La Metallurgia Italiana - Gennaio 2019 by aimnet3

Metallurgia Italiana

International Journal of the Italian Association for Metallurgy

n. 1 Gennaio 2019 Organo ufficiale dellâ&#x20AC;&#x2122;Associazione Italiana di Metallurgia. Rivista fondata nel 1909

e c

c c

1st announcement call for papers

10th european conference on continuous casting 2020

Bari . Italy 17-19 June 2020

e c

www.aimnet.it/eccc2020

c c Organised by

The 10th European Continuous Casting Conference - ECCC 2020 - will be organised by AIM, the Italian Association for Metallurgy, in Bari (Italy) on 17-19 June 2020 with focus on the status and future developments in the casting of steel. The ECCC is a unique forum for the European continuous casting community to exchange views on the status and the future development of the continuous casting process. The Conference program is abreast of the latest developments in control and automation, advanced continuous casting technologies, application of electromagnetic technologies and mechanical devices to improve the core microstructure, the lubrication issues for improving the surface qualities. Steel metallurgical issues will be addressed as well as their physical and numerical simulation. The exchange of experience in operational practice, maintenance and first results from the recently commissioned plants will integrate the program. The Conference aims at promoting the dialogue among the delegates with industrial and academic background and among the participants in former Conferences and new members of the continuous casting community.

Topics • Trends of innovation in casting technologies • New developments and advanced technologies for the casting of slabs, blooms and billets • Ladle and tundish recent metallurgical solutions for steel cleanness • Flow control, refractories and clogging • Mold lubrication and heat transfer • Product quality control: Surface quality and internal soundness • Numerical simulation and modelling (solidification, metallurgy, fluid flow, validation) • Safety and environmental aspects • Continuous casting technologies and circular economy • Operational practice and maintenance • Measurement, automation and process control • Post-processing of semi-finished products (Scarfing, machining and heat treatment) • Modernization and new implementations • Industry 4.0, machine learning and digitalisation

Call for Papers - Abstract Submission Prospective authors wishing to present papers are invited to submit, by 31 October 2019, a tentative title and an abstract of about 300 words (in English), specifying a maximum of two topics for each proposal, to the Organising Secretariat (aim@aimnet.it). The abstract should provide sufficient information for a fair assessment and include the title of the paper, the author’s names and contact details (address, telephone number and e-mail address). The name of the presenting author should be underlined. A poster session might be organized as well. There are two ways to submit papers: • fill in the form on the Conference website at: www.aimnet.it/eccc2020 • send the requested information by e-mail to: aim@aimnet.it.

Contacts ECCC 2020 Organising Secretariat AIM - Associazione Italiana di Metallurgia Via Filippo Turati 8, 20121 Milan - Italy Tel. +39 02 76021132 / +39 02 76397770 aim@aimnet.it - www.aimnet.it/eccc2020

La Metallurgia Italiana

Metallurgia Italiana

International Journal of the Italian Association for Metallurgy

n. 1 Gennaio 2019 Organo ufficiale dell’Associazione Italiana di Metallurgia. Rivista fondata nel 1909

International Journal of the Italian Association for Metallurgy Organo ufficiale dell’Associazione Italiana di Metallurgia. House organ of AIM Italian Association for Metallurgy. Rivista fondata nel 1909

n. 1 Gennaio 2019

Anno 111 - ISSN 0026-0843

Direttore responsabile/Chief editor: Mario Cusolito Direttore vicario/Deputy director: Gianangelo Camona Comitato scientifico/Editorial panel: Livio Battezzati, Christian Bernhard, Massimiliano Bestetti, Wolfgang Bleck, Franco Bonollo, Bruno Buchmayr, Enrique Mariano Castrodeza, Emanuela Cerri, Lorella Ceschini, Mario Conserva, Vladislav Deev, Augusto Di Gianfrancesco, Bernd Kleimt, Carlo Mapelli, Jean Denis Mithieux, Marco Ormellese, Massimo Pellizzari, Giorgio Poli, Pedro Dolabella Portella, Barbara Previtali, Evgeny S. Prusov, Emilio Ramous, Roberto Roberti, Dieter Senk, Du Sichen, Karl-Hermann Tacke, Stefano Trasatti Segreteria di redazione/Editorial secretary: Valeria Scarano Comitato di redazione/Editorial committee: Federica Bassani, Gianangelo Camona, Mario Cusolito, Carlo Mapelli, Federico Mazzolari, Valeria Scarano Direzione e redazione/Editorial and executive office: AIM - Via F. Turati 8 - 20121 Milano tel. 02 76 02 11 32 - fax 02 76 02 05 51 met@aimnet.it - www.aimnet.it

Colata continua / Continuous casting Consideration of phase equilibria in fe-Al-Ti-O system and its importance in steel cleanliness during casting process Y.-B. Kang, J.-H. Lee 5 Flux-steel reaction of caO-siO2 and caO-Al2O3-based mold fluxes with high-Al steel J. Yang, J. Zhang, O. Ostrovski, C. Zhang, D. Cai 12 Influence of selected alloy additions on time mixing for pulse-step method of liquid steel alloying in the tundish A. Cwudziński, J. Jowsa 20 A novel methodology to evaluate surface cracking risk during strand straightening in continuous casting G. Poltarak, S. Ferro, C. Cicutti 28 Attualità industriale / Industry news Manifestazioni AIM

siderweb LA COMMUNITY DELL’ACCIAIO

Gestione editoriale e pubblicità Publisher and marketing office: Siderweb spa Via Don Milani, 5 - 25020 Flero (BS) tel. 030 25 400 06 - fax 030 25 400 41 commerciale@siderweb.com - www.siderweb.com La riproduzione degli articoli e delle illustrazioni è permessa solo citando la fonte e previa autorizzazione della Direzione della rivista. Reproduction in whole or in part of articles and images is permitted only upon receipt of required permission and provided that the source is cited. Reg. Trib. Milano n. 499 del 18/9/1948. Sped. in abb. Post. - D.L.353/2003 (conv. L. 27/02/2004 n. 46) art. 1, comma 1, DCB UD Siderweb spa è iscritta al Roc con il num. 26116

Influence of fc-Mold on flow pattern and entrapment of inclusions in continuous casting strand edited by: S. Wang, W. Chen, X. Zhang, L. Zhang 39 Development and use of indicators in the algorithms to detect defects automatically on the slabs by utilising images taken from the hot slabs on line during continuous casting edited by: P. Hooli, H. Suopajärvi 46 Scenari / Experts' Corner State of the Art in High-temperature Mold Simulator Study of Initial Solidification in Continuous-casting Mold edited by: P. Lyu, W. Wang 56 Atti e notizie / Aim news Calendario eventi internazionali Rubrica dai Centri Corso "Gli acciai inossidabili"

69 72

lâ&#x20AC;&#x2122;editoriale La Metallurgia Italiana The 7th International Congress on Science and Technology of Steelmaking (ICS 2018) was held in Venice/Mestre from June 13 to 15, 2018. AIM, the Italian Association for Metallurgy organized this event for the first time in Italy. 400 participants representing all contents were attracted. Almost 200 oral and 11 poster presentations were accepted. The sessions of ICS 2018 covers all topics related to steelmaking. 34 papers Johannes Schenk Montanuniversitaet Loeben, Austria

where dedicated to continuous casting technology, with special focus on fluid flow and solidification, mold fluxes and slab quality control. Their content addresses the achievements in research and application in the recent time and the future trends of the technology developments. The key note lecture of Prof. Mapelli, Politecnico Milano was dedicated to lubrication in the casting mold. A review of the presented papers gives evidence that numerical modelling of different physical aspects of the casting processes is becoming more and more a standard tool in the research and process control. The characterization of high temperature properties of the cast steel and mold flux were in the focus of various papers. A number of papers were dealing with initial solidification and solidification structure. A selection of outstanding papers of the continuous casting session of the ICS 2018 are published in this issue of La Metallugia Italiana.

La Metallurgia Italiana - n. 1 2019

Colata continua

Consideration of Phase Equilibria in Fe-Al-Ti-O System and Its Importance in Steel Cleanliness during Casting Process Y.-B. Kang, J.-H. Lee

Phase diagram of oxide systems composed of FetO-Al2O3-TiOx was reviewed based on available experimental phase diagram data and CALPHAD thermodynamic database. Most recent updated thermodynamic database was used in order to predict phase equilibria of this oxide system under controlled oxygen partial pressure. In particular, phase equilibria of the oxide system in contact with liquid steel containing Al and Ti, with air, and with refractory for casting process were calculated. A series of experiments were carried out in order to validate the thermodynamic calculations. It was found that, apart from solid alumina, a liquid oxide composed of FetOAl2O3-TiOx could be formed at the interface between the refractory and the steel, when the liquid steel contains very low C (ultra low C steel grade). The formation of liquid oxide seems to be responsible for clog materials during continuous casting of ultra low C steel. Moreover, this steel can be reoxidized in a tundish by entrapped air. This also generates a liquid oxide mixed with a solid oxide, which significantly affect cleanliness of the steel. Oxidation behavior of the Al-Ti containing steel in the tundish and in the nozzle refractory is comparably discussed.

KEYWORDS: CLEAN STEEL – FE-AL-TI-O – CLOGGING – PHASE EQUILIBRIA

INTRODUCTION Ultra Low C (ULC) steel is one of typical steel grades that require high level cleanliness. This is partly due to susceptibleness of this steel to oxidation, as this steel contains extremely low C. In general, the ULC steel is deoxidized by Al. Once it is exposed to an oxidizing atmosphere, the steel is reoxidized and the oxidation product remains in the steel. Usually, the oxidation product is considered as non-metallic inclusion, and is known to be harmful to quality of the steel product and to continuous casting process by causing nozzle clogging. Therefore, cleanliness of ULC steel is seriously deteriorated by the reoxidation. This becomes more serious when the ULC steel is further alloyed by Ti (Ti-ULC). Ti is added in order to bind interstitial elements such as C and N. This enhances formability of the Ti-ULC steel sheet and allows it to be shaped easily. The Ti-ULC steel is used for outer panel of automobiles. Therefore, cleanliness of the steel should be maintained. Unfortunately, Ti-ULC steel suffers difficulties in keeping its cleanliness. Defect on cast slab is often observed which is known to contain some portion of Al and Ti oxides (1). The defect might have been non-metallic inclusion in liquid steel during the process of RH degasser – tundish – continuous casting (CC). Or, it may be a separated portion of clog material during the CC of the steel. It is well known that nozzle clogging is extremely serious for the casting of Ti-ULC steel (1). In order to keep the cleanliness of the Ti-ULC steel and to minimize defect generation and nozzle clogging, it is necessary to understand how the Ti-ULC steel is reoxidized during the process. La Metallurgia Italiana - n. 1 2019

Previous efforts have been focused on transient evolution of the non-metallic inclusions in the steel after Ti is added (26). Due to local inhomogeneity of the steel, composition, shape, and phase of inclusions vary during the alloying. This was suggested as a possible cause of defect generation and nozzle clogging. However, the root-cause is still unclear. Moreover, stable oxide phases which may form by deoxidation - reoxidation of Ti-ULC steel are controversial. As the understanding of the formation of various oxides is a key to understand the oxidation behavior of Ti-ULC steel, and further to resolve issues relevant cleanliness of the steel, it is necessary to look at phase equilibria of relevant oxide systems. In the present article, the system is confined in the Fe-Al-Ti-O system. In particular, as the oxides are mostly in contact with liquid steel, more attention was paid to the phase equilibria under reducing condition. The present article first reviews phase equilibria of the Fe-Al-Ti-O

Youn-Bae Kang and Joo-Hyeok Lee Graduate Institute of Ferrous Technology, Pohang University of Science and Technology, Rep. of Korea

Continuous casting and its sub-system in order to know what kinds of phase are stable and may appear as non-metallic inclusion or clog material. Reoxidation product of Ti-ULC steel is then interpreted from the phase diagram information. Its consequences on steel cleanliness and nozzle clogging are discussed. PHASE EQUILIBRIA OF FE-TI-AL-O AND ITS SUB-SYSTEM Only one oxide, Al2O3 is known. It is most well-known inclusion in steel.

Ti-O System Ti is one of representative transition metals. Ti presents as Ti2+, Ti3+, and Ti4+ in oxides depending on phase, temperature, and oxygen potential. Fig. 1 shows a part of Ti-O phase diagram (from Ti2O3 to TiO2) (8). Apart from Ti2O3 and TiO2, Ti3O5 and socalled Magneli phases (TinO2n-1, n > 3) are seen.

Fig. 1 â&#x20AC;&#x201C; A part of Ti-O phase diagram, from Ti2O3 to TiO2 (8). Fe-Al-O System Apart from Al2O3, FeAl2O4 is a stable oxide phase. This is known to form in liquid steel at very dilute Al concentration. Fig. 2(a) shows the FetO-Al2O3 phase diagram at metallic Fe saturation (9). This is close to a condition of the oxides in steel of extremely dilute solutes. Fig. 2(b) shows a well-known Al deoxidation equilibria in liquid Fe, down to low Al concentration (10). It can be read from these diagrams that Al2O3 is a stable oxide phase in liquid steel, but at extremely low Al concentration (below 1 ppm), FeAl2O4 can form.

Fe-Ti-O System Similar to the Fig. 2(a), the FetO-TiOx phase diagram section at metallic Fe saturation is shown in Fig. 3(a) (9). This phase diagram should be seen that the oxide phase is in equilibrium with steel of dilute solute concentration. Contrary to the Fig. 2(a), liquid oxide is likely to exist near steelmaking and casting temperature. Ti deoxidation equilibria is shown in Fig. 3(b) that is usually showing the deoxidation equilibrium with solid oxide phases (11). However, it is necessary to consider the equilibria at extremely low Ti concentration where liquid oxide should be stable (12).

Fig. 2 â&#x20AC;&#x201C; Fe-Al-O phase diagram: (a) FetO-Al2O3 diagram section at metallic Fe saturation (9) and (b) Al deoxidation equilibria in liquid Fe (10). Double circle represents co-saturation of Al2O3 and FeAl2O4. 6

La Metallurgia Italiana - n. 1 2019

Colata continua

Fig. 3 – Fe-Ti-O phase diagram: (a) FetO-TiOx diagram section at metallic Fe saturation (9) and (b) Ti deoxidation equilibria in liquid Fe (11). Fe-Al-Ti-O System In order to know stable oxide phases in equilibrium with liquid steel containing both Al and Ti, an oxide stability diagram can be used. Fig. 4 shows one of the oxide stability diagrams of the present system reported so far (13). Similar to other versions, it can be seen that equilibrium oxide phase in general Ti-ULC steel that contain a few hundred ppm of Al and Ti is Al2O3. It has been controversial whether liquid oxide and/or Al-Ti complex oxide (Al2TiO5) exist in the steel or not (5,14-16). Nevertheless, it is agreed that the Al2O3 is final equilibrium inclusion phase in the steel. The liquid oxide and Al2TiO5 may appear under special

condition (9). Kang and Lee pointed out that the liquid oxide is composed not only of Al2O3 and TiOx, but also FetO (9). Al-Ti-O System Since Ti is a transition metal, stability of the Ti containing oxide depends strongly on oxygen potential. Fig. 5 shows phase diagram of Al-Ti-O (a) in air and (b) under reducing condition (16). Not only stable Ti oxide changes from TiO2 to Ti2O3 at steelmaking temperature, but also stability of liquid oxide and AlTiO5 change. This means that non-metallic inclusion and clog 2 material in Ti-ULC steel are sensitive to local oxygen potential.

Fig. 4 – An Fe-Al-Ti-O oxide stability phase diagram (13)

Fig. 5 – Al-Ti-O phase diagram (a) in air and (b) under reducing condition (16)

La Metallurgia Italiana - n. 1 2019

Continuous casting From the phase diagrams shown above, it is seen that stability of oxide phases depends on oxygen potential. Therefore, the oxygen potential relevant to liquid steel processing should be considered. Moreover, it is seen from Fig. 3(a) and Fig. 4 that a liquid oxide containing FetO can form when a liquid steel containing Ti is oxidized. REOXIDATION OF TI-ULC STEEL AND ASSOCIATED OXIDE FORMATION Reoxidation and inclusion formation in liquid steel Regarding inclusion evolution in Ti-ULC steel relevant to steel cleanliness, a number of previous researches reported that stable inclusion in the steel is Al2O3. However, other inclusions that contain Ti could form as intermediate inclusion phase after Ti alloying (2-6). Composition and phase of these inclusions were not known accurately due to small size of the inclusion embedded in the steel matrix. Usually, Al and Ti concentrations in the inclusions were taken into account, but Fe is often neglected. However, under some condition where oxygen potential is higher than that in bulk liquid steel, oxidation of Fe can also occur simultaneously with Al/Ti oxidation. Sasai and Matsuzawa carried out a series of experiments by oxidizing Fe0.15Ti, and Fe-0.1Al-0.15Ti steel using Ar-O2 gas mixture (17). The oxygen partial pressure was varied from 0.04 to 0.23 bar, simulating reoxidation of ULC steel in tundish. This was an excessive oxidizing condition. They observed oxidized product on the surface of the steel, and reported that a liquid oxide formed when the steel contained Ti. By EPMA, it was found that the liquid oxide was composed of FeO-Al2O3-TiO2. Once this liquid oxide enters into the liquid steel, it could be easily reduced by Al or Ti, then could be transformed to Al-Ti complex Fe-Al-Ti-O oxide. Further reduction by Al results in Al2O3

formation at equilibrium state, according to the oxide stability diagram (Fig. 4). Mizoguchi et al. analyzed alumina inclusions in 0.015Ti-0.03Al ULC steel (18). They reported that the alumina was often observed as clusters. They proposed that FeO suspended in the liquid steel works as a binder at the moment of collision of individual alumina inclusion. They also suggested possible sources of the FeO: oxygen contamination of ferroalloy additives, residual steel adhering to the refractory surfaces, and air entrapment. The experiment of Sasai and Matsuzawa would correspond to the case of air entrapment (18). The reoxidation of Ti-ULC in tundish by air can cause a formation of liquid oxide containing FeO (17). This liquid oxide may work as the binder to form the alumina clusters in liquid steel, as was proposed by Mizoguchi et al. (18). Reoxidation and clog formation in SEN Nozzle clogging during continuous casting of Ti-ULC steel is more serious than that of Ti-free ULC steel. Although there are various sources causing nozzle clogging (19), that of Ti-ULC steel is still unclear. This is because the clog material is not only alumina inclusion but also a mixture of skull and inclusions (1). Recently, the present authors proposed that reoxidation reaction between nozzle refractory and Ti-ULC at the interface may be a reason of the nozzle clogging (9). The reoxidation occurs due to CO(g) generated from the nozzle refractory. It was known that there is a carbothermic reaction between SiO2 and graphite inside nozzle refractory. This results in formation of CO(g). The CO(g) moves at the surface of the nozzle through available pores. In an Al killed ULC steel, the CO oxidized Al to form Al2O3 that adheres on the surface (inner wall) of the nozzle. Schematic explanation of the adhesion mechanism by Matsui et al. is shown in Fig. 6 (20).

Fig. 6 â&#x20AC;&#x201C; Adhesion mechanism of alumina inclusions on SEN surface (20) 8

La Metallurgia Italiana - n. 1 2019

Colata continua However, if an ULC steel contains both Al and Ti, then the oxidation reaction product is not just Al2O3. A series of thermodynamic analyses including the phase diagram calculations shown in Fig. 4, the present authors proposed that a mixture of Al2O3 and FetO-Al2O3-TiOx form. The latter is a liquid at the casting temperature. It can also be seen in Fig. 4 that increasing O concentration due to the reoxidation by CO(g) lowers concentrations of Al and Ti near the interface between the nozzle and the liquid steel. According to the Fig. 4, the FetO-Al2O3-TiOx forms. The thermodynamic prediction was experimentally validated by oxidizing a number of Fe-Al-Ti alloys representing Ti-ULC steel (9). It was found that surface of the oxidized alloy was covered

by Al2O3 and the liquid FetO-Al2O3-TiOx. Fig. 7 shows an example of the oxidized surface, clearly showing the mixture of two phases. It was proposed that the liquid oxide may work as a precursor of the clog material in the SEN. The liquid oxide can work as a binder between the nozzle and the liquid steel, as a liquid oxide containing FetO easily wets both to refractory and liquid steel. Once the mixture adheres to the nozzle refractory, FetO in the liquid oxide would be reduced by Al/Ti in the liquid steel or by graphite in the refractory and leaves reduced Fe. Fig. 8 shows a schematic representation of the mechanism proposed by the present authors (9). Due to heat extraction through the nozzle wall, the reduced Fe may be solidified (9).

Fig. 7 â&#x20AC;&#x201C; Oxidized surface of Fe-0.125Ti-0.05Al alloy by CO gas (9).

Proposed mechanism m for r TiTi-ULC Ti cloggin clogging log in ng SiO iO O2 a and C in refractory react to form CO(g). fo CO CO(g) O(g) moves m through refractory po pore. CO CO(g) O(g g) o oxidizes xidizes Al/ A /Ti/Fe Tii simultaneously. sim simu mu u aneoussly.. ulta Fe etO-Al Al2O3-Ti TiO T iO Ox(l) + Al Al2O3(s (s) s s) form and attach to fo o th the inner ner wall. w walll. Th The he liquid oxide containing contai g Fe F et O works as a binder to refractory/inclusion/liquid re refract racttory/inclusion/liquid steel. FeO eO O iss gradually reduced by Al in liquid steel or C in refractory refractory, ry, forming form orm m ng reduced Fe and TiO min Ox Al2O3. Al

Fig. 8 â&#x20AC;&#x201C; Schematic figure showing clogging mechanism of Ti-ULC steel casting (9). La Metallurgia Italiana - n. 1 2019

Continuous casting SUMMARY Phase diagrams of the Fe-Al-Ti-O and its sub-system were reviewed in order to identify what kinds of oxide phases can form in ULC steel containing Al and Ti. Depending on composition, temperature, and oxygen potential, stable phase in the liquid steel varies. In particular, FetO containing liquid oxide can form during reoxidation when the liquid steel contains Ti. This is not the case when the liquid steel does not contain Ti. Such reoxidation can occur in a tundish and in a SEN. Tundish reoxidation by air generates a FetO containing liquid oxide. This would be entrapped into liquid steel, and cleanliness of the steel becomes low. The other reoxidation occurs at the interface between nozzle refractory. A mixture of FetO containing liquid

oxide and solid Al2O3 forms. This is thought to initiate deposition of clog material on inner wall of the SEN. It results in low productivity due to clogging, and bad product quality once the clog material is separated and enters into the liquid steel stream. It is necessary to minimize such reoxidation in order to keep cleanliness of the Ti-ULC steel. It was shown that phase diagram information can give us some insight to identify what is a reoxidation product, which is related to the steel cleanliness. ACKNOWLEDGEMENT Financial support from POSCO is well appreciated.

