Computer Simulation and Optimization of VOD Treatment
Računalniška simulacija in optimiranje VOD obdelave
N. Smajič, Inštitut za kovinske materiale in tehnologije, Lepi pot 11, 61001 Ljubljana
Mathematical model and the software CAPSS (Computer Aided Produc tion of Stainless Steel) developed as a pad of URP-C2-2566 research program were used for computer simulation of EAF-VOD-CC stainless steelmaking technology line. Basic aim of model testing was to optimise VOD treatment with emphasis on obtaining the maximum productivity at the lowest possible thermal load of VOD ladle and EAF tap temperature. It was concluded that only computer controlled oxygen blowing can secure maximum productivity at the acceptable thermal load of VOD ladle and lowest EA furnace tap temperature.
V okviru petletnega raziskovalnega programa URP-C2-2566 izdelani model in računalniški program CAPSS (Computer Aided Production of Stainless Steel) je bil uporabljen za računalniško simulacijo EOP-VOD-KL tehnologije za izdelavo nerjavnih jekel. Osnovni namen modelnih poskusov je bil optimiranje VOD obdelave s posebnim poudarkom na zagotavljanju maksimalne produktivnosti VOD naprave ob najmanjši možni toplotni obremenitvi VOD ponovce in minimalni temperaturi preboda. Ugotovili smo, da le računalniško programirano pihanje zagotavlja maksimalno produktivnost ob še sprejemljivi toplotni obremenitvi VOD ponovce in minimalni temperaturi preboda.
1 Introduction
The production in metallurgical industries was for a very long time based dominantly on experience. However, the development of theory of metallurgical processes and metallurgical thermodynamics couplcd with small-size and yet povverful computers has made it possible to combine the experience with theoretical knovvledge, which can result in a signilicant improvement of operation perfomiance and raising techno-economic production parameters to an essen-tially higher lcvel.
The application of computers in steelmaking has become indispensable due to high competition and ever increasing demands for lovver priče, better quality and narrower toler-ances of steel properties such as strength, ductility, hot and cold deformability, chemical composition, corrosion resistance, etc. The concept of TQC (Total Quality Control) which excludes any refuse is based on absolutely reliablc control of quality. At present TQC is obligatory for ali high priced products. Consequently, it is imperative also in the manufacture of stainless steel since the value of a heat of high quality stainless steel can amount up to 300,000 USS. Naturally, there is no room for any allowable refuse despite the fact that technological regulations at present rec-ognize allowable waste. Conventional technologial regulations deal vvith imaginary "average" heat and "normal" con-ditions. For that reason and because of great value, each heat of stainless steel must be regarded as unique i.e., it must be processed in a specific way taking into account actual initial state and conditions (chemical composition, temperature, VOD ladle state, pumping sy.stem state, etc.). There is no such thing as standard process parameter. Every de-viation of initial conditions from some imaginary standard can and must be compensated for by adequate adaptation of process parameters. Such a complex assignment can suc-cessfully be carried out only by computer aid.
The work was canied out as a part of the projeet Stainless Steel vvhich is aimed at the reduetion of production costs, improvement of the quality of stainless steel, development of new high quality and extra clean e.g. superfer-ritic stainless steels and the optimization of EAF-VOD-CC tehnological line operation.
The research was concentrated to vacuum oxygen de-carburization rate, which should be inereased as much as possible since VOD treatment is serious bottleneck of the technological line particularly when producing ELI (Ex-tra Low Interstitials) steel. In addition better control of melt temperature and lovver carbon, nitrogen and phospho-rus content especially in čase of extra clean steel must also be considered in any optimization of VOD treatment.
2 Computer simulation and model tests
Main part of the research was carried out in the form of computer simulation of VOD treatment which is a key part of the technological line.
The model tests performed by use of CAPSS (Computer Aided Production of Stainless Steel) vvere undertaken
•	to inerease productivity by shortening of VOD treatment,
•	to improve techno-economic production parameters and
•	to determine most rational and expedient measures to be taken in order to improve the present technology
2.1 CAPSS
The PC oriented software CAPSS has been developed in Mathematical Modelling and Computer Simulation Department of IMT (Institute of Metals and Technologies) Ljubljana on the basis of a sophisticated mathematical
model elaborated through extensive theoretical studies and investigations1-10. The model has been devised primarily for thermodynantical analysis of the Fe-Cr-C-Si-Mn-Al-Ti-O-N system in ntolten state which is essential in the production of stainless steel.
