CMC - Automaker Automotive Media special edition

Automaker

CMC Special Edition

CONTROLS, MEASUREMENT & CALIBRATION CONGRESS

Platinum Sponsor

Knowledge Partner: 速 Progress through Research

Manpower

- 600

Investments

- USD 60 Millions

Accreditations - ISO 9001,14001 & ISO / IEC - 17025

Core Areas Research & Development Engine Development Alternate Fuels Noise Vibration & Harshness (NVH) CAE / FEA / CFD Structural Dynamics / Fatigue Automotive Electronics Material Testing

Homologation Vehicle Evaluation Emission Certification Active and Passive Safety EMI / EMC

速 Progress through Research

The Automotive Research Association of India

An ISO 9001, ISO 14001, OHSAS & NABL Accredited R & D Institute, with the Ministry of Heavy Industries & Public Enterprises, Govt. of India

S. No. 102, Vetal Hill, Off Paud Road, Kothrud, Pune 411038 (India) P. B. No. 832, Pune - 411004 Phone : +91-20 - 30231111, 30231101 Fax: +91-20 - 30231104 Email : director@araiindia.com

Website : www.araiindia.com

WELCOME

Automaker Editorial

William Fone Editorial Director Wf@automaker.tv Alessandro Bassinelli Contribuitor

CONTROLS, MEASUREMENT & CALIBRATION

Marcos Cardoso Lima Costa Contribuitor

CONGRESS

Dr. Gotthard Rainer Technical Chairman

conteNT

Send Your Article Wf@automaker.tv

AGENDA CMC Congress

Subscribe to our Magazine www.automaker.com.br

Group

SPEAKERS CMC Congress

www.automaker.tv

AVL List GmbH, Austria

Automakermedia

Changing Calibration Paradigms: Innovative Ways To Increase xCU Calibration quality

Thomas Fone CEO

Tata Motors Ltd

Global Modelling Technique’ to Model Multiple Engine Variants for M&HCV BSIII Application

Mauricio de Araujo Almeida Director

Ashok Leyland Ltd, India

Doris Fone Financial Director

Virtual and Experimental Optimization of In-Cylinder Residuals with Modified Cam using Gas Exchange Analysis

4 6 10 18 23

Knowledge Partner:

® Progress through Research

The Automotive Research Association of India’ (ARAI), established in 1966, is a leading automotive R&D organisation set up by the automotive industry with the Government of India. ARAI is affiliated to the Ministry of Heavy Industries and Public Enterprises and is recognized by the Department of Scientific and Industrial Research.

An ISO 9001 certified organization, ARAI is accredited by National Accreditation Board for Testing & Calibration Laboratories (NABL). ARAI is also accredited for ISO 14001 and OHSAS 18001.

The state-of-the-art Laboratories of ARAI are well equipped with the most advanced facilities in the areas of Emission Evaluation; Noise, Vibration & Harshness (NVH); Computer Aided Engineering (CAE); Safety; Engine Development; Structural Dynamics; Vehicle Evaluation; Material Evaluation; Calibration and Automotive Electronics. An experienced and well-trained human resource of six hundred plus is ARAI’s main strength.

ARAI offers R&D services in the areas of engine development; alternate fuels; NVH – Noise, Vibration & Harshness; computer aided engineering; structural dynamics; automotive electronics and materials.

ARAI offers comprehensive certification and homologation services for entire range of automotive vehicles, systems and components, CEV’s and generators. It covers the areas of vehicle evaluation, emission, safety, materials, EMI/EMC etc. ARAI also helps vehicle manufacturers for export homologation.

ARAI assists the Government of India in formulation of automotive industry standards and harmonisation of regulations.

ARAI offers Post Graduate and Doctoral Programmes with reputed International Institutes in the field of Advanced Automotive Technology.

With its pursuit towards advancement of automotive technology to create safer, cleaner and affordable mobility solutions, ARAI envisions to become a global automotive R&D Organisation by providing R&D services to the automotive industry world over and be the preferred choice of customers.

SPEAKERS Dr. Gotthard Philipp Rainer Vice President, AVL List

Since Joining AVL in 1978 as a Finite Element Specialist, Dr. Rainer successful career developed by first becoming responsible for CAD, CAE and Mechanical Testing in various managerial roles, then in 1996 he was appointed Vice President of AVL list Gmbh, where he is the Head of the “Advanced Simulation Technologies” business unit. Prior to joining AVL he was an Assistant Professor at TH Darmstadt, and later whilst at AVL he also continued lecturing at the Johannes Kepler University in Linz, and Technical University of Vienna. (1992 – 1997). From 2000 Dr Rainer has been active on a number of organisational and chairman roles at conferences, including Virtual Engine (Germany) Virtual Powertrain, Germany, India, Brazil and China, Controls Measurement and Calibration (Brazil and India). And Engineeringdesign.in (Bangalore). He has published More than 60 Papers in the field of Numerical Simulation, Virtual Development, CAD, CAxIntegration, Workflow, etc.

Additional Information Keynote lectures at various conferences Member of VDI (German Society of Engineers) 1993 Best Technical Paper, MacNeal Schwendler, World Users Conference, Washington) Award for paper “Simulation and Testing Methods to Master NVH and Sound Design”; SIAT Conference, 21st-24th January 2009, Pune, India

Mr Andrew Freeman Director Cummins INC

Since Joining Cummins in 2003, Andrew has held roles at, MAN Diesel & Turbo, Caterpillar Inc, & Perkins Engines. Today Andrew is a Director at Cummins Inc and Leader of a global technical centre of over 500 professional engineers supporting new product development and current product tailoring. Responsible for leading Technical Functional Excellence, Technical Quality and Technical Operations for the India Area Businees Organisation. Additional responsibility as corporate Design Functional Excellence leader allows him to practice and develop influential leadership and mentoring skills across the company. His specialties: Building teams, process development, influential leadership, mentoring, thinking outside the box, empathising.

Mr, Neelkanth V. Marathe Senior Deputy Director, ARAI

Mr Neelkanth V. Marathe is a graduate in mechanical engineering from University of JabalpurIndia. He started his professional career with Automotive Research Association of India (ARAI), Pune. Moving through several responsible positions, he presently heads the Power Train Engineering division and is responsible for engine design and development projects.

With more than 30 years of experience, Mr. Marathe has led several major engine R&D projects, many of them have been commercialized successfully.

His individual strength and expertise lies in cylinder head & port design, crank-train design, valve-train design, diesel engine emission reduction, etc for I C Engines used in variety of applications.

Mr Marathe and his team has recently designed a state-of-art, 3 cylinder CRDI diesel engine with a high specific power of 75 kW per liter for Passenger Car application. A range of spin-off designs are available for other ratings and applications.

Mr Marathe has authored several technical papers in International and National conferences. He is an active member of “Standing Committee on Emissions for Genset Diesel engines”. He is also Chairman of the BIS TED-2 committee for “Prime-movers and Transmissions”.

