et al., 2006).

RESULT AND DISCUSSION Calculation of drag reduction Table 2 lists our experimental data on laser energy addition in hypersonics (Oliveira, 2008; Salvador et al., 2006), which will be input into the calculations of the reduction in drag and heating. The drag reduction on the blunt body occurs mainly because of the decrease in p02 after air spikes are induced by the laser beam upstream in relation to the flight path. A measure is done indirectly by measuring the stagnation pressure in the shock layer behind the bow shock, that is, p0SL . In fact, the magnitude of p0SL decreases after adding laser energy as shown in Table 2. Thus, before calculating drag reduction, we will calculate the reduction in stagnation pressure, and then compare it to the experimental one. The amount of laser energy added by unit of mass is defined as the ratio between the rate of laser energy, i.e., laser power and mass flow ρVA , in which A is the laser spot size. As a measure of A, we multiply the confocal parameter and the Rayleigh half diameter (assuming a Gaussian laser beam) (Siegman, 1986), as shown in

Fig. 3. Note that the laser spot size is not exactly the size of the laser– induced air-spike, which grows rapidly over time. Fortunately, the confocal parameter and the Rayleigh half diameter depend only on the focal length of the lens (200 cm) employed and the laser wavelength (10.6 µm). Hence, the laser spot size can be obtained and, consequently, the laser energy added by unit of mass. Before adding the laser energy, the hypersonic airflow is isentropic (region 1). Hence, from isentropic airflow properties in appendix A.1 of Anderson’s book (Anderson, 2003), for a given M1 we obtain T01/T1 and then T01 (once T1 is known beforehand). Eq. 4, assuming that the specific heat capacity at constant pressure of air is constant (1004,5 J/kg.K), gives T02. Now, from the Rayleigh airflow table in appendix A.3 of Anderson’s book (Anderson, 2003), for a given M1 we have p1/p*, T1/T* , and T01/T0*. Let us rewrite T02/T0* as (T02/T01) (T01/T0*). With this and, again, from the Rayleigh airflow table, for a given T02/T0* we obtain directly M2. Now, we calculate the reduction in stagnation pressure, i.e., the ratio between the stagnation pressure after and before laser-induced air spike. The stagnation pressure before adding laser energy is obtained directly from normal shock relations (since M1 and p1 are previously known). To obtain the stagnation pressure after adding laser energy, we first need to determine p2 (see Fig. 2). To do so,

Laser Beam

Confocal parameter

Rayleigh half diameter Figure 3. Illustration of laser spot size to estimate the laser energy per unit of mass.

Table 2. IEAv data on laser energy addition in hypersonics. Enthalpy T1* ρ1* Mach [unit] [K] [kg/m3]

Low Low Low High High

132 132 132 1079 1079

0.1019 0.1019 0.1019 0.0093 0.0093

6.9 6.9 6.9 5.7 5.7

* Calculated from Hypersonic Shock Tunnel Real Gas (HSTR) software (Oliveira, 2008)

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.43-48, Jan.-Mar., 2013

Speed [km/s]

Laser power [MW]

Stagnation pressure reduction [%]

1.6 1.6 1.6 3.4 3.4

1.35 1.85 1.77 1.5 1.7

31 43 48 54 45

Calculation of The Vehicle Drag and Heating Reduction at Hypervelocities with Laser-Induced Air Spike

we rewrite p2 as (p2/p*) (p*/p1) p1, in which p2/p* and p*/p1 are obtained directly from the Rayleigh airflow table for M2 and M1 respectively. Also, the stagnation pressure after adding laser energy is obtained directly from normal shock relations (since M2 and p2 were obtained previously). Figure 4 shows the reduction in total pressure against laser power for low and high enthalpy flows. Linear fits represent trends in the data. When there is no laser energy addition, the reduction in total pressure is null. Note that the laser-induced DEAS is more effective to reduce the total pressure as the laser power increases, which is in accordance with Eq. 9. Also, the rate and degree of total pressure reduction is more modest for calculations in comparison with experiments. This is because our methodology assumes no interaction between the bow shock ahead of the blunt body and the LSD wave generated (see Fig. 5), which lowers somehow the strength of the curved bow shock. The drag reduction is calculated as follows. With the Rayleigh airflow table, for a given M2 we obtain p2/p* and T2/T*. Let us rewrite p2/p1 and T2/T1 as (p2/p*) (p*/p1) and (T2/T*) (T*/T), respectively. Thus, p2/p1 and T2/T1 are both obtained. The supersonic solution of the Rayleigh flow implies that both freestream density and temperature will increase while the Mach number will decrease as the laser beam heats the freestream. Finally, with Eq. 16 we calculate D2/D1, as shown in Fig. 6. Linear fits represent trends in the data. When there is no laser energy addition, the reduction in drag is null. Calculations show that laser-induced DEAS is more effective to reduce vehicle drag when laser power and flow enthalpy (flow speed) increase. Also, the rate of drag reduction is more modest for low enthalpy. Experimental data on drag

1.00

Figure 5. Interaction between bow shock and LSD wave in hypersonic flow.

1.00 0.90

0.80

Low enthalpy calculations

0.80

0.70

High enthalpy experiments

0.70

0.60

High enthalpy calculations

0.50

Linear (Low enthalpy experiments)

0.40

Reduction in drag

Stagnation pressure reduction

reduction still need to be obtained to contrast with the tendencies shown in Fig. 6 in spite of the fact that the reduction in stagnation pressure (see Fig. 4) already suggests that the drag should also decrease with laser energy addition.

Low enthalpy experiments

0.90

0.30

Linear (High enthalpy experiments)

0.20

0.10

Linear (High enthalpy calculations) 1.50

2.00

Laser power [MW]

Figure 4. Reduction in stagnation pressure versus laser power for low and high enthalpy flows.

Linear (Low enthalpy calculations)

0.40

0.20

1.00

High enthalpy calculations

0.50

0.30

0.50

Low enthalpy calculations

0.60

Linear (Low enthalpy calculations)

0.00 0.00

Linear (High enthalpy calculations)

0.10 0.00 0.00

0.50

1.00

1.50

2.00

Laser power [MW]

Figure 6. Reduction in vehicle drag versus laser power for low and high enthalpy flows. J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.43-48, Jan.-Mar., 2013

Rêgo, I.S., Toro, P.G.P., Minucci, M.A.S., Chanes Júnior, J.B.,Costa, F.J. and Oliveira, A.O.

1.00 Low enthalpy calculations

0.90

Reduction in heat flux

0.80

High enthalpy calculations

0.70 0.60

Low enthalpy experiments

0.50

Linear (Low enthalpy calculations)

0.40 0.30

Linear (High enthalpy calculations)

0.20

Linear (High enthalpy experiments)

0.10 0.00

0.00

0.50

1.00

1.50

2.00

Laser power [MW]

effectively to reduce the convective heat flux, that is, increasing qL forces M2 → 1 (chocked flow) and thereby decreasing the ratio. Again, note that the rate and degree of heating reduction is more modest for calculations in comparison with the experiments (Salvador et al., 2006) due to the fact that our methodology neglects the interaction between bow shock and LSD wave. Our previous hypersonic DEAS experiments revealed that the distance between the blunt body and the laser-induced air spike, as well as the angle of incidence of the laser beam relative to the flow direction, influence the degree of reduction in drag and heating (Oliveira et al., 2008; Salvador et al., 2006). However, our calculations do not consider this fact for the sake of simplicity.

Figure 7. Reduction in heating versus laser power for low and high enthalpy flows.

Calculation of heat reduction The reduction in convective heat flux to the airframe of the blunt, body occurs mainly due to the deceleration of the freestream, as observed in our DEAS experiments. Also, our calculations show a deceleration of the freestream when the laser beam heats the hypersonic flow. The reduction in convective heat flux is calculated as drag reduction, except for the fact that now we use Eq. 17. Figure 7 shows heating reduction as function of laser power with increasing enthalpy. Calculations show that laserinduced DEAS is more effective to reduce heating to the bluntbody airframe as both laser power and flow enthalpy increase. This is because of the term (M2/M1)3 in Eq. 17, which contributes

CONCLUSIONS The hypersonic age of air and space travel will be a reality by 2030. Since 1992, Brazil seeks hypersonic technologies such as scramjet engines, waverider design and advanced techniques for controlling airflow at hypervelocities. We applied the Rayleigh flow relations and sonic flow conditions to determine the degree of reduction in vehicle drag and heating at hypervelocities when laser energy is added to the freestream. Calculations show the same physical trends observed in experiments, clearly indicating that laser-induced air spikes are in fact beneficial to enhance the flight performance of hypersonic vehicles.

REFERENCES Anderson J.D., 2007, “Introduction to Flight”, 6th edition, Science Engineering & Math, pp. 257-263. Anderson J.D., 2003, “Modern Compressible Flow: With Historical Perspective”, 3rd edition, McGraw-Hill, pp. 102-111. Heiser, W.H. and Pratt, D.T., 1994, “Hypersonic Airbreathing Propulsion”, 5th edition, AIAA Education Series, pp. 1-27. Minucci M.A.S. et al., 2000, “Experimental Investigation of an Electric Arc Simulated “Air-Spike” in Hypersonic Flow”, AIAA paper 00-0715, 38th Aerospace Sciences Meeting & Exhibit, Reno. Minucci M.A.S. et al., 2005, “Laser-Supported Directed-Energy ‘Air Spike’ in Hypersonic Flow”, Journal of Spacecraft and Rockets, Vol. 42, No. 1, pp. 51-57. Myrabo L.N. and Lewis J., 2009, “Lightcraft Flight Handbook: LTI-20”, Apogee Books, pp. 49-66. Myrabo L.N. and Raizer Y.P., 1994, “Laser Induced Air-spike for Advanced Transatmospheric Vehicles”, AIAA paper 94-2451, 25th AIAA Plasmadynamics and Laser Conference, Colorado.

Oliveira A.C., 2008, “Investigação Experimental da Adição de Energia por Laser em Escoamento Hipersônico de Baixa Densidade”, PhD Thesis (in Portuguese), INPE, São José dos Campos. Oliveira A.C. et al., 2008, “Bow Shock Wave Mitigation by LaserPlasma Energy Addition in Hypersonic Flow”, Journal of Spacecraft and Rockets, Vol. 45, No. 5, pp. 921-927. Salvador I.I. et al., 2006, “Experimental Analysis of Heat Flux to a Blunt Body in Hypersonic Flow with Upstream Laser Energy Deposition - Preliminary Results”, 4th International Symposium on Beamed Energy Propulsion, AIP Conference Proceedings, Vol. 830, pp. 163-171. Siegman A.E., 1986, “Lasers”, 1st edition, University Science Books, pp. 663-679. Toro, P.G.P., 1998, “Experimental Pressure and Heat Transfer Investigation over a Directed-Energy Air Spike Inlet at Flow Mach Number of 10 to 20, Stagnation Temperature of 1000K, with Arc Power up to 127 kW”, PhD Thesis, Rensselaer Polytechnic Institute, Troy.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.43-48, Jan.-Mar., 2013

doi: 10.5028/jatm.v5i1.208

Thermal Protection System and Trajectory Optimization for Orbital Plane Change Aeroassisted Maneuver Antonio Mazzaracchio1

Abstract: The aim of this paper was to identify, for a specific maneuver, the optimal combination between the trajectory and the associated heat shield configuration, namely the locations and thicknesses of the ablative and reusable zones, that maximize the allowable payload mass for a spacecraft. The analysis is conducted by considering the coupling between the trajectory’s dynamics and the heat shield’s thermal behavior while using a highly representative model of the heat shield. A global optimization procedure and original software were developed and implemented. The analyzed mission considers an aeroassisted transfer from two low Earth orbits with an assigned orbital plane change maneuver for a given delta wing vehicle equipped with a heat shield consisting of both ablative and reusable materials. The results indicate that the aeroassisted maneuver is more convenient than a “full propulsive” maneuver in the analyzed case, even considering the increased vehicle mass due to the presence of the heat shield. Keywords: Aeroassisted maneuver, Heat shield, Optimization, Orbital plane change, Thermal protection system.

INTRODUCTION One of the most innovative concepts introduced in recent decades to meet the new, more stringent cost requirements of space missions is to make use of aeroassisted orbital maneuvers. These maneuvers can significantly reduce the propulsion requirements and travel time of a mission in favor of a higher available payload mass allocation. Several studies have shown that using the atmosphere for assistance can benefit various classes of orbit (Walberg, 1985). Clearly, the spacecraft must be designed with an efficient aerodynamic configuration to properly utilize the atmosphere for the required global energy management. Aeroassisted maneuvers are an extension of purely gravityassist maneuvers, but use a closer approach to a celestial body to encounter its atmosphere. Essentially, a gravity-assist maneuver depends on the size and mass of the planet and how closely it can be approached, whereas an aeroassisted maneuver depends also on the atmospheric properties and on the vehicle’s aerodynamic characteristics. A spacecraft, which necessarily encounters portions of atmospheric flight during a mission, must normally be equipped with a thermal protection system (TPS) to protect its structures, equipment, and payload from aerodynamic heating. The aerothermal conditions to which the vehicle is subjected are generally very severe and, in some cases, extreme. The optimal design of a spacecraft and its trajectory is always a compromise between the conflicting interests of performance, safety and cost. Obviously, minimizing the TPS mass in compliance with the security constraints is a

1.Astronautical, Electrical and Energetic Engineering Department/Sapienza University of Rome – Rome – Italy Author for correspondence: Antonio Mazzaracchio | Via Salaria 851 | 00138 Rome – Italy | E-mail: a_mazzaracchio@hotmail.com Received: 06/12/12 | Accepted: 12/02/13

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.49-64, Jan.-Mar., 2013

Mazzaracchio, A

primary requirement. In fact, any savings in the TPS mass can be translated into an increased payload mass. The problem becomes more structured when considering aeroassisted maneuvers, as it also takes into account the mass of propellant required during the various phases of the maneuver itself. The search for increased benefits at the propulsive level from an aeroassisted maneuver implies greater use of the atmospheric phase of flight and a need for a more massive TPS. The question driving this paper is whether the increased TPS mass may cancel the propulsive benefits of the aeroassisted maneuver. The TPS design is closely related to the trajectory that must be followed. It is evident that a change in the planned trajectory of the vehicle generates, from the perspective of global optimization, a variation to the TPS mass as a result of the modified aerothermal and environmental conditions. The most critical parameters commonly considered in the design of a spacecraft’s TPS are the “peak heat flux”, the “peak dynamic pressure”, and the “total heat load” (time integrated heat rate). The first two drive the choice of material for the TPS, and the last defines the thickness of the material. A fourth parameter (although not specific to the TPS) is the “peak deceleration”, which is extremely important in the case of manned missions or payload limitations on the tolerable deceleration values. During the phases of the mission in which they carry out their task of thermal protection, ablative TPS materials lose mass and thus engender a change in the shape of the vehicle. Similarly, changes that occur in the shape of the heat shield influence the vehicle’s trajectory. Therefore, the dynamic and thermal problems are essentially coupled. However, they are generally treated separately because of the extremely high computational cost of solving the coupled problem. The usual procedure is, first, to optimize the trajectory and, then, to identify the configuration and the size of the heat shield that will ensure compliance with the imposed thermal constraints. More specifically, the optimized trajectory is first designed based on predefined heating limits. Then, a TPS able to withstand these thermal fluxes and thermal loads is designed. However, this decoupling technique generally leads to a nonoptimal solution. Additionally, simple behavioral models are commonly employed in the ablative analysis. In this paper, the problem is instead solved by coupling the dynamic and thermal analyses. The detailed thermal analysis is performed at the stagnation point, and numerical

and experimental approximations are used to calculate the heat entering the residual vehicle surface. A new tool for trajectory and heat shield optimization is developed for the conceptual development of a spacecraft and its mission. An aeroassisted maneuver executing a change of orbital plane between two circular orbits of the same radius is analyzed. The spacecraft is equipped with a heat shield consisting of ablative and reusable materials. The thermal models implemented are highly representative, and a genetic algorithm-based optimizer is used. The remainder of this introduction provides a historical perspective on trajectory and TPS optimization methods. A review of the model and the governing equations of the problem is presented in next chapter. Afterwards, the optimization procedure is discussed along with the heat shield model adopted. A case study on a delta wing vehicle is presented, and the relevant results and analyses are discussed. The final chapter offers a summary, conclusions, and recommendations for future improvements. A historical review of the methodologies used in the combined optimization of trajectories and TPSs is described by McGuire et al. (2004) and is briefly summarized hereinafter. In 1974, Garcia and Fowler introduced a new parameter for optimizing the trajectory and TPS of the Space Shuttle. In place of the simple heat load, which is well known to be the major driver for the mass of the heat shield, they used an “objective function” expressed as an integrated function of the heating rate at the stagnation point and of the angle of attack, with limits on the surface temperature. For the design of the HL-20 vehicle in 1993, Powell defined a guidance and control scheme to maintain a constant entering heat flux at the stagnation point by varying the bank angle but keeping the angle of attack constant. Powell also left the total heat load unconstrained. The trajectories obtained from this method were used by Wurster and Stone to calculate the relevant aerothermal database, which was in turn used as the main criterion in the selection of TPS materials. In 1998, Hill et al. generated optimal trajectories for the X-33 using the temperatures of ten different zones of the vehicle as constraints. They employed an aerothermal database environment, created by Prabhu et al., using computational fluid dynamics (CFD) techniques. This comprehensive database was created by decoupling the thermal and dynamic problems regardless of the trajectory employed.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.49-64, Jan.-Mar., 2013

Thermal Protection System and Trajectory Optimization for Orbital Plane Change Aeroassisted Maneuver

In two independent works, also in 1998, both Windhorst et al. and Chou et al. generated optimal trajectories with the aim of minimizing the heat load at a specific point on the TPS surface. Also in 1998, Allen et al. created an integrated multidisciplinary tool (consisting of aerothermodynamics, trajectory generation, and heat shield sizing modules) for the optimization of both trajectories and TPSs for planetary probes, in which the heat load was minimized, leaving the user to decide the TPS material distribution. In 2001, Nishioka and Ogasawara, using a method similar to that introduced previously by Garcia and Fowler, derived an expression that depends on the geometry of the vehicle and minimizes the heat load in the reentry trajectories of the Space Shuttle. Also in 2001, to generate optimal trajectories minimizing the mass of the TPS for a crew transfer vehicle (CTV), Saunders et al. adopted a methodology similar to that of Hill in 1998. They used several aerothermal performance constraint curves specific to each material. These curves correlated the altitude and speed of the vehicle so as not to exceed specific temperature limits, but left both the total heat load and the flight time unconstrained. There is also a large body of literature that has addressed the problem from a multidisciplinary point of view that also involves the design of the vehicle. Examples of such approaches can be found in works by Menees (1983), Windhorst et al. (2004), and Joshua et al. (2008).

Aeroassisted orbital plane change Figure 1 shows a classic scheme for an aeroassisted maneuver that changes the inclination of the orbital plane. The strategy is to use a combination of propulsive maneuvers in space and aerodynamic maneuvers in the atmosphere. More precisely, one assumes that the propulsive phases of the maneuvers are all concentrated in three impulses, and, therefore, the part of the flight within the atmosphere takes place without the use of propulsion. It is also assumed that the change to the orbital inclination occurs entirely during the atmospheric portion of the flight. The first propulsive impulse is a deorbit impulse, varying the speed by ΔV1 from the initial LEO of altitude HA to enter the atmosphere along an elliptic orbit segment. The second is a boost impulse upon exiting the atmosphere and is expressed by a speed variation of ΔV2 to achieve the final LEO by ascending once more along an elliptic orbit segment. The third and final impulse is a circularizing impulse, which varies the speed by ΔV3 to circularize the vehicle’s path within the final LEO of altitude HB, which is equal to HA by assumption. It is useful to describe in more detail the various phases of the mission as follows. Initially, the vehicle is moving on a circular orbit of radius RA with a speed VA, around the Earth,

∆V3

MODEL AND GOVERNING EQUATIONS

∆V2

The aim of the present research, as already indicated, was to identify optimal trajectories, subject to heat flux constraints, as well as other possible limitations, while leaving the total heat load unconstrained. The objective is to minimize the sum of the masses of the TPS and the propellant necessary to accomplish the desired orbital maneuver. The vehicle has a given size, shape, and initial total mass, and it must carry out an assigned variation to the inclination of its orbital plane between two circular low Earth orbits (LEOs) of the same radius.

∆V1

∆i

Figure 1. Schematic of aeroassisted orbital plane change maneuver.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.49-64, Jan.-Mar., 2013

Mazzaracchio, A

(1)

where: RA=R⊕+HA (2) The deorbit is accomplished by applying the first impulse ΔV1 against the spacecraft’s motion. This impulse places the vehicle along an elliptic orbit with the perigee inside the dense layers of the atmosphere. The atmospheric region below the altitude Hatm, in which the aerodynamic effects are considered to be conventionally present, is denoted the “sensible atmosphere”. One can determine the speed Vi and the flight path angle γi of the vehicle’s trajectory at the atmospheric entry point (H=Hatm) as follows: 2

⎛ 1 (V − ΔV1 ) − 1 (3) + A 2 μ⎟ V i = 2 RA ⎝ R atm ⎛ ⎟ ⎝

⎧ R γ i = cos -1 ⎜ A (V A − ΔV1 ) ⎩R atmVi

⎧ ⎜ ⎩

(4)

1 1 − R B R atm ΔV 2 = 2 µ − Vu (7) 2 ⎛ RB 2 1−⎟ cos γ u ⎝ R atm Once the final altitude is reached, the third impulse is applied to circularize the final orbit. The expression for the third ΔV is the following:

ΔV 3 =

where:

1 1 − R B R atm ⎛ RB

1 −⎟

⎝ R atm

Ratm=R⊕+Hatm (5) Obviously, it is necessary that the applied ΔV1 be greater than the minimum ΔV1, min, for which there would be only a tangential trajectory to the edge of the sensible atmosphere. This ΔV1, min is given by the following expression:

1 1 − R atm R A μ − 2μ 2 RA ⎛ RA −1 ⎟ ⎝ R atm

cos 2 γ u

RB cos γ u (8) R atm

Aerodynamic heating The atmospheric flight reduces the vehicle’s kinetic energy, which is mostly converted into the entering heat flux. This heat flux is evaluated at the stagnation point through correlations for both the convective and radiative components. The Sutton-Graves correlation is used for the convective component of the heat flux:

(6)

⎛ ⎟ ⎝

During the atmospheric portion of flight, the vehicle performs the required change in orbital inclination, being optimally controlled by modulations of both the angle of attack α and the bank angle σ while subject to the heating constraints. During this phase, the vehicle’s speed decreases due to aerodynamic drag.

q con

⎛ρ = C c ⎟ atm ⎝ rn

⎛ ⎟ ⎝

Δ Vi ,min =

μ − 2μ RB

⎛

V 3 ⎟ 1− ⎝

hw h0

⎛ ⎟ ⎝

VA = μ R A

Because of the energy lost during the atmospheric crossing and the turn, a new impulse is necessary to achieve the final altitude. At the end of atmospheric flight, the vehicle is situated at an altitude Hatm again, is driven at a speed Vu , and has a flight path angle equal to γu. A second propulsion impulse (boost) is applied to enter an ascending elliptic orbit with the apogee equal to the radius of the final circular orbit (in this case, the same as the initial orbit). The required ΔV2, as a function of Vu and γu, can be found from the following expression:

⎛ ⎟ ⎝

which has radius R⊕. The expression of the circular speed is the following:

⎛ ⎟ ⎝

(9)

and includes the “hot wall” correction factor 1–(hw/h0). The radiative heat flux correlation is based on the TauberSutton relation: b qrad = Cr rna ρ atm f (V )

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.49-64, Jan.-Mar., 2013

(10)

Thermal Protection System and Trajectory Optimization for Orbital Plane Change Aeroassisted Maneuver

The heat flux is assumed to be independent of the angle of attack. Thus, the latter appears only in the guidance equations. This assumption is commonly acceptable for an engineering tool suitable for conceptual design phases (an example of this assumption can be found in a paper by Gogu et al., 2009). THERMAL MODEL The thermal analysis is performed with a onedimensional plane model (Fig. 2). The model for the ablative part of the heat shield and the relevant assumptions are thoroughly described in previous work by the author (Mazzaracchio and Marchetti, 2010). However, it is worth briefly recalling here the set of equations that form the basis of the model.

Convective heating

Radiative heating Transpiration blocking effect

Reradiative cooling

= x

∂ ⎛ ∂T − q R ⎟k ∂x ⎝ ∂x

+ (hd − h )

s r ρ c P

∂ρ ∂t

(12)

∂h ∂T + m g d ∂x t ∂x

In Eq. 12 , the coordinate system y is fixed, whereas the mobile coordinate system x moves with the receding surface; initially, the two coincide. The terms in Eq. 12 each represent, in order: the rate of storage of thermal energy, the net rate of thermal energy transferred by conduction and internal radiation, the energy-consumption rate from pyrolysis, the convective energy rate due to coordinate system movement, and the convective rate from the pyrolysis gases. Because of . the hypothesis of the internal opacity of the body, the qR term is null. For both virgin material and char, the local specific heat depends on temperature, whereas the local thermal conductivity depends on both temperature and pressure. The pyrolysis gas enthalpy hd also depends on temperature and pressure, whereas the quantity h, which is the partial heat of charring, is defined as a function of both virgin and charred material properties as follows:

ρ vhv ρv

− ρ chc −ρc

(13)

The Arrhenius equation is used to represent the pyrolysis decomposition in a large-scale model:

Melting zone Surface recession mechanical erosion surface chemistry

Gas cooling Mature char Conduction

Pyrolysis gases

Conduction Bond-line

⎛ ρ − ρc ∂ρ exp − BT = K cf⎟ ∂t ρ − ρ v c ⎝ ⎛ ⎟ ⎝

Boundary layer

Virgin ablative material

∂T ∂t

ρ cP

(11)

q tot = q con + q rad

Reaction zone

The internal energy balance is expressed as follows: ⎛ ⎟ ⎝

The value of the constant in Eq. 9 can be found in work by Havey (1982). The values of the constants in Eq. 10 and the tabulated function f(V) are given by Tauber and Sutton (1991). The total heat flux at the stagnation point is then the sum of both the convective and radiative contributions:

(14)

The internal decomposition transforms part of the solid ablative material into pyrolysis gases. Because of the assumptions of quasi-static one-dimensional flow and impermeability of the interface with the virgin material zone, the pyrolysis gases’ mass flow is related to the decomposition by the following simple expression:

Substructure Spacecraft interior

∂m g ∂y

∂ρ ∂t

(15)

figure 2. schematic of the ablation phenomena. J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.49-64, Jan.-Mar., 2013

Mazzaracchio, A

Q in = q c , blow + q rad + q comb − F σSB ɛ w (Tw4 − T∞4)

(16)

The first term, qc, blow, comes from considering the reduction of the convective component by the injection of pyrolysis gases into the boundary layer. A second order approximation, a function of the mass flow of the outgoing gases known as “transpiration theory”, is assumed to represent this “blocking effect”. Setting: h A = 0 (αc m c + α g m g ) (17) q c , w

the final expression for the net convective heat flux is given by the following equations: ⎧ A < 2.25 ⎪ 2 ⎪ q c , blow = (1 − 0.724 A + 0.13 A ) q con ⎨ (18) ⎪ ⎪ A ≥ 2.25 ⎩ q c , blow = 0.04 q con

The coefficients αc and αg are used to differentiate the molecular weight of the gases in the boundary layer from the molecular weight of the injected pyrolysis gases, respectively. The coefficient αc also takes into account the fraction of char that is removed mechanically and not due to ablation. In the presence of an oxidizing atmosphere, the heating due to combustion of the solid ablation products in the boundary layer must be considered, and complete combustion is assumed: q comb = m c Δhcomb (19)

Finally, the char recession rate ṡr is obtained from experimental data as a function of either the surface temperature or the total entering heat flux. As a consequence of char removal, the surface moves with respect to the fixed coordinate system. The distance between the initial position of the surface and the current position gives the thickness loss. This distance is given by:

sr =

∫

t 0

sr dt

(20)

The model of the reusable portion of the heat shield is a classic model used for heat transfer without phase changes and surface recession, and was derived from the ablative model by eliminating all of the irrelevant phenomena due to the pyrolysis. Equations 12 and 16 become respectively:

ρ cP

∂T ∂t

= x

∂ ⎛ ∂T ⎟k − q R ∂x ⎝ ∂x

(21)

Q in = q con + q rad − F σSB ɛ w (Tw4 − T∞4)

(22)

Obviously, the dependence from temperature and pressure of the thermal characteristics is preserved. Equation 21 is also used for modeling the heat transfer in the substructure. Atmospheric flight The vehicle is considered to be moving subject to a Newtonian inverse square gravitational law, neglecting the Earth’s rotation: g=μ/R2 (23) The three degree-of-freedom differential equations of motion in spherical coordinates are the following (Vinh et al., 1980):

dR = V sin γ (24) dt dθ V cos γ cos ψ = (25) dt R cos ϕ

dϕ V cos γ sinψ (26) = dt R dV D = − − g sin γ (27) dt m

dγ 1 ⎛ L cos γ V2 = ⎟ − g cos γ + cos γ (28) dt V ⎝ m R

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.49-64, Jan.-Mar., 2013

⎛ ⎟ ⎝

The conditions at the external ablating surface are determined by the convective and radiative components of the aerodynamic heating and by the thermo-chemical interactions with the warm gases in the boundary layer. The energy balance at the surface is expressed in the following relation:

⎛ ⎟ ⎝

Thermal Protection System and Trajectory Optimization for Orbital Plane Change Aeroassisted Maneuver

dψ 1 ⎛ L sin σ V 2 = ⎟ − cos γ cosψ tan ϕ dt V ⎝ m cos γ R

(29)

⎛ ⎟ ⎝

The aerodynamic lift and drag forces are given, respectively, by: L = m CL ρ S V 2 2

(30)

D = m CD ρ S V 2 2

(31)

Assuming a parabolic drag polar, the drag coefficient of the vehicle can be written as: CD=CD0+KDCL2 (32) where the lift coefficient CL is a linear function of the angle of attack: CL=CL,αα (33)

This expression can be evaluated using the values at the atmospheric exit, assuming an initial inclination of 0°, because, as suggested, the change of orbital plane is accomplished entirely during the atmospheric portion of the flight. Finally, a fourth-order Runge-Kutta integration is employed in the trajectory computation.

OPTIMIZATION Figure 3 shows a diagram of the procedure adopted here. The inputs to the method consist of the vehicle model and the requirements and constraints for the maneuver. The trajectory is integrated and the thermal analysis is performed as previously described. The process for determining the minimum thickness of the TPS is iterative. The optimization methodology used here is a genetic algorithm (GA) described hereafter.

and obeys the following limitation: 0≤CL≤CL, max (34) Concerning Eq. 33 and Eq. 34, it can be assumed that the vehicle always enters the atmosphere with the same attitude: zero incidence and a bank angle equal to -180° (upside down). The angle of attack cannot take negative values because the lift modulation is enacted through the bank angle. In fact, it should be noted that a negative CL value can result from either a negative pitch angle or from a vehicle flying upside down with a positive pitch angle. In this latter case, to get a downward force, one must fly upside down with a positive lift coefficient. A non-negative flight path angle is required at the atmospheric exit to ensure the completion of the maneuver: γu≥0 (35) To evaluate the change in the orbital plane inclination, it is possible to use the following relationship between inclination, heading, and latitude: cos i = cosϕ cosψ (36)

Orbital maneuver

Spacecraft model

Trajectory

Heat flux TPS thickness minimization

TPS ablative/reusable Fitness function GA optimization

Figure 3. Diagram of optimization procedure.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.49-64, Jan.-Mar., 2013

Mazzaracchio, A

Genetic algorithm Genetic algorithms, developed by Holland (1975), are a global search optimization method based on the fundamental principles of Darwinian evolution, such as natural selection and genetic mechanics. Such a technique is — and has been — successfully used as a global optimization method for complex problems in many research fields and is also used as a suitable method for spacecraft trajectory optimization (Gath and Calise, 2001; Rajesh, 2002; Igarashi and Spencer, 2005). Other methods that are widely used in trajectory optimization problems because of their efficiency are the “gradient-based” optimizers. The efficiency of these optimizers, however, is strongly influenced by the researcher’s ability to provide good initial guesses and precise derivatives of the “fitness function”. Without a good initial guess, the method may not converge or may produce a local minimum solution, which is often unacceptable. These fundamental aspects of the method involve a substantial degree of intervention by expert users. Conversely, GA-based optimizers do not suffer from such limitations as they generate their own starting points. Thus, these optimizers are virtually insensitive to the initial guess. They are also intrinsically robust to non-linearities and discontinuities in the objective function. Their primary disadvantage is that GA-based optimizers require a large number of iterations and, subsequently, a longer computing time due to their generally slow convergence. The following fact must be emphasized: a GA cannot find the optimal solution to a problem. Instead, it identifies the fittest individual in a population that has undergone a process of evolution for the improvement of the species. Evolution, even biological, does not aim to attain an optimum. The GA will use the population it has at its disposal to generate individuals that are above the average while also taking into account the constraints associated with the development. In other words, the GA does not perform a mathematical optimization in the strict sense of the term. Nevertheless, GAs are a highly robust and efficient optimization methodology. The GA adopted here refers to a mixed one-point/twopoints crossover operator together with a reproduction plan that provides a full generational replacement with elitism (Charbonneau and Knap, 1995; Charbonneau, 2002a and 2002b). The termination criterion corresponds to the maximum number of generations chosen and the main parameters adopted for the case studies presented here are summarized in Table 1.

The state variables chosen for the optimization procedure are: the time history of both the angle of attack and the bank angle, the ΔV1 used for deorbit, and, if required, the transit time in the atmosphere. In this study, the latter was left as a free parameter with an upper limit beyond which the mission is considered a failure. This assumption serves to exclude runs in which the vehicle flies by gliding up and down without enough energy to exit the atmosphere. The fitness function is defined as the sum of an objective function (which must be maximized in the present case) plus the contributions of one “reward factor” for each constraint. Such a scheme allows easier handling of the constraints than the classic “penalty function” in a maximization problem such as this. In fact, the various reward factors are directly summable to the objective function. The reward factor Rf, j chosen here (Fig. 4), where ‘j’ indexes the various constraints, is a stepped pyramidal function (Yeniay, 2005). The function assumes a zero value outside of the range of definition (the base of the pyramid), which is defined by a semi-extension ‘est-j’ around the desired Table 1. GA main parameters.

Number of individuals in population Number of generations Number of genes Crossover probability Initial mutation rate Minimum mutation rate Maximum mutation rate Relative fitness differential

100 200 5 0.85 0.005 0.0005 0.25 1.0

Rf, HF 8 6 4

•

step-j

abl Q max

goal-j

2 est-j 0 500

550

600

650

•

Q in, max 700 750 W/cm 2

800

850 900

Figure 4. Example of the stepped pyramidal reward factor for the heat flux constraint.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.49-64, Jan.-Mar., 2013

Thermal Protection System and Trajectory Optimization for Orbital Plane Change Aeroassisted Maneuver

Objective function, constraints, and fitness function The total initial mass of the vehicle is the sum of the propellant mass, the TPS mass, and the structural and payload masses. The goal of the current optimization problem is to perform the orbital plane change while minimizing the sum of the mass of the propellant and the mass of the TPS needed. The objective function, which must be maximized, is then given by the final mass of the vehicle, that can be defined as the performance index of the problem. Initially, the propellant consumption due to the deorbit impulse must be calculated. The vehicle’s mass when it enters the atmosphere can be found by applying the Tsiolkovsky equation for the impulse: m ve , i = m ve , ini e

−

ΔV1 g 0 I sp

(37)

This mass will be further reduced by mTPS, los during the passage through the atmosphere because of TPS ablation.