REFERENCES [1]

Basu S, Choudhary SK, Girase NU. Nozzle Clogging Behaviour of Ti-bearing Al-killed Ultra Low Carbon Steel. ISIJ International. 2004;44:1653–60.

[2]

Wang C, Nuhfer N, Sridhar S. Transient Behavior of Inclusion Chemistry, Shape, and Structure in Fe-Al-Ti-O Melts: Effect of Titanium Source and Laboratory Deoxidation Simulation. Metallurgical and Materials Transactions B. 2009;40B:1005–21.

[3]

Wang C, Nuhfer NT, Sridhar S. Transient Behavior of Inclusion Chemistry, Shape, and Structure in Fe-Al-Ti-O Melts: Effect of Titanium/Aluminum Ratio. Metallurgical and Materials Transactions B. 2009B;40:1022–34.

[4]

Wang C, Nuhfer NT, Sridhar S. Transient Behavior of Inclusion Chemistry, Shape, and Structure in Fe-Al-Ti-O Melts: Effect of Gradual Increase in Ti. Metallurgical and Materials Transactions B. 2010;41B:1084–94.

[5]

Matsuura H, Wang C, Wen G, Sridhar S. The Transient Stages of Inclusion Evolution During Al and/or Ti Additions to Molten Iron. ISIJ International. 2007;47:1265–74.

[6]

van Ende M-A, Guo M, Dekkers R, Burty M, Dyck JV, Jones PT, Blanpain B, Wollant P. Formation and Evolution of Al-Ti Oxide Inclusions during Secondary Steel Refining. ISIJ International. 2009;49:1133–40.

[7]

Sun M-K, Jung I-H, Lee H-G. Morphology and chemistry of oxide inclusions after Al and Ti complex deoxidation. Met Mater Int. 2008 Dec 1;14:791–8.

[8]

Kang Y-B, Jung I-H, Lee H-G. Critical thermodynamic evaluation and optimization of the MnO-"TiO2"-" Ti2O3" system. Calphad. 2006;30:235–47.

La Metallurgia Italiana - n. 1 2019

Colata continua [9]

Lee J-H, Kang M-H, Kim S-K, Kang Y-B. Oxidation of Ti added ULC Steel by CO Gas Simulating Interfacial Reaction Between the Steel and SEN during Continuous Casting. ISIJ International. 2018;58 1257-1266.

[10] Hayashi A, Uenishi T, Kandori H, Miki T, Hino M. Aluminum Deoxidation Equilibrium of Molten Fe–Ni Alloy Coexisting with Alumina or Hercynite. ISIJ International, 2008;48:1533–41. [11] Cha W-Y, Nagasaka T, Miki T, Sasaki Y, Hino M. Equilibrium between Titanium and Oxygen in Liquid Fe-Ti Alloy Coexisted with Titanium Oxides at 1873 K. ISIJ International. 2006;46:996–1005. [12] Pesl J, Eriç RH. High-temperature phase relations and thermodynamics in the iron-titanium-oxygen system. Metallurgical and Materials Transactions B. 1999;30B:695–705. [13] Kang Y-B, Lee J-H. Reassessment of Oxide Stability Diagram in the Fe–Al–Ti–O System. ISIJ International. 2017;57:1665–7. [14] Ruby-Meyer F, Lehmann J, Gaye H. Thermodynamic Analysis of Inclusions in Ti-Deoxidised Steels. Scandinavian Journal of Metallurgy. 2000;29:206–12. [15] Kim W-Y, Jo J-O, Lee C-O, Kim D-S, Pak J-J. Thermodynamic Relation between Aluminum and Titanium in Liquid Iron. ISIJ International. 2008;48:17–22. [16] Jung I-H, Eriksson G, Wu P, Pelton A. Thermodynamic Modeling of the Al2O3–Ti2O3–TiO2 System and Its Applications to the Fe–Al–Ti–O Inclusion Diagram. ISIJ International. 2009;49:1290–7. [17] Sasai K, Matsuzawa A. Influence of Steel Grade on Oxidation Rate of Molten Steel in Tundish. ISIJ International. 2012;52:831– 40. [18] Mizoguchi T, Ueshima Y, Sugiyama M, Mizukami K. Influence of Unstable Non-equilibrium Liquid Iron Oxide on Clustering of Alumina Particles in Steel. ISIJ International. 2013;53:639–47. [19] Ogibayashi S. Mechanism and Contermeasure of Alumina Buildup on Submerged Nozzle in Continous Casting. Taikabutsu Overseas. 1995;15:3–14. [20] Matsui T, Ikemoto T, Sawano K, Sawada I. Effects of Carbon and Silica in Submerged Entry Nozzles on Alumina Buildup. Taikabutsu Overseas. 1998;18:3-9.

La Metallurgia Italiana - n. 1 2019

Continuous casting

Flux-Steel Reaction of CaO-SiO2 and CaO-Al2O3-based Mold Fluxes with High-Al Steel J. Yang, J. Zhang, O. Ostrovski, C. Zhang, D. Cai

Advanced properties of high-Al steel such as high mechanical strength and formability make it attractive for the applications in the automobile industry. However, when conventional CaO-SiO2-based mold fluxes are used for the continuous casting of such steel, [Al] in the steel reacts with SiO2 in the fluxes. It changes the chemical composition and in-mold performances of mold fluxes, leading to unstable heat transfer and insufficient lubrication for the steel strands. CaO-Al2O3-based mold fluxes with low SiO2 content are under development, expecting to suppress the interfacial reaction between mold fluxes and high-Al steel. In this article, the effects of different CaO-SiO2 and CaO-Al2O3-based mold fluxes on the reaction kinetics were evaluated through both pilot and laboratory tests. It was found that there were significant changes in SiO2 and Al2O3 concentrations in CaO-SiO2-based mold fluxes, indicating a severe reaction between fluxes and high-Al steel. CaO-Al2O3-based mold fluxes showed a better composition stability than CaO-SiO2-based mold fluxes when reacting with high-Al steel. In the lab tests, CaO-Al2O3-based mold fluxes with different CaO/Al2O3 ratios were also evaluated. It showed that the increase of CaO/Al2O3 ratio accelerated the flux-steel reaction according to the change of [Al] concentration in the steel. CaO-Al2O3-based mold fluxes showed promising results when reacting with high-Al steel, but the fluxing agents, e.g. B2O3 and Li2O, may also react with [Al] in steel and cause an additional increment of Al2O3 in mold fluxes, which needs further research in the future.

KEYWORDS: HIGH-AL STEEL – MOLD FLUXES – INTERFACIAL REACTION – KINETICS – THERMODYNAMICS INTRODUCTION Advanced high-strength steels (AHSS) have high mechanical strength, good ductility and low weight, showing a great potential in the modern automotive industry. As a new generation of AHSS, twinning-induced plasticity (TWIP) steel attracts great interests in recent years for its extraordinary mechanical properties (1-5). As such, it is austenitic steel with high content of Mn, providing a complex strain-hardening behavior due to its moderate stacking fault energy (20 – 40 mJ/m2) at room temperature (1, 2). Practically, Al is an important component in TWIP steel. It reduces the density of TWIP steel, stabilizes the formation of deformation twinning, and avoids the transformation of austenite into ε-martensite at low strains (6). Conventional mold fluxes for the continuous casting of steel are based on CaO-SiO2 oxide system. However, [Al] in the molten high-Al steel inevitably reacts with SiO2 in the fluxes at high temperature according to the following equation, which causes the accumulation of Al2O3 and the reduction of SiO2 in molten fluxes (7-9):

The variation of flux composition may make the in-mold performances of mold fluxes undesirable and unpredictable in the casting of high-Al steel, increasing the likelihood of sticker breakout and surface defects. CaO-Al2O3-based mold fluxes with low SiO2 content is recently introduced to depress the driving force of reaction [1]. But the increase of Al2O3 and decrease of SiO2 concentrations would raise the liquidus temperature of mold fluxes significantly (10). Fluxing agents, e.g. B2O3 and Li2O, are normally added to reduce the liquidus temperature of mold fluxes. However, B2O3 is also reactive with [Al] in steel (10) and the use of Li2O is limited in industry due to its high cost.

Jian Yang, Jianqiang Zhang, Oleg Ostrovski University of New South Wales, Australia

Chen Zhang, Dexiang Cai

Baosteel Corporation Research Institute, China

[1]

La Metallurgia Italiana - n. 1 2019

Colata continua In this article, the reaction of CaO-SiO2 and CaO-Al2O3-based mold fluxes in the casting of high-Al steels was investigated based on the composition changes of mold fluxes and steel in both pilot and laboratory tests. The aim of this study is to deepen the understanding of the flux behavior in flux-steel reaction, promoting the design of new mold fluxes for the casting of high-Al steel.

signed mold fluxes. After the temperature of molten steel was stabilized at 1600 °C, a steel sample was acquired. The composition of this steel is indicated as Steel 1. Then, 4 kg Flux 1 was added on the top of molten steel to replace the nonreactive protection slags. Mold flux samples were taken at 900 and 1800 s after Flux 1 was melted; while the steel sample was acquired at only 1800 s whose composition is defined as Steel 2. Afterward Flux 1 was replaced by Flux 2. It was sampled at 600, 900, 1200, 1500 and 1800s, respectively; while the steel sample was acquired at 1800 s. As [Al] concentration dropped to a low level after the reaction between Flux 1 and Steel 1, an additional experiment was arranged for Flux 2 to react with a new steel with a high Al content whose composition is listed as Steel 3. The second set of pilot tests investigated the reaction between CaO-Al2O3-based mold fluxes and steel. The steel sample acquired before the introduction of Flux 3 was marked as Steel 4. Flux 3 was then added and sampled at 600, 900, 1200 and 1500 s. Steel composition was acquired at 1500 s, which was marked as Steel 5. Afterward, Fluxes 4 was used to react with Steel 5 and sampled at 600, 900, 1200, 1500 and 1800s, respectively. The final composition of Steel 5 was obtained at 1800 s.

EXPERIMENTAL PROCEDURE Pilot Tests Both CaO-SiO2 and CaO-Al2O3-based mold fluxes were investigated in the pilot tests. The compositions of mold fluxes and steels used in the pilot tests are given in Tables 1 and 2, respectively. The concentrations of B2O3 and Li2O in mold fluxes were determined using inductively coupled plasma (ICP) analysis; other components were analyzed using X-ray fluoroscopy (XRF). The concentrations of [Al], [Mn] and [Si] in steel were examined using ICP analysis. Two sets of pilot tests were carried out in this study. The first set tested the reaction between CaO-SiO2-based mold fluxes (Fluxes 1 and 2) and high-Al steel. Approximately 150 kg steel was charged in an induction furnace with the protection of non-reactive mold fluxes which were used to avoid oxidation before the introduction of the deTab. 1 – Compositions of Mold Fluxes in Pilot Casting. (mass%) Number

SiO2

Al2O3

CaO

MgO

TiO2

MnO

Na2O

Li2O

B2O3

Flux 1

42.0

6.1

28.4

1.8

3.3

7.4

Flux 2

43.3

3.2

28.2

0.2

4.6

8.6

1.1

8.5

3.4

2.5

18.1

Flux 3

15.3

16.9

40.1

2.4

1.8

2.0

5.0

1.9

9.3

1.7

Flux 4

8.1

28.8

28.3

1.5

1.1

0.1

2.8

8.5

15.8

Tab. 2 – Compositions of Steel in Pilot Casting. (mass%) Number

Steel 1

0.65

0.13

17.7

2.00

Steel 2

0.68

0.77

17.5

1.12

Steel 3

0.64

1.04

17.0

1.80

Steel 4

0.58

0.10

18.1

1.79

Steel 5

0.59

0.36

17.9

1.62

Laboratory Tests The flux-steel reaction between CaO-Al2O3-based mold fluxes and high-Al steel was also carried out in laboratory. CaO-Al2O3based mold fluxes were prepared by mixing reagent chemicals using an agate mortar. The mixed powders were then melted in a graphite crucible at

La Metallurgia Italiana - n. 1 2019

1400 °C for 20 minutes. Afterward, the melt was quenched in water, forming a highly porous glass. It was then crushed to coarse grains with an average diameter of ca. 2 mm. The composition of mold fluxes is listed in Table 3; while the composition range of high-Al steel is given in Table 4. The composition measurements used the same way as in pilot tests.

Continuous casting Tab. 3 – Compositions of Mold Flues in Lab Tests. (mass%) Number

SiO2

Al2O3

CaO

MgO

Na2O

Li2O

B2O3

Flux 5

6.9

36.3

36.8

1.8

5.9

2.1

9.5

Flux 6

6.9

17.7

54.1

1.7

5.5

2.1

11.3

Flux 7

6.9

15.1

57.4

1.8

5.0

2.2

11.0

Tab. 4 – Composition Range of High-Al Steel in Lab Tests. (mass%) Number

Steel 6

0.14-0.20

< 0.15

21-25

1.5-2.5

The flux-steel reaction was carried out in a MoSi2 resistance furnace as illustrated in Figure 1. A MgO crucible (40 x 35 x 81 mm) was used to accommodate steel sample and mold fluxes (8). An Al2O3 protection tube (40 x 35 x 360 mm) was placed above the MgO crucible. The temperature of the sample was measured and calibrated by two thermocouples in the bottom of the furnace. 300 g high-Al steel sample was heated at 1500 °C in the MgO crucible. Highly purified Ar-5H2 gas was flown into the protection tube with a flow rate of 1 L/min in the heating process. After the temperature was stabilized at 1500 °C

for 5 min, a quartz tube (8 x 6 x 1200 mm) was lowered to the crucible through a stainless steel tube (14 x 12 x 700 mm) to acquire steel sample and then quenched into water immediately. Afterward, 30 g coarse grained mold fluxes were added to the MgO crucible through the stainless steel tube. Mold fluxes were assumed to take 1 min to melt completely on molten steel after which zero time was set. Quartz tubes were used to sample liquid steel periodically and then quenched into water. The concentrations of [Al] and [Mn] in acquired steel samples were analyzed using ICP analysis.

RESULTS AND DISCUSSION Flux-Steel Reaction in Pilot Casting Figure 2 show the changes of SiO2, Al2O3, MgO, MnO, Na2O, and B2O3 contents during the reaction between Flux 1 and Steel 1. The concentrations of [Mn], [Al] and [Si] before and after the reaction were shown in Table 5. The significant changes of SiO2 and Al2O3 in mold fluxes were observed. SiO2 content decreased from 42.0 to 2.9 mass% in 1800 s, which was only 7% of the original concentration; while Al2O3 content increased from 6.1 to 47.7 mass% during the same time span. It indicates that the rapid reaction between SiO2 in Flux 1 and [Al] in Steel 1 took place, leading to the accumulation of Al2O3 in the

mold flux and [Si] in the steel. MgO content increased from 1.8 to 5.8 mass% during the experiment, which mainly came from the refractory. Na2O content decreased slightly from 7.4 to 5.9 mass%. It might attribute to the evaporation of Na2Ocontaining phases at high temperature or the reaction with [Al] in steel. MnO content decreased from 3.3 to 0.8 mass% in the experiment while [Mn] concentration slightly decreased by 0.2 mass%. The decrease of MnO in mold fluxes may be caused by the formation of [Mg, Mn]Al2O4 spinel which was reported by Kim et al.(11). This phase is thermodynamically feasible to form when the Al2O3 content is high, but SiO2 is low. Therefore, the flux/steel interface might be a preferred location for the

La Metallurgia Italiana - n. 1 2019

Colata continua precipitation of [Mg, Mn]Al2O4 spinel. The decrease of B2O3 content from 3.4 to 0.3 mass% was observed in mold fluxes, which may result from evaporation, or reaction with [Al]. B2O3

is thermodynamically possible to be reduced by [Al] according to the following equation:

[2]

Fig. 2 â&#x20AC;&#x201C; Concentrations of SiO2, Al2O3, B2O3, MgO, Na2O, MnO, and B2O3 in mold fluxes during the reaction between Steel 1 and Flux 1. Reaction Steel 1/Flux1 Steel 2/Flux2 Steel 3/Flux2 Steel 4/Flux3 Steel 5/Flux4

Time

Before reaction

0.65

0.13

17.7

2.00

After reaction

0.68

0.77

17.5

1.12

Before reaction

0.68

0.77

17.0

1.12

After reaction

0.67

1.09

18.1

0.21

Before reaction

0.64

1.04

17.0

1.80

After reaction

0.67

1.07

17.2

1.26

Before reaction

0.58

0.10

18.1

1.79

After reaction

0.59

0.36

17.9

1.62

Before reaction

0.59

0.36

17.9

1.62

After reaction

0.58

0.45

17.6

0.94

The composition change of Flux 2 during the reaction with Steel 2 is given in Figure 3. The composition change of Steel 2 is listed in Table 5. During the reaction between Flux 2 and Steel 2, SiO2 content decreased from 43.3 to 17.2 mass% in 1800 s, while Al2O3 content increased from 3.2 to 27.5 mass% due to the severe reaction. The reduction of SiO2 was less intense than that in the reaction between Flux 1 and Steel 1 although the initial SiO2 content was close. It may result from the lower [Al] concentration in Steel 2 as [Al] concentration was proved to strongly affect the reaction rate in Reaction [1] based on the La Metallurgia Italiana - n. 1 2019

previous investigation on reaction between CaO-SiO2-based mold fluxes and high-Al steel (8). The changes of Na2O and MgO were similar to those in the reaction between Flux 1 and Steel 1. The concentration of MnO sharply increased in the first 600 s and gradually decreased thereafter. It was noticed that the concentration of MnO in mold fluxes and [Mn] in steel both increased after the reaction between Flux 2 and Steel. There are two possible reasons. The increase of MnO may come from the oxidation of [Mn] by SiO2, which is possible when [Al] concentration is low in steel (8). But once MnO content reached a 15

Continuous casting threshold, it might be reduced by [Al] again. The second explanation is related to the [Mg, Mn]Al2O4 spinel. The extra MnO may come from the remnant [Mg, Mn]Al2O4 spinel formed in the previous experiment. The remnant disintegrated into Flux 2 since Al2O3 content decreased and SiO2 content increased in the initial stage of flux pool after the introduction of Flux 2, which moved the flux composition away from the saturation line of [Mg, Mn]Al2O4 spinel. As a result, MnO content increased in

the initial stage. After 600 s, the accumulation of Al2O3 and reduction of SiO2 pushed the composition close to the saturation line of [Mg, Mn]Al2O4 spinel that could initiate the formation of [Mg, Mn]Al2O4 spinel again, causing the minimization of MnO content in the later stage. But the decrease of MnO was not severe since the SiO2 content in Flux 2 was not as low as that in the reaction between Flux 1 and Steel 1.

Fig. 3 â&#x20AC;&#x201C; Concentrations of SiO2, Al2O3, MgO, Na2O, and MnO in mold fluxes during the reaction between Steel 2 and Flux 2. The composition change of Flux 2 in the reaction with Steel 3 is indicated in Figure 4; while the composition change of Steel 3 is listed in Table 5. SiO2 content decreased from 43.3 to 1.2 mass% during the experiment; while Al2O3 concentration increased from 3.2 to 52.1 mass% in the meantime. The increment

of Al2O3 and the decrement of SiO2 were both more obvious than those in the reaction between Flux 2 and Steel 2. It revealed that a high concentration of [Al] in steel could accelerate Reaction 1. A higher [Al] concentration would facilitate the accumulation of Al2O3 and the reduction of SiO2 in mold fluxes.

Fig. 4 â&#x20AC;&#x201C; Concentrations of SiO2, Al2O3, MgO, Na2O, and MnO in mold fluxes during the reaction between Steel 3 and Flux 2. 16

La Metallurgia Italiana - n. 1 2019

Colata continua In the 2nd set of experiments, two CaO-Al2O3-based mold fluxes were used to react with high-Al steel. Flux 3 was used to react with Steel 4. The composition change of Flux 3 in the reaction with Steel 4 is indicated in Figure 5, and the composition change of Steel 4 is listed in Table 5. The initial concentration of SiO2 in the flux was 15.3 mass%, which was much lower than those in Fluxes 1 and 2. This concentration decreased to 2 mass% after 1500 s; while Al2O3 content increased from 16.9

to 40.0 mass%. In the meantime, [Si] concentration increased from 0.10 to 0.36 mass%, and [Al] concentration dropped from 1.79 to 1.62 mass%. It seems that 15.3 mass% SiO2 in the initial composition of Flux 3 did not obviously retard Reaction 1. Considering the increment of Al2O3 in the mold fluxes, the reductions of Na2O and B2O3 were more likely from the reaction instead of evaporation.

Fig. 5 â&#x20AC;&#x201C; Concentrations of SiO2, Al2O3, B2O3, MgO, Na2O, MnO, and B2O3 in mold fluxes during the reaction between Steel 4 and Flux 3. The composition change of Flux 4 in the reaction with Steel 5 is given Figure 6, and the composition change of Steel 5 is listed in Table 5. Flux 4 is based on CaO-Al2O3 flux system where the initial concentration of SiO2 was as low as 8.1 mass% and CaO/ Al2O3 ratio was 1. SiO2 content only decreased from 8.1 to 5.6 mass% throughout the experiment, which means only 30% of SiO2 was reduced. It was much smaller than the decline of SiO2 content in the previous experiments. Al2O3 content increased from 28.8 to 42.4 mass% in 1800 s, which was also relatively insignificant compared with those in the previous experiments. The major source of the Al2O3 accumulation presented in Figure 6 was likely to stem from the reduction of B2O3 whose concentration dropped from 15.8 to 5.4 mass% during

La Metallurgia Italiana - n. 1 2019

the experiment. Aside from the contribution to the increase of Al2O3, the abrupt increase of MnO could attribute to the reaction between [Mn] and B2O3. According to the results, a low concentration of SiO2 (< 8 mass%) could kinetically benefit the restraint on the accumulation of Al2O3. The low addition of SiO2 requires a large amount of fluxing agent, e.g. B2O3, to decrease the liquidus temperature and adjust other properties. However, the reduction of B2O3 introduced an additional increment of Al2O3 in mold fluxes. Therefore, the balance between the reaction rate and melting property could be a critical factor in the design of CaO-Al2O3based mold fluxes.