Dr. P.A. Lakshminarayanan

Chief Technical Officer, Simpson & Co P. A. Lakshminarayanan studied at Indian Institute of Technology, Madras for his B. Tech., M.S. and Ph.D. degrees. He worked at Loughborough University of Technology and Kirloskar Oil Engines Ltd. for five and 20 years respectively. In 2002, he was called to head the Engine R&D in Ashok Leyland. In 2011, he became Chief Technical Officer at Simpson and Co. Ltd. With his teams, he has developed more than eight diesel and CNG engine platforms and 150 types of engines commercially successful for the efficiency and cost effectiveness. Two engines received prizes from the Institute of Directors (India). He has authored 50 research papers in journals and conferences of international repute.

Four of them received the prizes for integrity and quality of the contents from the SAE (intl.), Combustion Society (India), AVL (Graz) and AVL (Pune) in 1983, 1993, 2005 and 2010 respectively. He has co-authored a book on “Modelling Diesel Combustion” published by Springer Verlag (2010). His next book “Critical Component Wear of Parts in heavy Duty Engines” was published in 2011 by John Wiley International. He is a fellow of SAE and a fellow of Indian National Academy of Engineering.

SPEAKERS Mr. Prabhu Santiago Project Manager, AVL List

Since 2008, Prabhu has been a project manager at AVL, specialising in Engine Calibration, Methodologies and Tools. He graduated at IIT Madress (Engine Technology), and in 2006 Joined Delphi, as a Diesel calibration Engineer.

Mr. Kishore LM Manager - Research and Development, Mahindra & Mahindra.

In his current role, he is responsible for E/E Architecture validation for all M&M vehicles Mr.Kishore has about 12 years of experience in the field of automotive E/E architecture design & validation.

He holds a Bachelors Degree in Electrical Engineering from Madras University and Masters Degree in Software Systems from BITS, Pilani. He has few patents & publications in International/National conferences to his credit.

He has exposure to identify challenges and improvement areas in a complex and demanding environment like OEM’s and further strategize to improve by defining road maps, establishing infrastructures, building capability and competence of teams as well as enabling them to perform at peak levels.

He has also conceptualized improvement initiatives for meeting the ever growing complexities of the E/E architecture and demands by introducing processes and methods, adjusting them to the needs of an OEM as well as defining synergies with varied flavour of stake holders of the E/E system.

Able and energetic to take up complex projects, handle them with sheer commitment and deliver with visible or measurable results.

Dr Manoj Kumar AP

Research Scholar, National Institute of Technology Karnataka Dr Kumar, as a research Scholar at the esteemed National Institute of Technology has published a number of works including: Numerical Study on Selective catalytic reduction system to determine DeNOx efficiency in Diesel Engine,IJTARME,ISSN 2319-3182,vol1,Issue1,pp102-106,2012 Effect of Physical Parameters on DeNOx Conversion in Selective Catalytic Converter Used in Diesel Vehicles, Applied Mechanics and Materials Vol. 376 (2013), pp 13-16 The Effect of Selective Catalytic Reduction System on DeNOx Efficiency of Single Cylinder Diesel Engine, Springer, Journal of Mechanical Science and Technology, Under review The Experimental and Simulation Study of Selective Catalytic Reduction System in a Single Cylinder Diesel Engine Using NH3 as a Reducing Agent ,International Journal of Chemical Engineering Volume 2014 (2014), Article ID 350185.

Experimental and CFD analysis of Selective Catalytic Reduction System on DeNOx Efficiency of Single Cylinder Diesel Engine Using NH3 as a Reducing Agent, Procedia Technology, Elsevier, Accepted

Mr John Henry Kwee

Deputy Chief Engineer & Technical Specialist, FEV John prior to joining FEV India, began work in 2007 at the Institute for Combustion Engines, Aachen University (Germany), as research and project engineer for diesel exhaust aftertreatement , In 2012 he joined FEV GmbH, Aachen, as technical specialist for Commercial, Industrial and Large Engines, and since 2013 he has held his current role as Deputy Chief Engineer at FEV India

Mr. Josko Balic

Product Manager, AVL List Has been a member of AVL software department for over 15 years, focusing in his work on vehicle system simulation, software interfacing methods and software integration in cloud environment and real-time hardware systems.

In the research role Mr. Balic has been contributing to several European Union partnership programs involving AVL, Technical University of Graz and some of the leading OEMs. As a product manager, Mr. Balic is responsible for the further development of vehicle system simulation and software backbone tools including the products such as AVL CRUISE, AVL BOOST Realtime and AVL CRUISE M. Mr. Balic strongly believes that the future of software applications in the automotive industry lies in a closer integration of the software tool chain with the control functions development and hardware tests.

SPEAKERS Mr. J. Balaji

Manager Engines (Product Development), Ashok Leyland M.Tech graduate from IIT Madras specialized in “Engine Technology”. Having vast experience on engine testing and combustion development domain. Worked on conventional mechanical FIE engine to modern CRS engine with EGR and SCR technologies. In-depth knowledge on calibration tools, optimization tools, testing equipment’s and regulations.

Mr. Balaji Bandarum Manager, Ashok Leyland

I have graduate in Thermal Engineering with specialization in Thermodynamics & Combustion Engineering from Indian Institute of Technology – Madras. Currently, working as Manager in Engine - R&D, Ashok Leyland Technical Centre, Chennai. I have 4 years of experience on engine calibration and optimization. My areas of interest are combustion modelling, alternate fuels and fuel cells.

Mr. Ranjithkumar TR

Manager of Vehicle Calibration, Maruti Suzuki.

Since Completing Automobile Engineering in 2004, Ranjithkumar became a Engineer at Bosch Limited, Bangalore, before his current position as a manager at Maruti Suzuki India Limited, Gurgaon. He has 7 years of experience as an engine and vehicle calibrator worked on variety of SI engine projects.

Mr. Seetareddy Beeravalli

Deputy Manager, Continental Automotive Pursued Bachelor of Technology from Nagarjuna University, Andhra Pradesh in the year 2004. Did Diploma in Embedded systems design at CDAC in 2005 on special intreset. Then Joined Siemens information systems limited in year 2005 in Engine systems business unit as developer. Merged into Continental Automovtive compenents India Pvt Ltd in 2009 as Part of Engine systems department working on Gas exchange subsystem which includes especially variable valve timing and valve lift, exhaust recirculation and air path.

Mr. Shivaram Kamat

Domain Consultant, TATA Consultancy Services Received his Bachelors Degree in Electrical Engineering from Walchand College of Engineering, Sangli, India and the Masters Degree in Power Electronics (EE) from Indian Institute of Technology, Delhi, India, in 1987 and 1990, respectively. He worked for electrical-electronics manufacturing industry for 9 years in areas of Electrical Power system Protection, Generation Excitation and Control, X-Ray generators as design engineer. He has been working with Engineering & Industrial Services division of TATA Consultancy Services Limited, Pune, INDIA since 2000. His Current Role is Domain Consultant for Automotive Engineering projects. In Tata Consultancy Services Limited he has been working in the areas of Advanced Process Control, Automotive Control and Diagnostics, Model Predictive Control, Machine Learning, Data Analysis and Model Based Development. He has multiple publications in conferences and journals in these areas. He is Senior Member of IEEE since 2011 and contributes as reviewer for IEEE Transactions on Neural Networks and Learning Systems.