The mass loss due to ablation comes from both surface recession, and the material density change due to pyrolysis. Thus, the mass of the vehicle at the atmosphere exit will be the following: mve, u = mve,i – mTPS, los (38) At this point, the boost and circularization impulses are applied in sequence, and the final vehicle mass is obtained by two successive applications of the Tsiolkovsky equation: ΔV2 + ΔV3 − g 0 I sp (39) m ve , fin = m ve , u e

Compliance with the heat flux constraints and the required variation of the orbital inclination are ensured by the reward factors. These reward factors are added to the objective function with appropriate multiplicative weights wj, which are set to foster rapid convergence. The final expression of the fitness function ff of the genetic algorithm is the following:

ff = wm

m ve , fin m ve , ini

+ wΔi R f , Δi + w HF R f , HF (40)

The convenience of the aeroassisted maneuver must be assessed with respect to the classic “all propulsive” maneuver in which the required inclination change is made by a single propulsive impulse outside of the atmosphere. For circular orbits in which only the inclination is changed while all other orbital characteristics remain the same, the required ΔV is the following: ⎛ Δi μ ΔV ap = 2 sin⎟ (41) RA ⎝ 2 ⎛ ⎟ ⎝

value of the constraint ‘goal-j’ [‘goal-j’ ± ‘est-j’]. The range of definition is, in turn, divided into subintervals, each of which have an amplitude equal to ‘step-j’, and each characterized by a value of the “level of respect” (or, otherwise, the “level of violation”) of the constraint that is proportional to the distance from ‘goal-j’. This process allows the calculation of the first quantized contribution to the reward factor. A minor corrective addendum is then added within the same step. This addendum is a linear function of the distance from the boundary of the subinterval in question and generates the inclination of the step. This structure of reward factors used in the genetic algorithm allows for a significant jump in value between one level and the next. It also allows the appropriate ordering of two points belonging to the same subinterval of merit. To choose the parameters defining the stepped pyramidal function, the value of the function at the end of a subinterval must not exceed the initial value of the next level. The additive contribution to the fitness function from the reward factors will grow as the variable values approach the values of the constraints, highlighting the individual in question within the entire population of the genetic algorithm. When all constraints are simultaneously satisfied, the value of the reward factor for this solution is enhanced.

Consequently, the expression for the final mass of the vehicle is: m ve , fin , ap = m ve , ini e

−

ΔVap g 0 I sp

(42)

The developed optimization method was validated by comparison with the trajectories presented by Gogu et al. (2009).

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.49-64, Jan.-Mar., 2013

Mazzaracchio, A

HEAT SHIELD MODEL point of evaluation of the heat flux

reusable

ablative

(b). Ablative versus reusable TPS areas.

Figure 5. Schematic of TPS model.

1 0.8 q/q_nose

Mapping and dimensional evaluation of TPS The mapping and sizing of the TPS are carried out through the procedure detailed as follows (Fig. 5(a) and Fig. 5(b)). • The total area of the vehicle is calculated as the sum of the contributions of the nose cone, the body and the two wings. • The total area of the heat shield is reduced by a certain percentage from the total area of the vehicle to reflect the actual coverage of the TPS (Gogu et al., 2009). • A thermal safety factor (TSF) is multiplied to the calculated thickness of the TPS materials. • The nose cone is divided into spherical segment panels, the body into cylindrical panels and the wings into rectangular panels. • For each panel, a reduction factor for the entering heat flux ⋅ is calculated with respect to the stagnation point flux Q . For the nose cone and the body, this reduction is only considered longitudinally. For the wings, this reduction factor is expressed as a product of a transverse factor (on the leading edge) and a longitudinal factor. Figures 6(a) and 6(b) show the longitudinal and transverse normalized reduction factors taken from a work by Reuther et al. (2004) and slightly modified here.

(a). TPS panels schematic.

0.6 0.4 0.2 0

0.2

0.4

0.6

0.8

x/l (a). Normalized heat reduction factor: longitudinal [nose and body].

1 0.8 q/q_nose

The kinematic conditions of the mission and the properties of the atmosphere are such that the aerodynamic heating is significant. The use of a considerable heat shield is thus necessary, as is often highlighted. Considering the spacecraft as a lifting vehicle, one assumes the presence of a protective coating on a portion of the outer surface of the vehicle. In particular, only the lower areas of the nose, body, and wings are covered by the TPS, corresponding to about 63% of the vehicle’s total surface. Due to the limited allowable maximal heat flux and surface temperature for the reusable materials, the severity of the thermal environment does not allow for the adoption of a fully reusable heat shield. Conversely, the adoption of a fully ablative TPS may result in significant additional mass, as the density of most of these materials is considerably higher than the reusable ones. Moreover, due to the lower entering heat flux on some regions of the TPS, the adoption of ablative materials in these zones would be unnecessary. A hybrid TPS is thus adopted, in which the use of the reusable material is confined to the areas where the entering ⋅ reu . heat flux is lower than the specified operational limit Qmax

0.6 0.4 0.2 0

0.2

0.4

0.6

0.8

y/l (b). Normalized heat reduction factor: transverse [wings].

Figure 6. Normalized heat reduction factor.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.49-64, Jan.-Mar., 2013

Thermal Protection System and Trajectory Optimization for Orbital Plane Change Aeroassisted Maneuver

• To be conservative, the reduction factor is calculated

•

• •

•

in the middle of the upper edge of each panel (point of maximal entering heat flux). The type of material used for each panel, reusable or ablative, is chosen depending on the level of the entering ⋅ reu . heat flux compared to the value of Qmax Each panel contributes to the mass per unit area of the TPS in proportion to its thickness and type of material. The required thickness of each panel is considered to be directly proportional to the entering heat flux. This assumption allows a “reduced virtual equivalent surface” to be obtained as the sum over panels of the product of each panel surface with its respective reduction factor. For the reusable area, this reduction factor is again ⋅ reu ⋅ abl and Qmax . In fact, in this divided by the ratio between Qmax instance, the initial thickness is equal to the maximum thickness of the reusable material itself.

Finally, in evaluating the mass, the adopted scheme is equivalent to a TPS with a uniform thickness equal to the maximum, including the TSF, but distributed on a “reduced virtual equivalent surface”. At this point, a methodological clarification is needed. For maximal savings in the TPS mass, as already indicated, the reusable material is assumed to cover all of the areas where the values of heat flux permit. In all other zones, the ablative material will be employed. Nevertheless, the zones where the reusable material can be used are still not defined. The entering heat flux at the stagnation point, calculated from Eq. 11, is a function of the trajectory. For this reason, and because of the correlation with the heat flux entering the other areas of the vehicle, a mapping of the TPS in terms of reusable vs. ablative material is not definable at the beginning. However, at the end of the genetic evolutionary process, an optimal trajectory will be found that obeys the constraints imposed on the heat flux, among others. Thus, a TPS mapping consistent with this final situation can be adopted a priori. Then, as a result of these assumptions, the distribution of the entering heat flux is fixed in advance, making it possible to define the boundaries of the reusable material. Obviously, during the search for the optimal trajectory, in all of the runs in which the heat flux constraints are not respected, an inappropriate TPS configuration will be achieved. This is especially true for runs in which the heat flux value is greater than the constraint. In these cases, a part of the reusable heat shield will be overexposed to entering heat flux,

⋅ reu implying partial incompliance with the constraint value Qmax . Conversely, if the vehicle will be exposed to a smaller heat flux than the constraint, the ablative material will be used more than necessary, covering areas that could be covered with the reusable material. Note that in the unconstrained case, and precisely because of the absence of any indication concerning the entering heat flux, one assumes that the entire shield is ablative. Returning again to the operational details, the design parameter chosen for sizing the TPS is the temperature of the bond-line (TBL), i.e., the temperature of the adhesive junction layer between the heat shield and the substructure. It is required that TBL be within the specified design temperature. The required minimum thicknesses are determined by independently iterating on both types of material. The minimum thickness of the ablative material is determined at the stagnation point, whereas the minimum thickness of the reusable material is calculated at the points with an entering heat flux equal to its operational limit. The actual thicknesses of all other points of the heat shield are then considered to be linear with the local heat load. Therefore, because the exposure time is the same for all points, the required thicknesses are ultimately assumed to be linear with the maximum heat flux entering the point. Note that during the iterative process to determine the minimum thickness within a single run, the trajectories do not vary noticeably from one iteration to another because the initial mass is constant by definition. In effect, the savings in the TPS mass due to the decreased thickness of the heat shield are offset by the mass of the structures and the payload so as to leave the initial total mass of the vehicle unchanged. The developed thermal model and sizing tool were validated by the author (Mazzaracchio and Marchetti, 2010) by comparison with an industry standard high-fidelity ablation and thermal response program, namely the “Fully Implicit Ablation and Thermal” (FIAT) software of NASA Ames Research Center.

CASE STUDY Spacecraft The analyzed spacecraft was characterized by a delta-wing configuration with a high L/D, roughly similar to the Boeing X-37 vehicle. The dimensions and aerodynamic characteristics used were taken in part from Gogu et al. (2009) and NASA

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.49-64, Jan.-Mar., 2013

Mazzaracchio, A

Facts (2003). Other data were reasonable assumptions made by the author. The primary dimensions and characteristics of the vehicle are listed in Table 2, and the principal aerodynamic and propulsive parameters are listed in Table 3. Atmospheric flight Table 4 presents the orbit altitudes of the required maneuver and the assumed conventional limit of the Earth’s sensible atmosphere. The atmospheric properties were derived from the 1976 US Standard Atmosphere model. TPS materials The spacecraft had a supposed heat shield consisting of a material called phenolic impregnated carbon ablator, known as PICA-15, for the ablative zones and LI-900 for the reusable zones. PICA-15 was used in the Stardust probe, whose reentry capsule returned to Earth in 2006 with samples of cosmic and cometary dust collected during its mission. PICA is a relatively recent material that is characterized by a low density — well below that of the classic carbon phenolic — while at the same time having a high ablative capability for elevated heat fluxes. PICA also features a thermal conductivity lower than that of other ablative materials for equal entering heat fluxes. PICA consists of chopped carbon fibers as insulation with a phenolformaldehyde resin as an infiltrant. The high porosity of the composite is the reason for its low density and thermal conductivity (Tran et al., 1997). LI-900 is a typical reusable material designed for use on the Space Shuttle TPS. LI-900 comes in the form of a surface insulating tile and is used to minimize thermal conductivity while ensuring maximum resistance to thermal shock. It is made from 99.9% pure silica glass fibers, which occupy only 6% of the total volume (Williams and Curry, 1992). Both the ablative and reusable layers are bonded onto the substructure, which consists of a 12.7 mm thick carbon-carbon/aluminium honeycomb sandwich. The substructure is considered an integral part of the TPS.

RESULTS AND ANALYSIS The study was conducted by analyzing three different cases corresponding to three different values for the entering heat

Table 2. Vehicle’s dimensions and characteristics.

mve, ini

4898.7 kg

Vehicle length

lve

9.38 m

Vehicle body radius

1.00 m

Vehicle wing span

wsve

4.50 m

Vehicle wing cord

wcve

3.50 m

Vehicle reference surface

11.69 m2

Vehicle TPS total surface

STPS, ve

42.65 m2

Bond-line limit temperature

TBL, lim

450 K

TSF

Gross vehicle mass

Thermal safety factor

Table 3. Vehicle’s aerodynamic and propulsive characteristics.

CD0

Zero-lift drag coefficient

0.032

Induced drag factor

1.4

Lift coefficient derivative

CL,α

0.5699

Maximum lift coefficient

CL , max

0.4

Isp

310 s

Propellant specific impulse Table 4. Maneuver’s characteristics.

Initial LEO altitude

185.2 km

Final LEO altitude

185.2 km

Required inclination change

Δi

18°

Hatm

129.6 km

Atmosphere’s upper limit

⋅ abl flux constraint. The first two cases corresponded to Qmax equal 2 to 397 and 568 W/cm , which are equivalent, respectively, to 350 and 500 Btu/(ft2∙s). The third case was unconstrained. The study required a running time of approximately 160 days on a I7 Intel® processor with 8 threads and a clock speed of 3.2 GHz. The fittest case was identified from a population of about 3.5 million individuals. The “all propulsive” case for this mission provided, by mean of Eqs. 41 and 42, a value for ΔVap equal to 2438.20 m/s, which implies a consumption of 2703.51 kg of propellant and a net vehicle mass equal to 2195.19 kg. Table 5 shows a comparison of the main characteristics of the obtained optimal trajectories and presents the maximum values of the dynamic pressure and the load factor. These latter are presented only for verification purposes, as they are not subject to constraints; in any case, their observed values are compatible with the feasibility of the mission. All cases were characterized by similar values of ΔV1 and ⋅ abl . γi, whereas ΔV2 and Hmin decreased with increasing Qmax Concurrently, the flight time tfl and the total heat load HL ⋅ abl . increased with increasing Qmax

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.49-64, Jan.-Mar., 2013

Thermal Protection System and Trajectory Optimization for Orbital Plane Change Aeroassisted Maneuver

Table 5. Trajectory comparison.

Table 6. Heat shield comparison.

397

568

No Limit

⋅ Qin, max, W/cm2

397.00

568.00

668.66

ΔV1, m/s

31.74

30.24

31.73

ΔV2, m/s

1349.22

1278.05

1227.93

ΔV3, m/s

65.57

58.01

71.76

1579.17

1616.80

1642.28

Hmin, km

44.07

40.03

37.73

γi, °

-0.46

-0.44

-0.46

γu, °

0.84

0.77

0.89

HL, J/cm2

72471

88448

88620

q, kN/m2

57.75

104.81

143.77

n, g

4.33

5.46

5.05

tfl, s

Heating constraints ⋅ abl Qmax , W/cm2

397

568

No Limit

mTPS , kg

308.66

284.81

267.37

abl mTPS , kg

164.76

190.20

198.19

74.72 69.18

25.43 69.18

69.18

3.69

6.70

7.07

mTPS, los , kg

76.86

91.83

95.34

mTPS, prop , kg

reu TPS

, kg

ss mTPS , kg

Sr , mm

Table 6 presents the main differences among the three optimal TPS scenarios. As previously stated, a fully ablative heat shield is used in the unconstrained case. In this “not limited” case, the maximum observed heat flux is equal to 668.66 W/cm2, which is much lower than the operating limit of the employed material (PICA-15). As expected, the thickness of the ablative part, and consequently its mass, grow as the heat load increases; in contrast, the overall mass of the TPS decreases considerably. This apparent contradiction lies in the type of materials used. Obviously, as the heating constraint takes higher values, the ablative portion of the heat shield is extended. However, in the present work, which uses a low density ablative material, it is disadvantageous to use the reusable materials. In fact, even though they have a lower density, they require greater thicknesses. These factors combine to produce a lighter TPS when a fully ablative configuration is employed. Clearly, this behavior is not general, and different results are possible when denser ablative materials are used. The mass lost during the crossing of the atmosphere due to the pyrolysis phenomena should be noted. Its value in the unconstrained case is 48% of the entire ablative part, 36% of the entire TPS mass, and 2% of the gross vehicle mass. The surface recession is about 7 mm. Finally, in this unconstrained case, the mass fraction of the entire TPS is approximately 5.5% of the gross vehicle mass. Additionally, at the propulsive level, the total consumption of propellant also diminishes with the increasing value of the heating constraint. The unconstrained case is therefore the most beneficial case in terms of savings in both the propellant mass and TPS mass, when using these particular TPS materials. To obtain the maximum benefit, in principle, one must adopt the highest

1827.37

1741.88

1705.30

δTPS , mm

abl

87.50

89.79

90.34

reu

δTPS , mm Mass gain wrt “all propulsive” case, kg Gain wrt mve, ini, %

32.15

31.49

567.48

676.82

730.84

11.58

13.82

14.91

allowable heating constraint. However, in this case, higher load factors and dynamic pressures are to be expected and may restrict the maximum allowable heat flux. In Fig. 7, a comparison of all of the masses involved highlights the benefit of this maneuver in comparison with the “all propulsive” case. The unconstrained case allows a considerable payload mass gain of 730.84 kg, which corresponds to about 15% of the gross vehicle mass. Figures 8, 9, 10, and 11 show the altitude, the velocity, the total entering heat flux, and the vehicle’s mass, respectively, as functions of the flight time during the atmospheric pass. When the lowest thermal constraint is imposed, the vehicle must dive into the atmosphere earlier to perform the assigned maneuver, even if it will descend less than in the other cases (Fig. 8). More precisely, to achieve the desired change in the orbital plane, it will fly for a longer time in less dense layers of the atmosphere than in the cases with a higher flux limit. This is necessary to prevent violation of the heat constraint and because of the lower 3000 2500 Mass, kg

Heating constraints ⋅ abl Qmax , W/cm2

567.5 74.7 164.8 69.2

2000 1500 1000 500 0

676.8 25.4 190.2 69.2

730.8 198.2 69.2

2703.5 1827.4 Propellant

"all propulsive"

Substructure

397 W/cm²

1741.9 Ablative

Reusable

568 W/cm²

1705.3 Mass Gain

No Limit

Heat Flux Constraint Figure 7. Mass comparison.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.49-64, Jan.-Mar., 2013

Mazzaracchio, A

available maneuverability in this region. However, the total flight time is ultimately greater for the unconstrained case. This fact and the lower minimum altitude reached, makes the total heat load for this trajectory higher. Figure 9 shows that the reduction in the velocity is less important in the unconstrained case, justifying the requirement of a smaller ΔV2 at the atmosphere exit. Figures 12, 13, and 14 represent the coverage areas, the type of material and the thicknesses of the heat shield, respectively, for all three cases on a half-vehicle scheme. It should be noted that the assumptions have led to a theoretical thickness step on the border between the reusable and ablative parts.

4900 4850 Mass, kg

4800

Heating Constraints 397 W/cm2 568 W/cm2 Unconstrained

4750 4700

200 400 600 800 1000 1200 1400 1600 1800 Atmospheric Flight Time, s

Figure 11. Mass versus time.

140

Reusable - Thickness, mm

120

397 W/cm2 568 W/cm2 Unconstrained

40 20

Heating Constraints

200 400 600 800 1000 1200 1400 1600 1800 Atmospheric Flight Time, s

Figure 8. Altitude versus time.

Above 115 105–115 95–105 85–95 75–85 65–75 55–65 45–55 35–45 25–35 15–25 5–15 Below 5

Width, m

Altitude, km

100 80

Ablative - Thickness, mm

Above 45 40–45 35–40 30–35 25–30 20–25 15–20 10–15 5–10 Below 5

1 0

5 Length, m

⋅

Stagnation Point Heat Flux W/cm

Width, m

5 Length, m

Ablative - Thickness, mm Above 115 105–115 95–105 85–95 75–85 65–75 55–65 45–55 35–45 25–35 15–25 5–15 Below 5

700 600 500

Heating Constraints

300 100 0

397 W/cm2 568 W/cm2 Unconstrained

200

1 0

200 400 600 800 1000 1200 1400 1600 1800 Atmospheric Flight Time, s

Figure 10. Total entering heat flux versus time.

⋅

800

400

abl Figure 13. TPS configuration for the case with Qmax = 568 W/cm2.

Width, m

Velocity, m/s

1 0

Figure 9. Velocity versus time.

Above 115 105–115 95–105 85–95 75–85 65–75 55–65 45–55 35–45 25–35 15–25 5–15 Below 5

397 W/cm2 568 W/cm2 Unconstrained

200 400 600 800 1000 1200 1400 1600 1800 Atmospheric Flight Time, s

Ablative - Thickness, mm

Above 45 40–45 35–40 30–35 25–30 20–25 15–20 10–15 5–10 Below 5

Heating Constraints

abl Figure 12. TPS configuration for the case with Qmax = 397 W/cm2. Reusable - Thickness, mm

8000 7800 7600 7400 7200 7000 6800 6600 6400 6200 6000

5 Length, m

Figure 14. TPS configuration for the unconstrained case.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.49-64, Jan.-Mar., 2013

Thermal Protection System and Trajectory Optimization for Orbital Plane Change Aeroassisted Maneuver

SUMMARY AND CONCLUSION A new tool was developed for the optimization of aeroassisted maneuvers by coupling the dynamic and thermal problems, and taking into account the mass of the vehicle’s TPS. This tool, which is suitable for conceptual design activities, refers to highly representative thermal models for reusable and ablative materials, and is based on a genetic algorithm optimizer. A maneuver to change the orbital plane between two LEOs with the same radius was analyzed, taking the maximum allowable entering heat flux at the vehicle’s stagnation point as a parameter. The optimal trajectories for completing the specified mission were found along with the respective optimal configurations of the TPS.

The results of this case study indicate that the aeroassisted technique is much more convenient than “all propulsive” maneuvers, even when considering the increased mass due to the need to give the vehicle a heat shield. In fact, the total budget between the savings in the propellant mass and the increased TPS mass was positive and allows for a substantial increase in the payload mass of the vehicle. The most favorable case is the one with no limit on the entering heat flux and a full ablative heat shield. The most important future developments and improvements are the following: • Extension of the analysis to other types of missions. • Extension of the analysis to other types of materials. • Improvement of the genetic optimizer by coupling it with other optimization methods for faster convergence.

REFERENCES Charbonneau, P., 2002a, “An introduction to genetic algorithms for numerical optimization”, Boulder, Colorado.

Menees, G.P., 1983, “Thermal protection requirements for near-Earth aeroassisted orbital transfer vehicle missions”, AIAA Paper 83-1513.

Charbonneau, P., 2002b, “Release notes for Pikaia 1.2”, Boulder, Colorado.

NASA Marshall Space Flight Center, 2003, “X-37 demonstrator to test future launch technologies in orbit and reentry environments”, NASA Facts, FS-2003-05-65-MSFC.

Charbonneau, P. and Knap, B., 1995, “A user’s guide to Pikaia 1.0”, Boulder, Colorado. Gath, P.F. and Calise, A.J., 2001, “Optimization of launch vehicle ascent trajectories with path constraints and coast arcs”, Journal of Guidance, Control and Dynamics, Vol. 24, No. 2, pp. 296-304. doi: 10.2514/2.4712. Gogu, C., Matsumura, T., Haftka, R.T. and Rao, A.V., 2009, “Aeroassisted orbital transfer trajectory optimization considering thermal protection system mass”, Journal of Guidance, Control and Dynamics, Vol. 32, No. 3, pp. 927-938. doi: 10.2514/1.37684. Havey, K.A. Jr., 1982, “Entry vehicle performance in low-heat-load trajectories”, Journal of Spacecraft and Rockets, Vol. 19, No. 6, pp. 506512. doi: 10.2514/3.62293. Holland, J.H., 1975, “Adaptation in natural and artificial systems”, University of Michigan Press, Ann Arbor, MI. Igarashi, J. and Spencer, D.B., 2005, “Optimal continuous thrust orbit transfer using evolutionary algorithms”, Journal of Guidance, Control and Dynamics, Vol. 28, No. 3, pp. 547-549. doi: 10.2514/1.11135. Joshua, E.J., Lewis, M.J. and Starkey, R.P., 2008, “Analysis of optimal Earth entry heat shield/trajectory configurations”, 15th AIAA International Space Planes and Hypersonics Systems and Technologies Conference, Dayton, Ohio, paper no. AIAA-2008-2594. Mazzaracchio, A. and Marchetti, M., 2010, “A probabilistic sizing tool and Monte Carlo analysis for entry vehicle ablative thermal protection systems”, Acta Astronautica, Vol. 66, issues 5-6, pp. 821-835. doi: 10.1016/ j.actaastro.2009.08.033. McGuire, M.K., Gage, P., Galloway, E.T., Huyhn, L., Nguyen, J., Bowles, J.V. and Windhorst, R., 2004, “Trajectory and thermal protection system design for reusable launch vehicles”, AIAA Paper 2004-4490.

Rajesh, K.A., 2002, “Reentry trajectory optimization: evolutionary approach”, AIAA Paper 2002-5466. Reuther, J.J., Brown, J.L., Prabhu, D.K., McDaniel, R., Saunders, D. and Palmer, G., 2004, “External computational aerothermodynamic analysis of the Space Shuttle orbiter at STS-107 flight conditions”, 37th AIAA Thermophysics Conference, Portland, Oregon, Paper no. AIAA- 2004-2281. Tauber, M.E. and Sutton, K., 1991, “Stagnation-point radiative heating relations for Earth and Mars entries”, Journal of Spacecraft, Vol. 28, No. 1, pp. 40-42. doi: 10.2514/3.26206. Tran, H.K., Johnson, C.E., Rasky, D.J., Hui, F.C.L., Hsu, M.T., Chen, T., Chen, Y.K., Paragas, D. and Kobayashi, L., 1997, “Phenolic impregnated carbon ablators (PICA) as thermal protection systems for discovery missions”, NASA TM-110440. Vinh, N.X., Busemann, A. and Culp, R.D., 1980, “Hypersonic and planetary entry flight mechanics”, University of Michigan Press, Ann Arbor, MI. Walberg, G.D., 1985, “A survey of aero-assisted orbit transfer”, Journal of Spacecraft and Rockets, Vol. 22, No. 1, pp. 3-18. doi: 10.2514/3.25704. Williams, S.D. and Curry, D.M., 1992, “Thermal protection materials thermophysical property data”, NASA RP 1289. Windhorst, R., Galloway, E., Lau, E., Saunders, D. and Gage, P., 2004, “Aerospace vehicle trajectory design and optimization within a multidisciplinary environment”, AIAA Paper 2004-704. Yeniay, Ö., 2005, “Penalty function methods for constrained optimization with genetic algorithms”, Mathematical and Computational Applications, Vol. 10, No. 1, pp. 45-56.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.49-64, Jan.-Mar., 2013

Mazzaracchio, A

LIST OF SYMBOLS A = see Eq. (17) a,b = exponents, dependent on ρatm and V BT = activation temperature, K Cc = constant, dependent on the atmosphere CD = drag coefficient CD0 = zero-lift drag coefficient CL = lift coefficient CL,α = derivative of CL wrt angle of attack CL,max = maximum lift coefficient Cr = constant, dependent on the atmosphere cP = specific heat, J/(kg∙K) D = drag, N dt = integration time step, s F = exterior view factor f (V) = function, dependent on the velocity f = fitness function g = gravitational acceleration, m/s2 g0 = gravitational acc. at sea level, m/s2 H = altitude, m HA = altitude of the initial LEO, m Hatm = altitude of the sensible atmosphere, m HB = altitude of the final LEO, m Hmin = minimum flight altitude, m HL = heat load, J/cm2 h = enthalpy, J/kg h = partial heat of charring, J/kg hd = pyrolysis enthalpy, J/kg hw = wall enthalpy, J/kg h0 = total enthalpy, J/kg hsp = propellant specific impulse, s i = inclination, rad Kcf = collision frequency factor, kg/(m3∙s) KD = induced drag factor k = thermal conductivity, W/(m∙K) L = lift, N l = current length, see Figs.6(a) and 6(b), m l ve = vehicle length, m m = mass, kg • mc = char removal rate, kg/(m2∙s) • mg = pyrolysis gas mass flow rate, kg/(m2∙s) • mTPS = TPS mass, kg abl

mTPS = mass of the ablative part of the TPS, kg ss

mTPS = mass of the TPS substructure, kg reu mTPS = mass of the reusable part of the TPS, kg mTPS, los = TPS mass lost, kg

mprop = propellant mass, kg mve, fin= final vehicle mass, kg mve, fin, ap = ”all propulsive” case final mass, kg mve, i = vehicle mass at atmospheric entry, kg mve, ini = initial vehicle mass, kg mve, u = vehicle mass at atmospheric exit, kg n = load factor, g nr = decomposition reaction order ⋅ Qin = net total heat flux at surface, W/m2 ⋅ Qin, max = maximum entering heat flux, W/m2 ⋅ abl Qmax = max heat flux for ablative part, W/m2 ⋅ reu Qmax = max allowable heat flux for reusable part, W/m2 q = dynamic pressure, kN/m2 = net hot wall convective heat flux, W/m2 q⋅ c, blow

q⋅ comb = combustion heat flux, W/m2 q⋅ con = net hot wall convective heat flux, W/m2 q⋅ = internal radiative heat flux, W/m2 R

q⋅ rad = radiative heat flux, W/m2 q⋅ = total heat flux, W/m2 tot

R = radius, m RA = initial LEO radius, m Ratm = radius of the sensible atmosphere, m RB = final LEO radius, m Rf, `j` = reward factor for component ‘j’ R⊕ = Earth’s radius, m

rn = vehicle nose radius, m rb = vehicle body radius, m S = vehicle reference surface, m2 • sr = surface recession, m sr = char recession rate, m/s STPS, ve = vehicle TPS total surface, m2 T = temperature, K TBL = bond-line temperature, K TBL, lim = bond-line limit temperature, K Tw = wall temperature, K T∞ = freestream temperature, K t = time, s tfl = flight time, s V = velocity modulus, m/s VA = circular orbit speed in initial LEO, m/s VB = circular orbit speed in final LEO, m/s Vi = speed at atmospheric entry, m/s Vu = speed at atmospheric exit, m/s w`j` = multiplicative weight of component ‘j’ wcve = vehicle wing cord, m wsve = vehicle wing span, m x = mobile coordinate system, y-S, m; or current centerline abscissa, see Fig. 6 (a), m y = fixed coordinate system, m; or current wing ordinate, see Fig. 6 (b), m Δi = variation of the orbital inclination, rad Δhcomb = combustion heat per unit weight, J/kg ΔVap = “all propulsive” impulse, m/s ΔV1 = deorbit impulse, m/s ΔV1,min = minimum deorbit impulse, m/s ΔV2 = boost impulse, m/s ΔV3 = circularizing impulse, m/s α = angle of attack, rad αc = weighting factor for char mass loss αg = weighting factor for pyrolysis gases γ = flight path angle, rad γi = flight path angle at atmospheric entry, rad γu = flight path angle at atmospheric exit, rad δtps = initial thickness of the ablative part, mm δtps = initial thickness of the reusable part, mm εW = surface emissivity θ = longitude, rad μ = gravitational parameter, m3/s2 ρ = density, kg/m3 ρ atm = atmospheric density, kg/m3 σ = bank angle, rad σSB = Stefan-Boltzmann constant, W/(m2∙K4) ϕ = latitude, rad ψ = heading angle, rad Subscripts A = value at initial LEO ap = “all propulsive” B = value at final LEO BL = bond-line c = char fl = flight g = pyrolysis gas HF = heat flux i = value at atmospheric entry los = lost m = mass prop = propellant TPS = concerning Thermal Protection System u = value at atmospheric exit v = virgin material ve = vehicle w = wall Δi = inclination variation Superscripts abl = ablative ss = substructure reu = reusable

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.49-64, Jan.-Mar., 2013

doi: 10.5028/jatm.v5i1.162

Study by FT-IR Technique and Adhesive Properties of Vulcanized EPDM Modified with Plasma Renata Patrícia dos Santos1,2, Mauro Santos de Oliveira Junior1, Elizabeth da Costa Mattos3, Milton Faria Diniz3, Rita de Cássia Lazzarini Dutra1,3

Abstract: The surface of vulcanized ethylene propylene diene terpolymers (EPDM) was modified by Ar and N2, microwave generated plasma in order to improve adhesion properties. Surface modification was characterized by universal attenuated total reflectance Fourier Transform Infrared (UATR/FT-IR), because it presented lower interference of ingredients of EPDM formulation when compared with other techniques used for the attenuated total reflectance (ATR) to different crystals (ATR/KRS-5 and ATR/Ge). Oxygenated groups were introduced on the surface after treatments which were formed when the activated surface was exposed to the plasma gas. In treatments with nitrogen, oxygen groups and probable nitrogen groups were formed on the surface and could be identified by FT-IR. Reduction in the measurement of the contact angle and an increase in the work of adhesion and in the peel strength (EPDM X Polyurethane (PU)) were observed after the treatment resulted in the improvement of the adhesion properties of the modified surface. Keywords: EPDM, Plasma, Surface characterization, UATR/FT-IR.

INTRODUCTION In the aerospace industry, EPDM rubber became an interesting alternative as a thermal protection for rocket motors because of its low density, low processing cost and also because it does not produce toxic compounds while burning, showing advantages in relation to nitrile butadiene rubber (NBR), the copolymer, which is normally used for this application (Moraes et al., 2007a). However, the elastomers of the ethylene-propylene are apolar, presenting low adhesion. With the purpose of improving this characteristic, surface treatment is required (Rapra Review Reports, 2002), without modifying the bulk properties of the polymer (Hegemann et al., 2003; Viadaurre et al., 2002). Plasma treatments are alternatives to reach these characteristics, besides being harmless to the environment (Costa et al., 2008). The disadvantage of plasma is that it requires a vacuum system, thus increasing the cost of the treatment (Moraes et al., 2007b). The consulted literature mentions some studies in which plasma treatments were used in EPDM and the characterization of the effect on the surface of the elastomer, approaching, in some cases, the sensitivity of the techniques used (Moraes et al., 2007a; Grythe and Hansen, 2006; Awaja et al., 2009). ATR/FT-IR is a technique that has been used to characterize different polymers. An example is the improved adhesion between polyethylene and the acrylic

1.Instituto Tecnológico da Aeronáutica – São José dos Campos/SP – Brazil 2.Indústria Química Una Ltda. – Itaquaquecetuba/SP – Brazil 3.Instituto de Aeronáutica e Espaço – São José dos Campos/SP – Brazil Author for correspondence: Rita de Cássia Lazzarini Dutra – Avenida Marechal Eduardo Gomes, 50 – Vila das Acácias | CEP 12228-904 São José dos Campos/SP – Brazil | E-mail: rclzdutra@iae.cta.br Received: 19/09/12 | Accepted: 16/01/13

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.65-74, Jan.-Mar., 2013

Santos, R.P., Oliveira Junior, M.S., Mattos, E.C., Diniz, M.F. and Dutra, R.C.L.

acid monomer with grafting surface, which provided information to detect the formation of oxygen-containing species on the poly(fluoride) after the glow discharge treatment (Basak et al., 2010). The surface modification of the EPDM, treated in Ar/O2 RF plasma, was studied by Basak et al. (2010), and the presence of oxygenated groups was identified in the surface by ATR/FT-IR (ZnSe crystal) and XPS (X-Ray Photoelectron Spectroscopy). A long treatment led to the reduction of the adhesion strength; the author reports this behavior with the reduction of surface energy and the antiadherent nature of sulphur blooming. Grythe and Hansen (2006) studied the effect of Ar, O2 and N2 plasmas on EPDM films (casting) and differences were revealed through atomic force microscopy (AFM), to which film surface presented more roughness when treated with O2, and a smoother surface when treated with N2. The biggest change was the presence of oxygen in hydroxyl, carbonyl and carboxyl groups, observed by XPS. Aiming at the excellent properties of EPDM in the aerospace field, Moraes et al. (2007a) carried out comparative studies on the adhesive properties of untreated NBR versus the EPDM treated with N2/Ar and N2/H2/Ar RF plasma. The XPS analysis showed that polar groups containing nitrogen and oxygen were incorporated to the surface after treatment. Adhesion strength to the interface EPDM/polyurethane liner improved after treatment, reaching values of interface NBR/polyurethane liner, system used nowadays. Moraes et al. (2007c) also studied the adhesion properties of EPDM treated with N 2/Ar and O 2/Ar plasma reactive ion etching (RIE) by FT-IR. The functional groups formed on the surface were: C-N, C=N and C=O, besides C-O functional groups, due to the rubber oxidation in N 2/Ar; C-O, C=O, O-C=O, C-O-O and CO 3 were generated in O2/Ar. Other cutting edge techniques of FT-IR, such as UATR, have stood out for the analysis of different materials when compared to transmission techniques. Basically, studies show there are not significant differences in the fingerprint region, below 1,500 cm -1 for the spectrum obtained by the method of transmission and by UATR (Waltham, 2005; Abibi and Hequet, 2005). In the nondestructive technique UATR of internal reflection, used to analyze solids, powders, liquids and gels, an

IR beam passes through an ATR element, with high refractive index, consisting of ZnSe-diamond or KRS-5diamond, and reaches the sample surface (Fig. 1a). Another requirement of UATR is the good contact between the crystal and the sample surface. The analysis of probe strength (Fig. 1b) can be adjusted in order to obtain a more appropriate contact, once different pressure levels directly influence the intensities of the obtained ATR spectrum. In most of the mentioned studies and reviewed literature, it was observed that the use of FT-IR contributed to the characterization of the chemical species formed as a consequence of the treatment used on the polymeric surface. Spectrum of the EPDM and polyolefin treated in plasma was found with ATR, however, the use of UATR is not emphasized, the new-generation of the FT-IR technique (Moraes, 2007c; Weon and Choi, 2009).

Metal mounting plate: Diamond IRE 316 S/S or Hastalloy

ZnSe (or KRS-5) focusing element

(a)

(b)

Figure 1. Scheme of UATR (a) probe and (b) top plate (Perkin Elmer).