Continuous casting

Fig. 6 – Concentrations of SiO2, Al2O3, B2O3, MgO, Na2O, MnO, and B2O3 in mold fluxes during the reaction between Steel 5 and Flux 4. Flux-Steel Reaction in Laboratory Tests Three types of CaO-Al2O3-based mold fluxes (Fluxes 5, 6, and 7) were studied based on the reaction between these mold fluxes and a high-Al steel (Steel 6). Different CaO/Al2O3 ratios were applied to the studied mold fluxes. The changes of [Al] content in steel are compared in Figure 7(a). In Flux 5, [Al] content decreased from 1.65 to 0.53 mass% in 1200 s and 0.37 mass%

in 3600 s. In Flux 6, [Al] content decreased from 1.49 to 0.06 mas% in 3600 s. In Flux 7, [Al] concentration decreased from 1.74 to 0.63 mass% in the first 11 minutes, but no further sample acquisition was made due to a lab incident. It seems that reaction was most intense in first 1200 s. As [Al] concentration should be very low due to the strong affinity of Al and O, the flux of [Al] in steel/flux interface could be written as follows (8):

[3] where [Al%]t and [Al%]0 are [Al] concentration at time t and zero, kAl is the mass transfer coefficient of [Al] (m/s), A is the cross-section area (m2), ρsteel is the density of steel (g/ m3) calculated according to reference (12), Wsteel is the total weight of steel (g), t is reaction time (s). The relationship between -ln([Al%]t/[Al%]0) and t is plotted in Figure 12(b). As kAl is proportional to the slope of the curve, it is indicated that the decrease of CaO/Al2O3 ratio decelerated the oxidation rate

of [Al]. Besides, the oxidation of [Al] became rather slower as experiment proceeded in the reaction with the kAl decreasing up to 6.8x10-6 m/s when CaO/Al2O3 ratio = 1; while it kept relatively constant after 1 min with the kAl = 4.0x10-5 m/s during the experiment when CaO/Al2O3 ratio = 3. The oxidation rate of [Al] was also relatively constant in the first 660 s with the kAl = 5.7 x 10-5 m/s when CaO/Al2O3 ratio = 4.

Fig. 7 – Concentration change of Al in steel during the reaction between Steel 6 and Flux 5/Flux 6/Flux 7. 18

La Metallurgia Italiana - n. 1 2019

Colata continua The [Al] losses in the liquid steel was higher in the lab experiments compared with those in pilot tests due to the ratio of mold fluxes to steel was much higher in lab experiments. The change of [Al] concentration in lab and pilot tests may not be comparable in value but their tendency should be similar. CONCLUSIONS 1. CaO-SiO2-based mold fluxes showed a severe reaction when casting high-Al steel, which resulted in a significant increase of Al2O3 content and a decrease of SiO2 content during experiments. It makes the in-mold performance undesirable and unpredictable; 2. CaO-Al2O3-based mold fluxes showed better a stability in

composition when casting high-Al steel. The changes of Al2O3 and SiO2 contents were not as severe as those in CaO-SiO2-based mold fluxes. However, the variation of B2O3 may also affect the melting property of mold fluxes; 3. The increase of CaO/Al2O3 ratio could accelerate the reaction between high-Al steel and CaO-Al2O3-based mold fluxes based on the variation of [Al] content in liquid steel. ACKNOWLEDGEMENT CONCLUSIONS Financial supports by Baosteel-Australia Joint Research & Development Centre (BAJC) and Australian Research Council Industrial Transformation Research Hub are greatly acknowledged.

REFERENCES [1]

Steinmetz DR, Jäpel T, Wietbrock B, Eisenlohr P, Gutierrez-Urrutia I, Saeed-Akbari A, et al. Revealing the strain-hardening behavior of twinning-induced plasticity steels: Theory, simulations, experiments. Acta Materialia. 2013; 61(2):494-510.

[2]

Gutierrez-Urrutia I, Raabe D. Dislocation and twin substructure evolution during strain hardening of an Fe-22wt.% Mn-0.6wt.% C TWIP steel observed by electron channeling contrast imaging. Acta Materialia. 2011; 59(16):6449-62.

[3]

Grässel O, Krüger L, Frommeyer G, Meyer LW. High strength Fe-Mn-(Al, Si) TRIP/TWIP steels development-properties-application. International Journal of Plasticity. 2000; 16(10):1391-409.

[4]

Sato K, Ichinose M, Hirotsu Y, Inoue Y. Effects of Deformation Induced Phase Transformation and Twinning on the Mechanical Properties of Austenitic Fe&ndash;Mn&ndash;Al Alloys. ISIJ International. 1989; 29(10):868-77.

[5]

Pierce DT, Jiménez JA, Bentley J, Raabe D, Wittig JE. The influence of stacking fault energy on the microstructural and strainhardening evolution of Fe–Mn–Al–Si steels during tensile deformation. Acta Materialia. 2015; 100:178-90.

[6]

Dumay A, Chateau JP, Allain S, Migot S, Bouaziz O. Influence of addition elements on the stacking-fault energy and mechanical properties of an austenitic Fe–Mn–C steel. Materials Science and Engineering: A. 2008; 483-484:184-7.

[7]

Cho J-W, Blazek K, Frazee M, Yin H, Park JH, Moon S-W. Assessment of CaO-Al2O3 Based Mold Flux System for High Aluminum TRIP Casting. ISIJ International. 2013; 53(1):62-70.

[8]

Kim M-S, Lee S-W, Cho J-W, Park M-S, Lee H-G, Kang Y-B. A Reaction Between High Mn-High Al Steel and CaO-SiO2-Type Molten Mold Flux: Part I. Composition Evolution in Molten Mold Flux. Metallurgical and Materials Transactions B. 2013; 44(2):299308.

[9]

Kang Y-B, Kim M-S, Lee S-W, Cho J-W, Park M-S, Lee H-G. A Reaction Between High Mn-High Al Steel and CaO-SiO2-Type Molten Mold Flux: Part II. Reaction Mechanism, Interface Morphology, and Al2O3 Accumulation in Molten Mold Flux. Metallurgical and Materials Transactions B. 2013; 44(2):309-16.

[10] Blazek K, Yin H, Skoczylas G, McClymonds M, Frazee M. AIST Transactions-Development and Evaluation of Lime-Alumina Based Mold Powders. Iron and Steel Technology. 2011; 8(8):231. [11] Kim DJ, Park JH. Interfacial Reaction Between CaO-SiO2-MgO-Al2O3 Flux and Fe-xMn-yAl (x = 10 and 20 mass pct, y = 1, 3, and 6 mass pct) Steel at 1873 K (1600 °C). Metallurgical and Materials Transactions B. 2012; 43(4):875-86. [12] Thu Hoai L, Lee J. Density of Liquid Fe-Mn-C Alloys. Metallurgical and Materials Transactions B. 2011; 42(5):925-7. La Metallurgia Italiana - n. 1 2019

Continuous casting

Influence of selected alloy additions on time mixing for pulse-step method of liquid steel alloying in the tundish A. Cwudziński, J. Jowsa

In the continuous steel casting process, where the tundish performs the function of a device supplying liquid steel to the mould, the applying tundish to alloying process is quite interesting. Therefore the scientific aim of the present work is to obtain new basic information on chemical homogenization process of liquid steel with alloy additions in the tundish during using pulse–step alloying method. Authors checked the effect of the density of selected alloy additions on the process of their mixing with the liquid steel. Within the work, the following alloys additions were considered: Cu, Mn, Al, Co, Ni-Cr and Ti-Al. All alloy additions was introduced to liquid steel by pulse-step method. The authors employed the numerical modeling technique to demonstrate the process of alloy addition mixing during liquid steel flowed through the one strand slab tundish. Ansys-Fluent® program was used for numerical simulation. For the description of the alloy additions mixing with liquid steel time mixing was calculated.

KEYWORDS: TUNDISH – PULSE-STEP ALLOYING METHOD – PROPERTIES OF ALLOYS – NUMERICAL MODELING

INTRODUCTION In the continuous steel casting process, where the tundish performs the function of a device supplying liquid steel to the mould, the applying tundish to alloying process is quite interesting. Therefore the scientific aim of the present work is to obtain new basic information on chemical homogenization process of liquid steel with alloy additions in the tundish during using pulse–step alloying method. Density is one of the key properties of alloy additions, which is expected to determine their mixing with liquid steel [1-2]. Therefore Authors checked the effect of the density of selected alloy additions on the process of their mixing with the liquid steel during alloying by pulsestep method. Alloy additions lighter than liquid steel will float, while those heavier than liquid steel will flow down towards the metallurgical vessel (furnace, ladle, tundish) bottom. For this reason, in a typical secondary metallurgy (ladle furnace) treatment, through the gas injection system, a mixing process is stimulated, thus smoothing out the effect of spontaneous spreading of additives within the liquid steel. The ladle furnace enables also the correction of liquid steel temperature, which is necessary due to the need for delivering additional heat for the alloy addition melting process. By contrast, in the standard tundish, the mixing process is determined by the feed stream and flow control devices (FCDs), which should provide the optimum hydrodynamic conditions within the entire tundish. Tundishes are equipped with various flow control devices, which can generate additional energy to effectively improve the mixing in the tundish working space [3-6]. Obviously the most effective FCDs beyond tundish pouring zone are argon curtains. Steel

temperature correction is also an issue, because during steel casting, the liquid steel temperature lowers, and tundishes are not normally furnished with metal reheating systems. Therefore, the process of liquid steel alloying in the tundish should be preceded by an appropriate mass and thermal balance in order to eliminate any factors that might disrupt the continuous casting process. Within the work, the following alloys additions were considered: Cu, Mn, Al, Co, Ni-Cr and Ti-Al. All alloy additions was introduced to liquid steel flowed through one strand tundish by pulse-step method. METHODOLOGY OF SIMULATIONS It was assumed in computer simulations that the alloy additions would have a liquid form. Therefore, the process of alloy addition dissolution and melting in the steel was not taken into ac-

Adam Cwudziński, Jan Jowsa

Department of Metals Extraction and Recirculation, Faculty of Production Engineering and Materials Technology, Czestochowa University of Technology, 19 Armii Krajowej ave, 42-200 Czestochowa, Poland

La Metallurgia Italiana - n. 1 2019

Colata continua count. This simplification was possible, because industrial experiments made previously had shown that, with an appropriately small size of alloy addition pieces, the duration of alloy addition dissolution in the steel was short enough to not significantly disturb the macro-mixing process [7]. In computer simulations, two types of nickel differing slightly in density and greatly in viscosity were considered. Nickel no. 1 was the subject of investigations reported in studied [8-10], and its properties were assumed based on data provided in studies [11-12]. Whereas, the properties of nickel no. 2 were determined using relationships based on relationships correlated with the process temperature. Obviously, as increasingly accurate informa-

tion is entered to computer programs, the user gets a greater chance of obtaining more valuable results from the application point of view. Computer simulations were performed in the program Ansys-Fluent, using the Species model. The mathematical model and the boundary conditions are detailed in study [7]. The computer simulations were done for a sequence of casting of 0.225x1.5 m slabs at a speed of 0.9 m/ min and pouring temperature of 1823 K. To illustrate quantitatively the liquid steel chemical homogenisation process, mixing curves were recorded at tundish outlet. The level of chemical homogenisation was determined using the following relationship:

[1] where: Cf - final concentration of alloy at tundish outlet (wt%),CPSM - dimensionless concentration of alloy for PSM, Ct - temporary concentration of alloy (wt%), C0 - initial concentration of alloy (wt%). Based on the mixing curves, the dimensionless mixing time (DMT) was calculated. The DMT is defined as the time, after which the minimum required liquid steel chemical homogenisation level is maintained, which should amount to at least 95%. A reference level for calculating the mixing time required for achieving a chemical homogenization level of 95% was the alloy addition quantity recorded at the tundish outlet at the time of finishing the casting of the heat. The time interval was expressed by DTM defined by the ratio of the actual time to the average time. The average time for the tundish under examination was 726 seconds. TUNDISH AND PULSE-STEP ALLOYING METHOD Based on the results reported in study [8] it has been demonstrated that the pulseâ&#x20AC;&#x201C;step method reduces the time needed

for attaining the level of 95% chemical homogenization. It has also been found that the effectiveness of the chemical homogenization process is determined by: the equipment of the tundish and the location of feeding an alloy addition [8-10]. Therefore, alloy additions in the form of either pure metals or alloys were introduced to the tundish in one of the most advantageous zones, i.e. in the tundish pouring zone between the ladle shroud and the tundish rear side wall. The tundish is a part of the machine for continuous casting of slabs. Checks were also made in the study to see how selected alloy additions would behave in a tundish with a varying configuration of its internal space equipment. Tundish variant no. 1, in which the tundish was furnished with a Subflux Turbulence Controller (STC) and a low dam, is shown in Figure 1. In variant no. 2, the STC geometry and dam height were changed. While in variant no. 3 of tundish equipment, the same STC as in variant no. 1 was used, but the dam height was changed. A detailed description of the FCD is provided in studies [8-9].

Fig. 1 â&#x20AC;&#x201C; One strand tundish with alloy addition position and flow control devices for variant no. 1 La Metallurgia Italiana - n. 1 2019

Continuous casting Alloy additions were introduced to the steel first in a one-off batch, and then in a continuous manner in a quantity that allowed the correction of chemical composition to increase the concentration of the alloy addition by 0.056 wt%. The table 1 provides the physicochemical properties of the alloy additions discussed in the paper. For pure metals, the data shown in Table 1 were developed based on studies [11-15]. Whereas for the alloys NiCr and TiAl, the data given in studies [16-19]

were used. It was assumed that the both binary alloys contained the identical content of either constituents, i.e. 50% each. This composition assured that their liquidus temperatures were below the casting temperature of the liquid steel grade under consideration. The heat capacity values for considered alloys was calculated in the computer program FactSage. The additive in the form of steel scrap had properties similar to those of the steel grade being cast.

Tab. 1 – One strand tundish with alloy addition position and flow control devices for variant no. 1 Alloy addition

Density, kg/m3

Viscosity, Pa·s

Heat capacity, J/kg·K

Thermal conductivity, W/m·K

Diffusivity, m2/s

Aluminium

2100

0.00052

1180

8.6·10-09

Manganese

5470

0.00344

840

4.6·10-09

Copper

7660

0.00159

520

163

4.8·10-09

Cobalt

7690

0.00261

690

4.2·10-09

Nickel no. 1

7650

0.00470

556

5.3·10-09

Nickel no. 2

7790

0.00159

556

5.3·10-09

Nickel-chromium

6900

0.00400

806

4.5·10-09

Titanium-aluminium

3360

0.00185

1005

6.0·10-09

Steel scrap

7010

0.00700

750

4.8·10-09

RESULTS AND DISCUSSION MIXING CURVES AND TIMES FOR CONSIDERED TUNDISH VARIANTS From the computer simulations, an illustration of the chemical homogenization process was derived in the form of mixing curves (Fig. 2). In figure 2, the zone of 95% chemical homogenization is marked off with broken lines. The distribution of the mixing curves varies, depending on the alloy addition type under consideration and the tundish equipment variant. While the change in STC does not essentially change the hydrodynamic conditions, the change of the dam height has already a definite influence on the alloy addition spreading within the bulk of liquid steel. In the presented diagrams, the mixing curves feature characteristic peaks that take on the form of a gradually dwindling sinusoid. In the case of tundish equipment variant no. 1, 3 peaks occur, while tundish variants no. 2 and no. 3 have 5 peaks, each. The increase in the number of peaks indicates an intensification of the mixing process in the higher-dam tundish variants. Hence, the maximum alloy addition concentrations in these variants are lower. The position of the peaks in relation to the X axis changes very slightly, while their position relative to the Y axis changes significantly, especially for alloy additions, such as Al, TiAl and Mn. Figure 2 represents mixing times for 22

the alloy addition under discussion. The shorter mixing times, i.e. 0.39, 0.47 and 0.87 DMT, were obtained for Ni, Co and Cu, respectively. Whereas, the longest mixing times were obtained for Al and TiAl, amounting to, respectively, 2.5, 2.6 and 2.61 and 2.23, 2.38 and 2.34 DMT. The difference between the best and the worst mixing times is 2.22 DMT, which implies that the effectiveness of the alloying process will be determined not only by tundish equipment and the alloy addition feed location, but also by the type of the alloy addition itself. The obtained mixing times suggest that alloy additions should be considered in two groups, i.e. those of a density, respectively, lower and higher than the cast steel density. The first group will include Al, Mn, TiAl and NiCr, while the second group, Ni, Cu and Co. From the obtained results it has been found that particular tundish equipment variants influence differently on the both groups of alloy additions in terms of the mixing time. For example, the tundish according to variant no. 1 better influences on the alloying process, reducing the mixing time of lighter alloys. By contrast, in the case of alloy additions heavier that the cast steel grade, a definite increase in mixing time is noticed in the tundish variant no.1. This phenomenon is reverse in the tundish variants no. 2 and 3.

La Metallurgia Italiana - n. 1 2019

Colata continua a)

Fig. 2 â&#x20AC;&#x201C; Chemical homogenisation process for pulse-step liquid steel tundish alloying method: mixing curves for tundish variant no. 1, b) mixing curves for tundish variant no. 2, c) mixing curves for tundish variant no. 3, d) mixing time for considered tundish and alloy additions

La Metallurgia Italiana - n. 1 2019

Continuous casting INFLUENCE OF ALLOYS PROPERTIES ON LIQUID STEEL ALLOYING PROCESS The obtained mixing time results for individual alloy additions have demonstrated that the chemical homogenization process will influenced by their physicochemical properties. Figure 3 illustrates the results for the effect of alloy addition density on the chemical homogenization process. For the density, a distinct effect of this parameter on the mixing time can be seen. The higher the specific density an alloy addition has, the shorter the mixing time can be attained, regardless of the equipment of the tundish internal space. In the case of alloy additions of a density higher than that of the liquid steel, no effect of viscosity on the chemical homogenization process can be noticed. By contrast, for alloy additions lighter than the liquid steel, their viscosity

increases linearly with density, so it is hard to point out clearly which quantity influences the mixing time more intensively. Even though less viscous additives should more readily spread within the liquid steel, yet this process may effectively disturb the alloy addition density that is lower than that of the steel. In this connection, the mixing time results for nickel nos. 1 and 2 are shown in figure 4. Both alloy additions differed in kinematic viscosity. As can be seen from the results in figure 4, identical mixing times were obtained for an alloy addition with a different kinematic viscosity. With a similar density value, in the alloy addition viscosity range under consideration, the viscosity of the alloy addition was found not to be a quantity decisive of the chemical homogenization process.

Fig. 3 â&#x20AC;&#x201C; Influences of selected alloys properties for mixing time a) alloys density, b) alloys viscosity

Fig. 4 â&#x20AC;&#x201C; Mixing time for nickel with different value of kinematic viscosity

La Metallurgia Italiana - n. 1 2019

Colata continua BEHAVIOR OF AL (LIGHTER ALLOY) AND NI (WEIGHTIER ALLOY) IN THE LIQUID STEEL To more accurately assess the process of chemical homogenization as a result of the modified flow of liquid steel in the tundish working space and the effect of the physicochemical properties of alloy additions, detailed results are presented for steel movement and Al and Ni concentrations in tundish variant no. 1. Figure 5 shows the pattern of liquid steel flow in four characteristic planes. Three of them are arranged in parallel to the tundish bottom at three heights, namely 0.25, 0.57 and 0.918 m. While the other one is arranged perpendicularly to the tundish bottom. This plane intersects the tundish longitudinally through the pouring and the nozzle zones. Additionally liquid steel streams in the working volume was generated. In the alloy addition feed zone, steel streams are visible, which flow in clearly from the ladle shroud region towards the rear and side tundish wall. While the STC causes the stopper rod

zone feed steel streams to flow along the tundish until the zone limited by the front wall. On the other hand, immediately at the tundish bottom, distinct reverse streams form, which have two regions of intensive steel circulation. Whereas, at the midheight, the metal circulates from the stopper rod zone to the STC zone, dividing thereby the facility into a zone of reverse streams flowing in to the feed zone and a zone of streams feeding the stopper zone. An inflow of reverse streams to the STC zone can also be seen in the plane perpendicular to the bottom. These streams combined with the feed streams form a region of vertical steel circulation within the tundish working space. A strong liquid steel circulation region is also observed immediately under the alloy addition feed zone. While in the stopper rod system zone, the streams feeding the tundish nozzle are descending in character, with small recirculation zones located at the liquid steel free surface.

Fig. 5 â&#x20AC;&#x201C; Liquid steel flow paths in the tundish The liquid steel hydrodynamic pattern formed in the tundish working space should definitely influence the distribution of the introduced alloy addition. Figure 6 shows the distribution of two alloy additions, Al and Ni, in selected planes located in the tundish. The dark blue color denotes the alloy addition concentration (pure alloy addition). The time of 180 seconds corresponds approximately to the time of attaining the first peak on the mixing curve. Hence it can be seen that for Al, its degree of dissipation in the liquid steel is much smaller than that of Ni. The behavior of an alloy addition in the form of Al, as recorded within the tundish working space, is consistent with observations made by steel industry engineers concerning the difficulties associated with feeding Al or Ca to the tundish steel La Metallurgia Italiana - n. 1 2019

and their effective interaction with the steel being cast. The liquid steel hydrodynamic pattern formed in the alloy addition feed zone forces the alloy addition to move towards the tundish side wall. Hence, the distribution of the alloy addition in the central tundish part is fairly asymmetric. A definitely higher Al concentration is also visible in the upper tundish zone, which is caused, on the one hand, by ascending streams generated by the STC and, on the other hand, by the Al density being much lower compared to liquid steel. While the Ni density, being lower than that of liquid steel, contributes to a better interaction between the alloy addition and individual steel streams and its more efficient dispersion within the liquid steel bulk.