His research interests include modelling, optimization and control of various components in HEV, EV, Power Train and Energy Storage Systems.

Mr. Anuroopa Varsha

Manager Development, TATA motors

Completed his BE in Mechanical Engineering from Visvesvaraya Technological University, Belgaum and Masters in Thermal and Fluids Engineering from IIT Bombay. He currently works as Manager (Development) at ERC Engines, Tata Motors and has 4+ years of experience in Base Engine & Vehicle Calibration for Medium & Heavy Duty Commercial Vehicle Applications.

During his current stint in Tata Motors, he has acted as a liaison between the heavy commercial platform and various technical partners like Bosch, AVL, Shell and others. He has a keen interest in technology which he has been applying to make engine optimisation simpler and faster. His current test facility in TML fancies Real Time Controllers for CAMEO along with iLinkRT interface.

SPEAKERS Ms. Shruthi Ananthachar

Technical Sales Engineer, ETAS India Holds an MS in Software Science from BITS, Pilani and BE in Electronics and Communication from VTU. She has over 10yrs of experience in the Automotive Industry with expertise in Software development and project management.

Mr. Nithin Nath

Regional Product Manager, ETAS India Holds a BTech degree in Mechanical Engineering from College of Engineering Trivandrum.

He comes with 12yrs of experience in Automotive Industry and expertise in the field of Engine Management system calibration and project management.

Mr Chinmay Misra

Technical Marketing Engineer, National Instruments

A Certified LabVIEW Developer, and has been working as a Technical Marketing Engineer at National Instruments since July, 2013. He has served previously as an Applications Engineer and Staff Applications Engineer at National Instruments during the period July 2011 - July 2013.

Mr Chinmay is a graduate in Mechanical Engineering from IIT Kharagpur. An Institute Gold Medalist, he has been an Undergraduate Researcher at the Microfluidics Laboratory, IIT Kharagpur and a Research Assistant at Institute for Integrated Energy Systems, University of Victoria, Canada.

Mr. R Vijayalayan

Senior Team Leader, Mathworks

A Senior Team Lead for the Control Design Application Engineering in MathWorks India and specializes in the field of Control Design and Automation. He closely interacts with customers in different domains to help them use MathWorks products for Physical modeling and Control Design.

Prior to joining MathWorks, Vijayalayan was a Team Lead in the Embedded Systems Group of Cranes Software International Limited, where he gained hands-on experience in setting-up Rapid Control Prototyping and Hardware-In-Loop Simulation environments for automotive customers. He has also worked as a Scientist B in the Gas Turbine Research Establishment. Vijayalayan holds a Bachelor’s Degree from Manonmaniam Sundaranar University and a Master’s Degree in Control Guidance and Instrumentation from IIT Madras.

Mr. Sanjeev Bedekar

Thermal & Science Functional Excellence leader - CRTI, Cummins INC

Sanjeev holds a Bachelor’s Degree in Automobile Engineering from Pune University and a Master’s Degree in “Heat Transfer & Thermal Power engineering” from IIT Madras and having around 18 years of industry experience. I have been associated with Mahindra tractor R & D team (FES Mumbai) at start of my carrier working in RnD department for Engine design, CFD simulations & cooling system Design , Optimization groups. During his tenure at Mahindra Tractor, he worked closely for new engine development program with consultancy support from AVL, Graz. After that, he been associated with “Finite to Infinite” having an experience as partner in business for FEA/Thermal domain as Private Consultancy. He been co-author of a book “Practical Finite Element analysis” Later to this, he had diversified his profile in HVAC & Crash analysis simulation through association with organization Tata Faurecia. Thereafter, he joined Mahindra Engineering services working closely with Navistar USA Engine CFD Analysis team for engine CFD simulation. Before joining CRTI, he worked with Altair Engineering Pune as a Technical Manager for CFD Solver in technical support role for India & Asian market. At CRTI, he is in role as “ TSFE (Thermal Science Functional Excellence ) Leader” for supporting & enhancing Functional Excellence in different Cummins business units.

Mrs. Ujjwala Karle

Manager - Automotive Electronics Road Map and XEV Powertrain Roadmap, ARAI - Automotive Research Association of India

A graduate engineer in electronics with 20 years of experience in the field of Automotive Electric & Electronics controls development, evaluation & implementation. Worked on innovative &indigenous control implementation &solutions such as cost effective FI for small vehicles, India OBD implementation for an OE, Duel Fuel retrofit, ESP for SUV segment, Series hybrid demo vehicle program. Expertise in ECU verification validation, Calibration in simulated environment & actual engine/ vehicle more specific to powertrain & chassis control. Part of core technology roadmap team in ARAI & is responsible for Automotive electronics road map & XEV powertrain roadmap. Part of Indian EV standard preparation Working on advance engine controls, hybrid vehicle sub systems simulation, development &integration etc. More than 20 technical papers and couple of Indian patents.

Changing Calibration Paradigms: Innovative Ways To Increase xCU Calibration quality Dr. Nikolaus Keuth, Prabhu Santiago, Ignacy Santiago, Dr. Michael Kordon, Dr. Johann C. Wurzenberger AVL List GmbH, Austria

N

ew legislation in the automotive industry such as as CO2 limitation, “Real Driving Emissions” (RDE) or “In-Use Tests” have increased the complexity of the development of internal combustion engines and drivetrains. Passenger cars, commercial vehicles and non-road applications must comply with emissions legislation under diverse operating and environmental conditions. This situation is further exacerbated by the diversity of the vehicle portfolio.

The effort required to achieve an adequate system and calibration validation is enormous.

Conventional development processes require the application of large amounts of manpower, test equipment and budget in order to reach the legislative targets while maintaining and keeping the same system reliability, particularly when considering that the majority of effort takes place on the right of the development process, see diagram 1, and is heavily dependent on existing engines and vehicle, often located very late in the project timeline. Links to earlier development stages and front-loading is often very difficult due to the lack of models.

The limit of efficiency has been reached with conventional development processes; new approaches are necessary.

Diagram 1: V development process

MODELBASED CALIBRATION APPROACH

AVL List GmbH has set up a method for model-based development throughout drivetrain development as part of its “Smart Calibration” strategy. This approach is based on a powerful, real-time, semi-physical engine model in combination with a powertrain model and a virtual testbed with established automation and application tools.

As early as in the concept phase, the engine model is coupled with the vehicle model and the performance and emissions behavior under real driving conditions is simulated in a “Model-in-the-Loop” environment. This enables components such as emissions after treatment systems to be dimensioned more exactly, or before the first engine start, an emissions and performance pre-calibration to be done. The “Hardware-in-the-Loop” (HiL) testbed is calibrated and validated in the application phase. Modern dataset management systems enable a continuous monitoring of the calibration process, as well as a smooth transfer of datasets from the virtual to the real world, in order to ensure dataset quality. This method is the foundation of a new process and opens up new possibilities in series calibration.