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.65-74, Jan.-Mar., 2013

Study by FT-IR Technique and Adhesive Properties of Vulcanized EPDM Modified with Plasma

Therefore, in this analysis we present the study of surface modification by microwave plasma system, the EPDM rubber with known diene and ethylene content, in which new active groups incorporated on the surface were characterized by UATR, and the evaluation of the adhesion (EPDM versus PU) through T-peel strength of the adhesive joint and goniometry. EXPERIMENT Materials Samples of EPDM rubber with different diene (ethylidene-norbornene) and ethylene content, one with high diene content (3–6%) and approximate ethylene content of 50%, Nordel IP 4520, and the other with low diene content (<1%) and approximate ethylene content of 42%, Nordel IP 3430, were both supplied by the Dow Chemical Company. The EPDM rubbers were prepared by the company Flexlab Consultoria e Treinamento Ltda. And the formulations contained 100 phr (parts per hundred) of rubber, 1 phr of carbon black, 2 phr of paraffinic oil, 1 phr of Tetramethylthiuram Monosulfide (TMTM), 0.5 phr of Mercaptobenzothiazole (MBT), 0.5 phr of stearic acid, 2 phr of ZnO and 1 phr of sulfur. The rubber was vulcanized at 190°C for 12 min. Soxhlet extraction From 1.0 to 1.5 g of EPDM rubber wrapped in filter paper was used; 100 mL of acetone was added to a bottom flask. The sample, already wrapped in a filter paper, was placed in the Soxhlet extractor which was connected to the flask. The condenser was connected to the extractor for water circulation to foil the thermostatic bath during the entire extraction process. Extraction was carried out for 8 hours. The excess solvent was transferred to a flask and taken to the oven for 20 to 30 minutes for solvent evaporation, and the residue was analyzed by FT-IR. Ultrasonic cleaning The samples of EPDM rubber were ultrasonically cleaned at a 37 kHz frequency in a stainless-steel tank, under usual cleaning conditions (Peters, 1996). They were immersed in Unalimp WB 2151 solution (detergent) supplied by the chemical industry Una Ltda. It was

diluted (4:21) and taken to the ultrasonic tanks for 300 seconds at 60°C. After the cleaning, they were rinsed in water also at 60°C for 3 minutes. Samples were then taken to the oven at 64°C (Tecnal, model TE-394/2) for 20 minutes; then, they were treated in plasma, after approximately 18 hours of cleaning. Contact angle measurement Contact angle measurements were performed using a Standard Automated Digital Goniometer model 500-00 of Ramé-Hart Inc., with the software DROPimage Advanced for calculation. Obtained contact angle was 4 to 7, in different points, by dropping deionized water. Plasma treatment The plasma for this study was generated in a microwave-excited plasma reactor that consists of a rectangular section chamber 230 mm wide, 80 mm high and 500 mm deep with a door for observation and for insertion and withdrawal of samples. There is a mechanical vacuum pump for reduction of the internal pressure of the chamber (Edwards model E2M18). The electric discharge was obtained by magnetron (2.45 GHz). Experiments were performed using Ar and N 2 gas flow rate of 20 sccm (standard cubic centimeters per min). The times ranged from 10 to 300 seconds. T-peel strength measurements (EPDM x PU) For the evaluation of EPDM adhesion (after treatment) versus PU adhesive versus plasticized polyvinyl chloride (PVC), strips were used to prepare the joints. PVC was chosen because its interface with PU is already known. To prepare the PVC, its surface was cleaned with a paper tissue impregnated with acetone; after the solvent evaporated, polyurethane adhesive was applied. The same polyurethane adhesive was applied on the EPDM surface after the plasma treatment. After water evaporated, since the adhesive is an aqueous dispersion, thin solid adhesive film was formed and heated in hot air, at 64°C, and immediately placed in contact, EPDM x PU adhesive x PVC, both with adhesive and a pressure of 5 kg/cm 2 was applied for 30 seconds, forming an adhesive joint. Values of the T-peel strength EPDM/PU adhesive/ PVC system were obtained in Emic E 500; a peel rate of 100 mm/min was used.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.65-74, Jan.-Mar., 2013

Santos, R.P., Oliveira Junior, M.S., Mattos, E.C., Diniz, M.F. and Dutra, R.C.L.

Analysis by FT-IR spectroscopy The analysis was conducted in the IAE/AQI by using the FT-IR Spectrum One PerkinElmer spectrometer (resolution 4 cm-1, gain 1, 20 scans), by the ATR technique, with KRS-5 (spectral range of 4,000–400 cm-1) and Ge (spectral range of 4,000–700 cm-1) crystals, both at 45°, and using the universal accessory UATR (spectral range of 4,000–550 cm -1), containing ZnSe and diamond, 100 N pressure. The transmission FT-IR technique was used to analyze the extracted residue from the rubber (resolution of 4 cm -1, gain 1, spectral range 4,000–400 cm-1, 20 scans) as liquid film.

(a)

(b) (%) T

(c)

4000.0 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 600 400.0 cm-1

Figure 2. FT-IR spectrum of the surface of EPDM Nordel IP 3430, after being ultrasonically cleaned (a) UATR, (b) ATR/Ge and (c) ATR/KRS-5.

RESULTS AND DISCUSSION Choice of the FT-IR technique: Different FT-IR reflection techniques, ATR (KRS-5 and Ge, with spectral range of 4,000–400 cm-1 and spectral range of 4,000–700 cm-1, respectively) and UATR (spectral range of 4,000–550 cm-1), under specific conditions were used to analyze the EPDM surface after plasma treatment. This study aims to select the most appropriate accessory to characterize the modified surface. Among these techniques, according to Fig. 2 and 3, the UATR showed lower interference of the bands in the other ingredients of the formulation, regardless of the rubber grade (Nordel IP 3430 or 4520), showing only the rubber bands. Therefore, this technique was chosen to characterize the treated rubbers. B ands that characterize the EPDM rubbers are found in their wavenumbers, attributed to their vibrational modes: asymmetrical stretching (υa) of the CH 2 group at 2,920 cm-1, symmetrical stretching (υs) at 2,850 cm-1, asymmetrical (δa) and symmetrical bending (δs) of CH 3 at 1,460 cm-1 and 1,377 cm-1, respectively, and in plain bending or rocking (ρ) of CH2 at 721 cm -1 (Awaja et al., 2009; Babbit, 1978; Basak et al., 2010). Figure 4 shows the transmission FT-IR spectrum of the residues extracted from the rubbers (Nordel IP 3430 and 4520). The extracts were obtained by Soxhlet, in acetone and heated up (Smith, 1979). The attributions of the bands were possible because the composition of the rubbers is known, thus enabling to

(a)

(b) (%) T

(c)

4000.0 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 600 400.0 cm-1

Figure 3. FT-IR spectrum of EPDM surface Nordel IP 4520, after being ultrasonically cleaned (a) UATR, (b) ATR/Ge and (c) ATR/KRS-5.

find in the spectrum IR of the soluble ingredients in the used solvent (Dutra et al., 1995). Bands of the CH2 group in Fig. 4 (a) and (b) are attributed to the presence of paraffinic oil at 2,922 and 2,853 cm-1 and CH 3 at 1,463 and 1,380 cm-1. The C=O groups were observed in 1,739, and C-O in 1,243 cm-1 (Smith, 1979). These oxygenated groups indicate the possible presence of the ester group of zinc stearate formed during vulcanization, after the reaction of stearic acid with ZnO was added to the formulation (Awaja et al., 2009).

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.65-74, Jan.-Mar., 2013

Study by FT-IR Technique and Adhesive Properties of Vulcanized EPDM Modified with Plasma

(a)

(a) (b)

(c)

(%) T (b)

(%) T

(d)

(e)

(f)

4000.0 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 600 400.0 cm-1

Figure 4. Transmission FT-IR spectrum, obtained of residue of rubbers obtained after extraction in acetone - (a) Residue extracted of Nordel IP 3430 and (b) Residue extracted of Nordel IP4520.

CH absorption of aromatics was observed by the presence of MBT (1,597 cm-1), Ring-CH 1 and 2 was disubstituted (ortho) in 748 cm-1. The RSCH3 group was confirmed by the band in 1,319 cm-1. In this case, the presence of thioamides at 1,623 cm-1 could be attributed to the TMTM, besides deformation at 1,142 cm-1 of C-N (Smith, 1979). FT-IR analysis after ultrasonic cleaning The results of the surface analysis by UATR show that the ultrasonic cleaning removed part of the components that bloomed to the surface of the rubber. And it can be confirmed when compared to the spectrum of the rubbers cut in half, in the longitudinal direction, when the internal part was analyzed (Fig. 5).

4000.0 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800

550.0

cm-1

Figure 5. FT-IR spectrum obtained by UATR of rubber Nordel IP 3430 (a) before submitting it to ultrasonic cleaning (b) after cleaning, (c) rubber cut in half. Spectrum of rubber Nordel IP 4520 (d) before submitting it to ultrasonic cleaning (e) after cleaning and (f) rubber cut in half. (a)

(b)

(c)

(d) T (%)

(e)

(f)

FT-IR analysis of EPDM surface, after treatment with Ar and N2 plasma: Figures 6 and 7 present the FT-IR spectrum of the Nordel IP 3430 and Nordel IP 4520 rubbers, both untreated and treated with Ar, N 2 plasma, respectively. In the treatment with Ar (10 seconds and 120 seconds) plasma for Nordel IP 3430, Fig. 6 (b), (c) and (d), oxygenated groups were created and were justified by the presence of the bands OH at 3,500 cm -1 and -COOH/C=C between 1,520 and 1,560 cm-1. Now for the treatment of 300 seconds with Ar, there is was the formation of C-O (1,000 to 1,100 cm-1). Also, OH (650 to 769 cm-1) (Silverstein and

(g)

4000 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 cm-1

550

Figure 6. UATR/FT-IR spectrum of Nordel IP 3430: (a) Reference sample, rubber analyzed after ultrasonic cleaning, untreated, (b) rubber treated in Ar plasma, 10 seconds, (c) 120 seconds, (d) 300 seconds; and in N2 plasma for (e) 10 seconds, (f) 120 seconds, and (g) 300 seconds.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.65-74, Jan.-Mar., 2013

Santos, R.P., Oliveira Junior, M.S., Mattos, E.C., Diniz, M.F. and Dutra, R.C.L.

Webster, 2000) to Nordel IP 4520 in Ar plasma (10 seconds and 120 seconds), Fig. 7 (b) and (c), oxygenated groups were also formed and justified by the presence of the bands C-O (1,000 to 1,100 cm -1) (Smith, 1979) and OH. Although argon is an inert gas, the incorporation of oxygenated groups is common. During the treatment, free radicals are formed and react with oxygen when in contact with the atmosphere (Viadaurre et al., 2002). In the treatment with N2 plasma, oxygenated and possible nitrogenated groups were formed. Some nitrogenated groups absorb at wavenumbers as much as oxygenated groups, thus there is possibility of overlapping bands. The presence of bands was observed in the same groups, such as: OH at 3,500 cm-1 and/or N-H at a range of 3,200-3,500 cm-1 and -COOH (1,520 to 1,560 cm-1) (Figs. 6

(a)

(b)

(c)

(d)

T (%)

(e)

(f)

(g)

4000 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 550 cm-1

figure 7. uaTr/fT-Ir spectrum of nordel Ip 4520 (a) reference sample, rubber analyzed after ultrasonic cleaning, untreated, (b) rubber treated in ar plasma, 10 seconds, (c) 120 seconds, (d) 300 seconds; and in n2 plasma for (e) 10 seconds, (f) 120 seconds, and (g) 300 seconds.

and 7e, f, g), C-O (1,000 to 1,100 cm-1 ) and/or C-N(1,000– 1,200 cm-1) (Figs. 6f and 7g) (Peters, 1996; Babbit, 1978; Basak et al., 2010), OH (650 to 769 cm-1) (Silverstein and Webster, 2000) (Fig. 7g). The presence of nitrogenated groups must be confirmed by XPS, because there is a band overlap in the FT-IR analysis. In literature, it is possible to find that oxygen has high permeability, and could migrate to the layer below the surface (Viadaurre et al., 2002), possibly being identified by FT-IR, whereas nitrogen is a less permeable gas, modifying the surface in nanometer content, being detected by XPS (Moraes, 2007c). Contact angle measurement The contact angle technique was used to characterize the surface properties of untreated and treated samples of EPDM rubber Nordel IP 3430 and 4520. As soon as the treatments were conducted, by dripping deionized water over the rubber surface and using an appropriate syringe, the contact angles were measured after the treatments. In this case, water was chosen for the study of the surface, because the adhesive used in the T-peel strength measurements, which will be discussed later on, has a polar characteristic (polyurethane adhesive). However, several pure liquids may be used to estimate the surface free energy of a material by means of various thermodynamic theories (Rapra Review Reports, 2002). The results (Table 1) show reduced values of contact angle after all the treatments. This reduction indicates change in polarity, greater affinity for water, which consequently leads to the improvement of the adhesive properties of surfaces. As it is known, the work of adhesion (WA) was calculated with the Young-Dupré equation (Eq. 1), using the surface tension (γLV) and the contact angle of the liquid on the surface (Fleming et al., 2010). Thus, when there is complete wetting, it means the liquid spread completely on the surface, and the work of adhesion reaches its maximum value.

WA =

γ LV • cos (θ ) + 1

(1)

Samples treated with argon and nitrogen plasma for only 10 seconds showed reduction in contact angle for both Nordel IP 3430 (54.2° and 57.7°) and Nordel IP 4520 (52.6° and 50.8°), respectively.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.65-74, Jan.-Mar., 2013

Study by FT-IR Technique and Adhesive Properties of Vulcanized EPDM Modified with Plasma

plasma. An improvement was observed after treatments, except for Nordel IP 3430 at Ar/10 s. With the treatment time of 120 seconds (Ar and N2), the values of T-peel were closer between the two EPDM

T-peel strength measurements Figure 8 shows the T-peel strength values of EPDM/Polyurethane adhesive/PVC joints obtained from the EPDM rubbers both untreated and treated with

Table 1. Contact angle and work of adhesion of untreated and treated Nordel IP 3430 and 4520 by plasma. EPDM Rubber – Nordel IP 3430

EPDM Rubber – Nordel IP 4520

Treatment conditions (gas/time)

Contact angle (°)

Work adhesion (mN/m)

Contact angle (°)

Work adhesion (mN/m)

Untreated

122.4±0.5

33.8±0.8

117.9±0.2

38.7±0.2

Ar/10 s

68.2±1.1

99.8±1.3

65.4±0.8

103.2±0.9

Ar/120 s

51.7±1.6

117.9±1.6

49.3±1.7

120.2±1.7

Ar/300 s

52.2±0.6

117.4±0.6

44.5±1.2

124.7±1.0

N2/10 s

64.7±0.1

104.0±0.2

67.1±0.6

101.1±0.7

N2/120 s

60.7±1.5

108.5±1.6

51.6±0.7

118.1±0.7

N2/300 s

57.1±0.8

113.5±1.0

56.2±0.8

113.5±1.0

T - Peel (180°) - EPDM x PVC (joint with polyurethane adhesive)

0.35 0.30

Peel Strength (N/mm)

0.25 0.20 0.15 0.10 0.05 0.00 Untreated

Ar/10 s

N2/10 s

Ar/120 s

N2/120 s

Ar/300 s

N2/300 s

Nordel IP 4520

0.07

0.14

0.19

0.21

0.22

0.28

0.13

Nordel IP 3430

0.09

0.25

0.26

0.19

0.12

0.18

Figure 8. T-peel strength measurements of EPDM/PU adhesive/PVC joints, with Nordel IP 3430 and 4520, untreated and treated with Ar and N2 (10 seconds, 120 seconds and 300 seconds) plasma. J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.65-74, Jan.-Mar., 2013

Santos, R.P., Oliveira Junior, M.S., Mattos, E.C., Diniz, M.F. and Dutra, R.C.L.

grades, whereas with 300 seconds, also with Ar and N 2, a larger variation was shown. Plasma processes are highly dependent on the system, the parameters developed for one system usually cannot be adopted for another, and this may explain the different results between the rubbers (Chan and Ko, 1996). The treatment with N 2/10 seconds shows higher results than that with Ar/10 seconds, probably because N 2 is a reactive gas and Ar is an inert gas, and within a shorter time N2 can already activate the surface. However, long treatment with the same gas that created the functional group on the surface can abstract this formed group or even degrade the surface treated, because of the increased temperature in the ionic atmosphere (Viadaurre et al., 2002). Before and after the treatment adhesive failure on the EPDM surface (adherent) was observed, this for the EPDM/PU adhesive (Fig. 9). The adhesion strength in the material and intermolecular spacing refer sometimes to the surface energy. This means that cohesion (internal strength of the adhesive itself ) was higher than the EPDM/ adhesion (Petrie, 2000). The different values obtained in the T-Peel show a greater or lower affinity with the change in surface.

PVC

This observation was possible in this case because the formulation of the polyurethane adhesive has a fluorescent optical brightening agent that allows the visualization of the adhesive film under ultraviolet light (Figs. 9b and c). Relating the results of contact angle measurements/ T-peel strength/ FT-IR: According to Table 1, Nordel IP 4520 presented reduction in contact angle values when treatment time increased from 120 to 300 seconds with Ar plasma. However, the T-Peel strength was increased (Fig. 8), indicating improvement in the adhesives properties, although it was not possible to detect spectrometric changes by IR (Fig. 7d). The same behavior was not observed with Nordel IP 3430, which presented reduction in T-peel, and this can be explained by the difference between the two grades. A possible explanation is that the adhesives properties are not related only to the formation of the functional groups that, in most cases, were detected by FT-IR. They are also related to the alterations in the morphology of the rubber (Sánchez et al., 2001). Basak et al. (2010) related the improvement in adhesion results to the increase of polarity and superficial irregularity

Adherend Adhesive Adhesive Failure

EPDM Rubber

(a)

Adherend

(b)

(c)

Figure 9. (a) Scheme of adhesive failure after failures observed after T-Peel test and pictures of the sample: (b) Nordel IP 3430 and (c) Nordel IP 4520.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.65-74, Jan.-Mar., 2013

Study by FT-IR Technique and Adhesive Properties of Vulcanized EPDM Modified with Plasma

which contributes to a good mechanical anchorage of the sample treated in plasma. So, the best treatment for Nordel IP 4520 was with Ar (300 seconds). Usually, the duration of 120 seconds, as aforementioned, showed similar results for both rubbers. For the other treatments, in which oxygenated groups and/or possible nitrogenated groups were observed by FT-IR, the improvement in adhesive properties on the EPDM surface can be associated to the increase in surface polarity. Also, the differences between the T-peel strength values are due to the possible differences in the amount of these formed groups and/or change in the morphology surface.

CONCLUSION The cleaning of the rubbers’ surfaces was necessary to remove the whitish layer formed on them. The

evaluation of these surfaces before and after cleaning was possible by means of the FT-IR reflection technique. Among these characterization techniques by reflection of FT-IR, the use of ATR with Ge and KRS-5 crystals showed higher absorption, which is associated to the agents of the formulation ingredients. This led to the choice of the UATR accessory to eliminate the interference of these bands and detect possible alterations of the surface modified in plasma. The UATR analysis detected oxygenated groups on the surfaces treated with Ar and N2 plasma. Possible nitrogenated groups were formed in treatments with N2 plasma, however, nitrogenated groups must be confirmed by XPS. The treatment time of 120 seconds presented a good option for the treatments in Ar and N 2 plasma, because their results were similar with the two different grades of rubber.

REFERENCES Abibi, N. and Hequet, E., 2005, “Fourier Transform Infrared Analysis Of Trehalulose And Sticky Cotton Yarn Defects Using ZnSe-Diamond Universal Attenuated Total Reflectance”, Text Res J, Vol. 75, No. 9, pp. 645-652. Awaja, F. et al., 2009, “Adhesion of polymers”, Progress in Polymer Science, Vol. 34, pp. 948-968. Babbit, R.O., 1978, “The Vanderbilt Rubber Handbook”, R.T. Company, Inc. Vanderbilt, Norwalk, 641p. Basak, G.C. et al., 2012, “Surface modification of argon/oxygen plasma treated vulcanized ethylene propylene diene polymethylene surfaces for improved adhesion with natural rubber”, Applied Surface Science, APSUSC-20806, pp. 1-14. Basak, G.C. et al., 2010, “Characterization of EPDM Vulcanizates Modified with Gamma Irratiation and Trichloroisocyanuric Acid and Their Adhesion Behavior with Natural Rubber”, The Journal of Adhesion, Vol. 86, No. 3, pp. 314-315. Chan, C.M. and Ko, T.M., 1996, “Polymer surface modification by plasmas and photons”, Surface Science Reports, Vol. 24, No. 45. Costa, T.H.F. et al., 2008, “Caracterização de filmes de poliéster modificados por plasma de O2 a baixa pressão”, Revista Matéria, Vol. 13, pp. 66. Dutra, R.C.L. et al., 1995, “Importância da Preparação de Amostras em Espectroscopia no Infravermelho com Transformada de Fourier (FT-IR) na Investigação de Constituintes em Materiais e Compostos”, Polímeros, Vol. 5, pp. 41-47.

Fleming, R.R. et al., 2010, “Plasma treatment of polyacrylonitrile/ vinyl acetate films obtained by the extrusion process”, Polymer Bulletin, Vol. 66, pp. 282. Grythe, K.F. and Hansen, F.K., 2006.,“Surface Modification of EPDM Rubber by Plasma Treatment”, Langmuir, Vol. 22, pp. 6109-6124. Hegemann, D. et al., 2003, “Plasma treatment of polymers for surface and adhesion improvement”, Nuclear Instruments and Method in Physics Research B. Vol. 208, pp. 281–286. Moraes, J.H. et al., 2007a, “Surface improvement of EPDM rubber by plasma treatment”, Journal Phys Appl. Phys, Vol. 40, pp. 7747-7748. Moraes, J.H. et al., 2007b, “Surface modification of EPDM in r. f. plasma: process optimization and surface characterization”. Phys Stat Sol (a), Vol. 204, p. 957. Moraes, J.H. et al., 2007c, “Rapid surface modification of EPDM with oxygen and nitrogen plasmas: a comparative study”, J Optoelectron Adv M, Vol. 9, pp. 475-477. PerkinElmer, “ATR accessories – An overview”. Technical Note, FT-IR Spectroscopy. Peters, D., 1996, “Ultrasound in materials chemistry”, J. Mater Chem. Vol. 6, No. 10, pp. 1605. Petrie, E.M., 2000, “Handbook of adhesives and sealants”. McGraw-Hill, New York. Rapra Review Reports, Polyolefins”, pp.5.

2002,

“Adhesion

and

bonding

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.65-74, Jan.-Mar., 2013

Santos, R.P., Oliveira Junior, M.S., Mattos, E.C., Diniz, M.F. and Dutra, R.C.L.

Sánchez, M.D.R. et al., 2001, “Surface modifications of a vulcanized rubber using corona discharge and ultraviolet radiation treatments”, J Mater Sci, Vol. 36, pp. 5794. Silverstein, R.M. and Webster, F.X., 2000, “Identificação Espectrométrica de Compostos Orgânicos”, 6th Ed, LTC – Livros Técnicos e Científicos Editora, Rio de Janeiro, pp. 84. Smith, A.L., 1979, “Applied Infrared Spectroscopy”, John Wiley & Sons, New York.

Viadaurre, E.F.C. et al., 2002, “Surface Modification of Polymeric Materials by Plasma Treatment”, Mater Res, Vol. 5, pp. 37. Waltham, 2005, “FT-IR Spectroscopy Attenuated Total Reflectance (ATR)”, Technical Note. Catalogue PerkinElmer. Weon, J.I. and Choi, K.Y., 2009, “Surface characterization and Morphology in Ar-Plasma-Treated Polypropylene Blend”, Macromol Res, Vol. 17, pp. 891.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.65-74, Jan.-Mar., 2013

doi: 10.5028/jatm.v5i1.233

Maximization of Fundamental Frequency of Laminated Composite Cylindrical Shells by Ant Colony Algorithm Rubem Matimoto Koide1, Marco Antonio Luersen1

Abstract: The success in developing modern aerospace systems addresses competitive subjects as optimization, reduced costs, sustainability, environment, weight, and safety. There is an increased demand for lighter materials such as laminated composites. In order to match the demand of aeronautical companies, the shell structures are very important. The dynamic behavior of composite structures is also essential to improve the potential applications of these materials. The knowledge of the dynamic response of the cylindrical shell structures is an important issue in their design, in which the ply thicknesses are often preestablished and the ply orientations are usually restricted to a small set of angles due to manufacturing constraints. Obtaining the best stacking sequence of laminated shell may lead to a problem of combinatorial optimization. As this problem is hard to be solved, several techniques have been developed. Ant colony optimization is a class of heuristic optimization algorithms inspired by the behavior of real ants, related to their ability to find the shortest path between the nest and the food source. Thus, this paper deals with optimal stacking sequence of laminated cylindrical composite shells with the aim of maximizing their fundamental frequency using this approach. The ant colony algorithm was implemented in Matlab platform and was linked to Abaqus academic version to compute the structural response. Cylindrical shells with and without a cutout geometry were studied. Fundamental frequencies were maximized for both cases, and results were presented and discussed. Keywords: Frequency maximization, Laminated cylindrical shells, Ant colony optimization, Heuristic optimization.

INTRODUCTION Many successful applications of composite materials and the development of modern aerospace systems are to address competitive subjects, such as optimization, reduced costs, sustainability, environment, weight, and safety. The demand of aeronautical companies for lighter, safer, and less polluting airliners is increasing for the next generations of aircraft, as related by Irisarri et al. (2011), mainly to integrate primary structures such as the centre wing box, fuselage, and wings. As reinforced by Todoroki et al. (2011), for laminated composite structures, the stacking sequence optimization is indispensable. These structures are widely used for aerospace components due to their high specific strength and specific stiffness. Increase use of laminated composite shells has motivated research in shell dynamic behavior. Thus, Shafei and Kabir (2011) analyzed the optimization techniques for improvement of structural elements designs regarding vibration stability or modal frequencies. The dynamic behavior study about composite structures is essential for assessing their full potential, as concluded by Sharma and Mittal (2010). The analysis of natural frequencies of composite/shells plays an increasingly important role in the design of aerospace engineering applications (Sharma and Mittal, 2010). In this way, an optimum structural design aimed at improving stable vibration frequency or the dynamic response was created. This paper has dealt with optimal stacking sequence design of laminated cylindrical composite shells. The stacking

1. Universidade Tecnológica Federal do Paraná – Curitiba/PR – Brazil Author for correspondence: Rubem Matimoto Koide | Laboratório de Mecânica Estrutural (LaMEs)/Universidade Tecnológica Federal do Paraná | Avenida Sete de Setembro, 3.165 | CEP 80230-901 Curitiba/PR – Brazil | E-mail: rubemkoide@hotmail.com Received: 19/08/12 | Accepted: 25/11/12

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.75-82, Jan.-Mar., 2013

Koide, R.M. and Luersen, M.A.

sequence optimization is formulated as a combinatorial problem and is solved using ant colony optimization (ACO), which is a class of heuristic algorithms proposed to solve optimization problems. The idea was inspired by the behavior of real ants, related to their ability to find the shortest path between the nest and the food source. Ant colony heuristic has been applied successfully to different kinds of problems. Due to their capacity to solve hard problem as the stacking sequence of laminated composite, the ACO was simulated in this study. The developed ant colony algorithm was applied to problems of maximizing the fundamental frequency of cylindrical shells with and without a cutout. ACO algorithm was coupled with a finite element code in order to obtain the structural response. Optimized stacking sequence results were presented and the modal frequency of cylindrical shell composite was also computed. The present work may help designers to optimize the dynamic response of complex structures without analytical solutions, and the optimization method may be applied in many others composite parts.

ANT COLONY OPTIMIZATION APPLIED TO CYLINDRICAL SHELLS Ant colony optimization ACO is a metaheuristic in which a colony of artificial ants cooperates in finding good solutions. This technique is applied to difficult discrete optimization problems. As described by Dorigo and Stützle (2004), the main procedure of the ACO metaheuristic manages scheduling of the three components of ACO algorithms via the Schedule Activities construct: management of the ant activity, pheromone updating, and daemon actions. The pseudocode of ACO metaheuristic is presented by Dorigo and Stützle (2004) as; Procedure ACOmetaheuristic ScheduleActivities ConstructAntsSolutions UpdatePheromones DaemonActions % optional End-ScheduleActivities End-procedure

The ConstructionAntsSolutions is a procedure in which the solution of the optimization problem is built. Therefore, the artificial ants apply a stochastic decision rule in a problem defined by a construction graph. While the solution is being built, based on the pheromone trails and heuristic information, the ants evaluate the solutions for searching the optimal feasible candidate. The Update Pheromone procedure is based on the solution available and it influences the quantity of pheromone. It can be increased or decreased. The deposit of new amount of pheromone increases the probability of a good solution in the next decision. The decrease process is related to pheromone evaporation. It influences in a choice of candidate due to the reduction of pheromone trail for the solution available, so a bad candidate is not selected. This process actives the convergence of the algorithm. The DaemonActions procedure can be used as an optional process to optimize the algorithm. Local or global information can be used to decide or influence a search optimization procedure. The Ant Colony System (ACS) is one extension of the first ant algorithm. Additional mechanisms in the ACS are: it exploits the search experience accumulated by the ants through the use of an aggressive action choice rule, as explained by Dorigo and Stützle (2004); the pheromone evaporation and deposit are updated only in a best-so-far connection. ACS algorithm has a framework based on three rules that manage the optimization problem. In this variant, the first procedure is ConstructionAntsSolution or the pseudorandom proportional rule defined by Eq. 1: arg max l∈N K {τil [ηil]β}, q ≤ q0 i

[τil]α [ηil]β K , q ≤ q0 (1) pij = Σ l∈N K[τil]α[ηil]β i

where, q is a random variable uniformly distributed in [0,1], q0 is a parameter for the best possible move (0 ≤ q0 ≤ 1), k is an ant, α is a parameter that determines the relative influence of the pheromone trail, β is a parameter that determines the relative influence of the heuristic information, η is the heuristic information value, i, j are the initial and next choice or candidate,

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.75-82, Jan.-Mar., 2013

Maximization of Fundamental Frequency of Laminated Composite Cylindrical Shells by Ant Colony Algorithm

l is a candidate solution, K Ni is the feasible neighborhood of ant k, K pij is the probability that ant k chooses the next solution if q > q0. If q ≤ q0 , it means that the ant is exploiting the learned knowledge based on the pheromone trails and the heuristic information. If q > q0, the ant explores other tours or searches around the best-so-far solution. The second rule is the Global Pheromone Trail Update. In Eq. 2, the formulation for this update is:

τij ← (1 - ρ) τij + ρΔτijbs, ∀(i , j) ∈ T bs

(2)

where, ρ is the global pheromone evaporation rate (0 < ρ ≤ 1), bs Δτij is the amount of pheromone that ant k deposits on each best-so-far solution, T bs is a set of best connections. The amount of pheromone is defined due to the objective function, in this case the inverse of the best-so-far fundamental frequency. When this rule is applied, both the evaporation and the new pheromone deposits are updated only to the best-so-far ant. Local Pheromone Trail Update, the last rule, is applied during the tour construction. The pheromone evaporation and a new pheromone deposit are updated when an ant is exploiting or exploring the connection, according to the pseudorandom proportional rule. It is formulated by Eq. 3:

τij ← (1 - ξ) τij + ξ τ0 (3) where, ξ is the local pheromone evaporation rate 0 < ξ < 1, and τ0 is the initial pheromone trails value. The effect of the local updating rule is that each time an ant uses an arc(i, j) or connection, its pheromone trail τij is reduced, so that the arc becomes less desirable for the following ants as explained by Dorigo and Stützle (2004). Laminated cylindrical shells The laminated composite material consists of stacks of layers, each one usually composed by a matrix of polymeric

material and fibers oriented in a specific direction. Laminated composites provide the designer the ability to tailor the material according to the application. The structures formed by these materials present high stiffness/ mass and strength/mass ratios. The mechanical properties of composite laminates depend on the material of each layer, the number of layers, the thickness of each layer, and the fiber orientations in each layer. The ply thicknesses are often predetermined and the ply orientations are usually restricted to a small set of angles due to manufacturing constraints. This fact leads to problems of discrete or stacking-sequence optimization. Figure 1 shows the laminated composite, the stacking sequence, and an application in cylindrical shell. Composite cylindrical shell structures find wide applications in many products. These kinds of structures are in general efficient in engineering applications including aircraft, offshore platforms, silos, spacecraft, pressure vessels, pipelines, and many others industry applications. Considering these applications, many studies have been developed in this area. The general two-dimensional theory of laminated cylindrical shells was presented by Barbero et al. (1990). Heyliger and Jilani (1993) studied free vibration of laminated anisotropic cylindrical shells and presented numerical results for isotropic, anisotropic, and orthotropic composite shells. Soldatos (1994) approached a review of three dimensional dynamic analyses of circular cylinders and cylindrical shells. The analytical solutions were proposed by Lam and Qian (2000) for the free vibrations of symmetric angle-ply thick laminated composite cylindrical shell. Regarding noncircular cylindrical shells, Ganapathi and Haboussi (2003) have been developing free vibrations of thick laminates. Ganapathi et al. (2004) studied the free flexural vibration behavior of elliptical cylindrical shells. The effects of shell geometry, cross-sectional properties, layup and ply-angle on the natural frequency, modes of vibrations, were evaluated for noncircular shell structures. Poore et al. (2008) also presented a free vibration analysis of laminated cylindrical shells with a circular cutout. Although advances in vibration studies involved complex mathematics analysis, experimental tests and recently numerical optimization methods, these subject studies are still under development.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.75-82, Jan.-Mar., 2013

Koide, R.M. and Luersen, M.A.

PILIES PLIES

45 -45 0 0 -45

ANGLES

LAMINATED COMPOSITE

CYLINDRICAL SHELL

Figure 1. Laminated composite.

Since this optimization problem is very hard to solve, several techniques have been developed. Genetic algorithms (GA) have been successfully applied to solve these kinds of problems (Le Riche and Haftka, 1993). Recently, a new class of algorithms, the ACO, was developed to solve combinatorial optimization problems. It was inspired by observing the behavior of real ants (Dorigo and Stützle, 2004). Regarding the optimization of composite cylindrical shells, Hu and Wang (1992) studied these structures with and without cutouts, and Hu and Tsai (1999) analyzed the fundamental frequencies. Tabu-embedded simulated annealing was applied for the optimal stacking sequence designed by Rao and Arvind (2007). Stacking sequence optimization of laminated cylindrical shells for buckling and free vibration using genetic algorithm and neural networks have been investigated by Gharib and Shakeri

(2010). Mota Soares et al. (1995) researched the sensitivity analysis of the optimization of general thin shell structures made of composite materials. Chern and Chao (2000) studied the natural frequencies of laminates in spherical, cylindrical, planed, and curved panel geometries by a three-dimensional theory. They also presented the comparison of natural frequency via a three-dimensional augmented energy variation approach. The Layerwise optimization was applied for maximizing the fundamental frequency of laminated panels by Narita and Robinson (2006). Goupee and Vel (2006) proposed a methodology to optimize the natural frequencies of functionally grade structures. Rao and Lakshmi (2009) suggested an algorithm for solving multi-objective optimization problem. Hybrid scatter search algorithm was employed to solve stacking sequence optimization of hybrid fiber reinforced composite plate, cylindrical shell, and pressure vessels. A comparison of different shell theories considering nonlinear vibrations of laminated circular cylindrical shells was obtained by Amabili (2011). Therefore, this work addressed the development of an ACO algorithm applied to the layup design of composite cylindrical shells. Numerical tests were performed for the layup optimization of laminated cylindrical shells with holes aiming at maximizing the fundamental frequency. Ant colony optimizAtion Applied to lAminAted cylindricAl shells ACO was inspired from the studies of real ant colonies foraging behavior. Pheromone is a chemical product that ants deposit on the ground. Its concentration influences probabilistically the choice of best way. This kind of behavior is called stigmergy and is the mechanism that controls communication among the ants, and make them take the shortest paths. Pheromone matrix is the basic information processed by ACO to find the solutions following the way with more concentration of this substance. Artificial ants are used to construct solutions for laminated composite cylinders by choosing probabilistically the orientation of the laminate stacks. An artificial ant means a colony of artificial ones that search the best ways. Figure 2 presents the procedure of ConstructionAntsSolutions. The solutions are built on the past search experience based on the level of pheromone and the heuristic matrix, which brings information about the problem to be solved. The heuristic information is the parameter related to problem, for this case the relation in function of total weight of laminated was considered. The ACO procedure starts with a random

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.75-82, Jan.-Mar., 2013

Maximization of Fundamental Frequency of Laminated Composite Cylindrical Shells by Ant Colony Algorithm

selection for the layers’ orientations. After the artificial ants have finished building the first laminated configurations, the pheromone is released with evaporation and the new amount of pheromone is deposited, based on the best solutions found at local and global iterations. The arc(i,j) denotes the stacking sequence, qualifying the pheromone trail τij for local updating rule, in function of frequency value. The algorithm stops when the maximum number of iterations or objective function evaluations is reached. It is assumed that the laminate is symmetric and balanced with the allowable ply orientations (0/0/+ 45/- 45/90/90). Also, to prevent matrix cracking, a maximum number of four contiguous plies with the same orientation is imposed as a constraint (Le Riche and Haftka, 1993).