Continuous casting a)

Fig. 6 – Behavior of alloy addition in the liquid steel during selected period of time after alloy feeding: a) Al - 30 sec, b) Al - 90 sec, c) Al - 180 sec, d) Ni - 30 sec, e) Ni - 90 sec, f) Ni - 180 sec CONCLUSIONS Based on the computer simulations carried out, it has been found that: • The magnitude of mixing time needed for attaining a 95% chemical homogenization is influenced by the alloy addition type. • The lighter the alloy addition relative to the liquid steel, the harder it is distributed within the liquid steel bulk, in spite of formed hydrodynamic patterns existing in the tundish. • The mixing time as a function of the alloy addition assumes a relationship similar to a linear one, which implies that as the 26

alloy addition density increases, especially above the density of the steel grade being cast, the mixing time shortens. • The shortest alloy addition mixing times with the constant at a level of 0.39 DMT were obtained for the tundish according to tundish equipment variant no. 2 and for feeding alloy additions in the form of Ni, Co and Cu. ACKNOWLEDGEMENTS This scientific work has been financed from the resources of National Science Centre, Poland in the years 2017-2019 as the Research Project No. 2016/23/B/ST8/01135 La Metallurgia Italiana - n. 1 2019

Colata continua REFERENCES [1]

Tanaka M, Mazumdar D, Guthrie RIL. Motions of alloying additions during furnace tapping in steelmaking processing operations. Metall Mater Trans B. 1993; 24:639-648.

[2]

Mazumdar D, Guthrie RIL. Motions of alloying additions in the CAS steelmaking operations. Metall Mater Trans B. 1993; 24:649-655.

[3]

Chang S, Zhong L, Zou Z. Simulation of flow and heat fields in a seven-strand tundish with gas curtain for molten steel continuous-casting. ISIJ Int. 2015; 55:837-844.

[4]

Vargas-Zamora A, Morales RD, Díaz-Cruz M, Palafox-Ramos J, Barreto-Sandoval J de J. Inertial and buoyancy driven water flows under gas bubbling and thermal stratification conditions in a tundish model. Metall Mater Trans B. 2004; 35:247-257.

[5]

Merder T, Warzecha M. Optimization of a six-strand continuous casting tundish: industrial measurements and numerical investigation of the tundish modifications. Metall Mater Trans B. 2012; 43:856-868.

[6]

Chen D, Xie X, Long M, Zhang M, Zhang L, Liao Q. Hydraulics and mathematics simulation on the weir and gas curtain in tundish of ultrathick slab continuous casting. Metall Mater Trans B. 2014; 45:392-398.

[7]

Cwudziński A. Numerical simulations and industrial experiments of liquid steel alloying process in one strand slab tundish. Ironmak Steelmak. 2015; 42:132-138.

[8]

Cwudziński A. Pulse-step method for liquid steel alloying in one strand slab tundish. Ironmak Steelmak. 2015; 42:373-381.

[9]

Cwudziński A. Numerical, physical and industrial studies of liquid steel chemical homogenisation in one strand tundish with subflux turbulence controller. Steel Res. 2015; 86:972-983.

[10] Cwudziński A. Numerical simulation of the liquid steel alloying process in a one-strand tundish with different additions postitions and flow control devices. Metall Res Techn. 2015; 112:308(1-12). [11] Guthrie RIL. Engineering in process metallurgy. Oxford: Clarendon press; 1992. 64, 483. [12] Iida T, Guthrie RIL. The physical properties of liquid metals. Oxford: Clarendon press; 1993. 217. [13] Iida T, Guthrie RIL, Isac M, Tripathi N. Accurate predictions for viscosities of several liquid transaction metals, plus barium and strontium. Metall Mater Trans B. 2006; 37:403-412. [14] Valencia JJ, Quested PN. Thermophysical properties. ASM Handbook Committee. ASM Int; 2008. 468-481. [15] Meyer A. The measurement of self-diffusion coefficients in liquid metals with quasielastic neutron scattering. EPJ web of Conference. 2015; 83:01002(1-7). [16] Mukai K, Xiao f. Density of Ni-Cr alloy in liquid and solid-liquid coexistence states. Mater Trans. 2002; 43:1153-1160. [17] Zhou H, Wang HP, Chang J, Wei B. Surface tension of substantially undercooled liquid Ti-Al alloy. Phil Mag Lett. 2010; 90:455462. [18] Wessing JJ, Brillo J. Density, molar volume and surface tension of liquid Al-Ti. Metall Mater Trans B. 2017; 48:868-882. [19] Liu Y, LV X, Bai C. Evaluation model for viscosity of Fe-Ni-Cr alloys using gibbs free energy of mixing and geometric methods. ISIJ Int. 2017; 52:1296-1302.

La Metallurgia Italiana - n. 1 2019

Continuous casting

A novel methodology to evaluate surface cracking risk during strand straightening in continuous casting G.Poltarak, S.Ferro, C.Cicutti

A modeling methodology is developed in order to estimate the surface cracking of round bars during straightening in the continuous casting of steel. For this purpose, the termo-mechanical state of the bar is acquired by finite element models, and a material failure criterion is codified afterwards. The output of this methodology is a surface cracking indicator. Hot ductility tests are performed in a Gleeble system to obtain the brittle temperature range of the material. Several cases are modeled, taking into account different bar diameters, casting speeds and steel compositions. These cases are then compared to surface defect measurements performed in a specific steel-shop in order to evaluate the suitability of the model.

KEYWORDS: CONTINUOUS CASTING – CRACKING – MATHEMATICAL MODEL HOT DUCTILITY – PROCESS VARIABLES

INTRODUCTION Surface quality in continuous casting products represents a constant challenge in steelmaking research, both in high productivity scenarios and in the manufacturing of high specification materials, since surface cracking can generate defects in the rolled material and even its scrapping. In this article, a methodology is developed whose objective is to estimate the cracking risk in round bars during the straightening in the continuous casting machine. This methodology consists of the thermo-mechanical finite element modeling of the solidifying bar followed by the coding of a failure criterion. First, the heat transfer equations are solved from the meniscus to the billet cut, obtaining the distribution of phase and temperature along the strand. Then, a 3D analysis is carried out using an Eulerian model that provides stress, strain and strain rate distribution. These parameters feed a cracking criterion that calculates the surface cracking risk in the most stretched fiber. Finally, a numerical indicator is obtained in order to quantify how critical the casting conditions are. The developed methodology is flexible enough to analyze the influence of casting speed changes, bar diameter, steel composition and geometric distribution imposed by pinch-roll units. Model results are compared with statistical data obtained from the surface inspection of final products in a large number of heats. A reasonable agreement between model calculations and defect index in the final product is obtained.

cracking criterion. A thermal simulation is first carried out using a heat transfer model, which calculates temperature and phase distribution along the bar. Thermal results are then used as an input for the mechanical model, whose main output are strain and strain rate fields. Finally, a surface cracking risk indicator is calculated by the codification of a failure criterion. HEAT TRANSFER MODEL Heat transfer problem is solved numerically by an in-house model [1] under the hypothesis that longitudinal heat flow is much lower than the radial flow. Hence, the domain is a travelling transversal section of the casting bar where the 2D heat transfer equation is solved,

[1]

Guillermo Poltarak, Sergio Ferro, Carlos Cicutti

Tenaris R&D Steelmaking, Argentina

MODELING SEQUENCE A modeling methodology is developed, consisting of a thermomechanical simulation followed by the codification of a surface 28

La Metallurgia Italiana - n. 1 2019

Colata continua where â&#x201E;&#x2039; is the volumetric enthalpy, đ?&#x2018;&#x2DC; the thermal conductivity, and đ?&#x2018;&#x17E;đ?&#x2018;&#x2030; the volumetric heat generation. Eq. 1 has the following boundary conditions: đ?&#x2018;&#x2021;=đ?&#x2018;&#x2021;0 initial bar temperature at the meniscus (liquid steel) đ?&#x2018;&#x17E;â&#x2C6;&#x2122;đ?&#x2018;&#x203A;=đ?&#x2018;&#x17E;đ?&#x2018;&#x161;đ?&#x2018;&#x153;đ?&#x2018;&#x2122;đ?&#x2018;&#x2018; heat flux imposed at the mold (primary cooling) [2] đ?&#x2018;&#x17E;â&#x2C6;&#x2122;đ?&#x2018;&#x203A;=â&#x201E;&#x17D;(đ?&#x2018;&#x2021;â&#x2C6;&#x2019;đ?&#x2018;&#x2021;đ?&#x2018;?đ?&#x2018;˘đ?&#x2018;&#x2122;đ?&#x2018;&#x2DC;) convection and radiation to the environment at bar surface where đ?&#x2018;&#x17E;=â&#x2C6;&#x2019;đ?&#x2018;&#x2DC;â&#x2C6;&#x2021;đ?&#x2018;&#x2021; is the heat flux according to Fourierâ&#x20AC;&#x2122;s Law. Convective heat transfer is used to simulate air and water sprays heat losses, and also to model radiation to the environment, by the application of an effective convection coefficient h which takes into account both effects. Eq. 1 is solved numerically by the finite elements method, using an implicit time integration. Nonlinearities, such as phase changes and parameters dependence with temperature, require the implementation of an iterative algorithm. This thermal model has been applied to optimize casting operative variables [3-4] and was validated with plant measurements [5] taken from different steel-shops. MECHANICAL MODEL The mechanical simulation is carried out using a 3D finite element eulerian model [6] which considers the metal as a vi-

scoplastic fluid. Different material states, such as liquid steel, mushy zone, solid material and air are modeled by changing viscosity magnitude along the bar. The mesh is fixed and the freesurface of the material is defined by an auxiliary variable called pseudo-concentration, which is updated through the transport equation. After the free surface is determined, the velocity field at the material has to satisfy the equilibrium equations. The momentum equation leads to the principle of virtual power,

[2] where đ?&#x2018;&#x2030; is the material calculation domain, đ?&#x153;&#x17D; is the Cauchy stress tensor, đ?&#x153;&#x20AC;Ě&#x2021; the strain rate tensor, đ?&#x2018;&#x201C;đ?&#x2018;&#x2030; the volumetric forces vector, đ?&#x2018;˘Ě&#x2021; the material velocity vector, đ?&#x2018;&#x2020;đ?&#x153;&#x17D; the domain boundaries, and đ?&#x2018;Ą the surface forces vector. The model entry section is defined by imposing the initial pseudo-concentration corresponding to the bar diameter, and the casting speed. Rolls are considered as rigid surfaces. As the material cannot penetrate them, the velocity in the direction of the die is imposed as zero. Finally, the exit section is assumed to be completely solid, meaning that the velocity is uniform in that section.

Fig. 1 â&#x20AC;&#x201C; Boundary conditions of the mechanical analysis.

Material mechanical behavior is given as a function of temperature, strain and strain rate. A material law developed by Kozlowski et al. [7] is employed in the austenitic range (i.e. when temperature is lower than the corresponding to a 10%

content of ferrite, Eq. 3a), and a law proposed by Li and Thomas [8] in the ferritic range (when temperature is higher than the corresponding to a 10% content of ferrite, Eq. 3b).

[3a]

La Metallurgia Italiana - n. 1 2019

Continuous casting [3b] being đ?&#x153;&#x17D;đ?&#x2018;&#x17D; and đ?&#x153;&#x17D;đ?&#x2018;&#x201C; the equivalent stresses of each phase, đ?&#x153;&#x17D;0 an initial structure constant, đ?&#x153;&#x20AC; the equivalent strain, đ??ś and đ?&#x2018;&#x201C;đ??ś parameters which depend on carbon content, đ?&#x2018;&#x161;, đ?&#x2018;&#x203A;, đ?&#x2018;&#x17D;đ?&#x153;&#x20AC; and đ?&#x2018;&#x203A;đ?&#x153;&#x20AC; functions of temperature, đ?&#x2018;&#x201E; the activation energy constant, and đ?&#x2018;&#x2021; the temperature, previously calculated by the thermal model. Phase changes, solidus and liquidus temperatures are extracted from a previously developed segregation model [9] as a function of steel chemical composition. Thermal strains and bar ovalization are analyzed separately and are not included in this study. SURFACE CRACKING The predominant cracking mechanisms of solid steel between 700 and 1000 Â°C is originated by the precipitation of ferrite in the grain boundaries. This softer phase promotes intergranular cracks and is manifested macroscopically as a ductility loss,

whose effects are increased by micro-alloys such as niobium and vanadium [10]. Reduction of area tests, at controlled values of temperature and strain rate, are applied to measure this phenomenon. If the sample undergoes a constriction (smaller final section) the material is ductile and able to absorb a high deformation before fracture. If, however, the final section is similar to the initial, the material is brittle and breaks suddenly. In other words, the critical strain that the material can withstand before breaking is closely related to the reduction of area. While temperature and strain rate can be kept under control in the laboratory, the situation is different in the mill, since these parameters fluctuate along the bar. Schwerdtfeger [11] developed a failure criterion based on reduction of area trials, which estimates the critical strain that the material is able to resist as,

[4b]

where đ?&#x2018;&#x201C;đ?&#x2018;&#x201D;đ?&#x2018;&#x;đ?&#x2018;&#x17D;đ?&#x2018;&#x2013;đ?&#x2018;&#x203A;, đ?&#x2018;&#x201C;đ?&#x2018; đ?&#x2018;&#x2019;đ?&#x2018;&#x201D;đ?&#x2018;&#x;đ?&#x2018;&#x2019;đ?&#x2018;&#x201D; and đ?&#x2018;&#x201C;đ?&#x2018;&#x203A;đ?&#x2018;&#x153;đ?&#x2018;Ąđ?&#x2018;?â&#x201E;&#x17D; are parameters depending on the grain size, solute segregation at grain boundaries and oscillation marks, and đ?&#x2018;&#x2026;đ??´ is the reduction of area at each point. Although Schwerdtfeger proposed an expression to calculate the

reduction of area diagram based on several parameters, laboratory tests to evaluate the ductility of specific steel grades were performed. Finally, a â&#x20AC;&#x153;damage integralâ&#x20AC;? that takes into account the cumulative damage at each fiber of the bar is calculated,

[5]

where đ?&#x2018;&#x2013;=1,2,â&#x20AC;Ś,đ?&#x2018;&#x203A; represents each time step, Î&#x201D;đ?&#x153;&#x20AC;đ?&#x2018;&#x2013; is the strain increase at time đ?&#x2018;&#x2013;, and đ?&#x153;&#x20AC;đ?&#x2018;? đ?&#x2018;&#x2013; is the critical strain the material can withstand at that moment, according to Eq. 4. If đ??żđ?&#x2018; đ?&#x2018;˘đ?&#x2018;&#x;đ?&#x2018;&#x201C; exceeds a value of 1, surface cracking is likely to take place. EXPERIMENTAL PROCEDURE REDUCTION OF AREA DETERMINATION Gleeble tests In order to assess the reduction of area curves of two different steel grades, tension tests were performed in a Gleeble 3500 system. The composition of the tested materials is presented in Table 1. For these tests, 10 mm diameter cylindrical samples

were machined from hot rolled material. The samples were heated up to 1330 Â°C at a rate of 1.5 K/s and held at this temperature for 300 seconds in order to allow grain coarsening and precipitates dissolution. After that, samples were cooled down to testing temperature at a rate of 1.67 K/s, held for 300 seconds for thermal homogenization and deformed until fracture at a strain rate of 10â&#x20AC;&#x201C;3 sâ&#x20AC;&#x201C;1 in the temperature range of 700 to 1100 Â°C for both steels. Additional tests were performed at 800 Â°C for steel B, at strain rates of 10â&#x20AC;&#x201C;1 and 5 sâ&#x20AC;&#x201C;1. Both the applied forces and the displacements produced during tests were recorded. Hot ductility was assessed by measuring the reduction of area (đ?&#x2018;&#x2026;đ??´) of the broken specimens.

La Metallurgia Italiana - n. 1 2019

Colata continua Tab. 1 – Chemical composition of tested steels (wt %). COMPOSITION OF TESTED STEELS Steel

A (no Nb)

0.12

1.05

0.25

0.04

0.00

0.02

B (with Nb)

0.11

0.10

0.25

0.07

0.03

0.00

Curve fitting A curve fitting of the experimental data was performed in order to get a continuous variation of reduction of area depending on

both temperature and strain rate. The variation with strain rate was fitted with a hyperbolic tangent function (Eq. 6, Fig. 2).

[6] The variation with temperature at a fixed strain rate of 10–3 s–1 was fitted with a scaled beta function [12] for both steels, as

shown in Fig. 3, Then, a second scaling was performed using the strain rate formula previously explained (Eq. 6).

Fig. 2 – Effect of strain rate on the reduction of area at 800 °C.

Fig. 3 – Reduction of area at 10–3 s–1 versus temperature for the two studied steels. La Metallurgia Italiana - n. 1 2019

Continuous casting A reduced ductility is observed for both steel grades between 700 and 1100 Â°C although steel B shows an even more brittle behavior, which can be associated with the Nb precipitates that promote intergranular cracking. The temperature above which the material presents a fully ductile behavior (100% reduction of area) is called đ?&#x2018;&#x2021;đ?&#x2018;&#x2018;đ?&#x2018;?, taking a value of 1089 Â°C for steel A and 1112 Â°C for steel B. DEFECT INDEX IN ROLLED PRODUCTS In the plant under analysis, continuous casting round bars are rolled into seamless pipes and the entire production is inspected by Non Destructive Testing (NDT) techniques. Each time an imperfection is detected, it is classified according to a set of pre-established rules. This classification determines the most likely part of the manufacturing process where the imperfection might have been generated (steelmaking, casting, rolling, heat treatment, etc). In the present case, inspection results of more than 5000 heats were compiled. An index to assess the frequency of surface imperfections related to the continuous

casting process in each heat was defined. Then, the influence of steel composition and process conditions were evaluated and compared with model predictions. RESULTS AND DISCUSSION SURFACE CRACKING RISK CALCULATION Surface cracking risk is calculated using the procedure explained in Section 1. The analysis is applied to the same caster from which the defectology data is obtained. Modeled calibers cover the entire range of diameters produced in this machine ranging from 148 to 330 mm. A standard casting speed is considered for each caliber, and a speed reduced by 30%, in agreement with the speed defined for the start-up. The two steel grades indicated in Table 1 are evaluated. At first, bar temperatures are determined with the thermal model depicted in Section 1.1. Surface temperatures are shown in Fig. 4 for all the bar diameters cast at a standard speed as a function of distance to meniscus.

Fig. 4 â&#x20AC;&#x201C; Evolution of surface temperature for different bar diameters.

It can be appreciated that the temperatures are lower as the bar diameter increases, since larger sections cast at lower speeds. As seen in Section 2.1, the upper bound of the brittle temperature range Tdb lies around 1100 Â°C for the tested steels. If the material is straightened below that temperature, higher cracking risk is expected. After the thermal simulation, the mechanical model (Section 1.2) is applied to determine stress and strain fields. Surface strain and strain rate along the intrados of the bar (the most stretched fiber) are presented on Fig. 5. Only the 330 mm

diameter bar is shown as an illustrative example of the calculation. The strain after each straightening point can be easily compared with the analytical value đ?&#x153;&#x20AC;s n=(1/R1â&#x2C6;&#x2019;1/Rn)(d/2) where đ?&#x153;&#x20AC;đ?&#x2018; đ?&#x2018;&#x203A; is the accumulated surface strain where the radius changes to đ?&#x2018;&#x2026;đ?&#x2018;&#x203A;, đ?&#x2018;&#x2026;1 the first machine radius and đ?&#x2018;&#x2018; the bar diameter. Analytical estimations of strain rate are fairly rough, thus justifying the use of the model in order to determine this magnitude. Strain rate peaks can be observed in correspondence with the straightening points.

La Metallurgia Italiana - n. 1 2019

Colata continua

Fig. 5 â&#x20AC;&#x201C; Surface strain (green) and strain rate (red) versus distance to steel meniscus. At last, thermo-mechanical results and ductility tests data are processed together in order to estimate surface cracking risk, as explained in Section 1.3. Reduction of area depends on both temperature (Fig. 4) and strain rate (Fig. 5). In Fig. 6, it can be

noticed that the material is straightened with a brittle behavior, since surface temperatures drop below đ?&#x2018;&#x2021;đ?&#x2018;&#x2018;đ?&#x2018;?. Surface cracking risk is then calculated and its values are also shown in Fig. 6.

Fig. 6 â&#x20AC;&#x201C; Reduction of area (green) and surface cracking risk (red) versus distance from meniscus. The above procedure is then repeated for all the combinations of diameter, casting speed and steel grade, and the final value

of surface cracking risk is obtained for each case. Results are summarized in Fig. 7.

Fig. 7 â&#x20AC;&#x201C; Simulation resutls of surface cracking risk as a function of bar diameter, steel composition and casting speed. La Metallurgia Italiana - n. 1 2019

Continuous casting A clear cracking risk raise is observed with increasing bar diameters. In small sections, the effects of steel grade and casting speed are negligible. But on larger diameters, Nb alloyed steels present a higher cracking risk. This effect is more evident at reduced casting speed. CORRELATION BETWEEN MODELED RESULTS AND DEFECT INDEX IN ROLLED MATERIAL As indicated in Section 2.2, results of pipe inspection in a large number of heats were compiled and analyzed. At first, the influence of bar diameter and casting sequence were analyzed, as seen in Fig. 8. As the bar size increases, the defect index rises. This trend is in agreement with the results of the model predictions indicated in Fig. 7 and is produced because, in bigger sections, larger strains are generated at the

bar surface during unbending. As the casting speed gradually increases during sequence starts, the first heat of sequence is normally cast at an average lower speed than the rest of the heats. In general, lower casting speeds promote a drop in the strand surface temperature, which can reach đ?&#x2018;&#x2021;đ?&#x2018;&#x2018;đ?&#x2018;? at the straightening point, increasing the risk of cracking. This fact is verified in the simulations performed, where an increment in the cracking risk is obtained when the casting speed is reduced by 30% (see Fig. 7). So, the larger defect index observed for the first heats of sequence could be partially explained by this effect. It is worth noting that smaller sections are usually cast at higher speeds, so the strand surface temperature remains relatively high during unbending and the effect is less important, as shown in Fig. 7 and 8.