10 |

Paper

Diagram 2: Workflow for model-based calibration

As can be seen in diagram 2, the models are constructed at a very early concept phase using component specifications and development targets. By employing semi-physical modeling approaches, this model, together with knowledge gained from calibration data from earlier

development stages, can be used to create first pre-calibrations for the ECU for the first engine start. The models are adapted and improved in steps using measurement data in order for them to be used for virtual component and system development. In a final step, the semi-physical models are extended

using DoE data so that they can reliably and accurately predict even emissions under a range of environmental conditions and can be employed for virtual calibration in a MiL/SiL/HiL environment. In order to be able to use such a workflow, a consistent model, modeling and testing landscape is necessary.

simulation. The engine model used in Smart Calibration is based on a semi-physical approach. By combining physical and empirical components, the complexity of the

model is reduced and the trade-off between computational speed, model accuracy and configuration effort is eliminated.

COMBUSTION ENGINE

An accurate virtual representation of the combustion engine is a pre-requisite for the efficient and meaningful application of a powertrain

Automaker CMC Special edition | 11

The zero-dimensional charge cycle is calculated physically by filling and emptying containers connected by throttle elements. The intake is modeled using a special approach that enables simplified configuration.

Further, the thermal inertias of the gas transporting components are simulated, as well as the inertia of the turbocharger, in order to be able to simulate realistic behavior during transient engine operation.

The combustion model consists of a network of submodels that follow the process of combustion chronologically. For computational speed reasons, a

crank angle based calculation of combustion and emissions is not considered. However, the most important characteristics of combustion, such as start of combustion, MFB50% and peak pressure are calculated.

Input values to the empirical parts of the combustion model that represent the highly complex combustion events are predominantly calculated physically. This makes it possible to cover a large part of the parameter space, provides the possibility to react to hardware changes and also results in a reduction in the number of input values required for

the empirical model.

The engine model is thus able to make quantitative predictions without the availability of measurement data of fuel consumption, temperatures and pressures, and NOx. If measurement values are available, a handful of parameters are available to adjust the model to reflect the available data. This increases the accuracy so that calibration tasks can be performed. Emissions such as soot, CO and hydrocarbons must currently still be modeled empirically out of the measurement data of the corresponding engine.

EXHAUST AFTERTREATMENT

The models for the exhaust system consist not only of the catalytic components and particle filter, but also the injection lines, injectors and sensors. The exact measurement of the thermal conditions in the exhaust system forms the basis of the calculation of emissions at the tailpipe. The application of models for the simulation of dynamic drive cycles brings with it the necessity of second-by-second calculation of emissions after each exhaust

system component.

Apart from the catalytic conversion, storage phenomena of emissions species are also considered. These are soot in the particle filter, NOx in the storage catalytic converter, or ammonia in the SCR catalyzer and also, for example, fluid urea-water solution or adsorbed hydrocarbons.

In order to guarantee an efficient workflow despite the large number of model parameters, it is necessary to have available a

coordinated methodology, from the selection of the necessary model characteristics, via the characteristic measurements, to parameter identification.

In the concept phase, the model parameters for the various components are accessed from a database that is continuously finetuned during the development project, in order to increase accuracy.

PHYSICAL BASED PLANT MODELING ON SYSTEM LEVEL The engine model used in this study is part of a broader system level simulation framework. Diagram 3 gives an overview of this framework with the distinction into the main physical domains of drivetrain, engine including combustion pollutant formation and exhaust after treatment, cooling and

lubrication. The individual domains on their own and also their interactions are controlled by various types of control functionality that interact with the plant model through a layer of actuators and sensors. The xCUs are typically controlled by a driver model that follows given driving profiles.

It turned out that the following list of requirements was essential for the MiL and HiL office environments in order to efficiently support engine development and calibration.

These requirements have been driving the development of the applied tool ever since.

These are: Multi-physical Consistency Scalability Customizability Openness and integrateability Engineering driven Multi-physical simulation tools are capable of comprising the different physical effects taking place in the different domains of drivetrain engine cooling etc. by dedicated models and numerical methods.

Especially the application of individually tailored integration rates together with conservative coupling techniques allows the running of numerically highly efficient models with real-time capability, as this is a key requirement to support HiL based calibration activities.

12 |

Paper

Consistent plant modeling is essential to support the development process from the early concept phase to the late function development and calibration phase. Model consistency enables the handshake between different development teams, each adding a focused value to a common model that accompanies the hardware development as virtual twin brother. Consistent handling of models and model parameters additionally allows a thorough tracking of model changes during its lifetime in the development process.

Scalable plant modeling in a consistent overall framework is the answer to requirements that also scale throughout the development process. On the one hand, highly predictive plant models are needed in the early concept phase and on the other hand highly accurate models are a must to support calibration. Both goals are addressed by a tailored combination of first order principles, physical, semi-physical and fully empirical approaches. Here it turned out that models featuring both a maximum of physical depth and running real-time are the best choice.

Diagram 3: Schematic of a system engineering plant model comprising the domains vehicle, engine and cooling

Customizable plant models are the key to efficiently support the dynamic advancing requirements in modern engine development programs. System simulation tools offer libraries of fundamental components that can be used to assemble entire vehicle, engine or cooling topologies. These components are typically parameterized only on the high level of the user-interface. Access to specific parts of the component source code on a fundamental programming level additionally allows modifying existing components or designing new ones. A comprehensive numerical solver basis together with a smart equation manager taking care of all fundamental balancing rules enables the calibration engineer to focus on the modeling of physical correlations.

Open and integrative simulation platforms target, amongst others, the fulfillment of the requirement of maintaining existing models built in different tools by typically enabling co-simulations. A second way of combining models is model-import where the equation of the “sub model” is integrated in time using the numerical solver of “main-model”. Here, the Functional Mockup Unit (FMU) standard is used to define interfaces.

This approach is chosen to embed customized cylinder and exhaust after treatment models into an air path network built out of library components. Besides openness in an office application, it is essential to port plant models to various hardware platforms for HiL and also test-bed applications.

Engineering driven plant model development can be seen as a generic requirement covering all previously mentioned functionalities. The application of tools (software and hardware) in daily engineering work allows the simple tailoring of them to the right engineering needs.

WORKFLOW-BASED CALIBRATION In order to be able to rollout the models in calibration, emphasis was placed during development of the methodology that the tool chain is also consistent for calibration engineers. Three distinctive groups were identified: the model creator, the model configurator and the calibration engineer. The three groups differ in their expertise and the way they use the models.

Automaker CMC Special edition | 13

Diagram 4: Workflow-based tool chain

As shown in diagram 4, the model creator creates the model out of powertrain components. The model configurator adapts the finished models to the individual application using a configuration workflow. To do this, the configurator can only access parameters and elements that the model creator has released. The models and parameters are distributed via a centralized storage area to the individual development environments.

Particular emphasis is placed in the HiL and SiL/MiL environments. The calibration engineer works in these virtual environments to perform a calibration of the xCU. The HiL environment is set up such that it is as compatible as possible with the downstream, environments, so that the tasks and tests can be transferred to them as simply as possible. In the SiL/MiL environments, emphasis has been placed in the workflow-based calibration approach that leads the calibration engineer through the workflow and thus solution-oriented, see diagram 5.