In order to validate the finite element code solution, the first numerical studies were a frequency calculation obtained by finite elements, compared to the results obtained by Poore et al. (2008). The problem solved was a cylindrical shell with radius (R=0.05)inch, length (L=2R), thickness (h=0.02R). The material property values were: E22=1.0x106 psi, E11=25E22, G12=G13=0.5E22, v12=0.25, and ρ=1.0. Simply supported (SS3) boundary conditions were adopted along the edges of the shell. Three different stacking sequences were simulated. Table 1 presents the comparison of the normalized fundamental frequency, which is defined as Eq. 4:

ω = ω (R2/h) √ (p/E22),

(4)

where, ω is the fundamental frequency. optimizAtion problem The optimization problem is formulated as Eq. 5:

NUMERICAL RESULTS The case studies considered herein are the maximization of fundamental frequency of composite cylindrical shells. ACO was implemented in Matlab platform and was linked to finite element (FE) code Abaqus in order to compute the frequency response. The ply angles orientation considered for the layers are 0º, ±45º, and 90º. The laminates are considered symmetric and balanced. The first step of simulation is done by ACO, which defines the ply angles and save them in a file that was read in script wrote in Python. This script is connected to Abaqus to send the stacking sequence and receive the computed results. The results are sent again to ACO to compute the optimization. Finite element solution is processed for the fundamental frequency, and the correspondent value is returned to ACO and so on, for searching the maximal fundamental frequency.

plies

01=angle 1 02=angle 2

Xijij nijij pijij

03=angle 3

laminated

Figure 2. ConstructionAntsSolutions procedure.

Find: θk, θk Є {02, ±45, 902}, k=1 to n (5) Maximize: Fundamental frequency Subject to: - Symmetric and balanced laminate - Maximum number of contiguous plies with the same orientation=4 where, θk is the orientation of each stack of the laminate, and n is the total number of stacks. Each stack is composed of two layers with the same orientation, but opposite signs (for instance, ±45º) to guarantee balance. Still, considering the symmetry of the laminate, n corresponds to the total number of layers NL divided by four (n=NL/4). Graphite/epoxy material was used. Their properties are, E1=128 GPa, E2=11 GPa, G12=G13=4.48 GPa, v12=0.25, and p=1500 kg/m3. Figure 3 shows the laminated cylindrical shell. The dimensions of the cylindrical shell structure are: radius (R=0.1 m), length (L=0.2 m), and ply thickness (h=0.125 mm) Simply supported edges (SS3) as boundary conditions were adopted for the simulations. Table 1. Normalized fundamental frequency comparison. layup

ω (poore et al., 2008)

ω (Fe code)

[90°/0°]2

19.473

19.582

[±45°]S

28.206

29.000

[±60°]S

25.909

26.515

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.75-82, Jan.-Mar., 2013

Koide, R.M. and Luersen, M.A.

cylindricAl shell withoUt cUtoUt Firstly, the fundamental frequency was obtained considering all the 64 plies oriented with the same angle (0º, ±45º or 90º). The results are showed in Table 2 and the corresponding cylindrical geometry with the finite element mesh is presented in Fig. 4. In the finite element analysis, the laminated cylindrical shells are modeled by eight-node thick shell elements with six degrees of freedom per node using reduced integration. The optimization was carried out on a single processor computer with 3.0 GHz Intel Core 2 CPU including 4GB memory, and the total computational time needed was about 75 hours (150 evaluations, per hour). The optimized stacking sequence (θk), the corresponding fundamental frequency (fn), and the number of objective function evaluations (N) to reach the solution are presented in Table 3. Table 4 presents the comparison of the optimized frequency against the ones obtained by all plies with the same orientation (0°, ±45° or 90°). Despite the small gain of the optimized laminate compared to the one with all plies oriented at ±45°, the percentage gains against the 0° and 90° laminates were high, achieving 95.71 and 71.23%, respectively.

y z

Figure 5 presents the fundamental vibration modes of the cylindrical shells with the cutout for different stacking sequences. It can be noted that the modes shown are very similar to each other, except those for the stacking sequence [90°32]S . [032]s

[±4516]s

[9032]s

[904 4511 02 902 ±45]s (optimized stacking sequence)

Figure 5. Fundamental vibration mode for different stacking sequences (structure with the cutout). Table 2. Fundamental frequency for the structure without cutout.

Figure 3. Laminated cylindrical shell.

layup

Frequency (hz)

[032]S

1803.40

[±4516]S

3450.30

[9032]s

2061.20

Table 3. Optimized results for the structure without cutout.

3529.40 (Hz)

[904 4511 02 902 ±45]S

11250

Table 4. Comparison of frequencies (structure without cutout).

Figure 4. Finite element mesh of the cylindrical shell without cutout. J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.75-82, Jan.-Mar., 2013

layup

Frequency (hz)

difference (%)

[032]S

1803.40

95.71

[±4516]S

3450.30

2.29

[9032]S

2061.20

71.23

Maximization of Fundamental Frequency of Laminated Composite Cylindrical Shells by Ant Colony Algorithm

Table 5. Fundamental frequency for the structure with the cutout. layup

Frequency (hz)

[032]S

1777.20

[±4516]S

3311.90

[9032]S

1996.90

0.06 m

cylindricAl shell with A cUtoUt In this case, the cylindrical shell geometry is more complex than the precedent one. Figure 6 shows the cutout geometry, and Fig. 7 demonstrates the cylindrical shell with the cutout and the corresponding finite element mesh. Because we are interested in the vibration frequencies and not in obtaining the stress field or the maximum stress, no mesh refinement was done in the cutout region. Firstly, the fundamental frequency was obtained considering all the 64 plies oriented with the same angle (0°, ±45° or 90°). These results are showed in Table 5. The optimized results for the structure with the cutout are showed in Table 6. Table 7 presents the comparison of the optimized frequency against the ones obtained by all plies with the same orientation (0°, ±45° or 90°). Again, as in the previous case, despite the small gain of the optimized laminate compared to the one with all plies oriented at ±45°, the percentage increases against the 0° and 90° laminates with 90.67 and 69.69%, respectively. As the total number of evaluations used for this case was the same as the precedent one (n=10512) and the time for each finite element evaluation was also approximately the same, the total computational time remained similar to the previous case (about 70 h). Figure 8 presents the fundamental vibration modes of the cylindrical shells with the cutout for different stacking sequences. It can be noted that the modes are very similar to each other.

0.02 m

Figure 6. Cutout geometry.

Figure 7. Finite element mesh of the cylindrical shell with cutout.

[032]s

[±4516]s

[9032]s

Table 6. Optimized results for the structure with the cutout.

3388.60 (Hz)

[902 ±453 902 02 ±4510]s

[902 ±453 902 02 ±4510 ]s (optimized stacking sequence)

10512

Table 7. Comparison of frequencies (structure with the cutout). layup

Frequency (hz)

[032]S

1777.20

90.67

[±4516]S

3311.90

2.32

[9032]S

1996.90

69.69

Figure 8. Fundamental vibration modes for different stacking sequences (structure with the cutout). J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.75-82, Jan.-Mar., 2013

Koide, R.M. and Luersen, M.A.

CONCLUSIONS Cylindrical shells tailored with reinforced composite material were optimized. The fundamental frequency maximization was performed by an ACO algorithm, coupled with a finite element code to obtain the structural response. The ply angle orientations, i.e., stacking sequences, were obtained for structures without and with a cutout. The numerical results were presented and discussed in both cases.

We concluded that the ant colony metaheuristic can be efficiently applied to find optimal stacking sequences of cylindrical shells with cutout or complex geometries. As a further investigation, we can apply this method to multiobjective optimization, for example, maximizing the fundamental frequency and the buckling load. Another improvement of this approach would be to include the material as a design variable.

REFERENCES Amabili, M., 2011, “Nonlinear vibrations of laminated circular cylindrical shells: Comparison of different shell theories”, Composite Structures, Vol. 94, pp. 207-220.

Lam, K.Y. and Qian, W., 2000, “Free vibration of symmetric angleply thick laminated composite cylindrical shells”, Composites Part B, Vol. 31, pp. 345-354.

Barbero, E. J., Reddy, J N. and Teply, J.L., 1990, “General two-dimensional theory of laminated cylindrical shells”, AIAA Journal, Vol. 28, No. 3, pp. 544-553.

Le Riche, R. and Haftka, R., 1993, “Optimization of Laminate Stacking Sequence for Buckling Load Maximization by Genetic Algorithm”, AIAA Journal, Vol. 31, pp. 951-956.

Chern, Y. and Chao, C.C., 2000, “Comparison of natural frequencies of laminates by 3-D theory, Part II: curved panels”, Journal of Sound and Vibrations, Vol. 230, pp. 1009-1030.

Mota Soares, C.M., Franco Correia, V., Mateus, H. and Herskovits, J., 1995, “A discrete model for the optimal design of thin composite plateshell type structures using a two-level approach”, Composite Structures, Vol. 30, pp. 147-157.

Dorigo, M. and Stützle, T., 2004, “Ant colony optimization”, MIT, Cambridge, USA. Ganapathi, M. and Haboussi, M., 2003, “Free vibrations of thick laminated anisotropic non-circular cylindrical shells”, Composite Structures, Vol. 60, pp. 125-133. Ganapathi, M., Patel, B.P. and Patel, H.G., 2004, “Free flexural vibration behavior of laminated angle-ply elliptical cylindrical shells”, Computers and Structures, Vol. 82, pp. 509-518. Gharib, A. and Shakeri, M., 2010, “Stacking sequence optimization of laminated cylindrical shells for buckling and free vibration using genetic algorithm and neural networks”, Second International Conference on Engineering Optimization, Lisbon. Goupee, A. and Vel, S.S., 2006, “Optimization of natural frequencies of bidirectional functionally graded beams”, Structural Multidisciplinary Optimization, Vol. 32, pp. 473-484. Heyliger, P.R. and Jilani, A., 1993, “Free vibrations of laminated anisotropic cylindrical shells”, Journal of Engineering Mechanics, Vol. 119, pp. 1062-1077. Hu, H. and Tsai, J., 1999, “Maximization of the fundamental frequencies of laminated cylindrical shells with respect to fiber orientations”, Journal of Sound and Vibration, Vol. 225, No. 4, pp. 723-740. Hu, H. and Wang, S., 1992, “Optimization for buckling resistance of fiber-composite laminate shells with and without cutouts”, Composite Structures, Vol. 22, pp. 03-13. Irisarri, F.X., Laurin, F., Leroy, F.H. and Maire, J.F., 2011, “Computational strategy for multiobjective optimization of composite stiffened panels”, Composite Structures, Vol. 93, pp. 1158-1167.

Narita, Y. and Robinson, P., 2006, “Maximizing the fundamental frequency of laminated cylindrical panels using layerwise optimizations”, International Journal of Mechanical Sciences, Vol. 48, pp. 1516-1524. Poore, A.L., Barut, A.E. and Madenci, E., 2008, “Free vibration of laminated cylindrical shells with a circular cutout”, Journal of Sound and Vibration, Vol. 312, pp. 55-73. Rao, A. and Arvind, N., 2007, “Optimal stacking sequence design of laminate composite structures using tabu embedded simulated annealing”, Structural Engineering and Mechanics, Vol. 25, pp. 239-268. Rao, A.R.M. and Lakshmi, K., 2009, “Multi-objective optimal design of hybrid laminate composite structure using scatter search”, Journal of Composite Materials, Vol. 43, pp. 2157-2182. Shafei, E. and Kabir, M.Z., 2011, “Dynamic stability optimization of laminated composite plates under combined boundary loading”, Applied Composite Material, Vol. 18, pp. 539-557. Sharma, A.K. and Mittal, N.D., 2010, “Review on stress and vibration analysis of composite plates”, Journal of Applied Sciences, Vol. 10, No. 23, pp. 3156-3166. Soldatos, K.P., 1994, “Review of three dimensional dynamic analyses of circular cylinders and cylindrical shells”, American Society of Mechanical Engineers, Vol. 47, No. 10, pp. 501-516. Todoroki, A., Shinoda, T., Mizutani, Y. and Matsuzaki, R., 2001, “New surrogate model to predict fracture of laminated CFRP for structural optimization”, Journal of Computational Science and Technology, Vol. 5, No. 1, pp. 26-37.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.75-82, Jan.-Mar., 2013

doi: 10. 5028/jatm.v5i1.182

Architecture for ES Receiver Systems Targeted at Commercial Wireless Communications Warren Paul du Plessis1

Abstract: Modern electronic support (ES) systems are descended from systems intended for the detection of small numbers of high-power radar systems, and are thus not suitable for the low-power transmitters and dense signal environments typical of commercial communication networks. A new ES-system architecture is proposed to allow the detection of large numbers of low-power emitters and the estimation of their angle of arrival. The proposed architecture has a number of benefits including versatility, suitability for deployment on airborne platforms, and modularity. Keywords: Radio receivers, Electronic warfare, Surveillance, Radio communication, Wireless communication.

INTRODUCTION Commercial wireless communication systems, especially cellular networks, are increasingly being used by criminal, paramilitary and military operators. Examples of such use include rhino poaching (Beaudufe, 2012), guiding illegal immigrants (Lacey, 2011), and insurgent attacks (Strother, 2007). The use of commercial cellular phones to co-ordinate complex operations is motivated by the low cost and wide availability of reliable cellular communications. Even the United States’ military is evaluating the use of smartphones by its soldiers (Milian, 2012). The location and tracking of cellular systems is thus becoming increasingly important, as shown by the passage of legislation such as the Regulation of Interception of Communications and Provision of Communication-related Information Act (RICA) in South Africa (RSA, 2002). While the natural approach to achieve the detection and tracking of cellular phones would appear to be the use of the cellular network itself, this is generally not feasible. Firstly, a legal framework must exist to ensure that cellular network operators are required to provide the necessary information to security forces. However, privacy concerns in modern democracies mean that such a framework is difficult and time-consuming to establish — if it is possible at all. But even then, such a framework would only apply to network operators in one country, and many of the activities described above are perpetrated across the borders between nations. Nations are justifiably hesitant to grant other nations even limited access to information about and control over their industries (cellular network operators, in this case) and citizens, especially before criminal activities have been proved. It is thus unlikely that the required level of access to cellular

1.University of Pretoria – Pretoria/Gauteng – South Africa Author for correspondence: Warren Paul du Plessis | Department of Electrical, Electronic and Computer Engineering, University of Pretoria | Pretoria/Gauteng – South Africa | E-mail: wduplessis@ieee.org Received: 24/10/12 | Accepted: 21/12/12

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.83-90, Jan.-Mar., 2013

Plessis, W.P.

networks will be achieved within a reasonable time frame, if it can be achieved at all. This reality leads directly to a requirement for communications intelligence (COMINT) electronic support (ES) receivers which can detect and locate cellular phones. However, this simplesounding task is noticeably more challenging than it might appear. The success of cellular systems means that these systems have a large number of users, many of whom will be actively accessing the network at any given time. Separating individual users — let alone identifying criminal or paramilitary users — in such a dense signal environment is extremely challenging. Furthermore, cellular networks are designed to make sure that base transceiver stations (BTSs) and mobile devices transmit only the minimum power necessary to maintain a reliable connection. This means that all transmitters of interest will be operating at low power, further complicating their detection. Finally, cellular transmissions tend to be very short, and slow frequency hopping (in which BTS and mobile devices gradually change their operating frequency) is sometimes implemented to reduce the effects of fading. The probability of intercept (POI) of a receiver is thus reduced unless all frequencies of interest are monitored continuously. Furthermore, integration times are limited by the duration of the transmitted signal rather than by the receiver system requirements. Modern ES systems are descended from systems developed to detect and locate a relatively limited number of high-power radar transmitters. As described above, commercial cellular systems comply with neither of these assumptions, suggesting that traditional ES systems will not be effective in this role. There is a requirement for ES receivers specifically developed for COMINT of cellular communication systems. This paper describes the architecture of an ES system that is targeted at the detection and location of cellular phones. This architecture is based on the use of large numbers of relatively simple receiver elements.

available. For GSM systems operating in the E-GSM 900 and DCS 1800 bands, mobile devices are required to have a maximum transmit power of 33 dBm (2 W) and 30 dBm (1 W), respectively (3GPP, 2005). However, these specifications have a tolerance of ±2 dB under normal conditions and ±2.5 dB under extreme conditions (3GPP, 2005), so these values could be as low as 30.5 dBm (1.1 W) and 27.5 dBm (0.56 W) in the E-GSM 900 and DCS 1800 bands respectively, while still complying with relevant specifications. However, a far greater concern for a cellular network is the interference caused by mobile devices and BTSs which transmit more power than required for reliable communications. Modern cellular systems thus implement power control whereby the power transmitted by a device can be reduced to minimise interference. The GSM standard allows power to be reduced in 15 steps of 2 dB each, therefore enabling power reduction of 30 dB (3GPP, 2005). However, these values are subject to a tolerance of ±5 dB under normal conditions and ±6 dB under extreme conditions (3GPP, 2005), so power levels in the E-GSM 900 and DCS 1800 bands can be as low as -1 dBm (0.79 mW) and -6 dBm (0.25 mW), respectively. Dense Signal Environment In 2011, there were an estimated 5.6 billion mobile connections worldwide (Gartner, 2011) with a global population of 7 billion people (PRB, 2011). Africa had an estimated 649 million subscribers by the end of 2011, so roughly two out of three people in the continent have some sort of mobile connectivity (BBC News, 2011). This extremely large number of users of commercial cellular systems coupled with the limited bandwidth available for such systems (Lazarus, 2010) leads to very dense signal environments. Maximal use is made of the narrow bandwidths available for commercial communication systems by using a cellular approach whereby frequencies are reused at distances which are sufficient to minimise interference. Figure 1 shows how frequencies could

Challenges Associated with ES for Cellular Communications Challenges associated with ES for commercial cellular systems are considered below with the emphasis on the Global System for Mobile Communications (GSM) standard due to its widespread adoption. Low Signal Power The power transmitted by a mobile device is extremely low both as a result of device limitations and of power control. Mobile devices are small and are powered by batteries, limiting the power

Figure 1. Frequency reuse in an idealised cellular network.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.83-90, Jan.-Mar., 2013

Architecture for ES Receiver Systems Targeted at Commercial Wireless Communications

be reused in a highly idealised cellular structure, and while the reality will not be this neat, the same basic principles apply. Since ES systems will be expected to have ranges much greater than BTSs to minimise the number of systems required, it is clear that any ES system will detect a large number of signals which overlap in both time and frequency. To further complicate matters, newer systems like the Universal Mobile Telecommunications System (UMTS) and Long-Term Evolution (LTE), which are based on techniques such as carrier-division multiple access (CDMA) and orthogonal frequency division multiplexing (OFDM), are specifically designed to have users overlap in both time and frequency. The traditional means for deinterleaving radar signals are inadequate in such dense signal environments. Short Signals With Changing Frequencies Communication signals tend to be very short mainly as a result of the use of time-division multiple access (TDMA) to support multiple users or time-division duplex (TDD) to separate uplink from downlink. Furthermore, mobile devices only transmit when data are available both to reduce interference and to improve battery life. GSM uses a combination of frequency-division multiple access (FDMA) and TDMA to support multiple users (Eberspächer et al., 1999; Redl et al., 1995). FDMA is achieved by using a 200-kHz channel spacing over the available bands. TDMA is implemented by allowing each frequency channel to support eight logical channels with timeslots lasting 577 μs, and a frame of eight timeslots lasting 4.615 ms. However, a GSM burst is shorter than a full timeslot at 547 μs to ensure some robustness to timing differences caused by range. An ES system thus cannot use the long averaging typical of many ES systems. GSM implements slow frequency hopping whereby the channel frequency changes from burst to burst (Eberspächer et al., 1999; Redl et al., 1995). GSM also allows discontinuous transmission to reduce interference and improve battery life (Eberspächer et al., 1999; Redl et al., 1995). This approach can be particularly effective as normal speech has pauses accounting for approximately 50% of the total conversation. Together, these two characteristics mean that the POI of a GSM signal can be extremely low. Scenario Geometry The relative positioning of a mobile phone, BTS and ES receiver has a major effect on the ES receiver requirements. Two important scenarios are shown in Fig. 2.

Mobile 1 rr1 ES receiver

BTS rr2

rb1 rb2 Mobile 2

Figure 2. Scenario geometry showing two special cases.

Mobile 1 in Fig. 2 is extremely close to the BTS (rb1 is small) and uplink power control will ensure that this device transmits the minimum allowable power. The ES receiver is thus tasked with detecting the mobile’s extremely weak signal at long range (rr1). This type of scenario can arise both in urban and rural environments. Cells tend to be small in urban environments to accommodate high user densities (Eberspächer, J. et al., 1999), so mobile devices will always be near a BTS. In a rural environment, a spotter on a hilltop could be underneath a BTS positioned on the same hilltop. Mobile 2 in Fig. 2 is at the extreme edge of the BTS’s coverage area (rb2 is at its largest value). The uplink power control problem associated with mobile 1 is thus avoided, but the range from the mobile to the receiver (rr2) is extremely large. The ES receiver is thus required to detect the mobile at extremely long range, and importantly, the ES receiver is required to detect the mobile at a greater range than the BTS (rr2>rb2). This scenario will arise most frequently in rural environments, where cell sizes are maximised due to low user densities (Redl et al., 1995). A further more subtle problem related to scenario geometry occurs when two mobile devices operating at the same frequency at the same time are positioned at substantially different ranges to the ES receiver. In this case, the signal from the more distant mobile device tends to be masked by the signal from the nearer mobile device. Large Unused Frequency Bands GSM channels have a bandwidth of less than 271 kHz and have a mere 200 kHz spacing from 880 to 915 MHz and 925 to 960 MHz for the uplink and downlink of the E-GSM 900 band, respectively, and from 1710 to 1785 MHz and 1805 to 1880 MHz for the DCS 1800 band uplink and downlink, respectively (3GPP, 2005). Even the wideband code-division multiple access (W-CDMA) technology used by the universal mobile telecommunications system (UMTS) — the successor to GSM — only uses a bandwidth of 5 MHz (3GPP, 2012). However, even these figures do not show

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.83-90, Jan.-Mar., 2013

Plessis, W.P.

170

the whole picture because not all allocated frequencies will be used within a specific area. Cellular communications are thus characterised by the use of relatively small frequency bands separated by larger frequency bands, which contain no commercial cellular signals. Development of an ES Architecture for Cellular Communications Starting from the challenges previously described, an ES system architecture targeted at commercial communications systems is developed. Implications of Challenges The first major challenge associated with detecting commercial communication systems is the very low signal levels at a COMINT receiver. For a fixed signal level at a receiver, the detection of a signal is determined by the parameters listed below: • The receiver noise figure (NF). This parameter is largely dependent on the NF of the first receiver amplifier and the losses before this amplifier (Gonzalez, 1997). Reducing these values will decrease the noise floor of the receiver. • The antenna gain. Higher antenna gain means that a greater signal is received at the antenna output for a specific field strength at the receiver. However, high antenna gain inherently entails a narrow beamwidth, which will reduce the POI of short-duration signals due to the need to scan the antenna beam. • The signal-to-noise ratio (SNR) required for detection. This SNR value will vary depending on the required detection probability and the maximum allowable false-alarm rate. For example, a detection probability of 99% with a false alarm probability of 10-6 will require a higher SNR than a detection probability of 90% with a false alarm rate of 10-3. Simple detection algorithms will require higher SNR, while more advanced algorithms can achieve acceptable detection and false-alarm rates with SNR values below 0 dB (i.e., the signal is weaker than the receiver noise). However, these advanced algorithms are computationally expensive and require powerful signalprocessing hardware. • The path loss from the transmitter to the receiver. While most of the factors that determine the path loss — such as mobile height, operating frequency and environment — are fixed, the height of the receiver can be controlled. Figure 3 shows

Path loss (dB)

160 150 140 130 120 110 100

500

1000 1500 2000 Receiver height (m)

2500

Figure 3. Path loss as a function of receiver height when the mobile height is 1.5 m in an open rural area at a frequency of 900 MHz at a range of 50 km.

how the path loss reduces as the receiver height increases for a transmitter at a height of 1.5 m in an open rural area at a frequency of 900 MHz at a range of 50 km using the HataDavidson propagation model (TIA, 1997). The second significant challenge associated with commercial communications is the large number of simultaneous signals which will be intercepted by an ES receiver. The received signals need to be separated in some way in order to allow individual signals to be detected. As mentioned before, time and frequency are insufficient to isolate received signals, so another parameter (e.g., position or angle) is required to separate commercial communication signals. Furthermore, the position of a transmitter, or at least the angle to the transmitter, is useful information in its own right. However, the number of signals that can be independently located by many algorithms — e.g., MUSIC (Schmidt, 1986) and ESPRIT (Roy and Kailath, 1989) — is limited by the number of independent receiver channels. Related to both these challenges is the effect of stronger signals masking simultaneously-received weaker signals. The dynamic range of a receiver is one of the main factors contributing to the possibility of simultaneously detecting both strong and weak signals. Proposed ES System Architecture The proposed ES architecture is described next and considers the points highlighted previously. However, the proposed architecture also has a number of other advantages. The underlying concept of the proposed system is to integrate the antenna and the complete receiver front end into single unit, and is summarised in Fig. 4. A large number of

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.83-90, Jan.-Mar., 2013

Architecture for ES Receiver Systems Targeted at Commercial Wireless Communications

Antenna

RF front end ADC

Receiver elements

Channel processing

System processing and control Figure 4. Proposed ES system block diagram.

such integrated receiver systems will then be combined into an ES system to achieve good system-level performance. While this system resembles a channelized receiver (Adamy, 2001; Adamy, 2009), it differs in two important respects. Firstly, each receiver element contains a complete receiver which can be controlled independently. For example, this means that the narrowband receiver elements can be dynamically assigned to known frequencies of interest, while ignoring the bands between these frequencies — an arrangement that would be impossible with a channelized receiver in which the frequency of each channel is fixed in relation to the other channels. Secondly, the signals received by multiple receiver elements operating at the same frequency can be processed coherently in the proposed system. Coherent processing allows the achievement of antenna gain and that accurate phase-based direction finding (DF) can be performed. While the majority of spectrum-sensing algorithms are based on the use of a single sensor (e.g. Yücek et al., 2009), the coherent use of multiple sensors has been shown to be feasible (e.g. Haykin et al., 2009). A system based on the proposed architecture in Fig. 4 will have a number of benefits, including the following. • Large numbers of elements mean that large antenna arrays with high antenna gain can be constructed. The fact that the signal at each receiver is sampled means that multiple narrow, high-gain antenna beams can be formed simultaneously through signal processing, thereby overcoming the usual trade-off between high antenna gain and wide-area coverage. • The narrow, high-gain antenna beams which can be formed also have the benefit of reducing the number of signals that need to be considered simultaneously by restricting the received signals to a smaller angular sector.

• The large number of elements also offers the potential to

•

• • •

obtain more independent samples of the environment, thereby allowing better deinterleaving of signals. The integration of the antenna and the radio-frequency (RF) front end will help to reduce the receiver noise figure by allowing these two elements to be designed together for optimal performance and by reducing the loss between them. While cost considerations will inevitably limit each receiver to the use of analogue-to-digital converters (ADCs) with lower sampling rates, it offers the opportunity to use devices with higher dynamic range (more bits), therefore improving the system dynamic range. Each receiver element produces a digital output, removing the need for low-loss, phase-matched RF cables. A powersupply line, a low-frequency local oscillator signal, digital control lines and a digital output are all that each receiver element requires to function. The use of a large number of low data-rate streams of data is ideally suited to modern, highly parallel signalprocessing technologies. The concept is inherently scalable depending on the number of receiver elements used to construct a system. The large number of elements makes the system extremely versatile as outlined below. The large number of elements also leads to redundancy, which will improve system reliability. Furthermore, the modular nature of the system will simplify repairs, improving the system’s availability.

This architecture is a natural match to modern digital signal processing (DSP) hardware technologies (including

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.83-90, Jan.-Mar., 2013

Plessis, W.P.

field-programmable gate arrays – FPGAs – and graphics processing units – GPUs). DSP devices are increasingly achieving high performance through the use of large numbers of relatively low-performance processors operating in parallel. Such parallel processing is ideally suited to the proposed architecture, in which large numbers of relatively low-rate data streams are generated. It might even be possible to integrate a low-cost DSP device into each receiver element to perform channel-specific processing like calibration. However, the greatest benefit of the natural match between the proposed architecture and modern DSP hardware is that it will be possible to better exploit the full processing power of existing DSP technologies. This creates the opportunity to allow the development and implementation of complex detection algorithms, which will lower the SNR required for a mobile device to be detected. The fact that the system produces digital signals makes it a good match for deployment on an aerostat. Aerostats are tethered lighter-than air (LTA) systems which have a number of advantages for persistent wide-area surveillance (TCOM, 2012; Raven, 2012), and here allow the receiver height requirement previously expressed to be addressed. It will be possible to transfer the digital signals produced by the proposed system to the ground for further processing using a single digital fibre. This allows powerful DSP technologies to be used for processing while maintaining a low-weight airborne system and tether. It is unlikely that a single aerostat system would be sufficient for large-scale surveillance, so it is possible that unmanned aerial vehicles (UAVs), manned aircraft and ground-based systems will also be required. The modular nature of the architecture shown in Fig. 4 means that the same basic building blocks can be used for all these systems. For example, an aerostat would have many airborne elements with extensive ground-based signal processing, while UAV might have only a handful of elements with simpler processing, but both systems would use the same basic elements and DSP technologies. Such system

frequency

(a) Antenna gain.

reconfigurability allows the development of a family of systems which enable the unique requirements of each platform to be accommodated. Over and above the benefits already highlighted, the versatility of the proposed architecture is one of its main attractions. A number of examples of this versatility are shown in Fig. 5 along with comparisons to a wideband ES system. For example, an ES system with an instantaneous bandwidth of 800 MHz has been developed based on the digital RF memory (DRFM) technology described by Olivier et al. (2011). Figure 5(a) shows how high antenna gain can be achieved by allocating all receiver elements to the same frequency and performing coherent processing. The gain is determined mainly by the number of receiver elements (Lo, 1964), so a system comprising a smaller number of wideband receiver elements is unable to achieve comparable antenna gain. Figure 5(b) shows how a wide range of frequencies can be covered by a number of independent receivers. The key point is that commercial communication systems are allocated to relatively narrow bands, which are separated by wide frequency ranges. Having the ability to monitor a large number of narrowband channels can help improve the system POI by allowing all channels in use to be simultaneously monitored. Attempting to use wideband receivers to cover commercial frequency bands is inefficient as the majority of covered frequencies have no signals of interest. Figure 5(c) shows how a combination of these two approaches is also possible whereby high antenna gain can be achieved at certain frequencies while still allowing other frequencies to be monitored. A wideband system is simply too restricted to achieve similar performance. Cost Estimate They key to the success of such a system is that the cost of each receiver element should be as low as possible to ensure

frequency

(b) Frequency coverage.

frequency

Figure 5. Demonstration of the versatility of the proposed ES system architecture. J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.83-90, Jan.-Mar., 2013

Wideband receiver Narrowband receiver Known frequencies

Architecture for ES Receiver Systems Targeted at Commercial Wireless Communications

that systems can consist of large numbers of receiver elements. Cost estimates for a single receiver element is given in Table 1, in which the low cost of each receiver element is ensured in the following ways: • The system is limited to operation in the range of frequencies most desirable for mobile communications (below 3.5 GHz (Lazarus, 2010)). Current microwave monolithic integrated circuit (MMIC) technology means that the cost of system components in this range of frequencies is low. • The bandwidth of each ADC is relatively low, allowing cheaper devices to be used. • Mass production techniques including the automated assembly of receiver elements will further reduce the element cost. It should also be noted that the cost estimate in Table 1 is conservative for the following reasons: • It might be possible to use a cheaper substrate. • Further cost reductions might be possible if all the RF components could be integrated into a single chip. • It might be possible to use two mixing stages instead of three. • The use of etched filters rather than separate components could be viable. • The synthesisers specified include mixers, thereby removing the need for separate mixers. Table 1.Estimated cost of an individual receiver element. Component

Substrate

Type

Rogers RO4003C substrate, 64 mil, 18"x24" Low-noise amplifier Minicircuits PSA-5453+ (LNA) and RFMD SGC4563Z Limiter Minicircuits RLM-33+ Variable attenuator Minicircuits DAT-15R5-SP+ Synthesiser 3x RFMD RFFC5072 Mixers 3x Minicircuits LAVI-362VH+ Filters Minicircuits HFTC16+, HFCN-740D+, RHP-180+ ADC Analog Devices AD9446-100 Additional components Capacitors, resistors, regulators, etc. Etching and assembly Total

Cost (USD)

200 5 10 5 30 75 20 70 125 625 1165

The goal of realising low-cost receiver elements appears to be achievable. For example, Ettus Research manufactures a number of Software-Defined Radio (SDR) systems out of which the most expensive is the USRP N210, which sells for USD 2195 when combined with an Ettus WBX RF daughterboard and an antenna. This system comprises a 100 MS/s ADC, a 400 MS/s digital-to-analogue converter (DAC), a 50 MHz to 2.2 GHz RF front end including a receiver and a transmitter, and an FPGA capable of 32 billion multiply-accumulate (MAC) operations per second. This Ettus system is far more capable, and thus expensive, than the receiver elements proposed here, as it contains a transmitter and a receiver.

CONCLUSION Commercial wireless communication systems are becoming increasingly important due to their adoption by criminal, paramilitary and even military users. The use of information gleaned from cellular network operators to monitor and track mobile devices faces a host of legal and political challenges which are unlikely to be overcome in the near term, if ever. There is thus a requirement for ES systems designed to perform COMINT for commercial communication systems. Commercial wireless communications present major challenges to ES systems due to the low power transmitted and extremely dense signal environments. Furthermore, short transmission times, slow frequency hopping and discontinuous transmission can lead to an extremely low POI. Finally, isolating a single user among the millions of users of commercial communication services is a daunting task. A new system architecture which overcomes these difficulties is proposed. This architecture is based on the use of large numbers of simple, low-cost receiver elements to achieve high system performance. This approach has the potential to achieve high antenna gain, low receiver noise figure and is well-matched to modern signal-processing technologies allowing computationally-expensive algorithms to be implemented. The fact that digital signals are generated at each receiver means that the proposed architecture is well-matched to aerostat-based deployment. The modular

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.83-90, Jan.-Mar., 2013

Plessis, W.P.

nature of the proposed system enables a family of similar systems to be developed from the same basic building blocks. Finally, this new architecture is extremely versatile, allowing combinations of high antenna gain and wide spectral coverage to be achieved as required.

ACKNOWLEDGEMENTS The author wishes to express his sincere thanks to the anonymous reviewers for their many comments and suggestions which greatly improved this paper.