CAPSS integrates thermodynamic principles, industrial experience, theory of metallurgical processes and expert knowledge required for the economic and massive production of high quality austenitic, ferritic and superferritic conventional, ELC and ELI (Extra Low Carbon, Interstitials) stainless steel.
Thermodynamical data and published informations on VOD operation results were obtained from reference11-33. CAPSS vvas calibrated by a posteriori analysis of a number of heats which had been produced in steelvvorks Jesenice to determine some entpirical unnteasurable constants typical for the steelworks. CAPSS was developed for:
•	Off-line control of VOD (Vaeuum Oxygen Decarbur-ization) unit of EAF-VOD-CC technological line for the production of stainless steel
-	to reduce the operating costs
-	to inerease the productivity and
-	to secure the high quality of produced steel
•	Research and development purposes
-	to perform model tests by simulation of VOD operation in order to determine the effect of a change in process variables (oxygen blovving rate, oxygen and argon comsumption, lime ad-dition, etc.) or initial conditions (e. g. melt composition, temperature, etc.) on operation results (productivity, production costs, quality of steel produced, etc.)
-	to optimize production technology and
-	to develop new grades of stainless steel
•	Education and training purposes
-	for training of technical personnel in newly installed VOD units or steelvvorks and
-	for education of University students
2.2 Computer control ofVOD treatment
Advantages of the computer control of VOD treatment are nunterous and evident. Firstly, intmediate and continuous inspeetion of the process of vaeuum oxidation and development of melt temperature. It offers the possibility to stop oxygen blovving at exactly right tirne i.e., at desired final carbon content. The progress and rate of oxidation are continuous^ monitored and controlled in order
•	to prevent melt temperature from exceeding allovvable upper limit,
•	to assure the lovvest heat loading of ladle lining and
•	to attain the maximum productivity and the lovvest pos-sible loss of chromium.
Computer control is benelicial also for reduetion stage of VOD treatment since it
•	offers monitoring of the reduetion of chromium from slag and temperature changes,
•	assures exact calculalion of proper addition of lime and reducing means and
• helps to synchronize the operation of whole technological line EAF-VOD-CC and especially of VOD unit and consequent continuous casting machine.
For research and development purposes simulation of imag-inary or real but future heats knovvn also as model test is particularly useful. Similar model tests performed in the form of simulation of real past heats i.e., a posteriory anal-ysis is very instruetive in education and training of technical personnel. It can reveal errors and vvrong decisions made in the past during VOD treatment of a given heat. Simulation and analysis of previous heats can be extensive and include ali heats manufactured e.g. in the last month or year. Such an analysis is valuable for evaluation of the existent tech-nology. It can help to find out and determine oscillations of techno-economic parameters (productivity, specific con-sumption of energy and raw materials, etc.) due to instable technology, unsuitable technical regulations, poor perfor-mance of technical personnel or inadequate maintenance. Particularly valuable is analysis of extreme heats i.e., ex-ceedingly good and bad operation results. It can result in significant technological improvements and more suitable regulations. By ali means it is first and obligatory step to-vvard the introduetion of computer control of steelmaking.
3 Optimisation of VOD treatment bv computer simulation
3.1 ELI stainless steel
When producing ELI e.g., superferritic stainless steel deni-trogenization of melt proceeds under deep vaeuum (approx. 100 Pa) at a high carbon content. Of course, such a vaeuum can be made only in especial and additional stage of vaeuum treatment vvithout oxygen blovving. There can be one, tvvo or even more denitrogenization steps vvhich depends on initial melt temperature and antount of heat losses. The extent and kinetics of this intermediary denitrogenization treatment can also be determined by model tests. Therefore, the efficiency of one stage and the necessity for multistage denitrogenization treatment to obtain required total (C+N) content can be established. As regards the oxygen blovving technique there are tvvo methods namely, uniform and "smart" oxygen blovving. In final stage of oxidation decar-burization rate is very small since the most part of oxygen is uitilized for harmful oxidation of chromium. Therefore, it should be smart and useful to decrease oxygen blovving rate gradually tovvard the end of decarburization.
Uniform oxygen blovving is another blovving technique vvhich applies steady blovving rate e.g., 900 m3/h and does not take into account rapid and continuous inerease in the chromium/carbon activity ratio occuring in the final stage of decarburization. This change of thermodynamic conditions results in a drastically reduced decarburization rate and inereased chromium oxidation rate vvhich is accompanied vvith steep rise in melt temperature. Consequently, the final stage of decarburization is most delicate since VOD ladle lining must be prevented from thermal overloading. On the other side computer simulation in the form of model tests airned at the optimization of VOD treatment should also optimize the productivity. The most appropriate compro-mise betvveen high productivity and temperature limitation can be find out only by the use of computer simulation and model tests.