Diagram 5: Workflow-based calibration

APPLICATION AND RESULTS

The virtual testbed, whether HiL or MiL environment, is used for concept investigation, calibration and validation under the various conditions required by legislation. In particular, the robustness of the system can be ideally tested, related to component tolerance, ageing and different environmental conditions.

The broad and consequent use of powertrain models in development means a breakthrough in modelbased calibration. What has been considered impossible for many years is now possible.

14 |

Paper

CONCEPT PHASE

Simulation tools have been successfully applied to the development of new powertrain concepts for a long time.

However, up until now mainly steady state operating points under standard environmental conditions have been used for the evaluation of future concepts.

Due to the increasingly complex interrelationship between engine and exhaust aftertreatment system and the necessity to comply with legislative emissions limits under real world driving conditions, concepts must be evaluated within a broader scope. This is exactly the point where fast, dynamic

powertrain models are necessary and represent a useful extension to existing simulation tools. It is possible, in very early stages of development, to test the behavior of the combustion engine in the future vehicle under real driving conditions with various boundary conditions and also to estimate the effect of component ageing.

By observing all interactions between engine, exhaust aftertreatment, control unit, software and calibration, sensors and actuators, and the environmental conditions, concepts can be evaluated and optimized in all their entire

complexity.

The exact observation of the future application of a combustion engine in various vehicles makes it possible to cluster the application into individual groups as early as in the concept phase. By evaluating whether sections of a dataset from other applications can be reused, it is possible to calculate exactly what the calibration effort will be, thus creating a tight and optimal calibration phase without duplicated effort. By clustering the work well, the effect that the calibration effort increases directly proportionally to the number of vehicle variants can be avoided.

EMISSIONS AND FUEL CONSUMPTION UNDER REAL DRIVING CONDITIONS

Emissions legislation for passenger cars in the European market is still based on the certification of emissions and fuel consumption in the standard test cycle NEFC. This cycle is, apart from a few transitional phases, basically steady state and does not represent all real driving conditions of a vehicle. Conventional combustion engines are optimized such that they comply with emissions and fuel consumption limits during the legislative test cycle. Outside of the cycle, CO2 and driveability are optimized. Even the engine’s hardware is designed so that the emissions limits in the test cycle are complied with; large reserves for emissions reduction outside of the operating areas relevant to the cycle have not yet been built in.

In recent years, various institutes have measured the emissions and fuel consumption of passenger cars under real driving conditions, with sometimes shocking results. The European Commission is working on a draft that will end in an “In-Use Compliance” or “Random Certification Cycle” concept, in order to guarantee that the emissions are kept low throughout the entire engine map. Although the draft for the legislation has not been finalized, most manufacturers are optimizing their vehicle’s emissions over a large area of the engine map and under different environmental conditions.

The requirement of “Real Driving Emissions” in combination with the CO2 legislation and

the broad OEM vehicle portfolio are the drivers of even bigger challenges in the development process. This has experienced continually increasingly complexity over the last ten years, but is now, in some cases, on the limit of what is possible. The use of virtual testbeds can defuse this situation, since tests are independent of the currently prevailing environmental conditions and can be run on virtual test systems, thus allowing parts of the development to be relocated upstream to earlier stages. The use of virtual testbeds is thus a necessary extension to currently used test environments.

Up until now, emissions optimization was mainly limited to a steady state approach with a focus on the few transient conditions in the NEFC cycle. The optimization of the combustion engine for all driving and environmental conditions can no longer be achieved using this approach. In order to be able to keep quality and development costs at the same level under such legislative requirements, a paradigm shift is mandatory concerning the application and introduction of new test and simulation possibilities.

What is to be introduced for passenger cars in the future is already in force for commercial vehicles. Emissions limits must be complied with in the vehicle under real driving conditions, with different altitudes and temperatures – and this not just with new, but also with aged components.

To cope with this new challenge with high quality, parts of the calibration at AVL List are performed on a virtual test bed.

To do this, a virtual representation of the vehicle is created during the development phase: The vehicle specific engine installation and the drivetrain are taken into account and with them, the test bed development under standard conditions is supported as well as the calibration of correction and protection functions outside the standard environment conditions.

Virtual “In-use” tests are run, so that the cycle emissions can be calculated for the vehicle in real situations. The environmental conditions and the state of the components are varied, in order to simulate real operation with aged components. The complex evaluation of these simulated tests are done on the virtual test bed in the same manner as for a real vehicle using “AVL PEMS Post Processing”.

By pre-calibrating on a virtual testbed, the efficiency on real testbeds and in the vehicle can be increased considerably.

The work begins with datasets with a high degree of maturity and the in-vehicle tests are practically pure validation exercises. The following table shows a comparison between measurements and simulation on a commercial vehicle engine in the 10l class. The measurements were taken during a winter test at 2,200m.

Automaker CMC Special edition | 15

Table 1. Comparison between measurement and simulation Compressor exhaust Turbocharger speed Turbocharger speed low temperature low high pressure stage pressure stage pressure stage [°C] [Rpm] [Rpm] Simulation

84981

88002

131

Measurement

86000

80800

123

Deviation abs.

-1019

7202

8

Deviation [%]

-1.2%

8.2%

6.4%

Table 2. Comparison between measurement and simulation Compressor exhaust temp. High pressure stage

Exhaust temp. at high pressure turbine bank 1

Exhaust temp. at high pressure turbine bank 2

[°C]

[°C]

[°C]

Simulation

157

561

604

Measurement

156

541

582

Deviation abs.

1

20

22

Deviation [%]

0.4%

3.6%

3.6%

DATASET VALIDATION

One of the critical points in development is the validation of all vehicle variants under all of the different driving conditions, in order to guarantee that the development, I terms of quality and cost, is carried out under the same conditions as in the past. Compared to calibration, the role of validation is ever increasing.

Test equipment is expensive and the question arises as to whether simulation can replace or at least support real validation, in order to reduce test effort. This question requires an exact analysis of all current activities to be able to define where new methods could be inserted.

So far, the validation at the end of the calibration phase was a secondary activity with, compared to calibration, low effort. The new legislative requirements, however, make validation one of the most important points in the calibration process, since the quality of complex products can be ensured.

In order to better understand AVL’s approach, the following three main sources of errors in calibration can be noted: Process: errors that are implemented, although the technical solution is correct.

Complexity: errors that are caused by high system complexity. This means that each component is intrinsically correct, but is calibrated without considering the interaction of all the individual components together.

Time and resources: errors that are caused through insufficient time and resources.

AVL has developed a validation approach to address and hence avoid these three error possibilities, as described here: 1. Dataset management (Dataset merging, clustering, tracing).

2. Automated HiL dataset validation taking different component tolerance into account before releasing the dataset.

3. A wide-ranging, simulation-supported fleet validation with active search for critical events.

Known error sources are continually monitored, not just when limits are violated.

16 |

Paper

CONCLUSION

The methods of model-based development are being implemented on AVL List GmbH’s development platform. Great emphasis is being placed on consistency, model interchangeability, efficient model development and simplicity of model use. Diagram 4 shows the tool chain used, consisting of calibration data management (AVL CRETA), semi-physical modeling (AVL CRUISE™), automated test execution (AVL CAMEO and AVL PUMA), measurement data evaluation (AVL CONCERTO) and calibration and simulation (AVL FOX).