REFERENCES 3rd Generation Partnership Project (3GPP), 2005, “Technical specification group GSM/EDGE radio access network; Radio transmission and reception (Release 1999)”, Std. 05.05, Rev. 8.20.0. 3rd Generation Partnership Project (3GPP), 2012, “Technical specification group radio access network; User Equipment (UE) radio transmission and reception (FDD) (Release 11)”, Std. 25.101, Rev. 11.3.0. Adamy, D.L., 2001, “EW 101: A first course in electronic warfare”, Chapter 4, Artech House. Adamy, D.L., 2009,” EW 103: Tactical battlefield communications electronic warfare”, Chapter 4, Artech House. BBC News, 2011, “Africa’s mobile phone industry ‘booming’” [Internet]. Available from: http://www.bbc.co.uk/news/world-africa-15659983. Accessed October 24, 2012. Beaudufe, C., 2012, “South African success story under threat” [Internet]. Available from: http://www.sowetanlive.co.za/news/2012/05/07/ south-african-success-story-under-threat. Accessed October 24, 2012. Eberspächer, J., Vögel, H.J. and Bettstetter, C., 1999, “GSM switching, services and protocols”, 2nd ed., John Wiley & Sons, Ltd. Ettus, 2012, “Ettus Research” [Internet]. Available from: http://ettus. com/. Accessed October 24, 2012. Gartner, 2011, “Gartner says worldwide mobile connections will reach 5.6 billion in 2011 as mobile data services revenue totals $314.7 billion” [Internet]. Available from: http://www.gartner.com/it/page. jsp?id=1759714. Accessed October 24, 2012. Gonzalez, G., 1997, “Microwave transistor amplifiers”, Chapter 4, 2nd ed. Prentice Hall. Lacey, M., 2011, “Smugglers guide illegal immigrants with cues via cellphone” [Internet]. Available from: www.nytimes.com/2011/05/09/ us/09coyotes.html. Accessed October 24, 2012. Lazarus, M., 2010, “The great radio spectrum famine”, IEEE Spectrum, Vol. 40, No. 10, pp. 26-31. Lo, Y.T., 1964, “A mathematical theory of antenna arrays with randomly spaced elements”, IEEE Transactions on Antennas and Propagation, Vol. 12, No. 3, pp. 257-268. Milian, M., 2012, “U.S. government, military to get secure Android phones”, Cable News Network (CNN)” [Internet]. Available from:

http://edition.cnn.com/2012/02/03/tech/mobile/governmentandroid-phones/index.html. Accessed October 24, 2012. Olivier, K., Cilliers, J.E. and du Plessis, M., 2011, “Design and performance of wideband DRFM for radar test and evaluation”, Electronics Letters, Vol. 47, No. 14, pp. 824-825. Population Reference Bureau (PRB), 2011, “2011 World Population Data Sheet” [Internet]. Available from: http://www.prb.org/Publications/ Datasheets/2011/world-population-data-sheet.aspx. Accessed October 24, 2012. Raven Aerostar, 2012, “TIF-25K | Tethered aerostats | Raven Aerostar” [Internet]. Available from: http://ravenaerostar.com/products/ aerostats/tif-25k. Accessed October 24, 2012. Redl, S.M., Weber, M.K. and Oliphant, M.K., 1995, “An introduction to GSM”, Artech House. Republic of South Africa (RSA), 2002, “Regulation of Interception of Communications and Provision of Communication-Related Information Act”, Government Gazette, Vol. 451, No. 24286, pp. 2-96, Act 70 of 2002. Roy, R. and Kailath, T., 1989, “ESPRIT – Estimation of signal parameters via rotational invariance techniques”, IEEE Transactions on Acoustics, Speech and Signal Processing, Vol. 37, No. 7, pp. 984-995. Schmidt, R.O., 1986, “Multiple emitter location and signal parameter estimation”, IEEE Transactions on Antennas and Propagation, Vol. 34, No. 3, pp. 276-280. Strother, T., 2007, “Cell phone use by insurgents in Iraq”, Urban Warfare Analysis Center [Internet]. Available from: http://www. babylonscover twar.com/Terrorist%20Groups/Weapon%20 Systems/Cell%20Phone%20Use%20by%20Insurgents%20in%20 Iraq.pdf. Accessed October 24, 2012. TCOM, 2012, “TCOM aerostats – over 30 years of reliable performance” [Internet]. Available from: http://www.tcomlp.com/aerostats.html. Accessed October 24, 2012. Telecommunications Industry Association (TIA), 1997, “A report on technology independent methodology for the modeling, simulation and empirical verification of wireless communications system performance in noise and interference limited systems operating on frequencies between 30 and 150 MHz”, version 20. Yücek, T. and Arslan, H., 2009, “A survey of spectrum sensing algorithms for cognitive radio applications”, IEEE Communications Surveys & Tutorials, Vol. 11, No. 1, pp. 116-130.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.83-90, Jan.-Mar., 2013

doi: 10.5028/jatm.v5i1.191

An Asynchronous Interface with Robust Control for Globally-Asynchronous LocallySynchronous Systems Duarte Lopes de Oliveira1, Eduardo Lussari1, Sandro Shoiti Sato2, Lester de Abreu Faria1

Abstract: Contemporary digital systems must necessarily be based on the “System-on-Chip” (SoC) concept. Especially in relation to the aerospace industry, these systems must overcome some additional engineering challenges concerning reliability, safety and low power. An interesting style for aerospace SoC design is the GALS (Globally Asynchronous, Locally Synchronous) paradigm, which can be used for Very Large Scale Integration – Deep-Sub-Micron (VLSI_DSM) design. Currently, the major drawback in the design of a GALS system is the asynchronous interface (asynchronous wrapper – AW) when being implemented in VLSI_DSM. There is a typical AW design style based on asynchronous controllers that provides communication between modules (called ports), but the port controllers are generally subjected to essential hazard, what decreases the reliability and safety of the full system. Concerning to this main drawback, this paper proposes an AW with robust port controller that shows to be free of essential hazard, besides allowing full autonomy for the locally synchronous modules, creating fault tolerant systems as much as possible. It follows the Delay Insensitive (DI) model interacting with the environment in the Generalized Fundamental Mode (GFM) without the need to insert any delay elements. Additional delay elements, although proposed by some previous work found in literature, are not desirable in aerospace applications. The proposed interface allows working on Ib/Ob mode, showing the DI model is more robust than the QDI model and, therefore, it does not need to meet isochronic fork requirements nor timing analysis. Once an interface presenting similar properties was not found in literature, the proposed architecture proved to have great potential of implementation in practical VLSI_DSM designs, including the aerospace ones, once it overcomes the main engineering challenges of this kind of industry. Keywords: Aerospace systems, Reliability, Low power, Asynchronous controllers, GALS.

Introduction Contemporary digital systems are usually implemented on Very Large Scale Integration (VLSI) and must necessarily be based on the “System-on-Chip” (SoC) concept. The reason for that is to satisfy the ever-growing demand for higher performance, reusability and low-power requirements (De Micheli, 2009; Muller-Glaser et al., 2004). Especially in relation to the aerospace industry, these systems must overcome some additional engineering challenges concerning reliability, safety, high complexity and the unavailability of component failure data, generating fault tolerant systems as much as possible (Sues, et al., 2005; Bertuccelli, 2008). SoC circuits are composed of functional modules, which can be the intellectual property cores (IP-cores) from many different vendors. These IP-cores are pre-designed, verified, tested and optimized for high-performance, providing both cost and development time reduction. Once SoC circuits are implemented in deepsub-micron (DSM) technologies (VLSI_DSM) (for example, 70 nm, 500M transistors for chip and f=2,5 GHz), delays caused by wires prove to be big when compared to the gate timing, and the difference between minimal and maximum delays in the gates is significant (Jain et al., 2001; Martin et al., 2006). Therefore, when SoC circuits are implemented using only a global clock signal, they are subjected to speed and power penalties (clock skew, distribution networks etc.), thus making timing analysis very complex (Friedman, 2001). Besides that, the harsh environment found in aerospace applications, with high temperature variations, can make this time analysis even more difficult.

1.Instituto Tecnológico de Aeronáutica – São José dos Campos/SP – Brazil 2.ETE Ferraz de Vasconcelos – São Paulo/SP – Brazil Author for correspondence: Lester A. Faria | Praça Marechal Eduardo Gomes, 50 – Vila das Acácias | CEP 12228-901 São José dos Campos/SP – Brazil | E-mail: lesteraf@gmail.com Received: 14/11/12 | Accepted: 18/01/13

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.91-102, Jan.-Mar., 2013

Oliveira, D.L., Lussari, E., Sato, S.S. and Faria, L.A.

Asynchronous project methodologies (Martin et al., 2006; Myers, 2004) can naturally eliminate such challenges by removing the clock signal from the design. Different classes of asynchronous circuits may be used to implement SoCs, which can be built from completely asynchronous modules, but these kinds of circuits are not a widely accepted solution. The main reasons for that refusal are: a) lack of reliable tools for asynchronous design; b) difficulties from hazard-free designing and testing; c) limited culture on asynchronous design; and d) lack of asynchronous IPs (Hardt et al., 2000). The aerospace industry imposes many additional challenges to the design of dedicated systems, such as the high complexity of systems; main power generation systems; missions’ profiles and environment; high demand for new technologies; high reliability and safety requirements; unavailability of component failure data; component sizes; and especially tight schedules, what leaves no room for errors. Any problem in an aerospace system leads to big losses of aircrafts (or spacecraft), crews, missions and revenues. In this context, reliability and robustness are important, leading to lower maintenance cost and lower failure frequency. The objective is always to maximize system performance, while satisfying constraints that ensure a reliable operation (Sues et al., 2005; Bertuccelli, 2008). Concerning to this special situation and the features of both synchronous and asynchronous systems, intermediate solutions were proposed between “totally synchronous” and “totally asynchronous”, such as the Globally Asynchronous, Locally Synchronous methodology (GALS). The term GALS was first used by Chapiro (1984), in his PhD thesis. A GALS system consists of many synchronous functional modules that communicate in the asynchronous form. In this paper, we refer to the GALS systems as digital systems partitioned in functional modules (that may be IPs), which carry their own individual clock signals and are unrelated between modules. An asynchronous communication scheme is provided for the communication between different modules with different clock domains. In order to handle the asynchronous communication between these modules, an interface circuit has to be added around each one of the synchronous modules, which is called an asynchronous wrapper (AW). The AW term was first used by Bormann et al. (1997). This local interface may be built by using local clocks, FIFOs, asynchronous controllers (Input Ports, Output Ports) etc. Techan et al. (2007)

Data _in Input port controller

Asynchronous Wrapper Locally Synchronous Module

Data_out Output port controller

Local clock generator

Figure 1. Asynchronous wrapper.

show different styles for asynchronous interfaces dedicated to GALS systems. Figure 1 shows a generic interface with a synchronous module as an example. GALS systems have been successfully used in many implementations, including the Application Specific Integrated Circuit (ASIC) (Gurkaynak et al., 2006; Amini et al., 2006; Miller et al., 2005) and Field Programmable Gate Array (FPGA) (Jia et al., 2005; Kumala et al., 2006; Yuan et al., 2005). Currently, FPGA devices have shown to be a common choice for implementing digital circuits (Muller-Glaser, 2004), growing considerably in recent years. High-performance FPGAs, with up to 50 million gates, can be easily found nowadays, therefore allowing complex digital systems, such as GALS, to be programmed on them (De Micheli, 2009) and to be implemented in CMOS technology, DSM. Asynchronous interfaces that use communication ports are of main interest, once they allow removing the asynchronous handshake scheme from the synchronous modules, allowing the synchronous module to be developed using standard techniques of synchronous design. Although the GALS methodology has solved problems related to the global clock signal, the communication between modules is already performed in the asynchronous paradigm, therefore being subjected to all its inherent problems. Implementations of ports: different approachs Different kinds of ports have been synthesized in the logic synthesis style (Myers, 2004). As an example, the ports proposed by Amini et al. (2006) have been specified in Signal Transition Graph (STG), which is a Petri-net-like speficification (Chu, 1987), being synthesized

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.91-102, Jan.-Mar., 2013

An Asynchronous Interface with Robust Control for Globally-Asynchronous Locally-Synchronous Systems

in the Petrify tool (Cortadella et al., 1997). These ports must meet the isochronic fork requirement (Myers, 2004), but the realization of this requirement in VLSI_DSM presents a high level of difficulty. Furthermore, the STG specification, as well as its synthesis method, is not familiar to synchronous paradigm designers. The ports proposed by Muttersbach et al. (2000), Muttersbach (2001), Reddy Ravi (2001) and Pontes et al. (2007) were specified in Extended Burst-Mode (XBM) and Burst Mode (BM). These ports were implemented, respectively, in 3D (Yun et al., 1999) and minimalist (Fuhrer et al., 1999) tools. They interact with the environment in the generalized fundamental mode (GFM), requiring a timing analysis and being subjected to essential hazard, especially in the DSM technology. Concerning to this last drawback, the insertion of delay elements may be a possible solution (VLSI_DSM), but it degrades the testability and cycle-time of the system. The insertion of delay elements is not adequate when implementing GALS in FPGA as well, because these devices (FPGAs) are not designed to favor the insertion of delay elements. Avoiding essential hazard in ports controllers: increasing the system’s reliability The XBM specification is quite interesting when describing port controllers, once it is not only “familiar” to synchronous paradigm designers, but also because the method that synthesizes ports described by XBM shows to be simpler when compared to the synthesis by STG (Myers, 2004). Yun et al. (1999) and Nowick (1993) proposed the insertion of delay elements on the feedback wires in order to avoid essential hazard in burst-mode controllers. Oliveira et al. (2008) proposed a sufficient condition that guarantees essential hazard-free operation on burst-mode controller without the need for extra delay elements, when mapped on VLSI_DSM or any type of LUT-based FPGA. The absence of delay elements is highly desirable when considering FPGA devices (difficulties in implementing this kind of elements) and, furthermore, in aerospace applications, in which the harsh environment must change the behavior of electronic components. This paper proposes robust port controllers for asynchronous interfaces used in GALS style. They are completely free of essential hazard and are described

in the XBM specification. The robust controller design for asynchronous interfaces is proposed as a solution to the increasing demand for high reliablility aerospace electronic systems. The paper also shows that the method proposed by Oliveira et al. (2011) to synthesize BM controllers free of essential hazard is improved for XBM controllers. These proposed ports are implemented in the following architectures: “Huffman machine with feedback output” and “standard RS”. The use of both architectures enables a better performance of the system, besides being more reliable and providing safer operation for aerospace applications. A new AW for GALS with robust ports is also proposed. Once it is known that a major drawback in the design of asynchronous wrapper is the synthesis of these ports, the proposed AW proved to be very important and robust. These ports are easily implemented both in VLSI_DSM and LUT-based FPGA. Other advantages of this wrapper are: 1) total autonomy to the locally synchronous modules, when interacting with the proposed AW; and 2) its ports interact with the environment in the mode I b/O b, thus not requiring timing analysis and being more robust than the GFM mode. In this mode, a new input burst is immediately accepted when all signals of output burst change their values. All of these achieved features make the proposed architecture a good option for aerospace implementations, once it increases the reliability of the full system, overcoming some of the main challenges in this kind of industry. Different styles of GALS design Once the synchronous modules of a GALS system operate at different frequencies and/or different phases, the communication between them is subjected to metastability (Ginosar, 2003). Metastability occurs when a specific signal violates the setup time or the hold time of the memory element, and during any time the output voltage assumes an intermediate value that leads the circuit to achieve a random logic value. Metastability may occur in a timing window defined by the sum of “setup” and “hold” times. So, the GALS design style is determined according to the treatment of metastability, since there different ones in literature. Techan et al. (2007) propose specific taxonomy to classify these styles, in which they basically can be classified into three main styles: a) weak synchronous interface; b) pausible clock interface; and c) asynchronous interface.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.91-102, Jan.-Mar., 2013

Oliveira, D.L., Lussari, E., Sato, S.S. and Faria, L.A.

Weak synchronous interface This style has three variants: a) heterochronous; b) mesochronous; and c) plesiochronous. In the heterochronous style (footer), the clocks of the synchronous modules run on different nominal frequencies (Techan et al., 2007). On the other hand, in the mesochronous style (from Greek, meso means average), the clocks show the same average frequency, but with different unknown phases, which are generated by the same oscillator (Techan et al., 2007). Finally, in the plesiochronous style (from Greek, plesio means “almost equal”), the clocks operate with equal nominal frequency, but being generated by different oscillators (Techan et al., 2007). These styles always require timing analysis, starting from the knowledge of the clocks and using FIFO as a base, phase adjusters and, sometimes, synchronizers. The advantage of these styles is to enable low latency and high frequency clocks. On the other hand, there is the need for a rigorous timing analysis. Figure 2 shows a mesochronous interface that uses a phase adjuster (timing recovery circuit – TRC).

Common Clock Reference Delay_wire2

Delay_wire1

Clock’

Clock Sending Module

Data Delay_ data

TRC

Data_sync

Receiving Module

Figure 2. Mesochronous with TRC.

Data_out Locally S_Req A_Req Synchronous Output Module

S_Ack

Port

Req_S

Asynchronous interface This style uses circuits known as synchronizers and handshaking signals. The synchronous modules have clocks running freely at different frequencies, without any prior knowledge about their timing. Data are synchronized from one clock domain to another. Some examples of data synchronizers are the well known “two registers”, or “double latches” (Mullins et al., 2007), or some other more elaborated synchronization schemes, such as the “synchronization pipeline” (Sjogren et al., 2000) and “FIFOs” (Dobkin et al., 2006). The proposed synchronizers do not totally eliminate failure due to metastability, once the probability of failure different of zero percent remains (Dobkin et al., 2006). The “two register” synchronizer presents as advantages its simplicity and robustness, but as a disadvantage there is an increasing area, power, and especially a high penalty in latency times, which leads to an increase of two clock cycles. Figure 4 shows the architecture of asynchronous interface as an example.

A_Ack

Ack_S

Clock

Pausible clock interface This style, firstly proposed by Chapiro (1984), tackles the problem of metastability by interrupting the clock signal. When data are ready for transmission, the clock is interrupted, enabling data synchronization. The synchronous modules have pausible clock signals. Most often, these clocks are locally generated using a ring oscillator and a mutualexclusion circuit, or arbitrator, which properly generates the pause and restart of the clock (Yun et al., 1999). The potential advantages of this style are the robustness in the treatment of metastability and power reduction. On the other hand, the weakness of this style is the possibility of “deadlock” and “jitter” (Mullins et al., 2007). Different architectures have been proposed for pausible clocks, for example, the one involving FIFO (Techan et al., 2007). Figure 3 shows an architecture involving pausible clock as an example.

1 Mux 0

Empty

Full

Delay_1

Locally Synchronous Module 1

Delay_2

Delay_3 Figure 3. Pausible clock for FPGA described by Techan et al. (2007).

Data Wr_en

Wd_clk

Rd_valid FIFO

Data Rd_en

Locally Synchronous Module 2

Rd_clk

Clock1

Figure 4. Asynchronous interface based on FIFO.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.91-102, Jan.-Mar., 2013

Clock2

An Asynchronous Interface with Robust Control for Globally-Asynchronous Locally-Synchronous Systems

Communication controllers (ports) GALS systems require asynchronous communication links, which can require two kinds of communication protocols: two or four stages handshaking. The ports can work as active (generating the “request” signal) or passive (generating the “acknowledge” signal). In GALS design there are two types of communication controllers: a) port of “demand”, b) port of “poll” (inquiry). In the port of demand, the data being transferred are immediately required after the communication. Therefore, in this type of controller the clock must be immediately stopped (paused) and reactivated (restarted) when communication is done. In the port of poll, the clock is not stopped immediately. It defines when it is “safe” to send the data. The clock is stopped (paused) only in cases when there is the need for additional time, in order to troubleshoot eventual metastability. XBM-EHF Specification: Condition BM is a kind of specification based on a state transition graph which was first proposed by Davis et al., (1979), later formalized by Nowick (1993), and improved by Yun et al. (1999) as XBM. It allows multiple input changes and is usually used to describe Mealy Asynchronous Finite State Machines (FSM). These machines interact with the environment in GFM. In GFM, a new input burst can only occur if the controller is stable (with no activity in the ports or in the lines). The XBM specification supports the BM specification, introducing two kinds of input signals: a) conditional signal that is sensitive to level, showing nonmonotonic behavior; and b) “directed don’t care signals” that can activated concurrently with the output signals. In this paper, the XBM specification is illustrated with the benchmark Biufifo2dma of the HP (see Fig. 5), with four inputs (cntgt1,dackn, fain,ok), two outputs (dreq,frout) and initial state 0. The description faindackn+/ frout+ in transition 4→3 means that the output (frout: 0→1) will follow the input burst (fain: 1→0 AND dackn: 0→1). Signals not enclosed in angle brackets and ending with + or — are “terminating”. Signals enclosed in angle brackets are “conditionals”, which are level sensitive with non-monotonic behavior. The input signals dackn, fain and ok are transition sensitive signals (TSS). The level sensitive signal cntgt1 is used to describe the mutual exclusion between transitions 2→5 and 2→4. The “directed

don’t care signal” fain* in transition 2→4 means that fain may either change its value or remain in its previous value. All state transition should have at least one signal called “compulsory”. A compulsory signal is an input signal that, in the previous state transition, is not directed to don’t care. A TSS input signal in a XBM specification is considered as a context signal in a transition A→B if it does not change its value during such transition (it is not on the label). On the other hand, it is considered as a trigger signal if it is labeled during this transition. The input burst of each state transition can be represented by an input transition cube (ITC). For example, the ITC in state transition 0→1 on Fig. 5 is cntgt1, dackn, fain, ok=2102 (the number 2 means“don’t care”). In this example, ok is a trigger signal, while dackn and fain are context signals (whose values are 1 and 0, respectively). Definition 1.1: Let A and B be a pair of total states in a XBM specification, and Ib/Ob be the input/output burst for the A→B transition. Let Es be one “terminating” input (Es ∈ Ib). Es is considered as an essential signal if it is a context signal on all transitions that address state A and is a trigger signal on the transition A→B. For instance (see Fig. 5), there is not an essential signal in state transitions 0→1, 4→3 and 3→2 because they are trigger signals on transitions 5→0, 2→4 and 2→5. Signal ok is essential on transition 1→2, because it is a context signal on transition 0→1. On transitions 2→4 and 2→5, dackn is essential signal. Lemma 1.1 – (proof is presented by Oliveira et al. (2008)) A XBM specification is essential hazard-free (XBM-EHF) only if for each state transition labeled by Ib/Ob, if Ob≠∅, there must be, at least, one essential signal.

ok- fain- dackn+ /

0 ok+ / frout+

1 fain+ / dreq+ frout-

<cntgt1-> fain* dackn- / dreq-

fain+ / dreq+

<cntgt1+> fain* dackn- / dreq-

faindackn+ / frout+ 4

Figure 5. Extended burst-mode specification of Biufifo2dma.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.91-102, Jan.-Mar., 2013

Oliveira, D.L., Lussari, E., Sato, S.S. and Faria, L.A.

As an example, Fig. 6 shows the HP-mp-for-pkt benchmark described by a BM specification. On all transition labels there is at least one essential signal. Therefore, it is a BM-EHF specification. Figure 7 shows the state flow map of HP-mpfor-pkt. As shown by Oliveira et al. (2008), which applies the rule generalized Ungle to check for essential hazard, this states flow map is subjected to essential hazard. The essential hazard depends on the code that the don’t-care assumes, when held the logic coverage free of logic hazard. Essential cube condition Lemma 1.1 is a necessary and sufficient condition for an essential-hazard-free specification, but not for hazard-free implementation. The super-state concept will guarantee the latter condition. According to Oliveira et al. (2008), the concept of super-state is presented. It is

AckPB - Req- / Allocoutbound + Ack -

0 Ackout+ / Allocoutbound - TRS+

AckPB+ / AllocPB - Ack +

Ackout - Req+ / TRS -AllocPB +

Figure 6. BM-EHF Specification.

Ac kPB Req

AllocPB TRS

Allocoutbound=0 Ack=0 Allocoutbound=1 Ack=0 Allocoutbound=1 Ack=1 Allocoutbound=0 Ack=1

00 01 11 10 00 01 11 10 00 01 11 10 00 01 11 10

Ackout=0 00 01 11 1000 0010 0100 0001 0010 0010 2 0010 0100 0 1000

Ackout=1

00 01 11 0001 1 0001 0001

0001 0001

1000

1000 0100 0100 0100 0100

Figure 7. State flow map: BM spec. subject to essential hazard.

used to obtain an implementation EHF. To simplify the implementation EHF, in this article we generalize the concept of super-state introducing the idea of essential cube. Definition 1.2: Consider an input burst Ib (a, b, ..n) and an output burst Ob (x, y,..m). We call a super-state the set of single total states defined by all 0/1 combinations of a subset SIb of the input burst signals, keeping all the remaining input signals and all the output signals constant. Definition 1.3: Consider a XBM-EHF specification and a super-state F of the state transition T, so that FÎ XBM-EH, whereas T is labeled by Ib/Ob. We call essential cube of transition T all the total states related to the 0/1 combinations of input burst and output burst (Ib/Ob). Whereas the states not reachable in the cube T are encoded with the value of context signals, and trigger signals are don’t-care. A super-state XBM flow map is derived from a XBMEHF specification by applying definition 1.2 to all total states. The essential cube is composed of 2N states, in which N is the total number of input signals plus the output signals that are labeled in a state transition. Figures 8a-d are part of the flow map for the BM specification described in Fig. 6. Cells in blue are used to compose super-states and essential cubes (applying the definition 1.3). For example, the 0→1 transition (see Fig. 8a,b and 9a,b) creates superstate 1 composed of two total states: AckPB Req Ackout Allockoutbound Ack AllocPB RTS=[0010001, 0000001]. State 0010001 is the final total state. Figure 9b shows the essential cube of the state transition 0→1, in which the next total states in blue are not reachable and belong to the essential cube. Due to the delays of gates and wires, the state totals which not are reachable can become reachable. For example, the 1→2 transition (see Fig. 8c,d and 9c,d) creates super-state 2, composed of four total states: AckPB Req Ackout Allockoutbound Ack AllocPB RTS= [1000010, 1100010 ,0100010,0000010]. State 0100010 is the final total state. Figures 9b,d respectively, show the essential cubes of the transitions 0→1 and 1→2. Lemma 1.2 and theorem 1.1 show the robustness of our controls. Lemma 1.2 – Let T (B→A) be a state transition of XBMEHF specification labeled by Ib/Ob and let an input signal any Is ∈ Ib and an output signal any Os ∈ Ob. If T is described by an essential cube, then in whatever order and whatever the time of arrival of Is and Os activation the total generated states belong to the essential cube T, and they all lead to the final total state A.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.91-102, Jan.-Mar., 2013

An Asynchronous Interface with Robust Control for Globally-Asynchronous Locally-Synchronous Systems

Proof: As a cube essential, T consists of 2N total states, in which N is the sum of the signals that compose the input burst (Ib) and the output burst (Ob). As the next states not reachable in the transition B→A are encoded in the way in which the signals of Ib and Ob are don’t-care, then whatever combination of activations of the signals Is and Os in T, the total states generated will belong to the cube essential T and lead to final state A, therefore the cube essential is free of essential hazard. Theorem 1: The XBM-EHF specification has an EHF implementation in the “Huffman machine architectures with feedback output” or “standard RS” if ∀ state transition T (B→A) ∈ XBM-EHF, all your activation is covered by the cube essential T. Proof: Lemma 1.2 says that if the XBM specification is EHF, then whatever the state transition T ∈ XBM has a cube essential T. As the context signals in T transition remains as a constant value in all the next states, which are reachable and not reachable in the cube essential, then regardless of the delays of gates and wires of the architectures, the activation of the next state belongs to the cube essential. As essential cube is EHF according to lemma 1.2, then the implementations on both architectures are EHF. Asynchronous Wrappers: Architecture The main objective of the proposed architecture is to provide a weak interface interaction between the locally synchronous module (LSM) and the asynchronous interface. Figure 10 shows the two different variables, “data available” and “data accept”, as the only ones used for communication between LSM and the interface. When data available=‘1’, data is ready to be transmitted, while when data accept=‘1’ the

0 Ackout+ / Allocoutbount- TRS+

Acko ut Allocoutbound TRS

AllocPB=0 Ack=0

1 (a)

Ackout- Req+ / TRS- AllocPB+

kout AllocPB TRS Req 00 00 Allocoutbound=0 01 01 2 Ack=0 11 10 (c)

00 01 11 10

AckPB=0 Req=0 1 0 01 01 1 01 0 10 01 (b)

AckPB=0 01 11 10 10 1 10 01 01 10 10 2 (d)

Figure 8. Part of BM specification: a) transition 0→1; b) flow map; c) transition 1→2;d) flow map.

data was received. Our architecture is based on the architecture proposal described by Reddy Ravi (2001). Figure 11 shows the architecture of the proposed output communication control, which implements the weak interaction between the interface and the LSM, while Figs. 12 and 13 show the proposed input and output asynchronous wrapper, respectively, with the insertion of a gated clock generator. Finally, Fig. 14 shows the full proposed AW that receives and transmits data. Ac AckPB=0 ko ut Req=0 Allocoutbound TRS 0 1

Ac AckPB=0 ko ut Req=0 Allocoutbound TRS 0 1

00 00xx 01 AllocPB=0 Ack=0

01 00xx 01

00 00xx 01 1

AllocPB=0 Ack=0

11 00xx 01 Essential 0 super state 10 10 01 Ac

Essential Cube

(a1)

AckPB=0 00 01 11

Re q AllocPB TRS

01 00xx 01 11 00xx 01 10

ut Re q AllocPB TRS 1

Allocoutbound=0 01 01 10 01 01 Ack=0 11 00xx 10 00xx 00xx 10 00xx 10

(a2)

00 00xx 10 00xx 00xx

Essential super state

AckPB=0 01 11

00 00xx 10 00xx 00xx 1

Allocoutbound=0 01 01 10 01 01 Ack=0 11 00xx 10 00xx 00xx Essential Cube

00xx 00xx

10 00xx 10

(b1)

00xx 00xx

(b2)

Figure 9. Part of BM flow map with ESS and EC: a) 0→1; b) 1→2.

Asynchronous Wrapper

DATA Locally Synchronous Data Module Available

LCLK

DATA RT

Data Accept

Output

Input

Interface

(a)

Locally Synchronous Module

LCLK

(b)

Figure 10. Locally synchronous module with weak interface: a) output; b) input. Communication control Data Available

Latch

Reset

SET

CLR

En_ D

Output PORT (AFSM)

RT AT A_CLK R_ CLK LCLK

Figure 11. Communication control of output. J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.91-102, Jan.-Mar., 2013

Oliveira, D.L., Lussari, E., Sato, S.S. and Faria, L.A.

DATA TRANSMITTER

CLK

OUTPUT INTERFACE Data Locally Available Synchronous D Q Latch Module Q

D SET Q Reset

GCLK

En_ D

Output PORT (AFSM)

CLR Q

R_CLK

Stop

A_CLK

Gated-Clock Generator CLK

A_CLK GCLK

Figure 12. Proposed output asynchronous wrapper.

Figure 15. Timing diagram: gated-clock generator.

R1_CLK

DATA RECEIVER

INPUT INTERFACE RR

Input En_ D PORT (AFSM)

AR R_ CLK

Q SET D Q CLR Reset

A_ CLK G CLK

Locally Data Synchronous Accept Module Q D Latch Q

A1_CLK

R2_CLK GCLK

Synchronizer-2

Synchronizer-1

Gated-Clock Generator

Stop-1

A2_CLK

Gated-Clock

Stop-2

CLK

Figure 13. Proposed input asynchronous wrapper.

Figure 16. Architecture of the proposed gated-clock generator. Asynchronous Wrapper Latches

DATA

Data Accepts

Input Control

Locally Synchronous Module

DATA

A1_CLK

Stop1 D Q Latch Q

RT G CLK

R1_CLK

Stop2

Data available

Gated-Clock Generator

R2_CLK

Output Control AT

A2_CLK

CLK

Figure 17. Topology of the gated-clock

CLK

A_CLK

R_CLK

Figure 14. Proposed asynchronous wrapper: with I/O.

Gated-Clock generator In this paper, a gated-clock generator (GCG) composed basically by two synchronizers and a gated-clock is also proposed. Figure 15 shows the timing diagram of the proposed GCG with the activation and deactivation of signal GCLK. While Fig. 16 shows the architecture of GCG, Fig. 17 shows the topology of the gated-clock and Fig. 18 shows the topology of its synchronizer. The stopping (pause) of the GCLK signal occurs when RCLK switches 0→1 and after two clock cycles the signal “Stop” switches 0→1, thus determining the stopping (interruption) of signal GCLK.

GCLK

SET

CLR

SET

CLR

Stop

CLK

Figure 18. Topology of the synchronizer.

Design: Ports (AFSM) The input/output ports used in the proposed AW were previously proposed by Muttersbach et al. (2000) and Muttersbach (2001). They are described in the XBM specification (as shown in Figs. 19 and 20). The XBM specification of the input/output ports meets the essential signal concept, therefore XBM_EHF.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.91-102, Jan.-Mar., 2013

An Asynchronous Interface with Robust Control for Globally-Asynchronous Locally-Synchronous Systems

7 R

A _CLK - R R * /

R_CLK RT

En _D + R R + / R_CLK +

RR - / - AR -

_CLK

A_C

R_CLK RT

2 RR - / R _CLK - AR -

A_CLK - R R * /

Z=1

En _D + / R T +

A T+/ R _CLK+

Z=0

AT 00

00 101 01 101

011 011

10 000

010

000

010

10 100

En_D=1 00 01 11 100

110 10 100

111 111

110 110

100

010

000 00x

Z=1

En_D=0 01 11 0x0

10 000 00x

101

101 11 1x1 10

111 111 6 110 1x1 110

En_D=1 01 11

10 000

10 100

011 2 011 010 01x 010 En_D=1 00 01 11 100 1x0

10x

000

0x0 En_D=0 01 11

00 001 001 0x1

A _CLK En _D - / RT + 6

En_D=0 01 11

A T+ / R _CLK +

A _C LK A T 00 R_CLK RT

A T- / R _CLK -

A _CLK + / R T-

En_D=1 01 11

001

A_ C LK A T 00 R_CLK RT

A _CLK + / R T-

A T-/ R _CLK -

10 00 8 000 001

Figure 21. State flow map: output port subject to essential hazard.

A_CLK - En_D + / RT + 8

En_D=0 01 11

Figure 19. XBM Specification: input port described by Muttersbach et al. (2000) and Muttersbach (2001).

Z=0

A_CLK + / AR +

AT 00

00 000

A _CLK + / A R+

En _D - R R + / R _CLK +

00x 3

10 4 100 10x 100

1x0

010

110

Figure 22. State flow map: output port essential hazard-free. Figure 20. BM Specification: output port described by Muttersbach et al. (2000) and Muttersbach (2001).

A_ C LK A T 00 R_CLK RT

Procedure: synthesis of ports The ports designing method starts from the XBM description and is synthesized in four steps: • Use the algorithm of Yun et al. (1999) and derive the minimum set of XBM flow charts; • Encode XBM flow tables using the adjacency diagram (Unger, 1969); • For each coded XBM flow table, insert the essential superstates, as seen in the previous section; • Perform the logic minimization, logic-hazard-free, for each “non-input” signal in the “standard RS” and “machine Huffman with output fed back” architectures (Oliveira et al., 2008). Figure 21 shows the state flow map of the output port, with the introduction of a state signal ‘Z’ to solve conflicts, while Fig. 22 shows all the minterms (black and blue) used

Z=0

00 01 11 10

0 0

En_D=0

A _C LK A T 00 R_CLK RT

Z=1

00 01 11 10

x x x

11 0

10 0 0

10 x x

En_D=0

01 5

x x x

x x

00 0 0 0 00 x x

En_D=1

0 0 0

En_D=1

11 0 0

10 0 0 3

11 x

10 x x

F SET= En_D A_CLK AT R_CLK

Figure 23. Karnaugh map: coverage hazard-free of signal Z (FSET).

in logic coverage, which ensure the output port to be free of essential hazard. The output port was implemented in the architectures “Huffman machine with output feedback” and “standard RS”. Figures 23-26 show the logic coverage free of logic hazard using the Karnaugh maps. Finally, Figs. 27 and 28 show, respectively, the logic circuits of output and input ports.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.91-102, Jan.-Mar., 2013

100

Oliveira, D.L., Lussari, E., Sato, S.S. and Faria, L.A.

A_ C LK A T 00 R_CLK RT

Z=0

00 01 11 10

En_D=0

x x

A _C LK A T 00 R_CLK RT

Z=1

00 01 11 10

11 x

10 x8 x

10 0 0

En_D=0

0 05 0

0 06 0

0 0

En_D=1

00 x x1 x

0 x2 x

x x3

En_D=1

00 0 0

R_CLK

11 0

10 04 0

A_CLK AT En_D

10 x x

F RESET= En_D A_CLK AT R_CLK

Figure 24. Karnaugh map:coverage hazard-free of signal Z (FRESET). A_ CL R_CLK RT K A T 00

Z=0

00 01 11 10

0 0

00 01 11 10

0 0 x

10 0 0

1 1 x

1 1

00 0 0 x

En_D=1 01 11

10 0 0

1 1 2 1 1 1 En_D=1 00 01 11 0 x 0

x En_D=0 01 11

A _C L R_CLK RT K AT 00

Z=1

En_D=0 01 11 x

Figure 27. Logic circuit: free-hazard output port hazard-free.

1 10 0 0

A_CLK R R E n_D 4

R_CLK = AT + En_D Z R_CLK + En_D Z R_CLK

Figure 25. Karnaugh map: coverage hazard-free of signal R_CLK.

R _CLK Z

A_ C LK A T 00 R _CLK RT

Z= 0

10 0 x

11 0 En_D=0 01 11

A_C LK A T 00 R _CLK RT

0 10 0 x

1 1

0 0

En_D=1 01 11

00 1 1 1

Z= 1

En_D=0 01 11 0

1 1

10 0 x

x 0 En_D=1 00 01 11 0 0 x

Figure 28. Logic circuit: input port hazard-free.