Fig. 8 â&#x20AC;&#x201C; Effect of bar diameter and sequence position on defect index. Bar line type is similar than the convention used in Fig. 7: dotted lines represent reduced speed and solid lines stand for standard speed. In order to study the effect of steel composition on the defect index, two groups of low C steels were formed, one with Nb and the other without Nb. Since steels with Nb are more frequently cast in larger sections, only bars with a diameter

greater than 225 mm were considered for this comparison. As indicated in Fig. 9, steels with Nb exhibit a higher defect index, which is in agreement with model predictions shown in Fig. 7.

Fig. 9 â&#x20AC;&#x201C; Effect of Nb on defect index. Bar colors are chosen according to Fig. 7 since the red bar can be associated with steel A (with Nb) and the green bar with steel B (no Nb). 34

La Metallurgia Italiana - n. 1 2019

Colata continua Finally, the impact of caster type on the defect index was also analyzed. For this analysis, similar steel grades (low C, alloyed with Nb) cast in the same section (330 mm diameter) but in two different continuous casting machines (one vertical and

the other curved) were compared. As shown in Fig. 10, the defect index is higher in the material cast in the curved machine, which confirms the role of straightening in the potential formation of defects.

Fig. 10 â&#x20AC;&#x201C; Influence of caster type on defect index. CONCLUSIONS A thermo-mechanical model was developed, followed by the codification of a material failure criterion to assess the cracking risk of round bars during strand straightening. For the development of the cracking criterion, ad-hoc hot tension laboratory tests were carried out in low carbon steels with and without Nb. Results of these tests showed that the steel alloyed with Nb exhibits a wider and deeper low ductility trough. The developed methodology was applied to analyze several scenarios. Performed simulations showed that the surface cracking risk is higher as the bar diameter increases, since bigger sections are subject to larger surface strains and lower temperatures at unbending. Both of these effects are detrimental for the material. For similar bar sizes, a reduction in casting speed can increase the risk of cracking because the strand sur-

La Metallurgia Italiana - n. 1 2019

face temperature decreases and can enter into the low ductility region during straightening. As smaller sections are normally cast at higher speeds, the strand surface temperature remains relatively high during unbending, so the effect is not significant. However, a reduction of casting speed in bigger sections may have a detrimental impact. This effect is even more noticeable when Nb alloyed steels are cast. Results of industrial inspection of rolled material produced from a large number of heats were carefully analyzed. This evaluation showed that the index of surface defects related to the continuous casting process follows the same trend obtained in the simulated scenarios. Hence, the developed methodology can be used to investigate the role of process variables on the risk of transverse cracking in the continuous casting of round bars.

Continuous casting REFERENCES [1]

Gonzalez M, Goldschmit MB, Assanelli AP, Dvorkin EN, FernĂĄndez Berdaguer E. Modeling of the solidification process in a continuous casting installation for steel slabs. Metallurgical and Materials Transactions B, vol. 34, num. 4 (2002), p. 455-473.

[2]

Savage J, Pritchard WH. The problem of rupture of the billet in the continuous casting of steel. Journal of the Iron and Steel Institute, vol. 178 (1954), p. 269-277.

[3]

Vazquez M, Poltarak G, Ferro S, Campos A, Cicutti C. 8th ECCC, Graz, Austria, Jun-2014.

[4]

Ristorto I, Vazquez M, Fuhr F, Alicandro J, Sabugal J, Campos A. 20th IAS Steel Conference, Rosario, Argentina, Nov-2014.

[5]

Ferro S, Cardozo M. 5th SteelSim, Ostrava, Czech Republic, Sep-2013.

[6]

Dvorkin EN, Toscano RG. A new rigid-viscoplastic model for simulating thermal strain effects in metal forming processes. Int. J. for Numerical Methods in Engineering, num. 58 (2003), p. 1803-1816.

[7]

Kozlowski PF, Thomas BG, Azzi JA, Wang H. Simple constitutive equations for steel at high temperature. Metallurgical Transactions A, vol. 23A (1982), p. 903-918.

[8]

Li C, Thomas BG. Thermomechanical Finite-Element Model of Shell Behavior in Continuous Casting of Steel. Metallurgical and Materials Transactions B, vol. 35B, num. 6 (2004), p. 1151-1172.

[9]

Cicutti C, Boeri R. Analysis of solute distribution during the solidification of low alloyed steels. Steel Research International, vol. 77, num. 3 (2006), p. 194-201.

[10] Maehara Y, Ohmori Y. The Precipitation of AlN and NbC and the Hot Ductility of Low Carbon Steels. Materials Science and Engineering, vol. 62 (1984), p. 109-119. [11] Schwerdtfeger K, Spitzer KH. Application of Reduction of Area-Temperature Diagrams to the Prediction of Surface Crack Formation in Continuous Casting of Steel. ISIJ International, vol. 49, num. 4 (2009), 512â&#x20AC;&#x201C;520. [12] Hazewinkel M, editor. Encyclopaedia of Mathematics. Springer Netherlands, Hardcover ISBN 978-1556080104 (2001).

La Metallurgia Italiana - n. 1 2019

I D A

N A LI

A T I E

N IO

Z A CI

U L L

A T E M

O S AS

A I G R

sottoscrivi o rinnova la tua quota ti aspetta un ricco di eventi

2019

Via Filippo Turati 8 20121 Milano . Italy aim@aimnet.it +39 02 76 021132 +39 02 76397770

Le manifestazioni AIM AIM meetings and events CREEP - Modulo avanzato Corso - Centro ME Milano, 6 febbraio

BULLONERIA E TRATTAMENTI TERMICI GdS - Centro TTM c/o Vimifasterners - Reggiolo, maggio

TRATTAMENTO TERMICO DI MATERIALI E COMPONENTI PRODOTTI PER MANIFATTURA ADDITIVA Gds - Centri MP e TTM c/o BEAMIT, Fornovo di Taro, 7 marzo

ECHT 2019 - HEAT TREATMENT & SURFACE ENGINEERING FOR AUTOMOTIVE Convegno Internazionale Bardolino, Lago di Garda, 5-6-7 giugno

CAUSE E SOLUZIONI DEI DIFETTI NEI GETTI PRESSOCOLATI Gds - Centro P c/o Kilometro Rosso, Bergamo, 8 marzo

POLVERI E PROCESSI PER ALTE PRESTAZIONI GdS - Centro MP Milano, 6 giugno

TRADIZIONE ED INNOVAZIONE NELLE TECNOLOGIE DI LAVORAZIONE DEL TITANIO Gds - Centro MTA Milano, 12 marzo

RIVESTIMENTI - II MODULO - Rivestimenti spessi - Placcatura e Termospruzzatura Corso - Centro R Milano, 19-20 giugno

METALLURGIA FUORI FORNO Corso itinerante - Centro A 28-29 marzo, 3-4-10-11 aprile

MICROSCOPIA ELETTRONICA IN SCANSIONE SEM PER METALLURGISTI - II ed. Corso - Centro MFM Lecco, 26-27-28 giugno

MASTER PROGETTAZIONE STAMPI Corso - Centro P marzo/maggio

LEGHE DI ALLUMINIO GdS - Centro ML Bologna, 27 giugno

LEGHE DI NICHEL E SUPER LEGHE GdS - Centro ME Milano, maggio

PRESS & SINTER: EVOLUZIONE DI PRODOTTI, MATERIALI E TECNOLOGIE GdS - Centro MP giugno

DIFETTOSITA’ IN COLATA CONTINUA E LINGOTTI GdS - Centri A e FOR Roma, 17 maggio METALLI A SCALA MICRO NANOMETRICA: TECNICHE DI INDAGINEI GdS - Centro MFM Roma, 17 maggio PROVE MECCANICHE Corso - Centro CCP Milano, 29-30 maggio Torino, 5-6 giugno

XIII GIORNATE NAZIONALI SULLA CORROSIONE E PROTEZIONE Convegno – SEGR Palermo, 3-4-5 luglio DEFORMAZIONE DEI PRESSOCOLATI: CAUSE E RIMEDI GdS - Centro P Torino, 18 settembre “ADDITIVE METALLURGY”. MATERIALI METALLICI E FABBRICAZIONE ADDITIVA Corso - Centri MFM, ML, MP Milano, 18-19 settembre

Per ulteriori informazioni rivolgersi alla Segreteria AIM e-mail: info@aimnet.it oppure visitare il sito internet www.aimnet.it

La Metallurgia Italiana - n. 1 2019

Attualità industriale Influence of fc-mold on flow pattern and entrapment of inclusions in continuous casting strand edited by: S. Wang, W. Chen, X. Zhang, L. Zhang In the current study, the influence of Flow Control Mold (FC-Mold) on the continuous casting strand with a cross section of 1050 mm×237 mm was discussed. The nail board tipping and steel plate test method were employed to investigate the surface velocity and liquid level profile. Quantitative metallographic analyses of inclusions and statistics of level fluctuation were adopted to observe the influence FC-Mold on the cleanness of steel and flow field. The length and angle of hooks were also measured to evaluate the magnetic field parameters. The result shows that the area fraction of inclusion, average diameter of inclusion and the length of hooks decreased obviously with the application of FC-Mold. For the large inclusions at sub surface (> 10 μm), the area fraction, average diameter and number density of inclusion in the center position was less than the quarter position. The upper magnetic field intensity (BU) equals the lower magnetic field intensity (BL), 0.33 T, is the optimal magnetic field parameter in current condition. KEYWORDS: CONTINUOUS CASTING STRAND – FC-MOLD – NAIL BOARD MEASUREMENT – INCLUSIONS – FLOW PATTERN

Shengdong Wang, Wei Chen, Xubin Zhang, Lifeng Zhang School of Metallurgical and Ecological Engineering,University of Science and Technology Beijing, China Corresponding author: Lifeng Zhang zhanglifeng@ustb.edu.cn

INTRODUCTION Since the 1980 s, high production efficiency and product quality have become the target of steel enterprises due to the rapid development of steel equipment. The improvement of continuous casting speed can increase the production of billet, improve the economic efficiency of enterprises and become the main content of efficient continuous casting. However, large level fluctuation was often observed with high casting speed (1).In the continuous casting process, with the application of two-port nozzle and the injection of argon gas, the fluid flow could lead to level fluctuation, especially the formation of double-roll flow (2). The upper backflow with high speed and floating bubbles could cause large level fluctuation, which facilitated the entrapment of mold flux and even led to the occurrence of breakouts. Hence, proper decrease of the steel velocity upwards was necessary. These years, the application of electromagnetic brake (EMBr) has been reported to optimize the fluid flow in the mold (3). Flow Control Mold (FC-Mold) was the third EMBr, and impo-

La Metallurgia Italiana - n. 1 2019

sed two static magnetic fields in the mold. The upper magnetic field was above the outport of the submerged entry nozzle (SEN), and the lower one was below the outport of SEN. The upper magnetic field could brake the speed of the upper steel, and suppress the surface velocity and level fluctuation of liquid steel. As was reported (4, 5) that hooks could entrap inclusions, bubbles and mold flux in the mold, the application of EMBr could reduce the hook depth (6), and then improve the surface quality of slabs (7). Therefore, it is very important to study the influence of the FC-Mold on flow field. In the current study, the influence of FC-Mold on the continuous casting strand was discussed. The nail board tipping and steel plate test method were employed to investigate the surface velocity and liquid level profile. Quantitative metallographic analyses of inclusions and statistics of level fluctuation were adopted to observe the influence FC-Mold on the cleanness of steel and flow field. The length and angle of hooks were also measured to evaluate the magnetic field parameters.

Industry news METHODOLOGY In order to investigate the influence of FC-Mold on the surface level fluctuation, meniscus velocity magnitude and cleanness of steel during the high speed casting process, the industrial experiments for SPHC steel with a cross section of 1050 mm× 237 mm were carried out. The meniscus velocity, surface profile, quantitative metallographic analyses of inclusions and measurement of hooks are employed to evaluate the magnetic field parameters. The experimental condition with different magnetic field intensity is summarized in Tab. 1. Nail board tipping and steel plate test were used to count the meniscus velocity and surface level. The schematic diagram of this measurement is shown in Fig. 1(a) and Fig. 1(b). When casting reached a steady state, the nail board was inserted in the mold vertically. After 3-5 seconds, lift the nail board from

the molten steel pool. Then the meniscus velocity can be calculated through the skull diameter and height (8), as shown in Eq. [1]. The principle of nail board tipping and steel plate test is similar, while the steel plate test can be used to measure the surface profile commodiously. Three samples with the dimension of 50 mm× 50 mm× 237 mm were taken at the edge, the quarter and the center of wide face. The face locates at 2 mm below the loose side was used for inclusion detection. The observed face was ground, polished and scanned to detect > 10 μm inclusion by ASPEX. The scanned area was over 150 mm2, and the morphology, composition, number density and area fraction of inclusion were obtained. The face vertical to the oscillation mark, was also ground, polished and etched to reveal hooks.

Tab. 1 – Experiment condition

EXPERIMENT CONDITION FC-Mold

Case

Casting speed (m/s)

SEN outer angle

1.7

20° down

1.8

20° down

0.33

1.8

20° down

0.17

0.33

BU (T)

BL (T)

Without FC-Mold

[1]

(a)

(b)

(c)

Fig. 1 – Schematic diagram of (a) nail board tipping, (b) steel plate test and (c) inclusion analysis position

La Metallurgia Italiana - n. 1 2019

Attualità industriale VELOCITY AND TOP SURFACE PROFILE OF CONTINUOUS CASTING MOLD Fig. 2 shows the comparison of meniscus velocity with different magnetic field intensity. Compared with low speed casting, the velocity distribution in meniscus is disorderly in high speed casting process. The maximum velocity appears at the quarter of mold width and the flow direction is directed from narrow face to SEN, which are the typical characteristics of

the double roll. Without the influence of FC-Mold, the maximum velocity magnitude at left side and left side is 0.64 m/s and 0.7 m/s. After the application of FC-Mold, the maximum meniscus velocity will reduced to 0.60 m/s and 0.51 m/s respectively. It can be seen that the application of FCMold have an obvious effect on the reduction of meniscus speed. The FC-Mold also can improve the symmetry of the flow field.

Fig. 2 – Influence of FC-Mold on distribution of meniscus velocity Fig. 3 and Fig. 4 display the measurement results of surface profile through the nail board tipping and steel plate test respectively. Δh in Fig. 4 represents the level profile, which can be calculated as show in Fig. 1(b). A large fluctuation is observed near the narrow face and SEN, which induced by the ascension of the flow field in upper recirculation zone.

The floatation and broken of the argon bubbles will cause a drastic fluctuation near the SEN. It is investigated that the FC-Mold can flatten the surface profile due to the braking effect. Thus, with the application of FC-Mold can decrease the occurrence of the slag entrapment.

(a) Without FC-Mold

La Metallurgia Italiana - n. 1 2019

Industry news

(b) BU=0.33 T, BL=0.33 T

(a) Without FC-Mold

(b) BU=0.33 T, BL=0.33 T

La Metallurgia Italiana - n. 1 2019

Attualità industriale Fig. 5 illuminates the statistics of level fluctuation at different range. With the magnetic field intensity BU=0.33 T and BL=0.33 T, the percentage of level fluctuation dropped from 7.78% to 3.45%, with a reduction of about 55.7%. A remar-

kable reduction of the large level fluctuation will be promoted due to the application of FC-Mold, especially when the BU=BL=0.33 T.

Fig. 5 – Influence of FC-Mold on level fluctuation CLEANNESS OF CONTINUOUS CASTING SLAB Fig. 6(a), Fig. 6(b) and Fig. 6(c) shows the influence of magnetic field intensity on the area fraction, average diameter and number density of inclusion, respectively. The main nonmetallic inclusions in the slab is Al2O3. For the large inclusions at sub surface (> 10 μm), the area fraction, average diameter and number density of inclusion in the center position was less than the quarter position. At

quarter of left side, With the magnetic field intensity BU=0.33 T and BL=0.33 T, the area fraction, average diameter and number density of inclusion decreased 40%, 11.5% and 19.0%, respectively. While all of this increased at quarter of right side. It may be caused by the existing of some larger inclusion. Overall, the area fraction, average diameter and number density of inclusion decreased obviously with the application of FC-Mold.

(a)

(b)

Industry news HOOK CHARACTERISTICS Fig. 7 shows the average hook depth and length under different magnetic field intensity. It can be seen that the appropriate magnetic field parameter, BU=0.33 T and BL=0.33 T,

(a)

can inhibit the growth of hook. With the application of FCMold, the high temperature zone in mold will move up, leading an increased temperature near the meniscus. Therefore, the growth of hook will be inhibited subsequently.

(b)

Fig. 7 â&#x20AC;&#x201C; Influence of the FC-Mold on hook characteristics

CONCLUSIONS (1) With the application of FC-Mold, the maximum meniscus velocity will be reduced obviously. The FC-Mold also can improve the symmetry of the flow field. Overall, the area fraction, average diameter and number density of inclusion decreased obviously with the magnetic field intensity, BU=0.33 T and BL=0.33 T. Moreover, the growth of hook will be inhibited at this parameter. (2) The upper magnetic field intensity (BU) equals the lower

magnetic field intensity (BL), 0.33 T, is the optimal magnetic field parameter in current condition. ACKNOWLEDGEMENTS The authors are grateful for support from the National Natural Science Foundation of China (Grant No. 51725402, No. 51504020 and No. 51704018), and the High Quality steel Consortium (HQSC) at University of Science and Technology Beijing (USTB), China.

La Metallurgia Italiana - n. 1 2019

AttualitĂ industriale REFERENCES [1]

Li L, Wang X, Deng X, Wang X, Qin Y, Ji C. Application of High Speed Continuous Casting on Low Carbon Conventional Slab in SGJT. steel research international. 2014;85(11):1490-500.

[2]

Liu Z, Sun Z, Li B. Modeling of Quasi-Four-Phase Flow in Continuous Casting Mold Using Hybrid Eulerian and Lagrangian Approach. Metallurgical and Materials Transactions B. 2017;48(2):1248-67.

[3]

Wang Q, Zhang L. Influence of FC-Mold on the Full Solidification of Continuous Casting Slab. JOM. 2016;68(8):1-10.

[4]

Sengupta J, Shin H-J, Thomas B, Kim S-H. Micrograph evidence of meniscus solidification and subsurface microstructure evolution in continuous-cast ultralow-carbon steels. Acta Materialia. 2006;54(4):1165-73.

[5]

Sengupta J, Thomas BG, Shin H-J, Lee G-G, Kim S-H. A new mechanism of hook formation during continuous casting of ultralow-carbon steel slabs. Metallurgical and Materials Transactions A. 2006;37(5):1597-611.

[6]

Xubin Zhang LZ, Hao Wang, Shengdong Wang, Qiangqiang Wang, Wen Yang. Subsurface hooks in continuous casting slabs of low-carbon steel. Chinese Journal of Engineering. 2017;39(2):251-8.

[7]

Kollberg SG, Hackl HR, Hanley PJ. Improving quality of flat rolled products using electromagnetic brake (EMBR) in continuous casting. Iron and steel Engineer. 1996;73:24-8.

[8]

Rietow B, Thomas BG. Using Nail Board Experiments to Quantify Surface Velocity in the CC Mold. Urbana. 2008.

La Metallurgia Italiana - n. 1 2019

Industry news Development and use of indicators in the algorithms to detect defects automatically on the slabs by utilising images taken from the hot slabs on line during continuous casting edited by: P. Hooli, H. Suopajärvi With Reveal CAST technology it is possible to take images from the hot slabs (or any kind of semi-products) after cutting, during continuous casting. With special illumination and imaging techniques, the thermal radiation does not affect so the images taken from the surfaces of the slabs look like they were taken from the cold slabs. In usual Reveal CAST installation images are taken from all the slabs and if needed from all the surfaces. Automatic defect detection has been developed to assist in slab quality monitoring. Based on the use of algorithms to detect defects, it is possible to develop special slab quality indicators and trend graphs that are updated after every cast slab. With the aid of these graphs, it is easy to observe trend of defectiveness and plan maintenance accordingly, for instance. Based on the analysis of slab images, it has been shown that correctly made maintenance clearly reduces defects on the slabs. Typically defect rate is low for some period after maintenance until it starts to rise again. Condition of the casting machine has been found to have a surprisingly high impact on surface quality of the slabs. In this paper, an example is presented showing correlation between visually rated defects and automatically created indicators. In addition, comparison of defects on the inner bow side against outer bow side with different type of steel grades will be presented. Statistics of the different defects is shown with the reporting tools included in the Reveal CAST. The results presented in this paper are part of RFCS project called “SUPPORT-CAST”. Goals of the project and participants shall be presented. KEYWORDS: CONTINUOUS CASTING – SLABS – SURFACE QUALITY – AUTOMATIC DEFECT DETECTION – IMAGES

Paavo Hooli, Hannu Suopajärvi Sapotech Oy, Finland

INTRODUCTION There are variety of reasons to monitor the surface quality of the semi-products (slabs, blooms, billets) from continuous casting. Based on the observed surface quality, the slab/billet can be directed further down in the processing line or they can be directed to scarfing or grinding to get rid of surface defects [1]. The steel industry has slowly shifting towards automated surface quality inspection to get rid of manually conducted visual inspection, which is time consuming, labor intensive, unsafe and results in low amount of material for

statistical studies [2,3]. To realize the energy savings and productivity increase, the automatic defect detection should be able to reliably detect the defects from the slab surface. There are variety of methods and technologies developed to detect the defects from the hot slabs. The research in this field has been intensive during the past years [e.g. 1,4]. In addition, there are several commercial actors that provide their solutions for on-line automatic defect detection. Even though the need for slab surface quality monitoring is inevitable for product quality assurance and in deciding the

La Metallurgia Italiana - n. 1 2019

AttualitĂ industriale appropriate route for the produced slab, the benefits of automatic defect detection can be further exploited to evaluate the condition of the slab caster. When the occurrence of defects increases, the predictive maintenance can be planned. Furthermore, if defects found are combined with the slab caster on-line data and additionally data concerning caster conditions, reasons for defects can be analyzed. These kinds of evaluations have not been conducted in the literature earlier. Slab surface quality monitoring - Reveal CAST In this study, Reveal CAST, a slab surface quality monitoring solution, developed and sold by a Finnish company called Sapotech [5] was used as a basic tool to provide huge amount of continuous casting slab surface data to slab quality indicator and trend graph development. Reveal CAST has been built on top of IoT Reveal Technology Platform, combining the

latest software, machine vision, high speed imaging and illumination technologies. The basic working principle of Reveal CAST is shortly described below. Usual Reveal CAST installation comprises of top and bottom imaging of continuous casting slabs; however, it is possible to image all the surfaces or concentrate on the corners of the slabs. The basic set up of top and bottom Reveal CAST is presented in Fig.1. The imaging units (1 and 2) are installed to an optimal distance from the slab surface. The number of cameras in each surface depends on the required resolution. Laser illumination is used to illuminate the surface of the slabs (3). By using the laser illumination, the heat radiation from the high temperature slabs, which usually weakens the image quality, can be avoided. The control cabinet (4) takes care the overall control of the system, data acquisition and processing. It also provides electrical power to the rest of the system.