Diagram 6: The AVL Calibration Process

A solution has been developed specially for HiL that permits the calibration engineer to work in the same environment as for a conventional test bed, see diagram 7.

The SiL/MiL environment consists of a combination of the simulation environment (AVL CRUISE™) and the workflow-based calibration tool chain (AVL fOX™) in order to standardize calibration tasks.

As shown in this article, there is already a broad spectrum of applications that can be run on the basis of a powerful and consistent model-based tool chain. By applying modelbased calibration methodology, considerable cost savings can be made and a number of calibration tasks can be carried out that are not possible in conventional development environments for reasons of time and cost, such as IUPMR, RDE or component tolerance investigations, with many different variants, vehicles and cycles.

Diagram 8: V development process enhanced

Diagram 7: The AVL HIL System Setup

The AVL HiL environment not only permits the simple software validation of individual ECUs, but also complex calibration tasks with several networked control units, such as hybrid strategy calibrations. The MiL/SiL environment offers the advantage of being able to run many simulations and tests in a short space of time, thanks to the possibility to simulate faster than realtime. The ability of the models to simulate different environmental conditions, such as altitude, cold, humidity etc. means that calibration work can be efficiently carried out and analyzed, and later, in the vehicle, only needs to be validated.

All in all, 30-50% of all calibration tasks can be run on a model-based development environment, which demonstrates the high potential of such a tool chain using Smart Calibration methodology, and which is an important step towards the frontloading of calibration tasks.

Automaker CMC Special edition | 17

Global Modelling Technique’ to Model Multiple Engine Variants for M&HCV BSIII Application Anuroopa Varsha Tata Motors Ltd

1. INTRODUCTION

I

Indian market traditionally has always been very sensitive to fuel efficiency.

The introduction of strict emission norms has thrown up lot of challenges not only to meet them, but also to have better fuel efficiencies. Because of the higher commercials involved in the modern day electronic engines, higher stress has been put on to deliver better fuel efficiency than the existing solutions (rotary fuel pump or mechanical systems) to offset the increased cost. The

increased commercials also demand commonization of the engine hardware for all the derived sub ratings of the engine.

With the advancement of new technologies, the number of variables to be calibrated has also increased fourfold. Because of these complexities good calibration tools and methodologies are necessary.

In this work, the AVL CAMEO Global Modelling Technique was suitably applied to already existing two variants and two new variants which were deemed necessary.

The engine is six cylinders six litre

with medium duty common rail, high pressure line cooled EGR and waste gate turbocharger.

The whole operating range of the engine variants were mapped with this technique with special considerations given to the vehicle operating regions.

The Global Modelling Technique was found to be very effective tool to model and calibrate multiple engine sub rating derived for the same engine.

The predictions of the model were found to be very accurate.

2. OBJECTIVE

The following were the objectives.

Improve existing engine calibration of two variants for minimum fuel consumption based on new road load data and to further calibrate two new engine sub variants.

Include and optimize nodal points derived from actual road load cycle.

Complete the calibration process in minimum time and minimum utilization of test bed resources.

3. NEED FOR GLOBAL MODEL

Traditional methods such as One Factor At a Time (OFAT) and factorial methods are not suitable for modern day electronic engines as the entire variation space and any interaction effect are not covered in OFAT and the later would require practically impossible large number of operating points.

In DOE local modelling technique, the user needs to specify the operating points in terms of engine speed and torque (in this case 13 mode ESC points) and then a variation space (ECU variables) is constructed for each operating point. This method with 4th order D-Optimal test and with

18 |

Paper

5 ECU variables as variation space would require a minimum of 75 points (including additional and repetition points). For this case the total number of variation points would be 3675 (4X12X75 for 12 points of ESC + 75 points for one common idle point). With these many number of points only ESC regions of sub-rating of the engine is modelled and any additional point (such as derived from vehicle operation) would require further addition of points.

Global Modelling Technique provides alternative and effective methodology to model entire engine operating region and is the

most suitable method for modelling multiple sub rating engines. Global Model Technique is a process in which engine speed and torque/ injected quantity is also made as a variation in the test design. The advantages include the ability to predict engine response as a function of variation parameters throughout the modelled region, calibrate for transients and make different calibrations of the engine depending on the intended application. In this work about 3500 points were used to model the entire engine operating region for 4 sub-variants.

4. CALIBRATION PROCESS

Please refer to Figure 4.1 which shows the workflow used for the global modelling process. As stated earlier, base engine calibration was already available for two of the variants. With this base calibration, part load data for the engine was generated. These points were used as the start/safe point for the operating points in the global model.

Part load data generation throughout the engine operating region with the existing calibration.

Division of the engine operating region into 4 sub-regions based on emission and non-emission region and ECU variations chosen.

Screening test to determine the boundary points based on the limit imposed on specific NOx in g/kW.h Global model preparation based on the operating regions Test run of modelled regions and test data collection Global modelling and optimization Map generation Verification of the predicted optimized results Figure 4.1: Workflow for Calibration Process using Global Modelling Technique

Since the emission region in M&HCV application extends from ‘A’ speed to ‘C’ speed with torque level ranging from 100% to 25%, engine operating regions were divided into two parts based on emission and nonemission regions. The other two parts came from the ECU variation parameters. Thus the whole operating region was mapped using four separate regions. This is shown in Figure 4.2. Idling was modelled as a local point to complete the model.

Figure 4.2: Global Model Regions

The maximum and minimum operating envelope as function of engine speed and torque were created by running a screening test with a maximum and minimum limit imposed on specific NOx (in g/kWh) for engine speed at every 100 rpm and engine torque varying from 100% to 20% of maximum torque with a step of 20% of maximum torque by varying ECU parameters.

Global Model was then prepared for the four regions with a total of about 3500 points for the four regions. After completing the test run, model was prepared and validated.

Automaker CMC Special edition | 19

5. GLOBAL MODEL PROCESS

5.1. TEST RUN AND MEASUREMENT The Global Model test was parameterized with the following variation parameters i. Engine Speed ii. Engine Brake Torque iii. Rail Pressure iv. Injection Timing v. EGR valve position vi. Pilot Quantity vii. Pilot Separation

In order to generate a good model, the fidelity of the collected measurement data should be very high.

In order to achieve this, ‘Individual Channel Stabilization’ feature of CAMEO was used to stabilize the engine at every variation point by monitoring and stabilizing the selected channels. The drifts in measurement data of various measuring instruments was also checked at every repetition point. The test cycle was stopped if any drift in the measurement instrument was found. The measurement instrument was corrected/re-calibrated and test run was continued after this process.

5.2. MODELLING Special attention was given to the following responses which were modelled i. Mass flow rate of NOx, THC and CO ii. Fuel Flow rate iii. Smoke and Soot based on smoke meter readings The following work flow shown in Figure 5.2.1 was used to check for raw data consistency and create a model.