0 10 0

x 0

RT = A _CLK (En _D Å Z )

Figure 26. Karnaugh map: coverage hazard-free of signal RT.

DISCUSSION & SIMULATION Oliveira et al. (2011) present a list of advantages of the GALS system, which leads to the conclusion that GALS design can play a relevant role in the future of digital design in all kind of applications, including aerospace ones. However, a major drawback to this use is the asynchronous interface.

Focusing on this kind of application, the proposed hazardfree asynchronous interface proved to have a great potential, being highly desirable for the aerospace industry, once it overcomes the main challenges of this industry, thus increasing the reliability of the full system. In the treatment of essential hazard, our ports support any type of mapping either for VLSI_DSM or PLDs devices. It follows the Delay Insensitive model (DI) (Myers, 2004), restricted to interact with the environment in GFM, but without the insertion of any delay elements. This interface allows working in Ib/Ob mode, showing that the DI model is more robust than the QDI model, therefore not needing to meet isochronic fork requirements. An interface presenting similar properties was not found in literature. Figures 29 and 30 show simulations of I/O ports of the proposed AW, which show that the proposed architecture satisfies the XBM specification, are hazard-free and robust.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.91-102, Jan.-Mar., 2013

An Asynchronous Interface with Robust Control for Globally-Asynchronous Locally-Synchronous Systems

Name reset Aclk Rr En Ar Rclk /Rclk Z /Z

Value at 0 ps

0 ps

20.0 ns

40.0 ns

60.0 ns

100.0 ns

80.0 ns

120.0 ns

140.0 ns

160.0 ns

180.0 ns

200.0 ns

101

220.0 ns

A0 A0 A0 A0 A0 A0 A1 A0 A1

Figure 29. Simulation of input port.

Name

0 ps

40.0 ns

80.0 ns

120.0 ns

160.0 ns

200.0 ns

240.0 ns

280.0 ns

320.0 ns

20.0 ns

reset At Aclk En Rclk Rt Z

Figure 30. Simulation of output port.

CONCLUSION GALS systems implemented in VLSI_DSM are an interesting design style for SoCs, however, typical problems concerning the asynchronous interface, especially for AW design, proves to be major drawbacks. In relation to aerospace applications, in which reliability and safety are major constraints, these drawbacks are prohibitive. Concerning this situation, a new architecture to AW was proposed in order to overcome the previously discussed problems, showing to be a good option for those designers who need to implement GALS in VLSI_DSM, including for aerospace applications, once it improves the reliability of the system, thus eliminating essential hazards. The achieved results showed that the proposed architecture is completely free

of essential hazard and allows full autonomy for the locally synchronous modules. It follows the DI model, interacts with the environment in GFM without the need to insert any delay elements, as suggested by the previous papers found in literature, and allows working in Ib/Ob mode, proving to be more robust than the QDI model and, therefore, not needing to meet isochronic fork requirements nor requiring timing analysis. Since an interface presenting similar properties was not found in literature, the proposed architecture showed to have a great potential of implementation in all VLSI_DSM systems, including the aerospace ones, in which the harsh environment imposes additional challenges to the designers. Future work leads to a robust asynchronous interface for the implementation of GALS, involving FIFO and an application aimed for software-defined radio.

REFERENCES Amini, E., Najibi, M. and Pedram, H., 2006, “Globally asynchronous

Bertuccelli,

locally synchronous wrapper circuit based on clock gating”, Emerging

Uncertainty in Aerospace Systems”, Ph.D. thesis, MIT, September,

VLSI Technologies and Architectures, IEEE Computer Society Annual Symposium, Vol. 00, pp. 2-3.

L.F,,

2008,

“Robust

Decision-Making

with

Model

2008. Bormann, D.S. and Cheung, P.Y.K., 1997, “Asynchronous Wrappers

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.91-102, Jan.-Mar., 2013

102

Oliveira, D.L., Lussari, E., Sato, S.S. and Faria, L.A.

for Heterogeneous Systems,” Proc. Int. Conf. Computer Design (ICCD), pp. 307-314. Chapiro, D.M., 1984, “Globally-Asynchronous Locally-Synchronous Systems”, PhD thesis, Stanford University, October, 1984. Chu, T.A., 1987, “Synthesis of Self-Timed VLSI Circuits from GraphTheory Specifications”, Ph.D. thesis, Dept. of EECS, MIT, June, 1987. Cortadella, J. et al., 1997, “Petrify: A tool for manipulating concurrent specifications and synthesis of asynchronous controllers”, IEICE Trans. Inf. Syst., Vol.E80-D, No. 3, pp. 315-325. Davis, A.L. et al., 1979, “A data-driven machine architecture suitable for VLSI implementation”, In C.L. Seitz, editor, Proc. of the Caltech Conf. on Very Large Scale Integration, pp. 179-194. De Micheli, G. 2009, “An Outlook on Design Technologies for Future Integrated Systems”, CAD of Integrated Circuits and Systems, Vol. 28, No.6, pp. 777-789. Dobkin, R., Ginosar, R. and Sotiriou, C.P., 2006, “High Rate Data Synchronization in GALS SoCs”, TVLSI, Vol. 14, No. 10, pp. 10631074. Friedman, E.G., 2001, “Clock distribution networks in synchronous digital integrated circuits”, Proc. IEEE, Vol. 89, No. 5, pp. 665-692. Fuhrer, R.M. et al., 1999, “Minimalist: An environment for the Synthesis, verification and testability of burst-mode machines”, Technical Report, Columbia University, TR-CUCS-020-99. Ginosar, R., 2003, “Fourteen ways to fool your synchronizer”, Proc. of ninth Int. Symp. On Async. Circuits and Systems, Vancouver, British Colombia, Canada, pp. 89-96.

Muller-Glaser, K.D. et. al., 2004, “Multiparadigm Modeling in Embedded Systems Design”, IEEE Trans. on Control Systems Technology, Vol. 12, No. 2. Mullins, R. and Moore, S., 2007, “Demystifying Data-Driven and Pausible Clocking Schemes”, Proc. of ASYNC’07, pp. 175-185. Muttersbach, J., Villiger, T. and Fichtner,W., 2000, “Practical Design of Globally-asynchronous Locally-synchronous System”, Proc. IEEE 6th Int. Symposium Advanced Research in Asynchronous Circuits and Systems, pp. 52-59. Muttersbach, J. 2001, “Globally-Asynchronous LocallySynchronous Architectures for VLSI Systems”, Ph.D. Thesis, ETH, Zurich, 2001. Myers, C.J. 2000, “Asynchronous Circuit Design”, Wiley & Sons, Inc., 2004, 2nd edition. Nowick, S.M, 1993, “Automatic Synthesis of Burst-Mode Asynchronous Controllers”, PhD thesis, Stanford University, 1993. Oliveira, D.L. et al., 2008, “Burst-Mode Asynchronous Controllers on FPGA”, Int. Journal of Reconfigurable Computing, Vol. 2008, pp. 1-10. Oliveira, D.L. et al., 2011, “Synthesis of Robust Conrollers for GALS_ FPGA from Multi-Burst Graph Specification”, Proc. IEEE VII Southern Conference on Programmable Logic (SPL), pp. 123-129. Pontes, J. et al., 2007, “SCAFFI: an Intrachip FPGA asynchronous interface based on hard macros”, 25th Int. Conf. on Computer Design, pp. 541-546.

Gurkaynak, F.K. et al., 2006, “GALS at ETH Zurich: Success or Failure?”, Proc. 12th IEEE Int. Symposium on Asynchronous Circuits and Systems, pp. 150-159.

Reddy Ravi, A., 2001, “Globally-Asynchronous, Locally-Synchronous Wrapper Configurations for Ponint-to-Point and Multi-Point Data Comunication”, Masters of Science, University of Central Florida, 2001.

Hardt, W. et. al., 2000, “Architecture Level Optimization for Asynchronous IPs”, Proc. 13th Annual IEEE Int. Conf. ASIC/SOC, pp.158-162.

Sjogren, A.E. and Myers, C.J., 2000, “Interfacing Synchronous and Asynchronous Modules within a High-Speed Pipeline”, IEEE Transactions on VLSI Systems, Vol. 8, No. 5, pp. 573-583.

Jain, A. et al., 2001, “A 1.2 GHz alpha microprocessor with 44.8 GB/s chip pin bandwidth”, IEEE Int. Solid-State Circuits Conf. Tech. Dig., pp. 240–241.

Sues, R.H. et al., 2005, “Reliability-Based MDO for Aerospace Systems”, AIAA-2001-1521, pp. 1-8.

Jia, X., Vemuri, R., 2005, “Using GALS architecture to reduce the impact of long wire delay on FPGA performance”, Proc. of the Asia and South Pacific Design automation Conf., pp. 1260-1263. Krstic, M. et al., 2007, “Globally Asynchronous, Locally Synchronous Circuits: Overview and Outlook”, IEEE Design & Test of Computers, Vol. 24, pp. 430-441. Kumala, A. et al., 2006, “Reliable GALS Implementation of MPEG-4 Encoder with Mixed Clock FIFO on Standard FPGA”, Int. Conf. on Field Programmable Logic and Application, pp. 1-6.

Techan, P., Greenstreet, M. and Lemieux,G., 2007, “A Survey and Taxonomy of GALS Design Styles”, IEEE Design & Test of Computers, Vol. 24, pp. 418-428. Unger, S.H. 1969, “Asynchronous Sequential Switching Circuits”, John Wiley & Sons Inc. Yuan, L. et al., 2008, “Research on the Problems of Satellite Borne FPGA Based Finite State Machine”, 2nd Int. Symposium on Systems and Control in Aerospace and Astronautics (ISSCAA), pp. 1-4.

Martin, A.J. and Nystrom, M., “Asynchronous Techniques for Systemon-Chip Design”, Proc. of the IEEE, Vol.94, No. 6, pp. 1089-1120.

Yun, K.Y. and Dill, D.L., 1999, “Automatic Synthesis of Extended BurstMode Circuits: Part I (Specification and Hazard-Free Implementation) and Part II (Automatic Synthesis)”, IEEE Trans. on CAD of Integrated Circuit and Systems, Vol. 18:2, pp. 101-132.

Miller, S.P. et al., 2005, “A Methodology for the Design and Verification of Globally Asynchronous/Locally Synchronous Architectures”, NASA/ CR-2005-213912, pp. 1-35.

Yun, K.Y. and Donohue, R.P., 1999, “Pausible clocking-based heterogeneous systems”, IEEE Transactions on VLSI Systems, Vol. 7, No. 4, pp. 482-488.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.91-102, Jan.-Mar., 2013

doi: 10.5028/jatm.v5i1.232

Fuzzy Modeling of Annoyance Caused by Aircraft Noise Using Laeq and Laeq Metrics d

Tarcilene Heleno1, Jules Ghislain Slama1

Abstract: Airport noise presents a major environmental impact arising from aircrafts airport activities, being one of the most complex and difficult mitigation problems. From the environmental point of view, airports cause serious problems to the population. Annoyance caused by noise was evaluated from the Day-Night Level (DNL) metrics proposed by Schultz. However, acoustic comfort is assessed separately during the day and at night, because the noise adverse effects are different. In this context, this paper presents a study on sound impact assessment related to noise annoyance caused by aircraft and aims to establish a method to analyze the sound impact on inhabited communities that are in the vicinity of airports. Besides, it proposes reviews available in literature and discusses noise annoyance exposure caused by transportation systems in an evolutive context. The study was based on criteria of evaluation levels panned for comfort of the community according to the Brazilian Standard ABNT/ NBR 10151. Therefore, a fuzzy logic system was developed in order to establish a relationship between the percentage of people highly annoyed by adverse noise effects around the airport, Day Equivalent Sound Level (LAeq ) and Night Equivalent Sound Level (LAeq ) metric descriptors. D

Keywords: Fuzzy logic systems, Annoyance, Metrics, Aircraft noise.

INTRODUCTION Noise exposure arising from air traffic has been the main cause of conflicts between airports and nearby communities in most major cities around the world. A paper entitled “Synthesis of social surveys on noise annoyance” has been seminal in a series of social research papers that summarize the annoyance data associated between the Day-Night Level metrics (DNL) and the number of highly annoyed people (Schultz, 1978). The DNL metrics is associated to the average sound energy produced by the summation of aeronautical events over a 24-hour period, whence a 10 dB(A) penalty weighting is applied during the night. Until recently, the methodology used to assess the percentage of people who were highly annoyed by noise was inspired by Schultz’s proposal. On the other hand, a new methodology based on fuzzy logic evaluation of criteria levels that take into account the community comfort according to the Brazilian Standard ABNT/NBR 10151 was proposed, establishing a relationship between the percentage of highly annoyed people by adverse noise effects around the airport, and LAeq and LAeq metrics. D

Evolution of studies on noise annoyance caused by air transportation Research about exposure to aircraft noise began in the United States with the introduction of the jet aircraft in military bases in the 1950’s. This led to the publication of original scientific papers which examined the exposure of the population to aircraft noise, which, in turn, paved the way to community response around airports. In the 1970’s,

1. Universidade Federal do Rio de Janeiro – Rio de Janeiro/RJ – Brazil Author for correspondence: Jules Ghislain Slama | Departamento de Engenharia Mecânica/COPPE/UFRJ/C.P. 68.503 | CEP 21945-970 Rio de Janeiro/RJ – Brazil | E-mail: julesslama@yahoo.com.br Received: 02/02/12 | Accepted: 30/10/12

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.103-110, Jan.-Mar., 2013

Heleno, T. and Slama, J.G.

the increased transportation noise coaxed extensive studies on the relationship between traffic flow, noise emission and community reaction (Fidell et al., 1988). The Environmental Protection Agency (EPA) later related the complaint and annoyance data with the noise exposure level through of a dose-effect relationship, presenting an interpretation of the annoyance caused by sound levels with the DNL metrics, with terms widely used to evaluate the environmental noise in the community. Nevertheless, it is worth noting that EPA did not relate the noise effects to people’s health, stress and hearing loss, restricting instead its recommendations to interference and annoyance activities while using the term community reaction when referring to complaints (EPA, 1974). Finally, by using a linear regression of social surveys in the neighborhood of airports, EPA developed a function correlating population response and noise levels, thus yielding from the following equation (Eq. 1) the percentage of highly annoyed (HA) people. %HA = 1.8(DNL - 4.6) (1) But before this study, the approach to characterize adverse reactions to aircraft noise was centered on complaint prediction, complaints alone, and behavioral consequences.

% HAP

104

100 80 60 40 20 0

SCHULTZ

100

SCHULTZ 0.576 1.110 2.120 4.030 7.520 13.59 23.32 37.05 53.25 68.78 81.00

DNL DNL: day-night level; HAP: highly annoyed people.

Figure 1. Relation between day-night level metrics and % highly annoyed people.

this equation now allows to estimate %HAP, the percentage of HAP exposed to certain noise levels. %HAP = 0.8533DNL - 0.0401DNL2 + 0.00047DNL3 (2) The influence of Schultz’s analysis Afterwards, the publication of Schultz’s results was followed by an update of its original curve whence a wider database was used to predict the annoyance caused by the exposure to transportation noise. In this update, a quadratic function was chosen instead of Schultz’s cubic curve, as shown in Eq. 3 (Fidell et al., 1988). %HA = 78.9181 - 3.2645DNL + 0.0360DNL2 (3)

Schultz curve: Relation between DNL and highly annoyed people (HAP) In 1978, Schultz proved that aircraft noise in different cities may be interpreted in a dose-effect relationship. Such original research procedures were based on various noise transportation studies (air, rail and roadways), as well as on the fragmentation between neighborhoods evenly exposed to noise in different degrees, and significantly impacted by noise in adjacent areas. The curve shown in Fig. 1 yields a reasonable expression for the relationship between noise level and community response, and such graphic enables the quantification of the highly annoyed people percentage according to the level of exposure to noise in the community. Based on data from social surveys, Schultz proposed an equation (Eq.2), relating the percentage of HAP with DNL metrics. He describes his research group by using a third degree polynomial function in an informal approach rather than a relation derived from linear regression analysis, and

By reanalyzing the data from Fidell et al. (1988), Finegold et al. (1994) eventually published the results of a new meta-analysis that recommended the following equation to predict annoyance based on the final 400 data points and dose-response relationship, as shown in Eq. 4. %HA =

1 1+e(11.13-0.14DNL)

(4)

The controversy of Schultz’s analysis Although Schultz’s results may be regarded as conventional scientific knowledge, his study remained controversial for years. While widely accepted in literature circles, it was strongly based on his personal prediction curves for community annoyance caused by transportation noise (Kryter, 1982). Initially, Schultz’ surveys to assess annoyance were conducted in various manners, with the conversion of metrics to DNL noise levels, thus leading to much objection — especially

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.103-110, Jan.-Mar., 2013

Fuzzy Modeling of Annoyance Caused by Aircraft Noise Using Laeqd and Laeqn Metrics

concerning the adoption of a personal definition of annoyance as a variable depending on its dose-effect relationship. Beside his procedures, Schultz’s subjective judgment to determine values that best corresponded to the term highly annoyed encountered strict criticism from sociologists, who accused him of manipulating results and questioned his lack of scientific definition in coining the term ‘highly annoyed’, which eventually resulted in repeating procedures for the researcher in order to obtain the same results. Other researchers finally contested the choice of a single relationship synthesizing sources in transportation noise: Schultz’ curve represents the average community response to three sources of transportation noise (air, rail & roadways). Is it possible for the same curve to represent the same noise annoyance caused by different transportation sources? Miedema and Vos (1998) analyzed the level of noise annoyance caused by each source of transportation noise separately, listing the percentage of highly annoyed and DNL for each of them, as shown in Fig. 2. Such curves display a systematic difference between the three sources of transportation noise and show that, in general, aircraft noise causes more annoyance than that of roads or railways. These problems with Schultz’s analysis have been widely discussed in specialized circles. In this paper, we will emphasize the choice of DNL metrics; its capacity to estimate the percentage of HAP due to noise transportation; and its association with a weighted average sound energy produced

100

%HAP

80 60

105

by all aircraft events that occurred in the last 24-hour period. This metric is given by Eq. 5.

DNL = 10 log10

3600 . 24

LA(t)

1010 dt +

LA(t)+10

10 22 10

(5)

However, such metrics is not appropriate to represent annoyance in airports in which, despite the 10dB(A) correction, night time use restrictions exist, and because it is not the same as to separately consider the contribution of noise exposure between day and night-time.

METHODS AND DATA Fuzzy modeling to assess noise annoyance In the Brazilian legislation, acoustic comfort is assessed separately during day and night-time because adverse effects are different depending on the nature of the noise (ABNT/NBR 10151). Annoyance begins when comfort is lost, and noise exposure should be evaluated separately during daytime and night-time. Hence, if one is to consider the use of L Aeq D and L Aeq to represent annoyance during day and N night-times, this metrics will define a new airport zone and provide a clear set of application curves to urban authorities, which will find their use to produce a compatible urban zone. Such metrics will also be more appropriate to assess the adverse effects of noise. LAeq is defined as the average sound energy of events that D took place between 7h and 22h, as demonstrated in Eq. 6.

Aircraft

Road Traffic

LAeqD = 10 log10

3600 . 15

Pa2 (t) dt (6) P20

Railroads

0 40

DNL (dB)

LAeq is defined as the sound energy of events that took N place between 22h and 7h, as demonstrated in Eq. 7.

DNL: day-night level; HAP: highly annoyed people.

Figure 2. Value of highly annoyed people versus day-night level for each source of transportation noise.

LAeqN = 10 log10

3600 . 9

Pa2 (t) dt (7) P20

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.103-110, Jan.-Mar., 2013

106

Heleno, T. and Slama, J.G.

This paper presents an alternative approach based on fuzzy logic to estimate the percentage of HAP due to aircraft noise that uses LAeq and LAeq metrics as input variables, D N whence annoyance is the output. Revoredo and Slama (2008) have shown results on the use of the DNL metrics according to LAeq and LAeq metrics, as D N shown in Eq. 8.

DNL = 10 log10

LAeqD

LAeqN

15 . 1010 + 90 . 1010

(8)

Basic fuzzy logic concepts are presented in the next section. Principles and Elements of Fuzzy Logic Fuzzy logic was developed by Zadeh (1965). A fuzzy set F defined over a universe of discourse X is characterized by a membership function mF(x) which assumes values in the interval [0, 1]. A fuzzy set is a generalization of a classical set (crisp set) whose membership function assumes only two values, zero or one. Thus, if X is the universe of discourse and its elements are denoted by x, then a fuzzy set A in X is defined as a set of ordered pairs: A = {x, mA(x) / x ∈ X} Where mA(x) is the membership function of x in A, that assigns a membership degree between 0 and 1 to each element of X. Fuzzy conditional states the form IF A THEN B, where A and B are terms with a fuzzy meaning, e.g., “IF day night sound level is high, THEN people are highly annoyed”. These rules can be presented below:

So, each rule is a fuzzy conditional statement and different fuzzy relations can derive from this. The implementation of each rule is made by defining operators to process the input variables. The involvement function will set the output variable. Fuzzy logic system is the nonlinear mapping of an input data vector into a scalar output (Mendel, 1995). The fuzzy system is defined by the aggregation of rules that make up the fuzzy algorithm through the use of connectives such as “IF”, “THEN”. Linguistic modifiers or hedges From a fuzzy subset, through the linguistic modifiers, it’s possible to generate other subset of values for a linguistic variable from a small collection of primary terms. For the assessment of environmental impact, hedges or linguistic modifiers are a benefit that allows the adjustment to local values (Shepard, 2005). A hedge h can be considered as an operator that modifies the fuzzy set M(u), representing the meaning of u, into the fuzzy set M(hu) (Zadeh, 1973b). Applied to fuzzy sets, hedges are modifiers that concentrate or dilute the support range of the fuzzy set. In other words, the hedge changes the shape of the membership curve so that support set values have a different range for the inclusion in the fuzzy set. Thus, according to Zadeh (1973), the “very” operator acting in a fuzzy set labeled x, this effect is shown in Fig. 3. So, using the modifier “very” together with “not”, and the primary term annoyed, we obtain the fuzzy sets: very annoyed, not very annoyed. The modifier very is defined by: very x ≜ x2

Antecedent: x is to A’ as y is to B’ Rule (Ri): If x is Ai and y is Bi then z is Ci Consequently: z is C’i Where x and y are linguistic variables related to the fuzzy model and z is the output linguistic variable. A’, Ai, B’, Bi, Ci, C’i are fuzzy sets of the x, y and z. Fuzzy algorithm is an ordered sequence of instructions in which some of the instructions may contain labels of fuzzy sets (Zadeh, 1973).

µ 1 Annoyed

Figure 3. Effect of hedge “very”.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.103-110, Jan.-Mar., 2013

Very Annoyed

Fuzzy Modeling of Annoyance Caused by Aircraft Noise Using Laeqd and Laeqn Metrics

The symbol ≜ stands for “equal by definition”. The modifiers “plus” and “minus” were used with the intention to decrease the degree related to concentration and dilation. Thus, operators acting in a fuzzy set labeled x, “plus” and “minus” are defined by: Plus x ≜ x1.25 Minus x ≜ x0.75 Then: Plus Plus x = minus very x Thus modifier “highly” has been defined as: highly = minus very very = x0.75 x2 x2 = x3 Fuzzy logic and noise annoyance The fuzzy logic is characterized by its ability to infer conclusions and generate responses from vague information, which is ambiguous, qualitatively incomplete and inaccurate. Its use is very simple and natural, leading to the construction of understandable systems that are easy to maintain. The noise annoyance can be associated with objective variables, but mostly to subjective variables. So, noise annoyance can be modeled by a fuzzy system. In this paper, the fuzzy system will assign a crisp value to a particular annoyance situation. In other words, a real number that represents the percentage of people disturbed by aircraft noise. The fuzzy model developed here aims to assess annoyance caused by aircraft noise with LAeq and LAeq metrics. D

Table 1. Community’s response to noise. Values [dB(A)] by which the sound level exceeds the level criterion (Δ)

Estimated response of the community Category

Description

Null

No reaction

Small

Sporadic complaints

Moderate

Widespread complaints

Strong

Community Action

Very strong

Strong community action

107

ABNT/NBR 10151/1987 Standard uses LAeq to define the N sound levels during daytime and night-time. It introduces a classification that relates the estimated response of the community to noise as shown in Table 1. In essence, it is presented as a fuzzy system because each addition to level criterion corresponds to a response category. According to the table, some fuzzy linguistic terms were chosen. Thus, we will use these sound levels as a reference to estimate the annoyance through fuzzy logic, since the standard establishes a comfort level for different areas, such as: residential, mixed and industrial areas. Fuzzy model description The fuzzy model was developed from a computational tool provided by the Fuzzy Logic toolbox, MATLAB version 7.0. Fuzzy model is described below. Fuzzification The fuzzy system consists of choosing input and output variables and determining for each of them which fuzzy subsets are associated with linguistic variables. Linguistic variables are ideally suited to express the concepts found in environmental impact assessments (Shepard, 2005). Linguistic variables can be defined by linguistic terms. In others words, these variables contain descriptive fuzzy terms, which represent a range within the variable. In this work, LAeq and LAeq were defined as input variables, D N and annoyance is the output variable. The fuzzy linguistic terms were estimated according to the sound levels adopted by the ABNT/NBR 10151/1987. The linguistic terms used to represent the fuzzy subsets of LAeq and LAeq were “very low”, “low”, “medium”, “high”, D N “very high”. And the linguistic terms used to characterize output variables were “null”, “small,” “moderate,” “strong” and “verystrong”. Linguistic terms were defined according to the Brazilian standard ABNT/NBR 10151/1987. When sound level exceeds the level criterion, there is an estimated response of the community to noise. Membership function A triangular fuzzy set might be used to create a fuzzy number from a measured, crisp value, but otherwise it would not represent the underlying semantics of the variable.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.103-110, Jan.-Mar., 2013

108

Heleno, T. and Slama, J.G.

Sigmoid and bell-shaped membership curves are extensively used in the environmental impact assessment to represent the meaning of measured components in the existing environment and predicted changes for alternatives (Shepard, 2005; Eller, 2009). In the model proposed, triangular membership curves were used for input variables and bell-shaped membership curves for output ones.

The same was done to the second condition, Predominantly Residential Mixed Areas with LAeq =55 and LAeq =50. D N Figure 4 presents an example of the windows that open in the toolbox, in which where we insert the L Aeq and LAeq values D N and obtain the annoyance percentage. In this work, various combinations of LAeq and LAeq were simulated. On the other D N hand, we need the percentage of the HAP, so we applied the modifier “highly” in the annoyed function.

Implication and Aggregation Rules Fuzzy sets are created by the application of the rule base and aggregation methods. These fuzzy sets reflect the degree of truth contained in the model and how well the model’s rules respond to the model’s input data. According to the input and output variables, 25 rules were developed. To illustrate, we have: Rule 1: IF LAeq is very low and LAeq is very low, THEN the D N annoyance is null. Rule 2: IF LAeq is very high and LAeq is low, THEN the D N annoyance is very strong.

Hedge The hedge has been used to generate the subset “highly annoyed” from the subset “annoyed”. So, according to Zadeh’s proposal, the “highly annoyed” function corresponds to a cubic function applied to “annoyed people”.

By using this fuzzy system, the annoyance has been related to the response estimated population for different combinations of L Aeq and L Aeq in accordance with D N acceptable sound levels for each kind of residential areas of ABNT/NBR 10151. Defuzzification Latest, we have to choose the defuzzification method to convert fuzzy statements into a crisp value in order to obtain the percentage of annoyed people. The defuzzification process was performed by the centroid method. We consider the sound levels of ABNT/NBR 10151 as a condition of acoustic comfort, particularly in the case of residential use. In this work the ABNT/NBR 10151 received more emphasis for residential use: “Strictly Urban, Residential, School or Hospital Areas” – with LAeq =50 and LAeq =45 and D N “Predominantly Residential Mixed Areas” – with LAeq =55 D and LAeq =50. N In the first condition, Strictly Urban, Residential, School or Hospital Areas with LAeq =50 and LAeq =45, sound levels are D N compatible to a condition of acoustic comfort. While sound levels increase, they are no longer compatible with residential use and, gradually, annoyance appears.

RESULTS Analysis of results according to the Brazilian Standard ABNT/NBR 10151 Fuzzy model is a representation of the Standard ABNT/NBR 10151/2000, whose results relate sound levels that exceed the level criterion with the estimated response of the community. This standard was based on World Health Organization’s (OMS) results, present in the publication Guidelines for Community Noise (WHO, 1999).

Figure 4. Window of the toolbox Fuzzy with input and output values.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.103-110, Jan.-Mar., 2013

Fuzzy Modeling of Annoyance Caused by Aircraft Noise Using Laeqd and Laeqn Metrics

Schultz’s results were obtained from Eq. 3, which related the percentage of HAP to DNL metric. The results are presented in Tables 2 and 3. The curves have the same behavior, with a single horizontal shift of 5 dB (A) between them, as shown in Fig. 5. From a certain noise level, the percentage of HAP obtained from the fuzzy modeling grows at a rate higher than the Schultz’s curve, depending on the type of area. Table 2. Residential condition: Strictly Urban, Residential, School or Hospital Areas with LAeq =50 and LAeq =45. N

LAeq

HAP (%) Fuzzy

0.03

0.98

3.22

15.92

44.42

Table 3. Residential condition: Predominantly Residential Mixed Areas with LAeq =55 and LAeq =50. D

LAeq

HAP (%) Fuzzy

0.03

0.98

3.22

15.92

44.42

%HAP

LAeq

Schultz (DNL) LAeqD=50 and LAeqN=45 LAeqD=55 and LAeqN=50

The percentage of HAP (slowly increasing) is low for low levels, because noise levels are appropriate for this use. However, for higher levels the percentage of HAP grows significantly. Residents in Strictly Urban, Residential, School or Hospital Areas with LAeq =50 and LAeq =45 are used to very low D N noise levels. Therefore, this condition (blue curve) is more appropriate to assess the annoyance. Both curves show an abrupt increase of the curve inclination, which is coherent with the shape of noise perception by humans according to WHO for high levels. Usually, annoyance comes earlier, and the perception depends on environmental noise level and type of area. LAeq and LAeq D N levels are used to charactere a region in accordance with the assessment criterion level. The fuzzy model developed to evaluate the noise annoyance takes into account the daytime and night-time noise levels for each area. As a consequence, %HAP was obtained for various situations.

CONCLUSION

80 70 60 50 40 30 20 10 0 45

109

Sound Level - dB(A) Figure 5. Curves obtained from fuzzy model to %HAP in each condition.

The fuzzy system assessing aircraft noise annoyance is based on a set of logical conditions that follow noise level recommendations in residential areas defined by ABNT/NBR 10151. The estimated community response to noise considered in the previous version of this Standard was important to define the variables of the fuzzy model. The fuzzy model relates the LAeq and LAeq noise metrics D N and the percentage of HAP to evaluate the annoyance using fuzzy logic. For low values of noise levels, the predicted annoyance will be less than that proposed by Schultz, while for higher levels, the annoyance predicted will present a much greater value than that in Schultz’s model. The term “highly annoyed” has been used according to the definition proposed by Zadeh, and this tool is appropriate to adjust the values. The choice of triangular membership functions is valid to LAeq and LAeq variables due to the fact that such functions can D N be easily adjusted through an optimization process based on genetic algorithms. For the variable annoyance, bell-shaped membership curves were used, since they are more appropriate to

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.103-110, Jan.-Mar., 2013

110

Heleno, T. and Slama, J.G.

represent the noise annoyance in cases of the assessment of environmental impact. From the methodology presented herein, it is possible to develop fuzzy systems for environmental impact assessment from other sources of noise.

ACKNOWLEDGMENTS The authors would like to thank Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ), CNPq, and CAPES for the financial support for the development of this work.

REFERENCES ABNT, NBR 10151, 1987, “Acoustics - Assessment of noise in populated areas to ensure the comfort of the community”. Procedure, Rio de Janeiro. ABNT, NBR 10151, 2000, “Acoustics - Assessment of noise in populated areas to ensure the comfort of the community”. Procedure, Rio de Janeiro. Eller, R.A.G., 2009, “Modelo de geração de tarifa de ruído aeronáutico utilizando Lógica Fuzzy”, Tese (Doutorado), Instituto Tecnológico de Aeronáutica, São José dos Campos. EPA, 1974, “Information on Levels of Environmental Noise Requisite to Protect Public Health and Welfare with an Adequate Margin of Safety’’ U.S. Environmental Protection Agency, EPA/ ONAC, Washington DC. Fidell, S., Schultz, T.J. and Green, D., 1988, “A theoretical interpretation of the prevalence rate of noise-induced annoyance in residential populations”. Journal of the Acoustical Society of America. Vol. 84, pp. 2109-2113. Finegold, L., Harris, C.S. and von Gierke, H.E., 1994, “Community annoyance and sleep disturbance: Updated criteria for assessing the impacts of general transportation noise on people”. Noise Control Engineering Journal, Vol. 42, pp. 25-30. Kryter, K.D., 1982, “Community annoyance from aircraft and ground vehicle noise”. Journal of the Acoustical Society of America, Vol. 72, pp. 1222-1242.

Mendel, M.J., 1995, “Fuzzy Logic Systems for Engineering: A Tutorial”, IEEE. Miedema, H.M.E. and Vos, H., 1998, “Exposure-response relationships for transportation noise”. Journal of the Acoustical Society of America, Vol. 104, pp. 3432-3445. Schultz, T.J., 1978, “Synthesis of social surveys on noise annoyance”. Journal of Acoustical Society of America, Vol. 64, pp. 377-405. Shepard, R.B., 2005, “Quantifying Environmental Impact Assessments Using Fuzzy Logic”. Springer Science + Business Media, Inc. Slama, J., Revoredo, T. and Mora-Camino, F.A.C., 2008, “Is DNL appropriate for airport noise zoning”, Acoustics Paris, pp. 1967-1971. WHO, 1999, “Guidelines for Community Noise”. Geneva, Switzerland. Edited by B., Lindvall, T., Schwela. Retrieved in May 15, 2012, from http://www.who.int/docstore/peh/noise/guidelines2.html Zadeh, L.A., 1965, “Fuzzy Sets”. Information and Control, Vol. 8, pp. 338-353. Zadeh, L.A., 1973, “Outline of a new approach to the analysis of complex systems and decisions processes”. IEEE Transactions on systems, man and cybernetics. Vol. SMC-3, pp. 28-44.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.103-110, Jan.-Mar., 2013

doi: 10.5028/jatm.v5i1.165

IMFLAR: An Intuitive Method for Logical Avionics Reliability Nilson Silva1, Luís Gonzaga Trabasso2

Abstract: The continued growth of the general aviation fleet demands the need of forever improved preventive methods of failure analysis, in order to reduce the number of incidents or accidents. It has been proved that one possible solution to avoid unsafe conditions is the installation of new avionic systems. This article presents the method named IMFLAR — an Intuitive Method For a Logical Avionics Reliability, an analysis method for avionic systems installations based on a conceptual model of human factors and an artificial neural network application, giving an overview of these installations and analyzing the involved risk factors. This is a new preventive approach that establishes a relationship between unsafe characteristics observed during the installation of avionic systems and an operational database of incidents and accidents, in order to provide a framework to make aviation safer. Additionally, this article describes the steps to obtain the necessary parameters that ought to be used to avoid unsafe conditions for a modification that installs an avionics system in the aircraft. Keywords: Accidents prevention, Avionics system, Artificial neural network.