Model tests carried out by the use of CAPSS vvere planned to compare "smart" and uniform blovving (900 m3/h) to find out the advantages and deficiencies of each.
"Smart" blowing	Uniform 900 m3/h
"Smart" blovving	Uniform 900 m3/h
"Smart" blovving	Uniform 900 m3/h
Time min. —>
Figure 1. Influence of blovving technique on VOD productivity.
Slika 1. Vpliv načina pihanja na produktivnost VOD naprave.
As can be seen from fig. 1 "smart" blovving character-ized by continuously diminishing blovving rate tovvard the end of decarburization, as compared to uniform 900 nt3/h rate, results in a serious drop of productivity i.e., prolonga-tion by 40 mins. of the tirne required for decarburization from 1.2% C to 0.02% C.
Quite unexpectedly, fig. 2 shovvs that adaptation of blovving rate to the continuous change in thermodynamic conditions as applied by "smart" blovving does not reduce the extent of chromium oxidation to slag. There is no change in final chromium content vvhich clearly does not depend on oxygen blovving method.
Uniform and intense oxidation vvith 900 m3/h blovving rate results in the reduction of time required for VOD treatment by 40 mins. as can be seen from fig. 3. Hovvever, this increase in productivity could be too expensive since max-imum melt temperature vvould exceed prescribed 1700°C limit.
Fig. 4 shovvs that blovving technique exerts practically no influence on the volume and kinetics of denitrogeniza-tion. Because of a higher melt temperature at uniform blovving favorable influence of temperature on denitrogenization of stainless steel can be seen.
E
3
S o
■C
O
17.20
16.80 16.60 16.40 16.20
1450 ........................................................................................................................................................................................
0 20 40 60 80 100120140160180 200220
Time min. --->
Figure 3. Effect of blovving method on melt temperature. Slika 3. Vpliv načina pihanja na potek temperature taline.
3.2 Conventional stainless steel
The manufacture of ELI stainless steel e.g., superferritic steel differs from production technology for converttional stainless steel by intermediate stop of oxygen blovving fol-lovved by special usually 15 mins. long denitrogenization step at a high carbon content (approx. 1.0% C). Initial carbon content of melt planned for the manufacture of com-
The attention has ben focused at the productivity, extent of chromium oxidation to slag, maximum temperature at the end of oxidation and volume and kinetics of denitrogenization. Initial melt composition and temperature vvas 17.0% Cr, 1.20% C, 0.1%. Si, 0.5%. Mn, and 1550°C, respectively. Final carbon content should be lovver or equal to 0.02 vvt.% . C. VOD treatment should be as short as possible to obtain 1 high productivity of VOD unit hovvever, melt temperature should not exceed 1700° degrees. The results of model . tests are presented in figs. 1^}.	^
15
Figure 2. Effect of blovving method on the oxidation of chromium. Slika 2. Vpliv načina pihanja na obseg in potek oksidacije kroma.
220
0) 3
a
E
Q) K
15.60
0..........20........40........60.......80......YoO 120 140160 180 200 220
Time min. -- >
1500
A
C O
-Q i-<0 O
0 0 Iiiiiii iii iuii lintn li i li i n iii i ii i ii i ii i m i m i iiimiii 111 lil ni hi iiii n lil i iii li ni i ii li ii i ii i n hi i ii' iti ii i ti i m iif
0 15 30 45 60 75 90 105 Time min. — >
"smart"
uniform	controlled
A
E
S o
v. -C
O
15.60 liiuinniiiiiniiniimiiniiiiii................................................I .....................