Fig. 1 â&#x20AC;&#x201C; Reveal CAST installation at the exit of continuous casting line, after cutting of the slabs.

Reveal CAST provides accurate images from slab surface, which are stitched together as whole slabs for automatic defect detection and for manual defect marking in an extremely easy-to-use web-based browser interface. In addition, Reveal CAST can be equipped with 3D topographic measurement, full dimension profile measurements and infrared imaging, according to customer need. Intensive work has been done to develop automatic defect detection in Reveal CAST. Customized algorithms, depending on the imaged surface, are used to automatically detect defects. Sticking defects are one of the most harmful defects on the slab surface to deteriorate the slab quality. Defect de-

La Metallurgia Italiana - n. 1 2019

tection mechanism in Reveal CAST is based on reading combinations of shadows and lighter areas [6-8]. Based on the algorithmic defect detection, six different indicators, ranging from 0 to 1 are automatically created, each indicating different features of defects. These quality indicators are then displayed on the user interface (UI) of Reveal CAST (Fig.2). Those six quality indicators are shown on the left-hand side of the Reveal CAST UI for every slab. If Reveal CAST is integrated with the manufacturing execution system (MES), slab number and other related data concerning the slab can be displayed on Reveal CAST UI as well. This slab information is saved and can be revived at any time.

Industry news

Fig. 2 – Example image of the slab surface with Reveal CAST online monitoring during the casting process. MATERIALS AND METHODS The material contains a huge number of images taken from the slabs by Reveal CAST and selected images were visually inspected. In Reveal CAST there is a feature to set one or several types of defects to make search and get a list of slabs containing only those types of defects, as for example in this study to find only slabs with sticking defects among thousands of slabs. Indicators of automatically detected slabs with sticking defect were recorded and those slabs were analyzed also visually with images. Correlation between those was analyzed. In addition, some special features related to the sticking defects were observed like orientation. The procedures are more thoroughly described below.

Correlation of automatic defect detection indicators and defect seriousness In this study, the main purpose was to evaluate, how well the automatically created indicators described above are detecting the defects. This was done by visually inspecting and rating of sticking defects by a continuous casting expert with 30+ years of experience. Around 30 slabs were thoroughly inspected and rated in a scale of 0–5 (defect seriousness). Fig.3 (a-c) show examples of 3 different sizes of sticking defects and visual rating of their seriousness using images made by Reveal CAST. This visual rating was possible, because in Reveal CAST images of all the slabs are stored. This kind of inspection using real slabs would not be possible having photos except from defects but also from whole slabs.

Fig. 3 – Examples of visual classification: a=1.5; b=3.2; c=4.5 (Range 0–5) (note: whole widths of the slabs are shown). Development of trend graphs from automatic defect detection indicators Trend graphs can be produced from the defect information that is provided by Reveal CAST. Trends can be based on the found defect (e.g. sticking defects, transverse crack) or based on the automatically created slab quality indicators (six indicators described above) and they can be plotted against 48

specific time period or a certain number of slabs produced or only for certain steel grades. Trend graph describing the level of defects over some specific time period is very useful to plan the maintenance or to follow when some parameter in the caster has been changed. Aim is to see easily that level of defects is not increasing but remains low.

La Metallurgia Italiana - n. 1 2019

AttualitĂ industriale Inner bow and outer bow defects One feature to evaluate possible reasons for defects is to observe if defects are on both sides or only on one side of the slab. For example, a reason for longitudinal cracks can be subentry nozzle locating too near to the other side of the mould or there is something wrong with mould powder feeding. Scratches caused by stagnate rolls are typically only on the other side of the slab. Of the 4000 slabs, the search function in Reveal CAST was used to select slabs with a sticking defect. With images of found 80 slabs it was given index (0-5) for sticking defects. Orientation and location Orientation of defects can be useful to observe. For example, a sticking defect can be over whole width, on the middle part or left or right side of the slab, triangular with the tip in the middle of slab, triangular with the tip on the left or right edge of the wide face or defect can be on the small face. The orientation and location of a defect can give you a clue to troubleshoot the problem. If orientation or location is repeatedly

on the left side, then alignment on the left side of the mould may be wrong. Period used in this study was 3600 slabs with 22 slabs with sticking defect. Again, the search function in Reveal CAST was used. RESULTSAND DISCUSSION Capability of Reveal CAST to detect sticking defects Fig.4 shows the correlation between automatically created indicators and visually rated seriousness of defects in case of slabs with sticking defects. It can be seen, that there is a good correlation with these two indices leading to the conclusion that development of automatic indicators has been successful and that the underlying techniques and algorithms deployed are fit for purpose. Sticking alarm is needed, when there is a danger of break out. In case (a) in the Fig.3 above there is a small sticking, which did not create alarm and probably it was not needed. However, there is a need to find the defect to remove it by local grinding. With Reveal CAST it was possible to do. In addition, it is easy to make tuning of sticking alarm function with images.

Fig. 4 â&#x20AC;&#x201C; Correlation between automatic defect detection indicators and defect seriousness.

Trend graphs to evaluate the stability of the continuous casting Fig.5 shows an example of a trend graph in which a share of slabs with sticking defects is depicted against sequential casting periods. Every period is at least 200 slabs. Some periods are consisting of thousands of slabs in cases of long periods with very low level of defects. After every maintenance stop

La Metallurgia Italiana - n. 1 2019

(letter M in Fig.5) it is possible to notice improvement in slab quality in terms of lower number of sticking defects. The only exception is the maintenance stop depicted with MX. It was revealed afterwards that there were difficulties to install the new mould to the caster; the mould was not fully aligned. It can be seen, that the level of sticking defects rose rapidly to undesired level.

Industry news

Fig. 5 – Trend graph showing the share of slabs with sticking defect in sequential casting periods. Fig.6 shows the sum of indicators before and after the maintenance “MX”. It is possible to see that level and variation of indicators are increasing after that maintenance even though there were no major defects yet. Such a small rise in the level

of features can be difficult to notice with data from on-line measurements. Steel grades in the Fig.6 were similar standard grades so they are comparable.

Fig. 6 – Sum of indicators before and after the maintenance MX Trend graphs can be created automatically and up-dated after every slab. These kinds of graphs are useful when considering the need for maintenance. Appropriately done maintenance has clear influence on the level of defects found from the cast slabs. Increasing trend indicates the need for immediate check of caster conditions and maybe maintenance depending of reason found for defects. When operators learn to use the automatic indicators and trends properly, they can understand the relationship between the defects and the condition of the caster better, and make the corrective actions in time, before the number of defects starts to increase. Role of the condition of the caster is probably more significant than earlier expected. Data concerning conditions of the caster is often insufficient, especially when conditions are undergoing changes during a long production period. 50

Defect occurrence in inner bow side versus outer bow side It is known that certain defects are not equally occurring on inner and outer bow sides with the bow type caster. When having images taken from both sides of the slabs it was possible to study the occurrence of sticking defects in inner and outer bow sides simultaneously. Reveal CAST was put into service in two steps in the reference plant. In the first step the top surface of the slabs was imaged. When Reveal CAST was taken into use also on outer bow side, it was possible to make a comparison shown in the Fig.7. First, it shows the percentage of slabs with sticking defects for five different stainless steel grades, namely 304L, 304, 321, 316Ti and 316. It can be seen, that the least sensitive grade for sticking defects is AISI316. The highest La Metallurgia Italiana - n. 1 2019

AttualitĂ industriale sensitivity to sticking defects has the grades with Ti alloying (321,316Ti). That two grades have also the largest difference in the level of sticking defects between inner and outer bow side. In Ti alloyed grades there are clearly more inclusions (TiN), which are disturbing the performance of the casting

powder [9], especially on inner bow side in the straight mould curved kind of caster machine. Principally that phenomenon is known, but with Reveal CAST images and automatic defect detection it was possible to prove.

Fig. 7 â&#x20AC;&#x201C; Steel grades and indexes of sticking defects on inner and outer bow side and shares of slabs with sticking defect. As shown in Fig.8 (straight mould curved machine) sticking defects are very seldom only on outer bow side. Most often they are on both sides, but often more serious on inner bow side. Many defects are only on inner bow side, like trash, extra material or grease spots. So, having Reveal CAST only for inner bow side may be adequate starting point to improve

the quality of the slabs as also the caster performance. In the curved mould type caster, the occurrence of the defects only on outer bow side is probably even less common than with straight mould curved machine. These results comparing inner versus outer bow side could be useful to consider if planning a new caster design.

Fig. 8 â&#x20AC;&#x201C; Steel grades and shares of sticking defects on inner versus outer bow sides on slabs. Operators and shift managers of the caster have a role to improve quality and safety of casting process. The operators in the reference plant wanted to have a separate display showing the latest slab (Fig.9). The display has the inner and outer bow side surfaces of the last slab larger so that even smaller defects appear better. Trend graph is also one for operators to follow. When operators can see the result of their work shortly during their shift, it is motivating them to find if something is wrong. La Metallurgia Italiana - n. 1 2019

It can be noted in Fig.9, that indicators on the inner bow side are higher than outer bow side. It is typical, that on the inner bow side there are more features. For example, the white spot in case a) on inner bow side is caused by lubricant from bearings. If several spots appear in the slab images, there can be a failure in the bearings. The surface quality of the outer bow side is very good as indicators are showing. In the case b) indicators on the inner bow side are clearly higher and are meaning a sticking defect, as automatically marked (green 51

Industry news box). Sticking defect on the outer bow side is milder in its features but caused yellow color as warning on indicator A. With this display, showing the latest slabs in control room, operators can see features which can forecast bigger problems later. Concerning defects found on the slab surface, learning is much more efficient when operators see the slab surface

in control room as it is in the production line compared to situation where operators would have to go through numerical data from on-line measurements. On-line measurements need more interpretation concerning the phenomena causing defects or features on the surface of slabs.

Fig. 9 â&#x20AC;&#x201C; Screenshot from control room display showing the latest slab- cases a and b. Casting direction is marked with an arrow. Orientation of the sticking defects Fig.10 shows an orientation of sticking defect on the inner bow side of the slab. Orientation on the outer bow side is the same (the dotted line in the Fig.10 a). Sticking defect is not typically over the whole surface. Consideration of orientation is one example of how the images can be utilized. Orientation of the sticking defect can give a hint, on which side of the mould there may be problems. In the case in the Fig.10 sticking has taken place on the left-hand side. There is over-

flow of new shell on to the stuck shell. The right side has moved forward during sticking maybe somewhat lower speed than casting speed. In Fig.10 b there is also magnification showing detail from the defect. Laser light is coming from the right-hand side and the shadows on the overflow are on the left-hand side. It means that tips of the overflow are on the previous surface and little above. Even small details are available with Reveal CAST as in this case to study how liquid has flown downwards and frozen on previous shell.

La Metallurgia Italiana - n. 1 2019

Attualità industriale

Fig. 10 – Orientation of a sticking defect. The arrow indicates the direction of casting. Orientation of the sticking defects Fig.10 shows an orientation of sticking defect on the inner bow side of the slab. Orientation on the outer bow side is the same (the dotted line in the Fig.10 a). Sticking defect is not typically over the whole surface. Consideration of orientation is one example of how the images can be utilized. Orientation of the sticking defect can give a hint, on which side of the mould there may be problems. In the case in the Fig.10 sticking has taken place on the left-hand side. There is over-

Fig. 11 – Orientations of sequential sticking defects.

La Metallurgia Italiana - n. 1 2019

Industry news CONCLUSIONS Online monitoring by having images of slabs taken after cutting can be used for many purposes. Main aim is to keep good surface quality of semi-products. Research work can be done much faster, when results are possible to evaluate soon after trial. Images made with Reveal CAST are offering a new level for research to study factors affecting quality of cast semiproducts.

ACKNOWLEDGEMENTS Part of the results presented in this paper have been obtained in the context of the RFCS SUPPORT-CAST project: “Supporting Control by Inspection of Surface Quality and Segregation on Cast Products through Integration of Novel Online Monitoring and Advanced Modelling into an Accessible Cloud Access”. Participants of the project are: Swerea Mefos AB, Outokumpu Stainless AB, SIDENOR Investigacion Y Desarrollosa, Acciaierie di Calvisano S.p.A., VDEh-Betriebsforschungsinstitut GmbH and Sapotech Oy.

REFERENCES [1]

Veitch-Michaelis J, Tao Y, Walton D, Muller JP, Crutchley B, Storey J, Paterson C, Chown A. Crack Detection in" As-Cast" Steel Using Laser Triangulation and Machine Learning. InComputer and Robot Vision (CRV), 2016 13th Conference on 2016 Jun 1 (pp. 342-349). IEEE.

[2]

Neogi N, Mohanta DK, Dutta PK. Review of vision-based steel surface inspection systems. EURASIP Journal on Image and Video Processing. 2014 Dec;2014(1):50.

[3]

Xi J, Shentu L, Hu J, Li M. Automated surface inspection for steel products using computer vision approach. Applied optics. 2017 Jan 10;56(2):184-92.

[4]

Landstrom A, Thurley MJ. Morphology-based crack detection for steel slabs. IEEE Journal of selected topics in signal processing. 2012 Nov;6(7):866-75.

[5]

Sapotech. Reveal CAST, available from: https://www.sapotech.fi/solutions/reveal-cast/

[6]

Hooli P. Real time monitoring of hot steel slabs during casting and automatic slab surface quality assessment enabling fast utilization of quality information feedback – for planning of grinding, fast evaluation of trials and online feedback of maintenance needs. 9th European Continuous Casting Conference, 2017; Vienna, Austria. p.13

[7]

Moilanen J, Hooli P, Puukko E.Follow of gradual chances of features on slab surface by taking images of hot slabs during casting. Scanmet V conference 2016, Luleå, Sweden, pp. 7-8

[8]

Hooli P. Benefits of follow gradual changes of features on slab surface by taking images of hot slabs during casting with the aid of a tool called Reveal CAST. AisTech 2016, Pittsburgh, US

[9]

Hooli P. Study on the layers in the film originating from the casting powder between steel shell and mould and associated phenomena in continuous casting of stainless steel. Doctoral Thesis, 2007, pp. 39-41.

La Metallurgia Italiana - n. 1 2019

GIORNATE NAZIONALI SULLA

xiii edizione

PALERMO - 3-5 LUGLIO 2019

PRESENTAZIONE

Aree tematiche principali

La XIII edizione delle Giornate Nazionali sulla Corrosione e Protezione si terrà a Palermo dal 3 al 5 luglio 2019, presso il Campus dell’Università degli Studi di Palermo. Le Giornate rappresentano l’evento di riferimento a livello nazionale per la discussione ed il confronto sulle questioni scientifiche, tecnologiche e produttive, nell’ambito della corrosione e protezione dei materiali. In particolare, il Convegno prevede la presentazione dei risultati raggiunti da vari gruppi di studio e da numerose aziende del settore, in forma orale e poster.

• • • • •

Corrosione delle strutture metalliche esposte all’atmosfera Corrosione e protezione delle reti idriche interrate Corrosione negli impianti industriali Protezione catodica: progettazione, collaudo, gestione e monitoraggio Comportamento a corrosione di leghe di titanio, nichel e acciai inossidabili Corrosione delle opere in calcestruzzo armato Corrosione dei beni culturali Corrosione negli impianti Oil & Gas Degrado e rilascio dei biomateriali metallici Rivestimenti e trattamenti superficiali Inibitori di corrosione Impatto delle nuove tecnologie produttive sulla corrosione Tecniche di studio e monitoraggio della corrosione Meccanismi di corrosione Case histories

Gli interessati a presentare memorie scientifiche/tecniche dovranno inviare, entro i termini, titolo, nomi degli autori con relative affiliazioni e un breve riassunto. Le deadline e ulteriori informazioni sono disponibili sul sito dell’evento www.aimnet.it/gncorrosione

• • • • • • • • • •

spazio aziende e sponsorizzazione

segreteria organizzativa

presentazione di memorie

Sulla pagina www.aimnet.it/gncorrosione sono disponibili numerose proposte per le aziende interessate alla sponsorizzazione dell’evento o ad uno spazio per l’esposizione di apparecchiature, la presentazione dei servizi e la distribuzione di materiale promozionale.

Organizzate da

ASSOCIAZIONE ITALIANA DI METALLURGIA Via Filippo Turati 8 · 20121 Milano · Italy tel: +39 02 76397770 · +39 02 76021132 info@aimnet.it www.aimnet.it

Sponsorizzate da

gnC PALERMO 2019

Experts’ corner State of the Art in High-temperature Mold Simulator Study of Initial Solidification in Continuous-casting Mold edited by: P. Lyu, W. Wang Central South University, School of Metallurgy and Environment, Changsha, Hunan, China. Continuous casting technique has been widely used for the production of crude steel worldwide. However, surface defects of steel strand often occur during the continuous casting process, which need to be removed before rolling process, resulting in the loss of yield, productivity and energy. Additionally, the better surface quality of strand is required for the development hot charging rolling technique, thin slab casting technique and strip casting technique. Therefore, it is necessary to study initial solidification of molten steel in continuous-casting mold, as it is the origin of steel strand and determines the strand surface quality. Extensive past works related the initial solidification of molten steel in mold have been conducted, such as stress and strain state of shell, heat transfer between mold and shell, solute segregation of shell, fluid flow of molten steel, infiltration and entrapment of molten slag, entrapment of bubbles and inclusion in shell, and its research methods can be generally divided into four categories: plant experiment, pilot caster experiment, mathematical simulation and mold simulator. Compared with other methods, mold simulator has great advantage in the study of initial solidification in the view of economy and simulation. In this paper, the development history of mold simulator and extensive results from mold simulator study are summarized with the aim to provide direction for the future mold simulator study of initial solidification, and to improve the understanding of this important phenomenon.

KEYWORDS: CONTINUOUS CASTING – INITIAL SOLIDIFICATION – SURFACE QUALITY – HEAT TRANSFER – MOLD SIMULATOR

INTRODUCTION Continuous casting technique has been widely used for the production of crude steel worldwide; however, problems of strand quality often occur during the continuous casting process.[1] Defects of strand include internal and surface defects, and it is of great importance to eliminate these defects; as the defects tends to lead a loss of yield, productivity, energy, etc.[2] Additionally, the better quality of strand is required for the development of hot charging rolling technique, thin slab casting technique and strip casting technique. Compared with internal defects, surface defects such cracks show more severe damage to final product, as they are easily oxidized and can’t be re-welded during rolling.[3] 56

The importance of initial solidification of molten steel in continuous-casting mold has been widely acknowledged, because many surface defects originate from the initial solidification area during continuous casting.[4 ,5 ] Therefore, a clear understanding of initial solidification of molten steel is of great importance for the control of strand surface defects. Initial solidification behaviors of molten steel in continuous mold happen in meniscus area as shown in Fig.1, which include stress and strain state of shell, heat transfer between mold and shell, solute segregation of shell, fluid flow of molten steel, infiltration and entrapment of molten slag, entrapment of bubbles and inclusion in shell. Extensive past work has been conducted to study the

initial solidification of molten steel in continuous mold, and its research methods can generally be divided into four categories: plant experiment[1,6-8], pilot caster experiment[9-12], mathematical simulation and mold simulator experiment. It is not always convenient to conduct experiments on an industrial continuous caster or a pilot caster to study the effects of operational parameters on the initial solidification of molten steel due to the practical constraints, such as difficulties in acquiring real-time data, controlling of experiment conditions, harsh environment and expensive cost. For mathematical simulation, analytical expressions were obtained based on mathematical models to predict the growth of initial solidified shell, the formation La Metallurgia Italiana - n. 1 2019

Scenari of oscillation mark (OM) and slag infiltration between mold and shell.[13,14] Alternatively, mathematical models were solved through the numerical method to simulate initial solidification behavior of molten steel and the formation of possible surface defects developed during initial solidification.[7,15-17] However, results from mathematical simulation could not completely reflect the practical complexity of initial solidification behaviors of molten steel inside the mold, because mathematical models are established based on the hypothesis and it is difficult to acquire precise boundary conditions and physical property parameters. Because of the limitation of plant experiment, pilot caster experiment and mathematical simulation in the study of ini-

tial solidification of molten steel inside the mold, many different types of mold simulator have been developed to study various aspects of initial solidification. Compared with other methods, mold simulator has great advantage in the study of initial solidification in the view of economy and simulation. Experimental materials used in mold simulator could be organic substance, low-melting point alloy and steel. It is convenient and easy to study the effect of casting parameters on initial solidification inside the mold using organic substance or low-melting point alloy as a substitute of steel during the mold simulator test[11,18-25]; however, these cannot reflect the actual solidification behaviors of molten steel around meniscus, such as solidification structure of shell, infiltration and lubri-

cation of mold flux. Consequently, actual steel and mold flux were used as experimental materials in mold simulator for a better simulation of initial solidification of steel inside the mold. The major factor in designing the mold simulator is to ensure that the experimental process and apparatus can truly simulate the reality. As far, mold simulators can be divided into three types: dip type, static type and oscillatory type. In this paper, the development history of mold simulators and extensive results from mold simulator study are summarized with the aim to provide direction for the future mold simulator study of initial solidification, and to improve the understanding of this important phenomenon.