INTRODUCTION Nowadays, general aviation is the category that covers the largest number of aircrafts in operation (e.g., business aircraft, charter flight and agricultural). Aircraft accidents are responsible for innumerous losses, claiming lives and causing large material damages. The Brazilian General Aviation fleet is increasing, and there is a continuous technological advance that implies in crescent risk factors for incidents or accidents. For this reason, it is necessary the use of new prevention methods. These methods must analyze relationships among the triad man-machineenvironment: where man represents human factors; machine symbolizes aircraft, systems and ground/air facilities; and environment is related to climatic and environmental factors. The triad man-machine-environment summarizes some fundamental elements in investigation of aircraft accidents and is the theme of the Centre for Research and Prevention of Aeronautical Accidents of Brazilian Air Force (CENIPA – Centro de Investigação e Prevenção de Acidentes Aeronáuticos). At present time, electronic systems are employed in aircraft, also known as airborne systems or avionics (term composed by aviation and electronics). Avionics are useful means of preventing unsafe situations. Installing these systems may involve mechanical and structural aspects, such as: installation and affixation of components or antennas on the fuselage and the passage of electrical cable from pressurized to non-pressurized areas; electrical aspects, for example: interference, lightning protection; software; human factors, such as changes in the pilot’s panel, alerts, etc. The main objective of this work was to select parameters that describe the installation of avionics system and the

1.Agência Nacional de Aviação Civil – São José dos Campos/SP – Brazil 2.Instituto Tecnológico de Aeronáutica – São José dos Campos/SP – Brazil Author for correspondence: Nilson Silva | Avenida Cassiano Ricardo, 521, bloco B, 2º andar | CEP 12246-870 São José dos Campos/SP – Brazil | E-mail: nilson.silva@anac.gov.br Received: 01/10/12 | Accepted: 13/02/13

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.111-126, Jan.-Mar., 2013

112

Silva, N. and Trabasso, L.G.

risks involved. With this approach, it is possible to enhance the flight safety, avoiding possible unsafe conditions, and, as consequence, maintenance costs can be reduced, lives and aircraft be saved. A second objective was to classify these parameters through an artificial neural network that graphically displays the data, similar to a risk matrix. This article is organized as follows: the section “Related works and statistics” presents a literature review of related works. The avionics systems and their installation are described in section “Avionics systems”. The section “Mathematical modeling of risk matrix, SCHELL and SOM” discusses models in use, the SCHELL model (acronym of Cultural, Software, Hardware, Environment and Liveware, a model of interfaces defined as follow in this work), presenting a description of a risk matrix and artificial neural network SOM (acronym of Self-Organizing Maps) (Kohonen, 1994). The section “IMFLAR Method” describes the proposed method that employs the concepts previously discussed in a practical application, in order to determine the parameters that can be used in safety analysis of an avionics system installation and in its operational use. Finally, some conclusions and comments on the IMFLAR method are presented in the “Conclusions” section.

RELATED WORKS AND STATISTICS In order to provide a complete description of the IMFLAR: An Intuitive Method for Logical Avionics Reliability, it is worthy to describe some related works and accidents statistics that presents some concepts employed in IMFLAR method. As defined by Federal Aviation Administration (FAA) (2010), the term ‘Reliability’ is […] an expression of dependability and the probability that an item, including an aircraft, engine, propeller, or component, will perform its required function under specified conditions without failure, for a specified period of time”. This concept is useful for a complete analysis of avionics systems, their installation and the risks involved. In addition to this analysis, some historical and cultural aspects should be also considered. For example, with respect to the aviation history of the 20th century, Van der Velde (1995) shows that the quantity of onboard equipment in aircraft is increasing. Regarding the cultural aspect, according to the Brazilian Association of General Aviation

(ABAG – Associação Brasileira de Aviação Geral) (2011), in recent years, the General Aviation market has been growing in Brazil and, consequently, the market for avionics systems has been following this trend. This growing market may generate a great demand for avionics installations, which must be properly approved by Brazilian National Agency of Civil Aviation (ANAC – Agência Nacional de Aviação Civil). The growth in the number of installations implies the necessity of more investments in the training of pilots. Additional training is also required for installation and maintenance teams, in order to avoid unsafe conditions in these installations. In addition to historical and cultural contexts, it is also important to emphasize aspects of flight safety. Regarding prevention and statistics on aircraft accidents, several sources can be consulted. For example, the Aircraft Crashes Record Office (ACRO) collects and records worldwide data of aircraft disasters with more than six passengers, excluding helicopters, sampled a total of 19,980 accidents, where it can be verified the preponderance of human errors over technical failures and environmental conditions (ACRO, 2011). The National Transportation Safety Board (NTSB) has examined the impact of modern digital display panels (or Glass Cockpit) replacing the conventional analog equipment in general aviation, demonstrating that the employment of this panel configuration is increasing. This study also indicates that aircraft equipped with Glass Cockpit are more susceptible to accidents during instrument flight (NTSB, 2010). The Aircraft Owners and Pilots Association (AOPA) analyzed the historical evolution of the accident rate for General Aviation between 2001 and 2010, and their data show a tendency to decrease the total number of accidents during this period (AOPA, 2010). Despite this decreasing rate, Burin (2011) observes the fact that the CFIT, LOC and ALA types of accidents continue to dominate the statistics for turboprop aircraft. CFIT, LOC and ALA are the acronyms for Controlled Flight Into Terrain – CFIT, Loss of Control – LOC and Approach and Landing Accidents – ALA. They collectively define types of accidents, according to the taxonomy of the CAST/ ICAO Common Taxonomy Team - CICTT. CICTT is a working group composed of members of the International Civil Aviation Organization (ICAO) and the Commercial Aviation Safety Team (CAST), which includes aviation authorities and manufacturers, to develop a common taxonomy and definitions of accidents (ICAO/CAST, 2011).

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.111-126, Jan.-Mar., 2013

IMFLAR: An Intuitive Method for Logical Avionics Reliability

From these statistics, it can be noticed that the number of accidents in General Aviation has a tendency to decrease, partly due to the adoption of avionics systems. But, with the increasing complexity of systems and information to pilots, in the near future, it will become necessary to increase the safety of the fleet. Under requirements of the aircraft, many avionics systems are introduced to increase flight safety, such as collision avoidance systems with ground, obstacles or aircraft, just to name a few of them. Among these, this article describes the implications of using TAWS – Terrain Awareness and Warning System – at the prevention and reduction of CFIT accidents in Brazil. The Brazilian Civil Aviation Regulations (RBAC), which are gradually replacing the old Brazilian Aeronautical Regulations (RBHA), are the Brazilian requirements that involve operational aspects and certification, among others. Some requirements related to this work can be considered: the RBAC 21 provides general requirements for certification of aircraft; the RBACs 23, 25, 27 and 29 refer to the certification of airplanes and helicopters; the RBHA 43 and 145 refer to the installation and maintenance of systems; RBAC 26 and 39 refer to continued airworthiness. The main operational requirements are contained in the RBAC/RBHA 91, 119, 121 and 135 depending on the category of the aircraft and its type of operation, but other RBAC/RBHA, as applicable, can be found on the website of the ANAC. The expression Safety Assessment defines a large field of knowledge, covering flight safety, system analysis, risk analysis, failure prediction, mitigation of unsafe conditions, etc. There is a large variety of Safety Assessment methods, according Everdij et al. (2006). For the Safety Assessment aspects of systems and human factors, there are specific requirements within the RBAC 23, 25, 27 and 29 and their equivalents 14 CFR (Code of Federal Regulations) Part 23, 25, 27 and 29 of FAA: §2x.1301–Function and installation, and §2x.1309 – Equipment, systems, and installations. These requirements should be complemented, as applicable, with requirements that include other aspects, such as: human factors; noise; mechanical analysis; structural; electrical; flightcrew evaluation; airworthiness; operational aspects, etc. The Advisory Circulars (ACs) of the FAA are one of the means of compliance with the requirement 2x.1309 for aircraft Part 23 and Part 25. Other ACs and interpretative materials, such as Policy Files, can also be

113

found at FAA’s homepage, and a great amount of documents from other authorities can also be found on the Internet, such as the Transport Canada Civil Aviation (TCCA) and European Aviation Safety Agency (EASA), among others. In particular, companies interested in certification of avionics systems in Brazil should refer to the material available on websites of ANAC, Department of Airworthiness of ANAC (SAR–Superintendência de Aeronavegabilidade), and Aircraft Modification/Supplemental Type Certification of SAR (HST – Homologação Suplementar de Tipo). In a Safety Assessment analysis, it may also be consulted data from the accident investigation authorities, such as statistics, studies, materials and recommendations of the NTSB and CENIPA, organizations such as ICAO, AOPA, and International Air Transport Association (IATA), standardization organizations such as the Radio Technical Commission for Aeronautics (RTCA), Society of Automotive Engineers (SAE), Military Specifications and Standards (MIL), Aeronautical Radio Inc. (ARINC), manufacturers such as Boeing, Airbus and Embraer, and data from manufacturers’ systems and components. The approach presented herein differs from other Safety Assessment analysis by introducing an Artificial Neural Network to group interfaces in a model based on human factors, namely, the SCHELL model. The approach presented herein differs from other Safety Assessment analysis by introducing an Artificial Neural Network to group interfaces in a model based on human factors, namely, the SCHELL model. The SCHELL model (Keightley, 2004 apud Perezgonzalez and Perry, 2010) is a model of interfaces based on human factors derived from the SHEL and SHELL models developed by Edwards (1972) and Hawkins (1987). The SHEL model (Edwards, 1972) uses the concept of interfaces where the human element is surrounded by other elements, and it is an acronym for: Software – which is not part of a physical system, such as aeronautical regulations and requirements, maintenance procedures, manuals, etc.; Hardware – which represents the installation, equipment, components, cables, antennas and accessories; Environment – represents the physical factors, such as environmental conditions, weather, internal and external noise, among others; and Liveware indicates the human element: for example, the pilot that operates the equipment or the maintenance technician.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.111-126, Jan.-Mar., 2013

114

Silva, N. and Trabasso, L.G.

The other Liveware interface was introduced by Hawkins (1987) in the SHELL model, and it illustrates the interactions of human elements, such as the interaction between members of maintenance/installation groups with the design team. A similar model, the m-SHELL model of Kawano (Itoh et al., 2004 apud Kawano 2002) includes the element management; The SCHELL model (Keightley, 2004 apud Perezgonzalez and Perry, 2010) considered the Cultural element — that encloses all elements of human culture, such as management aspects, maintenance and certification cultures of airline companies, and some others cultural aspects. To categorize these interfaces of the SCHELL model according to the associated risk, the method presented in this work uses parameters’ weighting, obtained by an Artificial Neural Network based on Self-Organizing Maps (Kohonen, 1994). The Self-Organizing Maps or SOM are an application of Artificial Neural Network, created by Professor Teuvo Kohonen (1994). Braga (2006) points out that the development of SOM networks was based on the self-organization of the human brain. The SOM networks are self-organized topographic maps of input patterns, where the spatial location of neurons is determined by their degrees of similarity. Related to these Safety Assessment aspects, the proposed model presents a different approach from the methods cited by Everdij et al. (2006), as it improves the Risk Matrix method through a SOM network, adapted to the SCHELL model interfaces (Keightley, 2004 apud Perezgonzalez and Perry, 2010).

AVIONICS SYSTEMS Accidents, human factors and avionics systems Figure 1, known as 1:600 Rule, is the result of a research in the industrial environment in 1969 and illustrates the average percentage ratio between incidents and accidents (Bird and Germain, 1969 apud ICAO, 2005). Figure 1 shows that small events occur at a greater frequency than accidents and its severity is inversely proportional to the possibility of occurrence. However, a major risk factor is that the majority of incidents are not reported. Thus, small incidents that were not considered can become events that, by their turn, might trigger a serious accident.

Therefore, there is a critical aspect of human factors which includes the manufacture of the avionics system, its installation, operation and maintenance, and that depends on the personnel involved in reporting problems or faults. This is not always efficient. With respect to the human factors involved, safety barriers can be created if critical parameters of the installation of each system are analyzed and procedures to reduce the unsafe conditions are elaborated. Installation and maintenance of avionics systems through repair stations, service centers and other maintenance facilities in Brazil The variety of failures that may occur in the components of avionics equipment, combined with the diversity of aircraft manufacturers and suppliers, may make the categorization of unsafe situations impossible by an analysis method of internal components and their interconnections, for the existing range of avionics systems. Except for the case of an aircraft repair, the avionics systems installation usually occurs in two ways: through direct incorporation of the aircraft’s type design or by major type design change. The avionics incorporated into the project type are made by the manufacturers of the aircraft or under their supervision. However, in Brazil, the most common way to install avionics is through major changes at installer’s facilities. Thus, the unsafe factors for avionics systems are closely related, besides the manufacturing, installation and maintenance. The maintenance and installation stations have a vast knowledge of unsafe factors in the field and, in addition to the aircraft’s pilots, are the initial point that reports problems to manufacturers. On one hand, some low complexity aeronautical systems, properly approved, may be developed by Brazilian facilities,

Fatal accident

Serious accidents

Accidents

600

Incidents

Source: adapted from Bird and Germain, 1969 apud ICAO, 2005.

Figure 1. 1:600 Rule.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.111-126, Jan.-Mar., 2013

Approval of major modifications to install avionics systems in Brazil The PST group – Supplemental Type Program (Grupo PST – Programa Suplementar de Tipo) that belongs to Aeronautical Products Certification Branch (GGCP – Gerência-Geral de Certificação de Produto Aeronáutico) of SAR/ANAC is responsible for major changes certification in Brazilian registered aircraft, including the installation of avionics systems. The PST group has a database of public domain, which lists the major changes certified in Brazil, from where the data used in this work was collected. TAWS – Terrain Awareness and Warning System For years, Controlled Flight Into or Toward Terrain (CFIT) category of accidents, dominated the global statistics of accidents. As defined by the ICAO/CAST (2011), CFIT is a collision in flight or near collision with terrain, water or obstacle without indication of loss of control. Figure 2 shows the causes of CFIT for General Aviation in Brazil, between 1999 and 2008, according to CENIPA. Following the recommendations of ICAO for CFIT prevention, Brazilian regulations RBAC/RBHA 91, 135 and 121 establish requirements for the installation of Ground Proximity Warning System (GPWS) in the aircraft.

115

Contributing Factors in General Aviation CFIT (1999-2008) 80 69.2 67.8 70 57.7 60 50 42.3 40 33.8 33.8 30 20.9 21.1 18.3 20 14.1 11.3 10 0

Ju dg m Ps en P yc ho lan t (ju ni l d) o In n gi di ca g ( pl sc la a ip sp n) l ec we Ad ine t at ver in fl (ps he se y) r c m igh on et t (i Po O e th di o nd or ti ro er ) fli op Su ons log gh e te ra per (m ical xp tio vi et) sio er ien nal En as n (s c vi pe e up in r ct s( ) Ap onm th e o pl e pe air ica nta r cr tio l in aft ) n fo (e of rm xp ) Ph com atio ys n m (e io an n lo ds v) gi ( ca l a com sp m ec ) ts (p hy )

such as: LED lighting systems, power supplies, small control panels and supports etc. On the other hand, when an aeronautical facility are installing an avionics system that was not developed by its own means, this one does not have access to all the design parameters, being unaware of hardware and software details of the system. Consequently, various tests must be performed in a device in order to be certified for aviation usage. Another relevant factor is the current use of Integrated Modular Avionics (IMA). With IMA, the maintenance of avionics systems is basically performed by modules substitution or by the replacement of the whole equipment, which are then sent (the IMA or complete equipment) to their respective manufacturers. Probably, the Brazilian installers do not have the necessary conditions to execute the kind of repair that is performed by the IMA system manufacturer. Thus, for a case study on maintenance of avionics in Brazil, an emphasis on systems should be adopted, avoiding a deep approach on component failures, because it is likely that most Brazilian facilities will not have access to failures analysis of IMA components.

Percentage

IMFLAR: An Intuitive Method for Logical Avionics Reliability

Source: ICA 3-2. CENIPA, 2009 (adapted)

Figure 2. Contributing Factors in General Aviation – CFIT.

An enhanced version of the GPWS System is called Terrain Awareness and Warning System (TAWS), which uses geo-location systems like Global Navigation Satellite System (GNSS) and databases that include terrain and a large amount of man-made obstacles to generate alarms and alerts. The installation of TAWS and GPWS has contributed to a substantial reduction in the number of CFIT accidents. There are two classes of TAWS systems: class A TAWS, which has interconnection with a display, radio altimeter and/or air data computer, and class B TAWS, where the interconnection with a display is not mandatory. Figure 3 illustrates a typical class A TAWS system, with some of its main common components, such as altimetry and positioning inputs, data processor, display and visual and aural alerts. From direct observation of the data presented in the public database of modifications, the PST group analyzed data of the certification processes during the year 2009 and concluded that the TAWS/GNSS installations accounted for about 50% of the total number of modifications corresponding to approximately 80% of the cases of major changes in electrical and electronic systems, presented to be certified by ANAC. Due to these characteristics, it has been gathered a significant quantity of risk factors for this type of system. In addition to this analysis, samples of processes for installations of TAWS/GNSS systems were collected from two base years: 2005, with 313 samples; and 2009, with 378 samples. These base years were chosen because they show the two extremes of the development of this kind of installations: a pioneering phase, where there is a clear need for training of certifying companies — applicants,

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.111-126, Jan.-Mar., 2013

116

Silva, N. and Trabasso, L.G.

GPS Antenna

Sensors

- ADC - Barometric altitude - Radio altimeter

Processor

Display Aural & Visual Alerts TERRAIN

Figure 3. Class A TAWS – block diagram.

installation and maintenance facilities; and a maturation phase with some degree of experience of applicants and maintenance facilities. Furthermore, in these years, other researches were conducted, such as the observations on the meteorological adverse conditions. Although the years between 2005 and 2009 could also provide sample data, these data were not measured with all the examined characteristics, therefore they were not included in this analysis. Consequently, the application of the proposed method observe only the sampled data of 2005 and 2009, as it was the initial objective of this research, that is to verify the processes in relation to risk factors, climatic factors, the systems involved and the quantity of these systems. The main risk factors detected for TAWS/GNSS systems are listed in Table 3.

mathematical modeling of risk matrix, SCHELL and som Principles of risk analysis and risk matrix According to Greenwell and Knight (2003), the risk can be defined as the probability of an event to occur, multiplied by the anticipated cost derived from the occurrence of the event. Equation 1 models the total risk based on this definition. In this work, the cost is considered equivalent to an indirect measure of severity (consequence) of an accident. R=P·C (1) where R = total risk; P = probability; C = (cost) = severity. Basically, two criteria classify a risk: the likelihood and the severity. A risk matrix is a graphic arrangement of the risk. The origin of the risk matrix comes from the definitions contained in MIL-STD-882D (Department of Defense – DOD, 2000). Consequently, the risk matrix is a combination of likelihood

Risk Matrix Severity Likelihood Catastrophic Critical Marginal Negligible Frequent 1 3 7 13 Probable 2 5 9 16 Occasional 4 6 11 18 Remote 8 10 14 19 Improbable 12 15 17 20 Risk High (1-5) Serious (6-9) Medium (10-17) Low (18-20) Source: based on MIL-STD-882D (DOD, 2000)

Figure 4. Risk matrix.

and severity in order to classify the different levels of risk. Figure 4 illustrates an example of risk matrix. The Risk Matrix is usually employed in civil and military applications, because it is an easy method used in risk management; but this method has some inherent limitations, as identified by Cox (2008). These are: low resolution, error induction, inefficient reallocation of resources based on the categories provided in the matrix, and ambiguity of inputs and outputs. Application of the SCHELL model interfaces As seen in Fig. 2, the influence of human factors is preponderant over the other CFIT factors. Due to this, the SCHELL method for human factors was adapted for the IMFLAR method in order to analyze influence of human factors during the certification process. These influences are employed as IMFLAR input parameters into a reliability model for the analysis of the involved risk in an avionics system installation during a certain period of time. As previously described in this work, the preliminary SHEL model (Edwards, 1972), which was developed to analyze the human factors interfaces, was modified to include other additional interfaces. Itoh et al. (2004) use the concept of m-SHEL model from Kawano (Itoh et al., 2004 apud Kawano, 2002) in a table format, grouping the interfaces in pairs. A practical application of this format, adapted from the work of Itoh et al. (2004), is presented in the item 3 – “SCHELL indexes block” of section “IMFLAR Method”, where the interfaces used are: LC – cultural interface; LS – software interface; LH – hardware interface; LE – environmental interface; LL – interface of human relationships; L – the human element itself, such as the pilot, the certification expert, or the maintenance technician. SOM – Self-Organizing Maps A Self-Organizing Map, also known as SOM neural network, consists of just a few to even thousands of artificial

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.111-126, Jan.-Mar., 2013

IMFLAR: An Intuitive Method for Logical Avionics Reliability

neurons organized in a grid, or lattice, usually bi-dimensional. For a competitive process of learning, the winning neurons are organized in a topographic map, as in the areas of the human brain. This map is a statistical arrangement in terms of the inputs (Castro and Castro, 2001). The mathematical definition of SOM requires the definition of Euclidean vector. For an input signal x(t) as Euclidean vector of d dimension (Kohonen and Honkela, 2007) is defined according to Eq. 2: x(t)=[ξ1 (t), ξ2 (t), ξ3 (t), .... , ξd (t)] (2) where: x(t) = Euclidean input vector, consisting of d input signals; t = data index in a sequence; ξ(t) = input signals; d = dimension, or quantity of input signals. Each neuron of the SOM network is represented by a weight vector, according to Eq. 3: m = [m1, m2, m3, ..., md] (3) where: m=d-dimensional weight vector; d is the dimension of the input vector. The neurons of the network represent the input vectors in the best possible way. Each input vector is presented to all neurons of the network and the “winner” is the one with the closest weight, i.e., the neuron with more similarities to the input vector in question. In this work, the Euclidean distance is used to determine the “degree of similarity” between the input vector and weights of the neurons. The “winner” neuron c for an input vector x is the one that has its weight value m with the smallest Euclidean distance to the input vector x. The winner neuron is also called the Best Match Unit (BMU), around which the closer neurons are organized. The organization of neurons is a process of smoothing by similarity, defined by the function called Nc –neighborhood of the winner. After finding the BMU, the network is updated by an iterative process (adapted from Laurino, 2004; Kohonen and Honkela, 2007). According to Reyes-Aldasoro (1998), the process of neurons self-organization (called nodes of the grid) on a SOM network can be described in two steps: combination of inputs and neurons; and updating around winning neurons. These processes are modeled by Eq. 4 and Eq. 5: Combination:

117

Updating: ⎧ m i ( t + 1 ) = m i ( t ) + ∝ ( t ) [ x ( t ) - m i ( t ) ], i ∈ N c (5) ⎨ ⎩ m i ( t + 1 ) = m i ( t ), i ∉ N c where, for a time t: x = input; mi = any node; mc = winner neuron; α = sequence of gain; and Nc = neighborhood of the winner neuron (in this work, Nc is defined internally in the software utilized). Each neuron is connected to adjacent neurons by the neighborhood relation Nc, which determines the topology of the SOM map. The topology has two attributes: structure and shape (Vesanto et al., 2000). Figure 5 shows the structure and shape for the SOM network used in this work. Grouped neurons can also be presented in other graphic format, known as U-matrix (Vesanto et al., 2000 apud Francisco, 2004). A U-matrix shows, through the colored units of the map, the distances between clusters. The colors of a U-matrix vary according to a distance scale, from dark blue to red, where the dark blue color represents the 66 nearest cells (neurons), or groups. The lighter colors to red represent the separation of the clusters (according to Vesanto et al., 2000 apud Francisco, 2004). Figure 6 shows an example of a U-matrix, applying the data to be presented in this work to a SOM network through the tool SOM_TOOLBOX_2, for MATLAB®. This toolbox is shortly described in step 4 of IMFLAR.

Figure 5. SOM topology: hexagonal structure grid and plane shape.

6.74

3.6

0.47

x(t) = || x(t)-mc (t) ||= mini || x(t)-mini(t)|| (4)

Figure 6. U-matrix.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.111-126, Jan.-Mar., 2013

118

Silva, N. and Trabasso, L.G.

IMFLAR METHOD IMFLAR description IMFLAR: An Intuitive Method for Logical Avionics Reliability is a method that organizes data from the PST public database, combining them with CFIT accidents statistical data, taking into account the interfaces from the SCHELL model. These weighted data are grouped by a SOM neural network that organizes them into risk groups. These groups and the organized SCHELL data are then used for the development of a simulation model in Simulink/ MATLAB®. This reliability model of can predict a situation of increased risk based solely on the data presented, modeling human factors, environmental and machine involved with the installation of the avionics system under analysis. The IMFLAR overcomes each of the limitations of Risk Matrix, due to the flexibility of the SOM neural network employed and its presentation in graphical format. As an application of the proposed method, the installations of TAWS avionics in Brazil are analyzed through a Safety Assessment analysis. This proposed analysis exemplifies the relationship between a statistical data from an accidents category (CFIT) and the installation of an avionics system that prevents this kind of accident. An alteration of risk, during the operational life of the aircraft, with the modification that installs the avionics system under analysis can be simulated by a variation of a source block parameters in the model in the Simulink/MATLAB® presented. Thus, the IMFLAR method can simulate the risk variation during the course of the years for an avionics system. An application of this simulation can be the creation of an alternative solution to generate other maintenance tasks for the system under analysis. The following subsections present: the steps that show how to acquire data for the IMFLAR; and the procedures for the elaboration of a simulation model on Simulink/MATLAB®. Steps of IMFLAR – acquiring data Figure 7 shows a block diagram that describes the necessary steps to implement the proposed method. Each block is described by its input, processing of data by the block, and its output. Step 1 – the block of contributing factors for accidents • Input item: equivalent to the statistical causes of real accidents related to the conditions of use of the certified system. This paper uses the statistical data of the major

causes of CFIT in General Aviation - 1999 to 2008, according to the CENIPA data records (CENIPA, 2009). • Data processing item: these statistics of the CFIT causes are scaled by the percentage of occurrence, according to Table 1, resulting in five risk levels (named E - Exposure). The scaling reproduces the way levels of the risk matrix are elaborated. The proposed scaling is just an example and other subdivisions are also possible. However, for a smaller number of divisions, the method tends to be less accurate, and for the higher number of subdivisions, the number of parameters to be analyzed tends to be greater. The Level or Amplitude is defined as the amount, in percentage, of installations involved with the risk factors, and the Exposure means the exposure to risk. Table 1 shows the relationship between the Exposure (E) and the Level or Amplitude (A). Indexes A and E are determined by the application of Table 1 to the 11 contributing factors determined by the CENIPA, according to Fig. 2. For example, for the first contributing factor <Judgment (jud)>, corresponding to the Level or Amplitude A=69.2%, then 26% <A ≤80%, which corresponds, according to Table 1, to an Exposure E=4. • Output item: as output of this block, the initial four columns of Table 2 enumerate and present the 11 contributing factors described by the CENIPA, and show the respective indexes A and E for these factors, which are the inputs to the SCHELL indexes block, described by the other columns of Table 2. The SCHELL indexes block is presented in step 3. Step 2 – the risk factors block The IMFLAR introduces a new concept in the certification of an aeronautical system: generic unsafe conditions, observed during the certification process of major change that installs this system, are considered as parameters to be analyzed, which may cause future risk events, after installation and during the operation of this system.

11Contributing Contributing factors factors for accidents accidents 22Risk Riskfactors factors Aeronautical Aeronauticalproduct product

Figure 7. Steps of IMFLAR.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.111-126, Jan.-Mar., 2013

33 Schell Schell indexes indexes

4 Artificial Artificial neural neural network network

5 Safety Safety parameters parameters

IMFLAR: An Intuitive Method for Logical Avionics Reliability

• Input item: inputs of this block are the unsafe conditions,

obtained from a sample of certification procedures relating base years of 2005 and 2009, for the TAWS system. The TAWS system was chosen because of the greater amount of processes for modifying aircraft relative to other systems. It was observed in the initial introduction of this system in Brazil (2005) and in a more advanced stage (2009), corresponding to a period in which the applicants for modification processes achieved a stage of maturation of the presented projects. The main risk factors detected for TAWS/GNSS systems are presented according to Table 3. • Data processing item: the same scaling process adopted for the block of contributing factors for accidents. For example, if the training factor corresponded to a Level A=100% of affected processes in the base year of 2005, this implies an Exposure E=5. • Output item: Table 4 presents the indexes A and E for risk factors as inputs to the SCHELL indexes block. As the output of this block, the initial three columns of Table 4 enumerate the 11 contributing factors, as described by Table 1. Relationship between Exposure (E) and Level or Range (A). E

A [%]

1 2 3 4 5

A≤20 20<A≤40 40<A≤60 60<A≤80 A<80

119

the CENIPA as fm (m = 1-11) and present the indexes A and E for these factors, which are inputs to the SCHELL indexes block, described by the other columns of Table 2. The SCHELL indexes block is presented below. Step 3 – The SCHELL indexes block • Inputs items: are the outputs of the previous two blocks,

namely, contributing factors for accidents and risk factors, corresponding to the Exposure E in Tables 2 and 4. • Data processing item: as observed on the CENIPA statistics and in other studies of accidents, the influence of human factors is predominant over other risk factors of an accident. Therefore, the exposures E of Tables 2 and 4 are combined by the SCHELL method of interfaces based on human factors. For this work, these interfaces are named SCHELL indexes. The presentation of the interfaces, in table format, uses a concept similar to m-SHEL model of Kawano (2002) adapted from SCHELL model. This model works, therefore, with six interfaces: LC, LS, LH, LS, LL, L, where L is the first influence of the human factor (pilot). The other factors of these interfaces are: LC – human /cultural interface, for example, supervision (air traffic control) or training; LS – human/procedure interface, for example, planning; LH – man/machine interface, for example, the application of controls; LE – human/environment interface, for example, environmental conditions; LL – man/man interface, for example, interaction between pilots; L – man, that is, the pilot and his own condition, for example, lack of experience of the pilot in the aircraft. These indexes are weighted as follows:

Table 2. Contributing factors to CFIT in general aviation (1999-2008) and SCHELL indexes. nn

Contributing factors

A [%]

Judgment (jud)

69.2

Planning (plan)

67.8

Psychological aspect (psy)

57.7

Indiscipline in flight (ind)

42.3

Adverse meteorological conditions (met)

33.8

Supervision (sup)

33.8

Other operational aspects (oper)

20.9

Poor flight experience in the aircraft (exp)

21.1

Environmental information (env)

18.3

n10

Application of commands (comm)

14.1

n11

Physiological aspects (phy)

11.3

nn = parameter number (n=1 to 11); A = Level; E = Exposure; LC, LS, LH, LE, LL and L = SCHELL indexes

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.111-126, Jan.-Mar., 2013

120

Silva, N. and Trabasso, L.G.

Table 3. Survey of risk factors for TAWS Systems in 2005 and 2009. Factor number

Element

Quality for analysis

Unit

factor number f1 f2 f3

Process Weather Antenna Processor

Total of processes Influence of weather during the year Installation in pressurized aircraft Software update

Display

Erroneous data

Training

Insufficient training

Total of processes [units] Percentage of cases [%] Percentage of cases [%] Unit per system [pcs./System] Percentage of cases [%] Unit per system [pcs./System] Percentage of cases [%] Need of training [units] -1 Percentage of cases [%]

Base year 2005 2009

313 2.75 <20 0 0 0 0 1 100

378 2.00 28.2 1 2 1/4 25

Source: ANAC/PST Data were obtained from two base years: 2005 with 313 cases sampled, and 2009 with 378 cases sampled, with the analysis of the following aspects: 1 – Weather factor: indicates the percentage of processes that were, somehow, affected in the base year due to meteorological conditions. These data were obtained, indirectly, from the public database of processes of the PST, through the records of cancellations of inspections/tests and comparison with the weather on the date/place of inspection/test. 2 - Antenna factor: indicates the percentage of processes that were, somehow, structurally affected due to the installation of antennas in pressurized aircraft. These data were obtained, indirectly, from the public database of processes of the PST, through the records of processes of the prototype aircraft, which were pressurized. 3 - Processor factor: indicates the quantity per unit of TAWS equipment for each system affected by software update (“load”). These data were obtained directly from the quantity of TAWS equipment to be installed in the process, which is affected by the software update (“load”), per manufacturer’s service bulletin. 4 - Display factor: indicates the quantity per unit of TAWS equipment for each system affected by the presentation of erroneous data. These data were obtained directly from the quantity of TAWS equipment to be installed in the process, which is affected by the manufacturer’s service bulletin that corrects this erroneous data. 5 - Training factor: measures the variation of the learning curve regarding to the installation/operation of the system. This factor was obtained indirectly by the number of meetings proposed by the modification process applicants, where it was observed that the value of the number of meetings (f5) decreased, in absolute value, inversely proportional to the elapsed time (t) according to: f5 ≅ a | t-1 | where: f5 = number of meetings per year ≅ need of annual training; α = index for units comparison - where, to the research data: α = α [s. (units / year)]; t = elapsed time (it was considered an average time of four years between the second half of 2005 and the first half of 2009).

Table 4. Risk factors and SCHELL indexes. Factor per year (2005) A [%]

E LC LS LH LE LL

f1 Weather f2 Antenna f3 Processor f4 Display f5 Training

1 2 1 1 5

0 0 0 0 5

Factor per year (2009) A (%)

E LC LS LH LE LL

f1 Weather f2 Antenna f3 Processor f4 Display f5 Training

1 2 5 5 2

0 0 0 0 2

2.75 20 0 0 100 2.00 28.2 80 80 25

0 0 0 0 5 0 0 0 0 2

0 2 1 1 0 0 2 5 5 0

1 0 0 0 0 1 0 0 0 0

0 0 0 0 2

fm = factor number (m=1 to 5); A = Level; E = Exposure; LC, LS, LH, LE, LL and L = SCHELL indexes

(A) Each of the contributing factors of accidents in Table 2 is analyzed. For example, the factor n1 Judgment (Jud), E=4. For each factor, only one SCHELL index is chosen as predominant to avoid problems with the possible ambiguity of data in order to make the neural network grouping of data in a more efficient way. For the Jud factor, the L index was chosen. Then, the same value of E for L is assigned, therefore, L=4. For other factors, the weight is zero, so LC=LS=LH=LE=LL=0. In this case, the same analysis is also repeated for the other factors. (B) Similarly, each risk factor in Table 4 is analyzed. For example, for the factor f1 Weather of 2005, E=1. For the Jud Factor, the LE index was chosen. Then, the same value of E is assigned to LE, therefore LE=1. The weight of other factors

are set to zero, i.e., LC=LS=LH=LL=L=0. In this case, the same analysis is also repeated for the other factors. (C) The predominant indexes in Table 2 are multiplied by the predominant indexes of Table 4, as shown in Fig. 8, in order to obtain Table 5. (D) For Table 5 preparation, according to the previous procedure, ten auxiliary tables are generated (five tables for five risk factors in 2005, and five tables for five risk factors in 2009). Each table has 11 rows, referring to the contributing factors, and 6 columns, related to SCHELL indexes, totalizing 66 elements. Most of the rows of the auxiliary tables contain only zeros, and only the 24 remaining lines contain significant elements that were grouped in Table 5. • Output item: the output of the SCHELL indexes block is composed of the significant factors that are presented to the neural network SOM of the artificial neural network block. After the development of Table 5, it is verified that only 24 factors are considered significant. These significant factors, in a practical approach, represent the Vulnerability of the system. Vulnerability is defined, in this paper, as a potential factor in exposing the analyzed system to the risk of an aircraft incident or accident. Therefore, the function of the SCHELL indexes block is similar to a Vulnerability detector, according to this definition.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.111-126, Jan.-Mar., 2013

IMFLAR: An Intuitive Method for Logical Avionics Reliability

Table 5. Significant factors. LC

LS LH LE

0 0 0 0 0 0 0 0 0 0 0 0 0 0 10 0 0 0 0 0 0 4 0 0

0 0 0 0 0 0 0 0 0 0 0 20 0 0 0 0 0 0 8 0 0 0 0 0

0 0 0 0 2 2 1 5 1 5 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2 1 2 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Significant factors

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 20 0 15 15 0 10 5 8 0 6 6 0 4 2

5f1met 5f1env 9f1met 9f1env 5f2comm 9f2comm 5f3comm 9f3comm 5f4comm 9f4comm 5f5jud 5f5plan 5f5psy 5f5ind 5f5oper 5f5exp 5f5phy 9f5jud 9f5plan 9f5psy 9f5ind 9f5oper 9f5exp 9f5phy

Number

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

The 24 parameters in Table 5 are designed as the example below: 5f1met =

Relationship between the risk factor (f1) from 2005 and contributing factor n5 5 = base year of data collection (2005) f1 = risk factor Weather (refer to Table 4) met = contributing factor to CFIT “n5” - Adverse meteorological conditions (refer to Table 2)

Table 2 nn Contributing factors n5 Adverse meteorological conditions (met) Table 4

Factors per year (2005) f1 Weather

LE 2 LE 1

Table 5 Significant factors LE 5f1met 2 Equal

Figure 8. Example of Table 5 elaboration.