Check for any bend in the variation parameters (difference in set and demand channels) Analysis of the repetition point (For the measured channels which were not monitored during the test) Creating models for the required channels Analysis of the model quality using the statistical data of the generated model Outlier identification through analysis of normal probability distribution of residuals Analysis of outlier effect using measured vs predicted plots Figure 5.2.1: Raw Data Analysis and Model Generation Work Flow

Model statistics and normal probability distribution plot for residuals for one of the modelled region for fuel flow and mass flow rate of NOx are shown in Figure 5.2.2 and 5.2.3. The model regressors are very good (R2adj and R2Pred > 0.95). This is because of emphasis given on getting high accurate data as explained before. The model statistics for other modelled channels were also found to be good. Figure 5.2.4 shows the intersection plot for NOx flow rate and fuel rate for one of the operating point with the 95% significance interval.

Figure 5.2.2: Measured vs Predicted Graphic

20 |

Paper

Figure 5.2.3: Normal Probability Distribution Graphic for Residuals 5.3. OPTIMIZATION AND VERIFICATION

Optimization process involved applying stationary driving cycle algorithm to the ESC points for the selected sub variant with appropriate cycle weightage. The optimization was always run for minimum cycle SFC condition with cycle constraints on NOx, HC, CO and Soot. The local constraints (ECU Variation) for each operating point were also fixed with the maximum and minimum operating envelop of the engine operating space. Map smoothing was also used as a constraint. The predicted results were verified by running a verification test. Once the verification results were found to be at satisfactory levels, the whole of the emission region was calibrated to meet the random NOx tests.

6. RESULTS

The predicted and measured plots of NOx, Smoke, HC, CO and fuel flow are respectively shown in Figure 6.1, Figure 6.2, Figure 6.3, Figure 6.4 and Figure 6.5. As evident from these figures, good correlation was observed between predicted and measured data. The predictions for the Cycle NOx and fuel flow rate were spot on. The predictions for the other modelled responses were excellent.

Figure 5.2.4: Intersection Graphic for Fuel Flow (FB_VAL_A) and NOx (MF_NOx_71) mass flow rate

Automaker CMC Special edition | 21

Figure 6.1: NOx Comparison

Figure 6.2: Smoke Comparison

The other advantage which was also derived out of the Global Model was for shift in emission points of ESC test cycle. These shifts in emission points resulted from a requirement of higher low end torque (about 20% increase in torque over existing level). Since the emission points depends on the position of location of 50% engine power on full load performance curve, the higher low end torque resulted in shifting of emission points.

Because of the Global Model, this was easily accommodated, which would have not been possible otherwise.

Fuel efficiency improvement of about 8% was derived on one of the variant when compared with already existing traditionally calibrated variant. This improvement was also seen on the vehicle. Other subvariant showed 3-5% improvement in fuel consumption when compared with already existing calibration.

Figure 6.3: THC Comparison

Figure 6.4: CO Comparison

Figure 6.5: Fuel Flow Comparison

7. CONCLUSIONS

The Global Modelling technique was found to be a very useful tool to map the entire engine operating region. The four sub-variants of the engine were calibrated to meet Euro III/BSIII emission norms within a time period of 305 hours running two shifts a day, which also included the time taken to run the model points. Fuel efficiency improvement of about 8% was seen on one of the variant, which was also validated on the vehicle.

For other sub variants, fuel efficiency improvement of about 3-5% was seen. The shift in ESC Emission Points that resulted due to further increase in low end torque was also easily accommodated with the Global Modelling Technique.

22 |

Paper

Virtual and Experimental Optimization of In-Cylinder Residuals with Modified Cam using Gas Exchange Analysis J.Balaji, S.Palanikumar, L.Navaneetharao, M.V.Ganesh Prasad Ashok Leyland Ltd, India

S

tringent emission regulations for off-road, CEV (Construction Equipment Vehicles) diesel engines in India (BS-3 emission norms) calls for major engine design changes to achieve a drastic reduction (>50%) in NOx and (>25%) PM (Particulate Matter) as compared to BS-2 emission levels. When it comes to diesel emissions the trade-off between NOx and total particulate emissions, forms one of the major challenges that face manufacturers in meeting the ever tightening emission norms.

Hence independent treatments are needed for both NOx and PM produced in diesel combustion.

In CEV engines limits for NOx emissions are stringent and becomes the driving factor in development. Figure 1 show the progression of emission limits for off-road engines in India.

Nitric oxide (NO) and nitrogen dioxide (NO2) are usually grouped together as NOx emissions. Nitric oxide is the predominant oxide of nitrogen produced inside the cylinder. There are two basic approaches in diesel engine NOx control 1) Exhaust gas after treatment systems such as Selective Catalytic Reduction (SCR) or Lean NOx Trap (LNT) and 2) In-cylinder NOx control such as retarded injection timing, injection rate shaping, water emulsification of the fuel and Exhaust Gas Recirculation (EGR).The above NOx control technologies are complex and need high initial and running cost. A relatively simple, cost effective and durable technology is a requirement of CEV sector. One such technology is internal EGR (iEGR) where the exhaust residuals are trapped in the cylinder by modified valve operation with eliminated problems of external EGR. Internal EGR has a great potential in Figure 1. Off-road emission limits in India reducing NOx emissions in diesel engines, shown earlier by researchers. Internal EGR has been used in Diesel, Gasoline and Natural gas and even in HCCI engines and also serves different purposes.

There are different methods of internal EGR trapping [6] i.e. secondary intake valve opening, secondary exhaust valve opening, late exhaust valve opening, early exhaust valve opening and early exhaust valve phasing. Previous studies have used various actuating mechanisms to achieve internal EGR i.e. Variable Valve Timing (VVT), Compression Release Retarder (CRR), CRR with on/off valve [8], special camshaft with control valve, increased exhaust back pressure control, active valve train control and cam phaser. These mechanisms were controlled by electronic control systems by programmed look-up tables mapped in development stage.

These systems eventually offset the cost of external EGR components. A simple cost effective approach to achieve internal EGR would be a fixed secondary valve event achieved by mechanical camshaft. Fixed secondary valve opening has limitations on control of internal EGR rates. Also meeting the emission targets on a mobile engine application is challenging. Hence the main objective of the present work is to develop an optimum internal EGR valve event for off-road engine based on 2EVO and 2IVO concepts.

THERMODYNAMIC SIMULATION USING AVL BOOST

A 1-Dimensional thermodynamic simulation model has been developed using commercial software called AVL BOOST V.10 by linking various sub models such as air cleaner, turbocharger, intercooler, cylinder and pipes. Simplified turbocharger model was used with target boost pressure ratios. The thermodynamic state of the cylinder was estimated considering the interactions such as piston work, fuel heat input, wall heat losses, blow-by losses, and enthalpy of in/out flowing masses. AVL MCC combustion model was used for heat release, performance and NOx prediction. Injection rate was calculated using measured nozzle end pressure, in-cylinder pressure and nozzle dimensions such as hole diameter, no of holes and flow coefficients using Bernoulli flow equation. The fuel quantity input to the model was measured from the baseline experiments. Intake port swirl ratio and flow coefficients were measured on cylinder head using paddle wheel swirl rig at each valve lift.

NOx prediction in Boost is based on the model developed by Pattas et al based on the Zeldovich mechanism.

Schematic diagram of the thermodynamic model are shown in Fig.2.