Fig. 1 â&#x20AC;&#x201C; Meniscus area in continuous-casting mold.[26]

DIP MOLD SIMULATOR As early as in the 1980s, Machingawuta[27] established a dip mold simulator to study the heat transfer behavior and mineralogical constitution of mold flux film; during the simulator test, the water-cooled copper mold was dipped into the molten flux for a period of time to obtain mold flux film, and cuspidine was found as a primary crystalline phase in flux film. Since then, the similar simulators were also developed La Metallurgia Italiana - n. 1 2019

by Wen[28 ,29 ] and Wang[30,31] to study the properties of mold flux, and experimental process during the dip mold simulator test is shown in Fig. 2. Steel grades with different carbon content were used in dip mold simulator to study the influence of carbon content on a solidified shell structure at meniscus, and then the formation mechanism of hook-like structures near oscillation marks of continuous cast slab was proposed.[32] Force measurement device was introduced 57

Expertsâ&#x20AC;&#x2122; corner to dip mold simulator to simulate the formation of hot-tear cracks, which was called the submerged split chill tensile (SSCT) experiment, as shown in Fig. 3. [33-35] The SSCT experiment can help to investigate the mechanical behavior and hot tear crack formation of shell during initial solidification process. Strezov[36] developed a dip mold simulator with rapid insertion device and splash-proof copper substrate to obtain a thin shell similar to industrial steel strip, which can simulated the rapid solidification of steel during twin-roll strip casting.

In terms of strip casting, Bouchard[23,37] used dip mold simulator to investigate the effect of substrate condition and substrate temperature on heat transfer and surface quality of copper alloy cast, and implication of experimental results on strip casting was discussed. Without mold oscillation and shell extraction device, dip mold simulator only contains chilled plate that is immersed into steel or slag bath, which cannot mimic the dynamic nature of continuous-casting process of slab or billet.

Fig. 2 â&#x20AC;&#x201C; Experimental process during the dip mold simulator test.[30]

Fig. 3 â&#x20AC;&#x201C; The schematic of the SSCT apparatus.[33] STATIC MOLD SIMULATOR Static mold simulator, also called bottom-pouring mold, is actually similar to the dip mold simulator, with the exception that static mold simulator have the molten steel contained in the mold. In the 1980s, Wray[22] developed a bottompouring mold, in which liquid lead was pumped into mold

chamber, and then solidified shell formed against the mold wall, as shown in Fig.4. From this mold simulator study, three types of surface features, fine groove, wide-spaced wrinkle and surface lap, of chilled cast were recognized and their corresponding formation mechanisms were proposed. Linear, horizontal defects are commonly observed on chill-cast sur-

La Metallurgia Italiana - n. 1 2019

Scenari face, and all of these surface features can be termed ripple marks. Later, Stemple[38 ] used bottom-pouring mold with one glass mold face to investigate the influence of wave motion and casting variables of liquid Sn-Pb alloy on surface morphology of chilled cast and then surface wave mechanism for ripple mark formation was proposed. Tomono[39 ,40 ] used bottom-pouring mold with a quart window to observe the meniscus dynamics of molten steel, as shown in Fig.5. With the help of camera, overflow of liquid steel from solidified meniscus can be observed and so Tomono proposed an overflow mechanism of ripple mark (they called folding mark) on surface of steel billet cast. In order to study the air gap

and heat transfer between mold and cast, thermocouples and displacement meter were mounted in bottom-pouring mold by Nishida[41], as shown in Fig.6. Ripple marks not only were found on surface of ingot but also on surface of continuouscasting strand.[11,42] Recently, interest in ripple marks has been revitalized because it is believed that ripple marks also play a role in continuous-casting strand besides oscillation marks.[43] It should be emphasized that static mold simulator only could be used to investigate the initial solidification behaviors which are unrelated to mold oscillation, because static mold simulator did not be equipped with oscillation system.

Fig. 4 â&#x20AC;&#x201C; The schematic of static mold simulator.[22]

Fig. 5 â&#x20AC;&#x201C; Menisci during bottom pouring: (a) liquid steel in contact with chill and (b) liquid slag and steel in contact with chill.[39,40]

La Metallurgia Italiana - n. 1 2019

Expertsâ&#x20AC;&#x2122; corner

Fig. 6 â&#x20AC;&#x201C; The schematic of static mold simulator with thermocouples and displacement.[41]

OSCILLATORY MOLD SIMULATOR For the purpose of getting a better simulation of real initial solidification inside industrial continuous-casting mold, several researchers have developed the dip mold simulators in which the copper mold is equipped with oscillation motor and shell extractor. In this paper, these type of mold simulators are termed oscillatory mold simulators. In the early 1980s, NKK Corp. developed an oscillatory mold simulator for the study of initial solidification behaviors of molten steel inside mold, such as the effect of mold oscillation conditions on oscillation marks, formation mechanism of oscillation marks, and mold powder consumption.[44-46] But this simulator cannot be used to study the heat transfer behavior during initial stage of melt solidification because of the lack of temperature sensor embedded in mold wall. Based on the oscillatory mold simulator developed by NKK Corp., the mold simulator which can simulate the crack formation of steel shell during continuous casting was established by Tata group. And Santillana[47] used this apparatus to investigate the relationship between temperature and stress of shell when the cracks of shell formed, as shown in Fig.7. In the early 21th century, The US Steel R&T Center developed the oscillatory mold simulator system in which the copper mold was instrumented with high-sensitive thermocouples connected with quick acquisition system, and displacement meter for measuring motions of extractor. [48 ,49 ] In this study, temperatures acquired by thermocou-

ples were used to calculate an estimate of heat flux through mold hot face that contacts with shell using one-dimensional inverse heat conduction problem (1D-IHCP) model[50]. These allows the observation of the temperature and heat flux variations during on oscillation cycle. Fig.8 shows the Gaussian distribution both of pitch and depth of oscillation marks on strand surface produced by mold simulator and industrial caster. These experimental data show that there is reasonable similarity of surface features between mold simulator shell and industrially cast slab. Consequently, this type of mod simulator have a good simulation quality of industrial mold. A hypothesis about OM formation was proposed by Badri after the mold simulator study and the this hypothesis was confirmed by the fact that the positive peaks of heat flux derivative correlate quite well with the physical locations of OMs on shell surface, as can be seen in Fig.9. Then, the oscillatory mold simulator with three molds, called multi-mold simulator, was established by POSCO. Sohn[51,52] used the multi-mold simulator to investigate the effect of casting variables on heat transfer near meniscus for low and medium carbon steel. Also, the mold flux film retrieved after casting was analyzed, which suggested that the crystallization of mold flux is a dynamic process, and crystalline phase can grow when retained in the high temperature condition for a period of time.

La Metallurgia Italiana - n. 1 2019

Scenari

Fig. 7 – The oscillatory mold simulator by Tata group.[47]

Fig. 8 – Distribution of pitch and depth of oscillation marks for an ultra-low carbon steel stand produced by mold simulator and industrial caster.[48]

Fig. 9 – The relation between OM locations on shell surface and heat transfer.[49] La Metallurgia Italiana - n. 1 2019

Expertsâ&#x20AC;&#x2122; corner More recently, Wang[53-55] developed an oscillatory mold simulator and the initial solidification of molten steel inside mold was well studied in detail. The greatest improvement of the oscillatory mold simulator developed by Wang is that the heat fluxes through mold hot face is calculated by twodimensional inverse heat conduction problem (2D-IHCP) model[56 ], which allows to obtain a more accurate heat transfer information around meniscus. As shown in Fig.10(a), it is known that the heat flux near meniscus ( the area from 3 to 8 mm) is about 1.0 MW/m2, and reaches the largest value of about 1.5 MW/m2 at the position of 6 mm below the steel level. Heat fluxes during one oscillation cycle along the casting direction at the moments of 52.65, 52.80, 52.95, 53.10 and 53.25 seconds marked in Fig.10(a) with red lines were shown in Fig.10(b). It can be found that the variation amplitude of heat flux is more dramatic near the meniscus region than the other region. Result from PSD anlyisis of heat fluxes is shown in Fig.11.There are several heat flux signals with different frequency and intensity in the PSD contour map. The heat flux near the liquid slag surface (S14) has a stronger high-frequency signal that is around 1.67 Hz and its intensity reaches a peak value of 116.29 dB/Hz. This may be associated

with the oscillation-introduced heat transfer phenomena. In order to study the initial solidiifcation near mold corner, a right-angle copper mold with two cooling face was applied to oscillatory mold simulator system. [57] According to the fitting results in Fig.12, it is obvious that the average solidification factor of shell at the corner (2.766 mm/ s1/2) (Fig.12(a)) is larger than that at hot face (2.325 mm/ s1/2) (Fig.12(b)), which indicates that the cooling intensity of shell at corner is stronger. Through changing mold shape of mold simulator, effect of mold shape on the initial solidification of molten steel around the meniscus region was studied by Wang[58]. The results suggested that the designed chamfered structure would increase the thermal resistance and weaken the two-dimensional heat transfer around the mold corner, causing the homogeneity of the mold surface temperatures and heat fluxes. Oscillatory mold simulators have been successfully applied to investigate initial solidification of molten steel, which possess the advantages of low cost, convenient operation, reasonable similarity to actual caster. In most of cases, physical simulation needs mathematical model as an auxiliary tool in order to get more useful information about initial solidification.

Fig. 10 â&#x20AC;&#x201C; The heat fluxes through mold hot face during continuous casting (a) the contour map of heat fluxes and (b) the heat fluxes at different positions and moments during one oscillation cycle.[53]

La Metallurgia Italiana - n. 1 2019

Scenari

Fig. 11 â&#x20AC;&#x201C; analysis for the heat flux during continuous casting.[53]

Fig. 12 â&#x20AC;&#x201C; Thickness fitting of shell and its instantaneous solidification factor (Kt) at different positions: (a) thickness of shell at hot face and (b) thickness of shell nearby corner.[57] CONCLUSIONS The importance of initial solidification of molten steel in continuous-casting mold has been widely acknowledged, because many surface defects originate from the initial solidification area during continuous casting. A clear understanding of initial solidification of molten steel is of great importance for the control of strand surface defects. Research methods about initial solidification can be generally divided into four categories: plant experiment, pilot caster experiment, mathematical simulation and mold simulator. Compared with other methods, mold simulator has great advantage in the study of initial solidification in the view of economy and simulation. In this paper, the development history of mold simulator and La Metallurgia Italiana - n. 1 2019

extensive results from mold simulator study are summarized with the aim to provide direction for the future mold simulator study of initial solidification, and to improve the understanding of this important process. Oscillatory mold simulators have been successfully applied to investigate initial solidification of molten steel in mold, and possess a better simulation quality of initial solidiifcation in industrial mold than dip and static mold simulators. However, a lof of further work is required before mold simulators achieve their full potential. At present, initial solidification phenomena about heat transfer, solidification structure, mold flux behavior, and oscillation mark have been studied in detail using mold simulator. But, mold simulator studies on liquid 63

Expertsâ&#x20AC;&#x2122; corner flow, shell stress, bubble motion and inclusion motion have hardly been reported. Actually, stress state of shell, temperature distribution and flow field of molten pool during mold simulator test have some deviation from those inside industrial mold, which needs to be consider and solve in the future mold simulator study. In the future, mold simulator would be

the main research technique in the study of initial solidification; at the same time mathematical models would be as an indispensable auxiliary tool in order to get more useful information about initial solidification.

REFERENCES [1]

Mahapatra RB, Brimacombe JK, Samarasekera IV, Walker N, Paterson EA, Young JD. Mold behavior and its influence on quality in the continuous casting of steel slabs: Part i. Industrial trials, mold temperature measurements, and mathematical modeling. Metallurgical Transactions B. 1991;22(6):861-74.

[2]

Sengupta J, Thomas BG, Shin HJ, Lee GG, Kim SH. A new mechanism of hook formation during continuous casting of ultra-lowcarbon steel slabs. Metallurgical & Materials Transactions A. 2006;37(5):1597-611.

[3]

Brimacombe JK, Sorimachi K. Crack formation in the continuous casting of steel. Metallurgical Transactions B. 1977;8(2):489505.

[4]

Wolf MM. Initial Solidification and Strand Surface Quality of Peritectic Steels. Continuous Casting. Warrendale, PA: Iron and Steel Society; 1997. p. 1-111.

[5]

Nakato H, Ozawa M, Kinoshita K, Habu Y, Emi T. Factors affecting the formation of shell and longitudinal cracks in mold during high speed continuous casting of slabs. Transactions of the Iron & Steel Institute of Japan. 1984;24(11):957-65.

[6]

Mahapatra RB, Brimacombe JK, Samarasekera IV. Mold behavior and its influence on quality in the continuous casting of steel slabs: Part II. Mold heat transfer, mold flux behavior, formation of oscillation marks, longitudinal off-corner depressions, and subsurface cracks. Metallurgical Transactions B. 1991;22(6):875-88.

[7]

Takeuchi E, Brimacombe JK. The formation of oscillation marks in the continuous casting of steel slab. Metallurgical & Materials Transactions B. 1984;15(3):493-509.

[8]

Sengupta J, Shin HJ, Thomas BG, Kim SH. Micrograph evidence of meniscus solidification and sub-surface microstructure evo-

La Metallurgia Italiana - n. 1 2019

Scenari lution in continuous-cast ultralow-carbon steels. ACTA MATER. 2006;54(4):1165-73. [9]

Hanao M, Kawamoto M, Yamanaka A. Growth of Solidified Shell Just below the Meniscus in Continuous Casting Mold. ISIJ INT. 2009;49:365-74.

[10]

Suzuki M, Mizukami H, Kitagawa T, Kawakami K, Uchida S, Komatsu Y. Development of a New Mold Oscillation Mode for Highspeed Continuous Casting of Steel Slabs. ISIJ INT. 2007;31(3):254-61.

[11]

Tomono H. Elements of oscillation mark formation and their effect on transverse fine cracks in continuous casting of steel. Lausanne: EPFL; 1979.

[12]

Singh SN, Blazek KE. Heat transfer and skin formation in a continuous-casting mold as a function of steel carbon content. JOM. 1974;26(10):17-27.

[13]

Yamauchi A, Emi T, Seetharaman S. A Mathematical Model for Prediction of Thickness of Mould Flux Film in Continuous Casting Mould. ISIJ INT. 2002;42(10):1084-93.

[14]

Fredriksson H, Elfsberg J. Thoughts about the initial solidification process during continuous casting of steel. SCANDINAVIAN JOURNAL OF METALLURGY. 2002;31(5):292-7.

[15]

Lopez PER, Mills KC, Lee PD, Santillana B. A Unified Mechanism for the Formation of Oscillation Marks. Metallurgical & Materials Transactions B. 2012;43(1):109-22.

[16]

Moinet A. The free meniscus problem in the continuous casting of steel: a computational model of cast surface formation: Carnegie Mellon University; 2012.

[17]

Ramirez-Lopez PE, Lee PD, Mills KC. Explicit modelling of Slag Infiltration and Shell Formation during Mould Oscillation in Continuous Casting. ISIJ INT. 2010;50(3):425-34.

[18]

Ackermann P, Heinemann W, Kurz W. Surface quality and meniscus solidification in pure chill cast metals. Archiv FĂźr DassenhĂźttenwesen. 1984;55(1):1-8.

[19]

Yang J, Meng X, Zhu M. Experimental Study on Mold Flux Lubrication for Continuous Casting. STEEL RES INT. 2014;85(4):7107.

[20]

Tsutsumi K, Jun-Ichi O, Hino M. Inflow Behavior Observation of Molten Mold Powder between Mold and Solidified Shell by Continuous Casting Simulator Using Sn-Pb Alloy and Stearic Acid. ISIJ INT. 2000;40(6):601-8.

[21]

Miyoshino I, Takeuchi E, Yano H, Sakane J, Saeki T, Kajioka H. Influence of electromagnetic pressure on the early solidification in a continuous casting mold. ISIJ INT. 2007;29(12):1040-7.

[22]

Wray PJ. Geometric features of chill-cast surfaces. Metallurgical Transactions B. 1981;12(1):167-76.

[23]

Bouchard D, Hamel FG, Nadeau JP, Simard D, Bellemare S, Dreneau F, et al. Effects of substrate surface conditions on heat transfer and shell morphology in the solidification of a copper alloy. Metallurgical & Materials Transactions B. 2001;32(1):111-8.

[24]

Zhang H, Wang W, Zhou D, Ma F, Lu B, Zhou L. A Study for Initial Solidification of Sn-Pb Alloy during Continuous Casting: Part I. The Development of the Technique. Metallurgical and Materials Transactions B. 2014;45(3):1038-47.

[25]

Zhou D, Wang W, Zhang H, Ma F, Chen K, Zhou L. A Study for Initial Solidification of Sn-Pb Alloy During Continuous Casting: Part II. Effects of Casting Parameters on Initial Solidification and Shell Surface. Metallurgical and Materials Transactions B. 2014;45(3):1048-56. Metallurgical and Materials Transactions B. 2014;45(3):1048-56.

[26]

Meng Y, Thomas BG. Heat Transfer and Solidification Model of Continuous Slab Casting: CON1D. Metallurgical and Materials Transactions B. 2003;34B(5):685-705.

La Metallurgia Italiana - n. 1 2019

Expertsâ&#x20AC;&#x2122; corner [27]

Grieveson P, Bagha S, Machingawuta N, Liddell K, Mills KC. Physical properties of casting powders: Part 2 - Mineralogical constitution of slags formed by powders. IRONMAK STEELMAK. 1988;15(4):181-6.

[28]

Wen G, Sridhar S, Tang P, Qi X, Liu Y. Development of Fluoride-free Mold Powders for Peritectic Steel Slab Casting. ISIJ INT. 2007;47(8):1117-25.

[29] Wen G, Tang P, Yang B, Bai ZX. Simulation and characterization on heat transfer through mould slag film. ISIJ INT. 2012;52(7):1179-85. [30]

Liu Y, Wang W, Ma F, Zhang H. Study of Solidification and Heat Transfer Behavior of Mold Flux Through Mold Flux Heat Transfer Simulator Technique: Part I. Development of the Technique . Metallurgical and Materials Transactions B. 2015;46(3):1419-30.

[31]

Ma F, Liu Y, Wang W, Zhang H. Study of Solidification and Heat Transfer Behavior of Mold Flux Through Mold Flux Heat Transfer Simulator Technique: Part II. Effect of Mold Oscillation on Heat Transfer Behaviors. Metallurgical and Materials Transactions B. 2015;46(4):1902-11.

[32]

Yamamura H, Mizukami Y, Misawa K. Formation of a Solidified Hook-like Structure at the Subsurface in Ultra Low Carbon Steel. ISIJ INT. 2007;36(Suppl):S223-6.

[33]

Bernhard C, Hiebler H, Wolf MM. Simulation of Shell Strength Properties by the SSCT Test. ISIJ INT. 1996;36.

[34]

Rowan M, Thomas BG, Pierer R, Bernhard C. Measuring Mechanical Behavior of Steel During Solidification: Modeling the SSCC Test. Metallurgical & Materials Transactions Part B. 2011;42(4):837-51.

[35]

Bernhard C, Xia G. Influence of alloying elements on the thermal contraction of peritectic steels during initial solidification. IRONMAK STEELMAK. 2006;33(1):52-6.

[36]

Strezov L, Herbertson J. Experimental Studies of Interfacial Heat Transfer and Initial Solidification Pertinent to Strip Casting. ISIJ INT. 1998;38(9):959-66.

[37]

Bouchard D, Nadeau JP, Simard D, Hamel FG, Howes B, Paumelle C. Control of heat transfer and growth uniformity of solidifying copper shells through substrate temperature. Metallurgical & Materials Transactions B. 2002;33(3):403-11.

[38]

Stemple DK, Zulueta EN, Flemings MC. Effect of wave motion on chill cast surfaces. Metallurgical & Materials Transactions B. 1991;20(12):503-9.

[[39] Tomono H, Ackermann P, Kurz W, Heinemann W, ^editors. Elements of Surface Mark Formation in Continuous Casting of Steel. Solidification Technology in the Foundry and Casthouse; 1983; Coventry, England. Pub Place: The Metals Society; Year Published 40]

Tomono H, Kurz W, Heinemann W. The liquid steel meniscus in molds and its relevance to the surface quality of castings. Metallurgical & Materials Transactions B. 1981;12B:409-11.

[41]

Nishida Y, Droste W, Engler S. The air-gap formation process at the casting-mold interface and the heat transfer mechanism through the gap. Metallurgical Transactions B. 1986;17(4):833-44.

[42]

Singh SN, Blazek KE. Heat transfer and skin formation in a continuous-casting mold as a function of steel carbon content. JOMUS. 1974;26(10):17-27.

[43]

Jacobi H, Schwerdtfeger K. Ripple Marks on Cast Steel Surfaces. ISIJ INT. 2013;53(7):1180-6.

[44]

Kawakami K, Kitagawa T, Mizukami H, Uchibori H, Miyahara S, Suzuki M, et al. Fundamental Study and Its Application of Surface Defects of Powder Cast Strands. Tetsu- to- Hagane. 1981;67(8):1190-9.

[45] Takeuchi H, Matsumura S, Hidaka R, Nagano Y, Suzuki Y. Effects of Mould Oscillation Conditions on Oscillation Marks of Stainless Steel Casts. Tetsu- to- Hagane. 1983;69(2):248-53. 66

La Metallurgia Italiana - n. 1 2019

Scenari [46]

Tsutsumi K, Murakami H, Nishioka S, Tada M, Nakada M, Komatsu M. Estimation of Mold Powder Consumption in Continuous Casting. Tetsu- to- Hagane. 1998;84(9):9-16.

[47]

Santillana MB. Thermo-mechanical properties and cracking during solidification of thin slab cast steel. Delft: Technische Universiteit Delft; 2012.