Plus

121

Step 4 – artificial neural network block • Input item: inputs of this block are the significant factors yielded from SCHELL indexes block output and correspond to the input data of the SOM neural network. • Data processing item: in this block, the SOM tool was applied, a neural network which classifies significant factors. For this, the SOM_TOOLBOX_2 was used. This toolbox was created by the Laboratory of Computer and Information Science (CIS) of Helsinki University of Technology for MATLAB®. The SOM_TOOLBOX_2 groups and sorts the data into a graphical format, denominated Self-Organizing Map (SOM). This toolbox consists of a series of mathematical functions and graphs, which are elaborated using the MATLAB® functions. The SOM_TOOLBOX_2 has some specific program instructions for data collection, training the network and displaying the data. (A) The input function item: the reading of signals is performed by the instruction “som_read_data”. This function recognizes only files with extension “.data” created by the program SOM_PACK (also developed by Helsinki University of Technology), written in C language, which has no compiler for the Windows® platform. Thus, for use in the MATLAB® running under the Windows®, the input data must be converted to a text file in ASCII format, as shown in Fig. 9: (B) The combination and update functions item: once data is acquired, the SOM network is created, initialized and trained by the command “som_make” which performs combination and update functions. (C) The output function item: data grouped by the SOM are displayed through the “som_show” function in various graphical presentations, for example, by U-matrix or cellular maps, as shown in Fig. 10. • Output item: the output of this block is the data presented in map format by the function “som_show.” Data in Table 5 are grouped by the SOM tool, as shown in Fig. 10. This Figure also shows the corresponding U-matrix. Significant factors are, thus, grouped and mapped by the SOM neural network and these groups are named, in this work, as safety parameters. Therefore, the safety parameters are the critical design points that can imply in failures of the system under study and, in practice, are equivalent to the grouping of the system’s Vulnerabilities.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.111-126, Jan.-Mar., 2013

122

Silva, N. and Trabasso, L.G.

Step 5 – the safety parameters • Input item: data grouped by SOM neural network. • Data processing and output item: significant factors grouped by neural network are organized in Table 6. This grouping is essential to an engineering analysis of these parameters, because the significant factors are grouped into common areas. At the same time, the study by areas decreases the number of parameters to be examined. This reduces the cost of man-hour analysis of which parameters significantly affect project’s safety. The procedures for elaboration of simulation model on Simulink/MATLAB®

SL CL HL EL LL

1. The determination of the risk variation per time Figure 11 shows pairs of significant factors from Table 5 used as the boundaries of variation of the involved risk. This risk represents the workload for pilot, actions for operational/certification team from aviation authority, verifications for continuous airworthiness and periodic maintenance tasks, or other parameters to be investigated as appropriate. Each IMFLAR Group B, P, O, C and E is representative of the involved risk for the analyzed system (GPS/TAWS). The risk is composed by the sum of parcels of IMFLAR Group elements and each IMFLAR Group adds a combination of the Eq. 6 and Eq. 7 as applicable. These equations are fitting the values of the T, in order to combine them with the maximum values of each significant factor from Table 5.

0.0 0.0

0.0

0.0 0.0 0.0 P1

un=anT+bn

(6)

0.0 0.0

0.0

0.0 0.0

vn=cnT+dn

(7)

...

... ...

0.0 Pn

Figure 9. Example of SCHELL significant factors (P1 to Pn, n=1 to 24) for the analysis of the SOM network.

Labels

U-matrix

6.7

3.6

0.4

Figure 10. SOM Map and U-matrix.

where: T = time variation; an; bn; cn; and dn = adjustable values, in order to fit the equation with the extremes values of significant factors from the Table 5; and un, vn = risk variation per time (un and vn are different functions which variations are related with the IMFLAR Group as demonstrated as follow). Example for calculation of an and bn for u7: From Eq. 6: un=anT+bn ⇒ u7=a7T+b7; From Table 5: t=1 (2005) ⇒ T7 = (t/100) = (1/100) and u7=5f5comm =10; t=5 (2009) ⇒ T7 = (5/100) = (1/20) and u7=9f5comm=4. The values of (T7, u7)2005 and (T7, u7)2009 are used to solve the equation u 7=a7T+b7, in order to obtain the values of a 7 and b7. The variable T shows the behavior of involved risk (represented by IMFLAR Groups) per elapsed time. Thus,

Table 6. Safety parameters and IMFLAR groups. Factors grouped by SOM neural network

5f5plan, 9f5plan 5f2comm, 9f2comm, 5f3comm, 9f3comm, 5f4comm, 9f4comm 5f1met, 5f1env, 9f1met, 9f1env 5f5oper, 9f5oper 5f5jud, 5f5psy, 5f5ind, 5f5exp, 5f5phy, 9f5jud, 9f5plan, 9f5psy, 9f5ind, 9f5exp, 9f5phy

IMFLAR Group

Safety parameter

Symbol

PLAN COMM MET, ENV OPER PHY, PSY, JUD, IND, EXP

Planning Commands Environment Operational Behavior

P C E O B

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.111-126, Jan.-Mar., 2013

IMFLAR: An Intuitive Method for Logical Avionics Reliability

20 15 10

5f1met 9f1met 5f1efv 9f1efv 5f2comm 9f2comm 5f3comm 9f3comm 5f4comm 9f4comm 5f5oper 9f5oper 5f5plan 9f2plan 5f5jud 9f5jud 5f5psy 9f5psy 5f5ind 9f5ind 5f5exp 9f5exp 5f5phy 9f5phy

Figure 11. Significant factors.

the involved risk has three different functions per time depending on the analyzed group. It is not dependent on human factors in Group E (the extremes are matching = constant function). It is composed by a constant function and an increasing function for group C and the risk is decreasing for groups B, P and O. (A) IMFLAR Groups B, P, and O have a directly relationship with human factors and the Training factor from Table 3 and, for this case, the Eq. 6 is fitted by

123

Eq. 8. This represents the “man” interactions or “human factors” influence. (B) For the IMFLAR Group C the involved risk are rising with time (this is an accumulative risk which represents the software loads, deterioration of components and others aspects related to the maintenance and continuous airworthiness of system). This function was considered linear for an initial approach. This represents the “machine” factor. (C) For IMFLAR Group E, the acquired values of met=2*env (constants for the collected data) are replaced by the average curve (f1=-0.548t+45.84) that represents the “seasonal variability of rainfall in the São José dos Campos airport fields between the years 1982 and 2009”, according Corrêa et al. (2010). The site of São José dos Campos was elected because it represents the local were the most certification tests were performed for GPS/TAWS systems related to this research. The function f1 were divided by 100. This is necessary for adjusting the magnitude of the variables to a reasonable value of IMFLAR Group E. This represents the “environment” factor. The same procedure for calculation of an and bn for u7 can be utilized for the determination of other un and vn values. These calculations are summarized in Tables 7 and 8.

Table 7. IMFLAR application. Function

T=100/=f5 with a=100 (100%)

un=anT+bn Group

Factor

phy

3/80

5/4

f5=T=100/t

exp

3/40

5/2

f5=T=100/t

ind

9/80

15/4

f5=T=100/t

psy

9/80

15/4

f5=T=100/t

Risk per time

jud

3/20

f5=T=100/t

B=u1+u2+u3+u4+u5

plan

3/20

f5=T=100/t

P=u6

oper

3/40

5/2

f5=T=100/t

O=u7

Function

vn=cnT+dn Group

T=t

Factor

f2comm

f2=1

f3comm

f3=t

Risk per time

f5comm

f4=t

C=v1+v2+v3=f2+ f3+f4=2t+1

f1met

-0.01168

0.9168

v4=2f1/100

f1env

-0.00584

0.4584

v5= f1/100 =(-0.548t+45.84)/100

Risk per time E=v4+v5=2f1+f1=3f1

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.111-126, Jan.-Mar., 2013

124

Silva, N. and Trabasso, L.G.

Table 8. Summarization of equations from Table 7. Function

Description

Function number

un=anT+bn

Fit function of T=100/t from the Eq. 8

(9)

vn=cnT+dn

Fit function of T=t

(10)

f1=-0.548t+45.84

Function for factor per year =f1

(11)

f2=1

Function for factor per year =f2

(12)

f3=t

Function for factor per year =f3

(13)

f4=t

Function for factor per year =f4

(14)

T for Eq. (9) → f5=T=100/t

Time variation (T) and function for factor per year =f5

(15)

T for Eq. (10) → T=t

Time variation (T)

(15)

B=u1+u2+u3+u4+u5

Risk per time for B group

(16)

P=u6

Risk per time for P group

(17)

O=u7

Risk per time for O group

(18)

C=v1+v2+v3=f2+ f3+f4=2t+1

Risk per time for C group

(19)

E=v4+v5=2f1+f1=3f1

Risk per time for E group

(20)

Figure 12. Simulink/MATLAB® model.

2. The elaboration of the Simulink/MATLAB® model for the IMFLAR Figure 12 presents a Simulink/MATLAB® model that represents the elements of Tables 2 to 8 for the IMFLAR. The equations that model the blocks are on Table 8 and the elements of this model are on Tables 2 to 8. This model translates the representation of IMFLAR parameters to a functional block diagram in order to simulate and foresee the variation of risk per time, including a separated analysis for “human factors”, “machine” and “environment” influences as previously explained.

Figure 13 presents a graphical representation of IMFLAR Groups variations and the resulting risk R for the IMFLAR. 3. A practical application of Simulink/MATLAB® model for IMFLAR With the elaborated model, it is possible to test an individual parameter variation. Figure 14 represents the behavior of R related to a variation (step) on the O group signal. This variation is implemented by the block arrangement illustrated in Figure 14 and it might characterize an operational change. An example of this change could be the introduction

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.111-126, Jan.-Mar., 2013

IMFLAR: An Intuitive Method for Logical Avionics Reliability

25 B P O C E R

20 15

Unsafety Factor

10 5 0

10 12 14 Time (years)

125

O R

20 15 10 5 0 0

4 6

8 10 12 14 16 18 20 Time (years)

Figure 13. Simulink/MATLAB® model – IMFLAR groups and risk outputs.

Figure 14. Simulink/MATLAB® model outputs – operational variation.

of new rules for aviation in order to comply with the phases for Performance-Based Navigation (PBN) implementation in Brazil. This is a versatility example of the proposed IMFLAR.

may be used to foresee safety parameters, vulnerable areas of functionalities in future systems or simulate an unsafe operational condition as described in this work. Likewise, on any computational method used for Safety Assessment, an engineering analysis of the data grouped by SOM is necessary, in order to avoid errors. However, as described in this article, the application of this method is simple, the graphical grouping is easy to understand, and the presented results allows one to conclude that this method can be employed, not only as a Safety Assessment tool, but as an application for a predictive design where unsafe conditions are described by equations that are translated into the blocks of the simulation model and any safety parameters variation can be tested. Finally, the proposed IMFLAR can be considered an upgrade of other established methods, such as the Risk Matrix and the SCHELL model, overcoming some limitations of these methods and proposing new approaches to Safety Assessment and predictive design, showing potential of using this method for aeronautical applications and other engineering areas.

ConclusIONS The growing of the General Aviation fleet and the development of new avionics systems demand needs to increase flight safety, which is aeronautical authorities’ responsibility in partnership with manufacturers, maintenance teams and other aviation professionals. To do this, it may be necessary to develop other Safety Assessment methods using more adequate approaches to new patterns. Thus, an artificial neural network can be a viable alternative, since its effectiveness is appropriately certified. The proposed IMFLAR, on the SOM artificial neural networks, enables simulating logical decisions and groupings, similar to the human brain. The grouping of unsafe parameters turns the application of this presented method into a practical tool for an avionics system installation project. The grouping of these parameters is also one of the main advantages of the proposed method over other project methods, mainly because the safety parameters grouped by SOM are general and do not have a direct relation to unsafe conditions for any specific equipment from any manufacturer, but these parameters relate common failures of any functionality. Thus, it can be used to verify the vulnerabilities of a functionality, such as TAWS, which can be commercially supplied by various equipment, or may be present in future systems to be designed. Thus, this method can also be predictive and

Acknowledgment and Disclaimer The authors would like to thank the support of experts from the National Civil Aviation Agency – ANAC and Aeronautics Institute of Technology – ITA, in preparing this article. Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the ANAC or ITA.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.111-126, Jan.-Mar., 2013

126

Silva, N. and Trabasso, L.G.

REFERENCES Aircraft Crashes Record Office (ACRO), 2011, “Various Statistics”, retrieved in 17 Aug. 2011, from http://www.baaa-acro.com/ Statistiques%20diverses.html. Aircraft Owners and Pilots Association (AOPA), 2010, “AOPA General Aviation Trends Report”, Fourth Quarter 2010, AOPA Online, retrieved in 17 Aug. 2011, from http://www.aopa.org/whatsnew/trend.html. Agência Nacional de Aviação Civil (ANAC), “Aircraft Modification – Brazilian STC (CHST)”, PST/GGCP/SAR/ANAC, retrieved in 17 Aug. 2011, from http://www2.anac.gov.br/certificacao/CHST/CHSTE.asp Agência Nacional de Aviação Civil (ANAC), “Control of Supplemental Type Certification Processes”, PST/GGCP/SAR/ANAC, retrieved in 17 Aug. 2011, from http://www.anac.gov.br/certificacao/PST/ index_pst.asp Associação Brasileira de Aviação Geral (ABAG), 2011, “Anuário Brasileiro de Aviação Geral – 2011”, Brazil. Bird, F.E. Jr. and Germain, G.L., 1969, “Damage Control”, American Management Association, New York, 1969.

Francisco, C.A.C., 2004, “Rede de Kohonen: Uma ferramenta no estudo das relações tróficas entre as espécies de peixes” (dissertação de mestrado), Programa de Pós-Graduação em Métodos Numéricos em Engenharia, área de Concentração em Programação Matemática, Universidade Federal do Paraná, PR, Brazil. Greenwell, W.S. and Knight, J.C., 2003, “Risk-Based Classification of Incidents”, 2003 Workshop on the Investigation and Reporting of Incidents and Accidents. Hawkins, H.F., 1987, “Human Factors in Flight”, Gower Technical Press Ltd. International Civil Aviation Organization (ICAO), 2005, “ICAO Accident Prevention Programme”, ANB – Air Navigation Bureau. International Civil Aviation Organization/Commercial Aviation Safety Team (ICAO/CAST), 2011, “CICTT Occurrence Category Definitions”. Itoh, H., Mitomo, N., Matsuoka, T. and Murohara, Y,, 2004, “An Extension of M-Shel Model for Analysis of Human Factors at Ship Operation”, 3rd International Conference on Collision and Grounding of Ships (ICCGS 2004), Izu, Japan.

Braga, P.L., 2006, “Reconhecimento de voz dependente de locutor utilizando Redes Neurais Artificiais”. Escola Politécnica de Pernambuco. Universidade de Pernambuco, PE, Brazil.

Kawano, R., 2002, “Medical Human Factor Topics”, Saga Medical School, Saga, Japan.

Burin, J.M., 2011, “Leveling off”, Aerosafetyworld, Flight Safety Foundation, Feb. 2011.

Keightley, A., 2004, “Human factors study guide”, Palmerston North, Massey University, 190.216, New Zealand.

Castro, F.C.C. and Castro.M., 2001, “Mapas de Kohonen (SelfOrganizing Map). Redes neurais artificiais”, dissertação de mestrado, Pontifícia Universidade Católica do Rio Grande do Sul (PUCRS), Cap. 7, RS, Brazil.

Kohonen, T., 1994, “Self Organizing Maps”, Springer Series in Information Sciences, Springer, Espoo, Finland.

Centro de Investigação e Prevenção de Acidentes Aeronáuticos (CENIPA), 2009, Instrução do Comando da Aeronáutica. ICA 3-2. “Programa de prevenção de acidentes aeronáuticos da aviação civil brasileira para 2009”. Corrêa, C.S., Villaron, M.A., Araujo, R.T. and Pinto, M.L.A., 2010, “Variabilidade sazonal da precipitação no aeroporto de São José dos Campos – SP entre os anos de 1982 e 2009”, IN: CONGRESSO BRASILEIRO DE METEOROLOGIA, XVI., 2010, Belém. Anais do XVI CBMET, Belém, PA, Sociedade Brasileira de Meteorologia (SBMET), 2010. retrieved in 12 Sep. 2012, from http://www.cbmet2010. com/anais/artigos/310_40078.pdf. Cox, A.L., 2008, “What’s Wrong with Risk Matrices?”. Risk Analysis, 28 (2), 497-512. Department of Defense (DOD), 2000, “Standard Practice for System Safety, MIL-STD-882D, USA. Edwards, E., 1972, “Man and machine: Systems for safety”, Proceedings of British Airline Pilots Associations Technical Symposium, British Airline Pilots Associations, London, pp. 21-36.

Kohonen, T. and Honkela, T., 2007, “Kohonen Network”, Scholarpedia, Vol. 2, No. 1, pp. 1568. Laurino, L.S., 2004, “Rede Neural para classificação de caracteres”, Instituto de Informática, PPGC – Programa de Pós-Graduação em Computação, CMP135 – Arquiteturas Especiais de Computadores, Universidade Federal do Rio Grande Do Sul, RS, Brazil. National Transportation Safety Board (NTSB), 2010, “Introduction of Glass Cockpit Avionics into Light Aircraft”, Safety Study, NTSB/S-01/10. Perry, M.J. and Perezgonzalez, J.D., 2010, “SCHELL model”, Page revision: 17, last edited: 22 Aug 2010, 04:55 GMT, AviationKnowledge, retrieved in 17 Aug 2011, from http://aviationknowledge.wikidot. com/aviation:schell-model. Reyes-Aldasoro C.C., 1998, “A Non-linear Decrease Rate to Optimise the Convergence of the Kohonen Neural Network Self-Organising Algorithm”, ROCC99, Acapulco, Mexico. SOM_TOOLBOX_2, Laboratory of Computer and Information Science (CIS) Helsinki University of Technology, retrieved in 17 Aug. 2011, from http://www.cis.hut.fi/projects/somtoolbox/ Simulink/MATLAB MathWorks, Inc.

V.7,

R14/,

1984-2004,

The

Everdij, M.H.C., Blom, H.A.P. and Kirwan, B., 2006, “Development of a structured database of safety methods”, 8th International Conference on Probabilistic Safety Assessment and management (PSAM8), New Orleans, Louisiana, USA.

Van Der Velde, R.L., 1995, “Historical Perspective, One Hundred Years of Flight Testing”, Agardograph Agard AG, - rta.nato.int.

Federal Aviation Administration (FAA), 2010, “Developing and Implementing an Air Carrier Continuing Analysis and Surveillance System”, Advisory Circular (AC) 120-79A, U.S. Department of Transportation.

Vesanto, J., Himberg J., Alhoniemi, E. and Parhankangas J., 2000, “SOM Toolbox for Matlab”, Technical Report A57, Helsinki University of Technology, April 2000, retrieved in 17 Aug. 2011, from http:// www.cis.hut.fi/projects/somtoolbox/

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.111-126, Jan.-Mar., 2013

Thesis abstracts This section presents the abstract of most recent Master or PhD thesis related to aerospace technology and management.

A Contribution for the Pre-galactic Universe Study Eduardo dos Santos Pereira Instituto Nacional de Pesquisas Espaciais; São José dos Campos/SP – Brazil pereira.somoza@gmail.com Thesis submited for PhD degree in Astrophysics at Instituto Nacional de Pesquisas Espaciais, São José dos Campos, São Paulo, Brazil, in 2012. Advisor: Doctor Oswaldo Duarte Miranda. Keywords: Cosmology, Structures formation, Gravitational waves, Black holes, Astrophysical computing. Abstract: We developed a model based on the “PressSchechter-like formalism”, which allows the derivation of the cosmic star formation rate (CSFR). In particular, our CSFR produces good agreement with observational data for redshifts z<6. Using the CSFR, we studied the stochastic backgrounds of gravitational waves produced by the collapse of stars to produce black holes as remnants. We have obtained, for some models, Signal to Noise ratios (S/N)>10 from the correlation of two advanced LIGO detectors (Advanced LIGO or LIGO III). Another possibility of gravitational wave detector, for the next decade, is the Einstein Telescope (ET), which will increase the Signal to Noise ratios studied in this work by a factor of 10. Thus, gravitational wave astronomy can contribute greatly to the study of pre-galactic universe, helping to reconstruct the history of star formation at high redshifts. We also studied the growth of supermassive black holes and their connection with the CSFR. In particular, from seeds of 1000 solar Mass (at a redshift of ~20), we reconstructed the evolution of the comoving density of black holes just using Soltan’s argument. In order to reproduce the quasar luminosity function (QLF), our model shows that the mean radiative efficiency of the accretion disks should be a function of redshift. An interesting result is that the duty-cycle of quasars is maximum within the redshift range 8.5 to 11, which is within the observational uncertainties associated with the reionization of the Universe. Thus, perhaps mini-quasars may have had an important role during the reionization of our universe. This thesis also produced a framework called PyGravWC for the study of cosmology and gravitational waves

of cosmological origin. This framework is available to the community if respected the term of the GNU-GPL.

Tack Study in Prepregs of Epoxy Resin Reinforced with Carbon Fiber Eduardo Gouveia Martins Romão Faculdade de Engenharia de Guaratinguetá; Guaratinguetá/ SP – Brazil romao_eduardo@yahoo.com.br Thesis submitted for Masters in Mechanical Engineering at Faculdade de Engenharia de Guaratinguetá, Guaratinguetá, São Paulo State, Brazil, 2012. Advisors: Drs. Michelle Leali Costa and Edson Cochieri Botelho Keywords: Tack, Prepreg, Mechanical testing, Thermal analysis, Polymeric composites. Abstract: Tests to prepreg tack evaluation are very important in the aeronautic field due to the previous knowledge of this property can result in great economy and safety in the manufacturing of aeronautical structural components. However, there is no standardization in literature of a specific test for this type material. It is often necessary to use standards applied to other types of products, which may lead to unreliable results. Besides the methodology for tests, the storage conditions could affect directly the condition of adhesion between layers and, thereby, the quality of manufactured parts made by prepreg. Then, this dissertation aims to develop a methodology for evaluating the tack in prepregs of epoxy matrix reinforced with carbon fibers building a device specially designed for this purpose. Furthermore, with this device and the proposed method, a study of the freezer storage time influence in the prepreg carbon fiber/epoxy resin was also performed. As auxiliary techniques for this study, the curing of the prepreg by differential scanning calorimetry (DSC), the glass transition temperature and elastic modulus by dynamic mechanical analysis (DMA) were evaluated. The device developed shown to be effective to measure the separation energy for the fresh prepreg as storage for one year.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.127-128, Jan.-Mar., 2013

128

Thesis abstracts

Global Instability Analysis of Compressible Flow Elmer Mateus Gennaro Escola de Engenharia de São Carlos; São Carlos/SP – Brazil elmer@sc.usp.br Thesis submitted for the degree of PhD in Mechanical Engineering at Universidade de São Paulo, São Carlos, São Paulo State, Brazil, 2012. Advisors: Doctors Marcello A. Faraco de Medeiros and Vassilis Theofilis Keywords: Hydrodynamic instability, BiGlobal instability analysis, Eigenvalue problem, Arnoldi Method, Compressible flow. Abstract: The investigation of linear instability mechanisms is essential for understanding the process of transition from laminar to turbulent flow. This thesis has presented an algo-rithm for the numerical solution of the compressible BiGlobal eigenvalue problem. This algorithm explores the sparsity of the matrices resulting from the spatial discretization of the eigenvalue problem in order to improve the performance in terms of both memory and CPU time over previous dense algebra solutions. Both methods of spectral collocation and finite difference spatial discretization have been implemented and a performance study has been carried out in order to determine the best practice for the efficient solution of a general physical problem with sparse matrix techniques. A combination of spectral collocation and finite differences can further improve the performance. The code developed was then applied in order to revisit and complete the parametric analyses on global instability of the compressible swept Hiemenz flow initiated by Theofilis et al. (2004) and neutral curves have been obtained by this flow as a function of the Mach number in the 0<Ma<1 range. The present numerical results fully confirm the asymptotic theory values presented by Theofilis et al. (2004). This work presented a complete parametric study of the instability properties of modal three dimensional disturbances in the subsonic range for the flow configuration at hand. Up to the subsonic maximum Mach number value studied, it was found that an increase in this parameter reduces the critical Reynolds number and the range of the unstable spanwise wavenumbers.

The Influence of Erosion and Wear on the Accretion and Adhesion of Ice for Nano Reinforced Polymeric Composites Used in Aeronautics Omid Gohardani Cranfield University, Cranfield, United Kingdom; omid.gohardani.cranfield@gmail.com Thesis submitted for the degree of Ph.D. in Aerospace Engineering, at Cranfield University, Cranfield, United Kingdom, 2011. Advisor: Dr. David W. Hammond Keywords: Aircraft icing, Liquid erosion, Rain erosion, Nanotube composites, Polymer matrix composites. Abstract: The usage of polymeric matrix composites in aerospace applications has been significantly prevalent based on their desired material characteristics, which include higher strength, lower weight and heat resistance. With current advancements in nanotechnology, carbon nanotube reinforced polymeric matrix composites may enhance the operational usage of these advanced materials even further. In this study, a set of novel aerospace material candidates are characterized based on their mechanical properties, resilience to liquid erosion, wettability, and ice adhesion. The experimental evaluations presented allow for a preliminary ranking of the polymeric matrix composites and assessment of the influence of reinforcing carbon nanotubes. The role of erosion in particular is highlighted from both a historical viewpoint and based on empirical results for static and dynamic wettability and ice adhesion. Discussion of different ranking systems and fractography arising as a consequence of liquid impact is further addressed in this study. It is found that the candidate samples exhibit different physical parameters,

but

nominally

similar

erosion

resilience,

despite the presence of the reinforcing carbon nanotubes. The wettability of the experimental materials and their ice adhesion characteristics are further shown to be influenced by the presence of carbon nanotubes and largely dependent upon degradation of the material surfaces.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.127-128, Jan.-Mar., 2013

INSTRUCTIONS TO AUTHORS (Revised in March, 2013)

SCOPE AND EDITORIAL POLICY

PAPER SUBMISSION

The Journal of Aerospace Technology and Management (JATM) is the official publication of the Departamento de Ciência e Tecnologia Aeroespacial (DCTA), in São José dos Campos, São Paulo State, Brazil. The journal is quarterly published (March, June, September, and December) and is devoted to research and management on different aspects of aerospace technologies. The authors are solely responsible for the contents of their contribution. It is assumed that they have the necessary authority for publication. When submitting the contribution, authors should classify it according to the area selected from the following topics:

The manuscript should be digitalized using a Microsoft Word (.doc) software program and submitted electronically in English. See the instructions at http://www.jatm.com.br/ ojs/index.php/jatm/about/submissions#onlineSubmissions. If there is any conflict of interest regarding the evaluation of the manuscript, the author must send a declaration indicating the reasons so that the review process occurs fairly. After submitting the manuscript, the corresponding author will receive an e-mail with the Term of Copyright Transfer, in which the author agrees to transfer copyrights to the DCTA in case of acceptance for publication, thus being forbidden any means of reproduction (printed or electronic) without previous authorization of the Editor in Chief. If the reproduction is allowed, it is mandatory to mention the JATM. The author also declares that the manuscript is an original paper, its content is not being considered for publication in other periodicals and that all co-authors participated satisfactorily in the paper elaboration as to make public the responsibility for its content. The declaration must be printed, signed by the main author, and sent back by mailing to the following address: Instituto de Aeronáutica e Espaço/ATTN: Helena Prado/ Praça Marechal Eduardo Gomes, 50 – Vila das Acácias – CEP: 12228-901 – São José dos Campos/SP, Brazil. Papers already presented at conferences will be accepted if they were not published in complete form in the Annals of the conference or if they are extended with additional results or new findings. These articles will be evaluated as the others. Articles from guest authors will be published after approval of one specialist associate editor. The JATM does not publish translated articles from other journals.

• Acoustics • Aerodynamics • Aerospace Meteorology • Applied Computation • Astrodynamics • Ceramic Materials • Circuitry • Composites • Computational Fluid Dynamics • Defense Systems • Eletromagnetic Compatibility • Energetic Materials • Fluid Dynamics and Turbulence • Guidance Navigation and Control • Management Systems • Metallic Materials • Photonics • Polymeric Materials • Processing of Aerospace Materials • Propulsion and Combustion • Radars and Tracking Systems

PEER REVIEW

• Robotics and Automation • Structures • Synthesis and Characterization of Aerospace Materials • Thermal Sciences • Vibration and Structural Dynamics.

Manuscripts will be reviewed by at least two expert consultants, members of the Editorial Committee or external evaluators (ad hoc referees) in double blind peer review mode,

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.129-132, Jan.-Mar., 2013

ensuring complete anonymity. In case of disagreement on the results of the evaluation, the manuscript will be forwarded to a third reviewer, and it will be accepted for publication only if two approvals are received. The evaluators can accept the manuscript in the form it was submitted, they can reject it or request revisions. The manuscript that requires revision will be sent to the author that is supposed to submit a new version and, in the case the author does not agree with the suggestions, it is necessary to send a “letter to editor”, explaining the reasons. The Editor will approve after verifying in the new version the adherence to the reviewers’ suggestions or will send to another evaluation round if the changes have not been sufficiently addressed. Accepted manuscripts can be edited to comply with the format of the journal, remove redundancies, and improve clarity and understanding without altering meaning. Authors are also strongly advised to use abbreviations sparingly whenever possible to avoid jargon and improve the readability of the manuscript. All abbreviations must be defined the first time that they are used. The edited text will be presented to authors for approval.

MANUSCRIPT CATEGORIES Editorial: Any researcher may write the editorial on the invitation of the Editor in Chief. Editorials should cover broad aspects of Aerospace Technology. Such manuscripts are not submitted to peer review. Review articles: These should cover subjects that are relevant to the scope of the journal. Authors should bear in mind that they are expected to have expertise in the reviewed field. The article may be of any length required for the concise presentation of the subject. Original papers: These articles should report results of the scientific research. The article may be of any length required for the concise presentation and discussion of the data, but succinct papers are favored in terms of impact as well as in readability. Communications: They should report previous results of ongoing research and should not exceed eight pages. Thesis abstracts: The journal welcomes recent Masters and PhD thesis abstracts for publication. Such contribution will not be submitted to peer review.

MANUSCRIPT STRUCTURE Whenever is possible, articles should include the following subsections, however articles from some areas should follow their usual format. Title and names of authors: The title should not contain abbreviations. All authors should be identified with full name, e-mail, institution to which they are related, city, state, and country. One of them should be indicated as the author for correspondence and his/her full address is required. Abstract: They are limited to 250 words and structured into objectives, methods, results, and conclusions. Citations or abbreviations (except internationally recognized abbreviations, such as weights, measures, and physical or chemical ones) are not permitted. Keywords: Three to six items that should be based on NASA Thesaurus volume 2 – Access Vocabulary. Introduction: It should set the purpose of the study, providing a brief summary (not a review) of previous relevant studies, and stating the new advances in the current investigation. The introduction should not include data or conclusions from the work being reported. A final sentence summarizing the novel finding to be presented is permissible. Methodology: The authors are free to use any structure in this section to fit the objectives of the work, they could also rename it (e.g. Numerical analysis, Case study, and so on), and in some cases it may be advisable to omit it. Clear and sufficient information to permit the study to be repeated by others should be briefly given. Standard techniques need only to be referenced. Previously published methods may be briefly described following the reference. Results: This section should be a concise account of the new information that was discovered, with the least personal judgment. Do not repeat in text all the data in the tables and illustrations, but briefly describe what these data comprise. Discussion: The discussion should include the significance of the new information and relevance of the new findings in light of existing knowledge. Only unavoidable citations should be included. Citations to review articles are not encouraged in this section. In some cases, it may be advisable to merge with the previous section (“Results and Discussion”). Acknowledgements: This section should be short, concise, and restricted to acknowledgements that are necessary. The financial support received for the elaboration of the manuscript must be declared in this item.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.129-132, Jan.-Mar., 2013

References: Acceptable references include journal articles, numbered papers, books, and submitted articles, if the journal is identified. References must be restricted to directly relevant published works, papers, or abstracts that have been accepted for publication. References from private communications, dissertations, thesis, published conference proceedings, and preprints from conferences should be avoided. Self-citation should be limited to a minimum. Authors are responsible for the accuracy and completeness of their references. References in text: The references should be mentioned in the text by giving the last name of the author(s) and the year of publication. Either use “Recent work (Smith and Farias, 1997)” or “Recently Smith and Farias (1997)”. With three or more names, use the form “Smith et al. (1997)”. If two or more references have the same identification, distinguish them by appending “a”, “b”, etc., to the year of publication. Standards should be cited in text by the acronym of entity followed by the number, and they do not need to appear in the reference list. Reference list: References should be listed in alphabetical order, according to the last name of the first author, at the end of the article. Only citations that appear in the text should be referenced. Unpublished papers, unless accepted for publication, should not be cited. Work that is accepted for publication should be referred to as “In press”. It is recommended that each reference contains the digital object identifier number (DOI). References retrieved from the Internet should be cited by the last name of the author(s) and the year of publication, or n.d., if not available, followed by the date of access. Some examples of references are as the following ones: Coimbra, A. L., 1978, “Lessons of Continuum Mechanics”, Ed. Edgard Blücher, São Paulo, Brazil, 428p. Alves, M.B. and Morais, A.M. F., 2009, “The management of knowledge and technologies in a Space Program”, Journal of Aerospace Technology and Management, Vol. 1, No. 2, pp. 265-272. doi:10.5028/jatm.2009.0102265272 Paek, S.K., Bae, J.S. and Lee, I., 2002, “Flutter Analysis of a Wraparound Fin Projectile Considering Rolling Motion,” Journal of Spacecraft and Rockets, Vol. 39, No. 1, pp. 66-72. Bae, J.S., Kim, D.K., Shih, W.H., Lee, I. and Kim, S.H., 2004, “Nonlinear Aeroelastic Analysis of a Deployable Missile Control Fin,” Journal of Spacecraft and Rockets, Vol. 41, No. 2, pp. 264-271.

Clark, J.A., 1986, “Private Communication”, University of Michigan, Ann Harbor. EMBRAPA, 1999, “Politics of R&D”, Retrieved in May 8, 2010, from http://www.embrapa.br/publicacoes/institucionais/ polPD.pdf. Silva, L.H.M., 1988, “New Integral Formulation for Problems in Mechanics” (In Portuguese), Ph.D. Thesis, Federal University of Santa Catarina, Florianópolis, S.C., Brazil, 223p. Sparrow, E.M., 1980a, “Forced Convection Heat Transfer in a Duct Having Spanwise-Periodic Rectangular Protuberances”, Numerical Heat Transfer, Vol. 3, pp. 149-167. Sparrow, E.M., 1980b, “Fluid-to-Fluid Conjugate Heat Transfer for a Vertical Pipe-Internal and External Natural Convection”, ASME Journal of Heat Transfer, Vol. 102, pp. 402-407. Tables: Tables should be constructed using the table feature in the word processor or using a spreadsheet program, such as Microsoft Excel. They should be numbered in order of appearance in the text, using Arabic numerals. Each table should have a title and an explanatory legend, if necessary. All tables must be referenced and mentioned in the text as “Table” and succinctly described in the text. Under no circumstances should a table repeat data that are presented in an illustration. Statistical measures of variation (i.e., standard deviation or standard error) should be identified, and decimal places in tabular data should be restricted to those with mathematical and statistical significance. Authors should take notice of the limitations set by the size and layout of the journal. Therefore, large tables should be avoided. Figures: All illustrations, line graphs, charts, schemes, photographs, and graphs should be referred as “Figure” and submitted with good definition. Number figures consecutively using Arabic numerals in order of appearance. References should be made in the text to each figure using the abbreviated form “Fig.”, except if they are mentioned in the beginning of the sentences. Captions should be descriptive and should allow the examination of the figures, without reference to text. The size of the figures (including frame) should be 8 cm (one column) or 17 cm (two columns) wide, with maximal height smaller than 22 cm. Equations: Type them on individual lines, identifying them by Arabic numerals enclosed in parenthesis. References should be made in the text to each equation using the abbreviated form “Eq.”, except in the beginning of the sentences, where the form “Equation” should be used.

J. Aerosp. Technol. Manag., São José dos Campos, Vol.5, No 1, pp.129-132, Jan.-Mar., 2013

PRODUÇÃO EDITORIAL

Uma empresa do Grupo ZP Rua Bela Cintra, 178, Cerqueira César São Paulo/SP - CEP 01415-000 Tel.: 55 11 2978-6686 www.zeppelini.com.br

JATM is supported by:

Journal of aerospace technology and management

JOURNAL OF AEROSPACE TECHNOLOGY AND MANAGEMENT Vol. 5 N. 1 Jan./Mar. 2013 ISSN 1984-9648 ISSN 2175-9146 (online)

www.jatm.com.br

V.5, n. 1, Jan./mar., 2013

Journal of Aerospace Technology and Management