Automaker CMC Special edition | 23

Secondary intake lift strategy was studied in simulation and found 13% secondary intake lift to that of main lift was found suitable to meet the target internal EGR residuals of 10-12%.

Figure 2. 1D Thermodynamic model

EXPERIMENTAL SETUP

ENGINE - A six cylinder 4 stroke turbocharged intercooled diesel engine with a displacement of 0.96 liter/cylinder and producing a rated bmep of 10.2 bar has been chosen for this study. This direct injection diesel engine is fitted with an inline fuel injection pump with multiple hole injector. Detailed engine specifications are given in table 1.

Table 1. Engine Configuration Type

4- Stroke, Water Cooled

Bore x Stroke

104(mm) x 113(mm)

No. of Cylinders

6 Cylinder

Compression Ratio

17.5:1

Injection System

Inline Pump

Rated Power

160 HP @ 2400 rpm

Max Torque

570 Nm @ 1200 – 1800

Turbocharger

Pulse Turbocharger

TEST CYCLE - A steady state 8 mode test cycle according to ISO 8178 part-4 (1996) [11] (fig 3) for certifying the emissions of non-road diesel engine had been used for result validation.

Figure 4. Experimental set up

Figure 3. ISO 8178, C1 8 Mode Cycle

24 |

Paper

TEST SETUP - Engine was coupled to an AVL alpha 350 eddy current dynamometer by a propeller shaft.

Torque measurement was carried out by a strain gauge load cell and for speed measurement an inductive pulse pick up was used. Figure 4 shows the experimental set up. Engine was instrumented with AVL Indiset Advanced plus which has piezoelectric transducer for measuring cylinder pressure, strain gauge transducers to measure injection line pressures at pump side and injector side to estimate the injection rate. Gaseous exhaust emissions were measured using a Horiba 7000 series Exhaust gas analyser capable of measuring all the legal pollutants ie CO, THC, NOx. Particulate matter emission was measured using an AVL 472 smart sampler partial flow particulate measurement device.

Figure 5. Gas exchange analysis set up

AVL’s Gas Exchange & Combustion Analysis (GCA) was done additionally for internal EGR rate estimation. GCA is a software tool based on AVL BOOST combining the measurement and simulation to allow assessment of non-measurable in-cylinder parameters i.e. internal EGR rate.

Measured signals are the input for reduced thermodynamic simulation model, which calculates the mass flow rates across valves and thereby estimates the internal EGR rate.. Engine cylinder head was drilled and tapped to mount the Intake and exhaust port pressures sensors on the ports. For intake port pressure GH12D air cooled piezoelectric sensor and for exhaust port pressure GU21C water cooled piezoelectric sensor were used. Figure 5 & 6 shows the gas exchange analysis setup.

Figure 6. Intake and exhaust pressure sensor instrumentation

RESULTS AND DISCUSSION

Experiments were conducted at full loads and 8 modes of ISO 8178 C1 test cycle modes. Experiments with base camshaft showed at full loads both intake and exhaust pressures reduce as the speeds reduce from rated speed to max torque speed. Correspondingly calculated instantaneous mass flows show a reduction wrt speed. Part loads at C1 test cycle showed a lower exhaust flow rates at the part loads than at full loads.

The negative exhaust flow in the suction stroke is due to the effect of valve overlapping period. Refer figures 7 to 12 for base camshaft results. An internal residual level of 3 to 4% with the base camshaft observed which is correlating to simulation (fig 13).

Figure 7. Intake & Exhaust pressures at full load

Figure 8. Intake & Exhaust flow at full load

Automaker CMC Special edition | 25

Figure 9. Intake & Exhaust pressures at 2200 rpm

Figure 10. Intake & Exhaust flow at 2200 rpm

Figure 11. Intake & Exhaust pressures at 1500 rpm

Figure 12. Intake & Exhaust flow at 1500 rpm

Figure 13. Residual EGR % of base camshaft

Secondary intake opening strategy (2IVO) was selected for the study with 1.6mm lift at first. A new turbocharger was matched based on simulation results which increases the intake boost pressures by 30% which can be clearly seen in pressure and flow measurements of camshaft with 1.6mm lift. Intake mass flow showed a negative trend during the beginning of exhaust stroke which traps the residuals at intake manifold. With 1.6mm 2IVO an internal EGR residuals of 12% observed in the 6 modes of C1 test cycle. With 1.6mm lift NOx reduction of 32% with huge increase in soot emission which was mainly due to insufficient fresh air charge. Hence 12% iEGR levels with the present 1.6mm 2IVO camshaft were estimated to be too high to maintain the fresh air charge levels at mid speed. In order to increase the fresh air charge levels and reduce smoke levels further, secondary lift was reduced to 1.30 mm. An instantaneous mass flow reduction of 12% observed with 1.3mm lift and an overall internal residual reduction of 2-4% observed compared to 1.6mm lift (fig 20). Results show a good improvement in air excess ratio and improved smoke levels of the engine with 8-9% iEGR rate. Especially at lower speeds intake manifold pressure increased upto 0.2 bar. With 1.3mm secondary lift NOx reduction was found to be 18% and PM increase was only 22% which is well within the limits. C1 cycle results of NOx and PM (table 2) were close to the legal limits with 1.30 mm 2IVO and show a direction to achieve the targets. Further tuning of fuel combustion parameters can reduce the emissions to meet the BS3 CEV limits.

26 |

Paper

Figure 14. Intake & Exh pressures with 2IVO at 2400 rpm

Figure 15. Intake & Exh flow with 2IVO at 2400 rpm

Figure 16. Intake & Exh pressures with 2IVO at 2200 rpm

Figure 17. Intake & Exh flow with 2IVO at 2200 rpm

Figure 18. Intake & Exh pressures with 2IVO at 1500 rpm

Figure 19. Intake & Exh flow with 2IVO at 1500 rpm

Table 1. Emission Results Results

Base Engine

1.60 mm 2IVO

1.30 mm 2IVO

BS3 CEV Limits

CO(g/kWh)

0.586

2.276

0.794

4.500

NOx+HC (g/kWh)

5.47

3.700

4.490

3.800

PM(g/kWh)

0.132

0.502

0.162

0.270

CONCLUSION

Thus the tool Gas Exchange and Combustion analysis helps to measure the incylinder parameters with a reduced Boost model. Internal EGR rates were estimated with a very good accuracy levels and internal EGR rate targets were optimized for BS3 CEV off-road limits.

Automaker CMC Special edition | 27

AVL. A KALEIDOSCOPE OF SOLUTIONS FOR INTEGRATED POWERTRAIN SIMULATION. With highly precise system and component simulations supported by user-friendly and robust integrated simulation platforms, AVL Advanced Simulation Technologies (AST) provides workflow-guided solutions for engine, powertrain and vehicle development processes.

AVL AST offers tools for the conceptual and detailed design phases all the way up to powertrain calibration, Hardware-in-the-Loop (HiL) and testbed applications. By combining in-house and third-party software tools with the comprehensive expert knowledge of highly experienced engineers, AVL supports the development of products of the highest quality with optimized costs and shortest time to market.

Systematically and precisely – like a kaleidoscope. www.avl.com/ast