[48]

Badri A, Natarajan TT, Snyder CC, Powers KD, Mannion FJ, Cramb AW. A mold simulator for the continuous casting of steel: Part I. The development of a simulator. Metallurgical and Materials Transactions B. 2005;36(3):355-71.

[49]

Badri A, Natarajan TT, Snyder CC, Powers KD, Mannion FJ, Byrne M, et al. A mold simulator for continuous casting of steel: Part II. The formation of oscillation marks during the continuous casting of low carbon steel. Metallurgical and Materials Transactions B. 2005;36(3):373-83.

[50]

Beck JV. â&#x20AC;&#x153;IHCP1D: A Program for Calculating Surface Heat Fluxes from Transient Temperatures Inside Solids,â&#x20AC;? Version 5.31. Houston, TX: Beck Engineering Consultants Company; 1997.

[51]

Park J, Ko E, Choi J, Sohn I. Characteristics of medium carbon steel solidification and mold flux crystallization using the multimold simulator. MET MATER INT. 2014;20(6):1103-14.

[52]

Ko E, Choi J, Park J, Sohn I. Simulation of low carbon steel solidification and mold flux crystallization in continuous casting using a multi-mold simulator. MET MATER INT. 2014;20(1):141-51.

[53]

Zhang H, Wang W, Ma F, Zhou L. Mold Simulator Study of the Initial Solidification of Molten Steel in Continuous Casting Mold. Part I: Experiment Process and Measurement. Metallurgical and Materials Transactions B. 2015;46 (B):2361-73.

[54]

Zhang H, Wang W. Mold Simulator Study of the Initial Solidification of Molten Steel in Continuous Casting Mold: Part II. Effects of Mold Oscillation and Mold Level Fluctuation. Metallurgical and Materials Transactions B. [journal article]. 2016;47(2):92031.

[55]

Zhang H, Wang W. Mold Simulator Study of Heat Transfer Phenomenon During the Initial Solidification in Continuous Casting Mold. Metallurgical and Materials Transactions B. 2017;48(2):779-93.

[56]

Zhang H, Wang W, Zhou L. Calculation of Heat Flux Across the Hot Surface of Continuous Casting Mold Through Two-Dimensional Inverse Heat Conduction Problem. Metallurgical and Materials Transactions B. 2015;46 (B):2137-52.

[57]

Lyu P, Wang W, Zhang H. Mold Simulator Study on the Initial Solidification of Molten Steel Near the Corner of Continuous Casting Mold. Metallurgical and Materials Transactions B. 2017;48(1):247-59.

[58]

Lyu P, Wang W, Long X, Zhang K, Gao E, Qin R. Study of the Effect of Mold Corner Shape on the Initial Solidification Behavior of Molten Steel Using Mold Simulator. Metallurgical and Materials Transactions B. 2018;49(1):78-88.

La Metallurgia Italiana - n. 1 2019

Aim news Calendario degli eventi internazionali International events calendar 2019 March 10-14, San Antonio, USA TMS 2019 148th Annual Meeting & Exhibition June 5-7, Bardolino, Garda Lake, Italy ECHT 2019 - heat treatment & surface engineering for automotive

QUOTE SOCIALI AIM 2019 (ANNO SOLARE) Benemeriti (quota minima) 1.750,00 € Sostenitori (quota minima)

750,00 €

Ordinari (solo persona)

70,00 €

Seniores

25,00 €

Juniores

15,00 €

La quota dà diritto di ricevere la rivista dell’Associazione, La Metallurgia Italiana (distribuita in formato digitale). Ai Soci viene

June 13-15, Guangzhou, China 2019 China International Metal & Metallurgy Exhibition

riservato un prezzo speciale per la partecipazione alle manifestazioni AIM e per l’acquisto delle

June 25-29, Düsseldorf, Germany METEC & 4th ESTAD 2019

pubblicazioni edite da AIM. Per ulteriori informazioni, iscrizioni, rinnovi:

September 1-5, Stockholm, Sweden EUROMAT 2019

AIM, Via F. Turati 8 20121 Milano Tel.: 02 76021132/76397770,

September 9-13, Cracow, Poland EUROCORR 2018

fax: 02 76020551 e-mail: amm.aim@aimnet.it www.aimnet.it

September 30 - October 2, Graz, Austria 10th European Stainless Steel Congress, Science and Market 6th European Conference and Expo Duplex August, Xi’an, China 10th Pacific Rim International Congress on Advanced Materials an Processing (PRICM10)

La Metallurgia Italiana - n. 1 2019

Le Rubriche - Centri di studio Attività dei Comitati Tecnici CENTRO RIVESTIMENTI E TRIBOLOGIA (R) (riunione del 12 settembre 2018)

Manifestazioni in corso di organizzazione

Iniziative future

-Il coordinatore Timelli sta organizzando la GdS “Linee guida per la risoluzione dei difetti nei getti pressocolati” e sta cercando una location adatta, possibilmente presso una fonderia; nel frattempo, ha steso un programma preliminare, che presenta al CT per la discussione. Restano da definire alcuni relatori, mentre la data viene fissata per fine febbario 2019.

- Il secondo modulo del corso rivestimenti “Rivestimenti spessi (placcatura/ termospruzzatura)” si terrà a Milano il 19-20 giugno 2019; il pomeriggio del secondo giorno è prevista la visita ad una azienda del settore. La prima giornata ricalcherà quella del corso già effettuato nel 2014, mentre nella seconda gli interventi saranno limitati al mattino per consentire la visita tecnica nel pomeriggio. Stato dell’arte e notizie - Il CT ha discusso e approvato l’inserimento di tre nuovi membri, di cui due provenienti dall’ambito accademico.

CT PRESSOCOLATA (P) (riunione del 12 settembre 2018) Consuntivo di attività svolte -Durante il 37° Convegno Nazionale AIM una sessione è stata dedicata interamente alla pressocolata con 6 presentazioni di buon livello tecnico, che hanno riscosso apprezzamento nonostante la presenza di pubblico in sala fosse limitata, forse a causa di altre sessioni del convegno. -La GdS ”La produzione di getti per applicazioni strutturali” si è svolta il 9 novembre a Travagliato (BS) presso la Idra, con feedback molto positivi. I temi che hanno destato maggiore interesse si possono riferire a lubrificazione, provini, controllo presse, controllo parametri di iniezione, gestione difetti, trattamento termico, gestione del processo. Il coordinatore Valente ha preparato e presentato una sintesi dei questionari di soddisfazione ricevuti dai partecipanti e ne ha tratto spunti per ulteriori futuri approfondimenti.

La Metallurgia Italiana - n. 1 2019

Iniziative future -La GdS “Deformazione dei getti: cause e rimedi” si terrà presso il CRF a Torino il 18 settembre 2019. Scarpa ha affiancato Parona e Tatti nella coordinazione della giornata, che vedrà tra l’altro una discussione sulle tolleranze di assemblaggio, un inquadramento teorico sulle deformazioni e sulla loro misura e sulla simulazione, lasciando uno spazio finale per testimonianze delle fonderie. -Master Progettazione Stampi: viene fissata la riunione di un comitato ristretto per definire i moduli dedicati, in particolare, alle giornate di pratica e di co-design. -Si discute della possibilità di dedicare una GdS allo zinco, possibilmente entro novembre 2019.

CT TRATTAMENTI TERMICI E METALLOGRAFIA (TTM) (riunione del 29 novembre 2018) Consuntivo di attività svolte -La GdS “Trattamenti termici degli acciai per stampi a caldo e a freddo per il settore automotive” si è svolta ad Ivrea l’11 ottobre 2018 con 70 partecipanti. Il presidente Petta conferma il buon successo anche come approfondimento del corso già tenuto nella stessa sede. Il coordinatore Rivolta segnala che i questionari compilati hanno mostrato alti valori di soddisfazione, con qualche richiesta di approfondimento per le future GdS. -La GdS “Ottimizzazione dei trattamenti termochimici e dei processi meccanici nell’industria meccanica” (8 novembre

2018 a Provaglio d’Iseo) si è svolta presso la Gefran con circa 45 partecipanti: Petta ringrazia l’azienda per l’ospitalità e l’organizzazione di ottimo livello, oltre che per l’interessante e apprezzata visita in azienda. Morgano, coordinatore della giornata, traccia un bilancio della giornata, giudicata dai partecipanti tra il buono e l’ottimo nei questionari compilati. La giornata è stata dedicata a Giuseppe Rosso della Silco e a sua moglie Miriam, recentemente scomparsi in un tragico incidente. Manifestazioni in corso di organizzazione

-Il presidente Petta parla del convegno ECHT 2019 “Heat Treatment & Surface Engineering for automotive”, che si svolgerà a Bardolino del Garda (VR) dal 5 al 7 giugno 2019 con chairman l’ing. Morgano. Data l’importanza strategica in ambito internazionale della buona riuscita di questo convegno, Petta raccomanda a tutti i membri del CT di attivarsi per diffondere la notizia dell’iniziativa e raccogliere memorie e partecipazioni. -La GdS “Trattamento termico di materiali e componenti prodotti per manifattura additiva” è co-organizzata con il CT “Metallurgia delle Polveri e Tecnologie Additive”. In assenza del coordinatore Pellizzari, l’ing. Valentina Vicario, coordinatrice per il CT MP, in collegamento telefonico ragguaglia sullo stato di avanzamento. La manifestazione si terrà presso la Beamit di Fornovo Val di Taro nel marzo 2019: c’è un programma preliminare e sono già stati ipotizzati i relatori. Stato dell’arte e notizie -Nel corso della riunione è stata consegnata una targa al prof. Ramous per la lunga militanza nel CT TTM e per le attività formative offerte tramite AIM. -Cinque nuovi candidati hanno manifestato interesse a partecipare alle attività del CT: quattro provengono dal mondo industriale, uno dal settore accademico. Il presidente Petta li presenta e loro illustrano le aziende di appartenenza e le attività da loro svolte. Tutti vengono accettati e dalla prossima 69

Le Rubriche - Centri di studio riunione parteciperanno come membri effettivi.

CT ACCIAIERIA (A) CT FORGIATURA (F) (riunione congiunta del 5 dicembre 2018)

Consuntivo di attività svolte -I presidenti Mapelli e Rampinini relazionano sull’esito e la partecipazione alle varie iniziative che si sono svolte recentemente, in Italia e all’estero, sui temi connessi ai due CT riuniti. In particolare le manifestazioni AIM (GdS sul rischio aziendale, corso acciai inossidabili, Clean Tech) hanno avuto un esito molto soddisfacente dal punto di vista tecnico, anche se numericamente la partecipazione avrebbe potuto essere più ampia. Manifestazioni in corso di organizzazione -La GdS “Difettosità in Colata Continua e Lingotti”, organizzata dal CT Acciaieria, si terrà il 7 marzo a Brescia. C’è già un concreto elenco di presentazioni, a cui se ne stanno aggiungendo altre per completare la panoramica degli argomenti. -Il Corso itinerante “Metallurgia fuori forno” sarà effettuato tra fine marzo e inizio aprile 2019: si stanno definendo le disponibilità delle acciaierie ad ospitare le giornate del corso non ancora fissate.

Iniziative future -Sulla base di una idea lanciata durante la precedente riunione, si approfondisce la possibilità di dare vita ad un convegno, co-organizzato con altri enti, sulle leghe per l’automotive. Le tematiche spazierebbero dai lunghi ai piani, in acciaio o altre leghe, fino alla fonderia di ghisa e alluminio. Si potrebbe tenere verso fine 2019 in una città del nord Italia. Sarebbe quindi obiettivo della manifestazione coinvolgere tutta la filiera dell’automotive. -Rampinini propone una GdS sui “Costi di gestione” pensando alla problematica dei trasporti, che ora è di grande attualità. 70

Stato dell’arte e notizie - Viene approvato l’ingresso di un nuovo membro nel CT Acciaieria. CT MATERIALI PER L’ENERGIA (ME) (riunione dell’11 dicembre 2018) Manifestazioni in corso di organizzazione -La sessione base del Corso sul Creep è in fase di svolgimento, mentre il corso avanzato è previsto per il 6 febbraio 2019. La partecipazione è buona, ma un giudizio di sintesi sarà dato a fine corso. Iniziative future -La GdS “Leghe di Nickel e Superleghe”, che si terrà attorno alla metà di aprile 2019, non è ancora stata pianificata in dettaglio, anche se alcuni interventi sono già stati definiti. Si discute dell’impostazione generale da dare alla GdS per coinvolgere un pubblico il più ampio possibile. Si pensa di chiedere anche il supporto del CT Corrosione. -Rimane viva l’idea di organizzare una GdS sui materiali per l’eolico, ma la discussione è rimandata alla prossima riunione.

CT CORROSIONE (C) (riunione 14 dicembre 2018)

Manifestazioni in corso di organizzazione -La manifestazione “Giornate Nazionali sulla Corrosione e Protezione” si svolgerà a Palermo presso la cittadella universitaria dal 3 al 5 luglio 2019. La coordinatrice Santamaria fornisce molti dettagli circa la logistica della manifestazione (aule, catering, cena sociale a Mondello, spostamenti) e sulle sponsorizzazioni. La cerimonia iniziale avverrà presso lo storico palazzo Steri, con i saluti delle autorità e le relazioni degli ospiti stranieri, di cui è già confermata la presenza. Per sabato 6 luglio si sta organizzando una gita sociale in barca nei dintorni di Palermo. Per quanto riguarda la definizione dell’aspetto scientifico, occorre aspettare la presentazione degli abstract del Convegno (31 gennaio 2019). “La Metallurgia Italiana” dedicherà tutto il numero di novembre-dicembre a questo evento, come già fatto in passato. Iniziative future -Tra le possibili iniziative per il 2019 vengono citate la riedizione del “Corso di Corrosione” e/o di giornate di studio su macrotemi quali “PetrolchinicaOil&Gas” e “Corrosione nelle infrastrutture”, mirate principalmente agli aspetti applicativi e non solo teorici.

Stato dell’arte e notizie

- Gotti e Di Gianfrancesco riferiscono che i lavori del GdL Creep Italiano procedono con riunioni tenute regolarmente. Per quanto riguarda ECCC, Di Gianfrancesco segnala l’ingresso di 4 nuovi partecipanti dalla Francia e il prossimo convegno di Edimburgo a settembre. Bassani offre la possibilità di un contributo di AIM all’organizzazione di uno dei successivi convegni da tenersi in Italia Edimburgo a settembre. Bassani offre la possibilità di un contributo di AIM all’organizzazione di uno dei successivi convegni da tenersi in Italia.

-Il CT approva all’unanimità l’ingresso di due nuovi membri e una sostituzione.

La Metallurgia Italiana - n. 1 2019

gIO

AT AD

tradizione ed innovazione nelle tecnologie di lavorazione del titanio

IS TU

DIO

12 marzo 2019, Milano (c/o Centro Congressi Fast)

Il Centro di Studio “Metalli e Tecnologie Applicative” dell’AIM propone una Giornata di Studio, che riassume l’attuale stato dell’arte nel settore del Titanio e delle sue tecnologie di lavorazione. Uno degli scopi della Giornata vuole essere di riunire le competenze del mondo della ricerca con quelle degli operatori industriali. Tale confronto verrà introdotto da una panoramica iniziale sulle caratteristiche chimico –fisiche e metallurgiche del Titanio e sue leghe. Questa iniziativa segue quelle sviluppate dal nostro Centro di Studio negli anni precedenti, con lo scopo di recepire le richieste di informazione emerse nelle precedenti giornate, nonché di presentare le recenti innovazioni. Saranno presenti relatori provenienti sia dal mondo industriale, che dal mondo universitario e dei Centri di ricerca. Il Centro di Studio Metalli e Tecnologie Applicative con questa manifestazione ritiene di venire incontro alle esigenze di tutti coloro che, a vario titolo, vogliono conoscere, approfondire o discutere il variegato mondo delle tecnologie di lavorazione ed applicazione del titanio. Il programma completo è disponibile su www.aimnet.it

gIO

cause e soluzioni dei difetti nei getti pressocolati

AT AD

IS TU

8 marzo 2019, Bergamo (c/o Kilometro Rosso)

Questa Giornata di Studio organizzata dal Centro di Studio Pressocolata vuole offrire degli spunti per ottimizzare la qualità del prodotto nell’ambito delle varie realtà aziendali. La necessità di ottenere risultati costanti su grandi serie di componenti prodotti porta alla necessità di operare uno stretto controllo delle variabili che intervengono sul processo produttivo. I parametri da considerare, e la cui variazione può causare la comparsa di difetti e/o imperfezioni non presenti all’inizio del ciclo produttivo, sono molteplici. Per questi motivi la qualità dei getti, nel settore della pressocolata è, al tempo stesso, un argomento di essenziale attualità, di notevole criticità e di sfida. Gli strumenti di conoscenza su cui impostare le interazioni finalizzate alla qualità dei getti non mancano e la stessa AIM se ne è fatta carico negli anni: il manuale dei difetti nei getti pressocolati è stato alla base dei rapporti tecnici CEN sviluppati nell’ambito del progetto europeo STACAST. È noto come i getti pressocolati, rispetto ad altri ottenuti con differenti processi di fonderia, devono convivere maggiormente con la presenza di difetti, anomalie e imperfezioni nel prodotto finale. Questo misurarsi con la presenza di difetti è divenuto nel corso degli anni sempre più problematico; ai getti pressocolati sono richiesti, infatti, requisiti e prestazioni sempre crescenti. A questo fine, è necessario poter garantire il mantenimento di adeguati standard produttivi, anche su componenti prodotti in grandi serie. Risulta perciò essenziale non solo elaborare una classificazione sistematica e operativa dei difetti e individuarne le caratteristiche morfologiche più significative, ma soprattutto descriverne le principali cause, mettendo in evidenza i più efficaci interventi correttivi.

Il programma completo è disponibile su www.aimnet.it

DIO

Aim news Corso “Gli Acciai Inossidabili” L’Associazione Italiana di Metallurgia si pregia di un’ampia offerta formativa d’alta qualità, garantita da una lunga e vasta esperienza nell’organizzazione di corsi di formazione ed aggiornamento e dalla possibilità di potersi affidare per le docenze ai maggiori esperti del settore di riferimento sia di estrazione industriale che accademica. Uno dei fiori all’occhiello del catalogo di formazione AIM è il Corso sugli Acciai inossidabili, del quale nell’ottobre e nel novembre 2018 è stata organizzata a Milano la decima edizione, con il coordinamento dell’ing. Mario Cusolito e dell’ing. Sandro Fraccia. La settima, nonché penultima, giornata di Corso si è svolta presso l’azienda Eure Inox di Peschiera Borromeo e si è conclusa con la visita allo stabilimento. Il Corso, che ha cadenza triennale, tratta in modo esaustivo la metallurgia ed i trattamenti termici degli acciai inossidabili, le diverse famiglie e le relative proprietà, i processi produttivi, le lavorazioni ed i prodotti, la saldatura e la sinterizzazione. Ampio spazio viene inoltre dedicato alle norme di riferimento e alle specifiche di acquisto, oltre che ai criteri di progettazione e alla scelta del materiale in funzione degli impieghi. Vengono infine analizzati il mercato e le applicazioni degli acciai inossidabili. La decima edizione ha visto la partecipazione di 60 iscritti. Il questionario di soddisfazione è stato compilato da circa la metà dei partecipanti, che hanno confermato un ampio apprezzamento dell’iniziativa. Il prossimo appuntamento con il Corso sugli Acciai inossidabili è previsto per il 2021. Mario Cusolito

La Metallurgia Italiana - n. 1 2019

Atti e notizie

La Metallurgia Italiana - n. 1 2019

MODERNIZATION In tempi di rapida evoluzione dei mercati, si desidera restare al passo con la concorrenza. Noi ascoltiamo. Impariamo. Offriamo: i nostri specialisti vi aiuteranno a rinnovare il vostro impianto in modo piĂš veloce e flessibile per una maggiore efficienza e con minori costi operativi, aumentando la produttivitĂ ed ottenendo una migliore qualitĂ del prodotto. Qualunque sia la vostra visione, noi la realizzeremo per voi.

Leading partner in the world of metals

www.sms-group.com

www.aimnet.it/echt2019.htm

Heat Treatment & Surface Engineering for Automotive

Organised by

REGISTRATION FEES Early bird registration fees (by May 8, 2019) SPEAKER (presenter) SESSION CHAIRPERSON

MEMBER (*)

NON MEMBER

€ 480

€ 580

€ 590

€ 690

COMMITTEE MEMBER PARTICIPANT (non-presenter) STUDENT ** EXHIBITOR / SPONSOR

Patronized by

€ 360 € 630

€ 530

assofluid

REGISTRATION FEES Standard registration fees (after May 8, 2019) SESSION CHAIRPERSON COMMITTEE MEMBER PARTICIPANT (non-presenter)

NON MEMBER

€ 480

€ 580

COMITATO ITALIANO DEI COSTRUTTORI DI FORNI INDUSTRIALI

Unione P roduttori Italiani Viteria e B ulloneria

Sponsor and exhibitors

€ 740

€ 640

STUDENT ** EXHIBITOR / SPONSOR

FEDERATA

MEMBER (*)

€ 400 € 530

€ 630

* AIM, ASSIOT, ASSOFLUID, CICOF and UPIVEB Member ** Students will have to provide valid proof of student status. The social event on June 6 is not included in the student registration fee.

Additional ticket for Social event for accompanying persons: € 122 (22% VAT included) (Includes only the social event on June 6)

CONFERENCE REGISTRATION FEES INCLUDE: • • • • •

Admittance to technical sessions and to the exhibition Conference bag with electronic proceedings Social event on June 6 coffee breaks lunches

For non-members the fee includes AIM Membership for the second half of 2019 and for the year 2020.

EXHIBITION & SPONSORSHIP

registration informaation

5-7 June 2019

The Conference will feature a table-top exhibition that will represent many areas of industry with latest equipment, facilities and instruments, products and services in the field of heat treatment and surface engineering. Companies will be able to reinforce their participation and enhance their corporate identification by taking advantage of benefits offered to them as Contributing Sponsors of the Conference. Companies interested in taking part in the table-top exhibition or sponsoring the Conference may contact the Organising Secretariat.