0 15 30 45 60 75 90 105 Time min. —>
"smart"
uniform
controlled
Figure 4. Infiuence of blowing technique on denitrogenization. Slika 4. Vpliv načina pihanja na potek in hitrost razdušičenja.
mon stainless steel is therefore significantly lower in order to shorten VOD treatment. Final carbon conten of common steel is higher which also helps to reduce the time required for vacuum oxygen decarburization. Special denitrogenization stage is not necessary since low nitrogen content is not preseribed for common stainless steel. Except for ELC grades technological proeess for common stainles steel does not include (tnal vacuum degassing stage followed by re-duetion. Therefore vacuum decarburization treatment can be performed also by controlled blowing rate in addition to "smart" and uniform blovving. The need for additional i.e., controlled oxygen blowing appears as a consequence of vvider range wherefrom proeess parameters can be se-lected. The tirst aim of controlled blowing technique is to maximize the productivity of EAF-VOD-CC production line by shortening duration of VOD treatment which is the slowest stage and therefore VOD unit acts as a bottleneck. The problem is serious particularly in čase of UHP electric furnace coupled with VOD unit. Hovvever, maximization of productivity by optimization of VOD must take into account follovving requirements:
•	EAF tap temperature should be as low as possible,
» highest allovvable temperature at tire end of oxidation is 1700° C and
•	melt temperature at the end of VOD treatment should coiTespond to the requirentents of continuous casting (CC).
Exact solving of the problem is not possible. A compro-mise is needed between the contradictory requircments men-tioned. Computer simulation i.e., a series of model tests carried out during tapping, ladle transport and preparation for VOD treatment is only possible. Figs. 5, 6 and 7 present results of model test planned to determine most appropri-ate controlled blowing technique in order to optimize VOD treatment.
Figure S. Optimization of VOD productivity by seleetion of proper blouing technitjue for common stainless steel. Slika 5. Optimiranje produktivnosti z izbiro načina pihanja pri izdelavi klasičnega nerjavnega jekla.
As seen from fig. 5 "smart" blovving requires longest time for decarburization. There is no essential difference in decarburization time at uniform and controlled blowing technique (101 mins. and 102 mins., respectively).
1750
A
O o
a
E
<]>
1700
1650
1600
1550
1500 i
1 450 *—IIHMIlllMI........—um—TT------r--------------------
0 15 30 45 60 75 90 105 Time min.—>
-"smart"	uniform	controlled
Figure 6. Intluenee of blou ing method on chromium oxidation. Slika 6. Vpliv načina pihanja na oksidacijo kroma v žlindro.
In ali three cases chromiunt content at the end of treatment is the same as seen in fig. 6. This holds for both oxidation and reduetion stage. Hovvever, it can be seen that during processing chromium content of melt is on the average lovvest at controlled blowing because of lower initial temperature.
350
300
A	
1 1 E	250
CL	
0.	
š	200
C	
0)	
CD 0	150
i:	
§	
100
"V.
50 LriifiniiHiiiiiiiiiiiiii«
0 20 40 60 80 100 120140 160 180 200 220 Time min.—>
"Smart" blovving
Uniform 900 m3/h
Figure 7. Influence of blovving method on themial load of VOD ladle.
Slika 7. Vpliv načina pihanja na toplotno obremenitev VOD ponovce.
At uniform 900 m3/h oxygen blowing rate (fig. 7) melt temperature at the end of oxidation just exceeded the al-lowable limit (1700° C). Heat load of VOD ladle is lowest at controlled blowing which is optimized to obtain praeti-cally the same productivity (102 vs. 101 rnins.) at EAF tap temperature reduced by 20° C.
4 Conclusion
By the use of PC oriented software CAPSS which had been developed as a part of the research programme URP-C2-2566 computer simulation of stainless steelmaking EAF-VOD technology was carried out.
A series of model tests was perfornted to investigate the influence of three different oxygen blowing methods on the productivity of 90 ton VOD unit, chromium losses with slag, volume and kinetics of denitrogenization, and themial load of VOD ladle.
Based on the rcsults obtained following conclusions can be dravvn.
•	Common oxygen blowing technique known as "smart" blowing does not make it possible to reduce chromium losses. Hovvever, it has very negative influence on the productivity of VOD unit and consequently whole steelmaking EAF-VOD-CC technological line.
•	Uniform oxygen blowing with constant blowing rate does not inerease chromium losses to slag. By proper
seleetion of blowing rate this blowing method can re-sult in a higher productivity hovvever, VOD ladle lining can be thermally overloaded.
• Best results can be achieved by computer controlled blovving rate which ensures highest possible produc-tivity at lovvest tap temperature and thermal load of VOD ladle.
5 References
1 N. Smajič: Mathematical Model for EAF-VOD Stainless Steelmaking, Proceedings of 6. CENIM, Madrid, October 1985.
N. Smajič: Modelle Thermodinamique de 1'Elaboration de 1'Accier inoxydable suivant le Procede VOD, Proceedings of 25. JAS, St. Etienne,1986.
N. Smajič: Verifikacija matematičnega modela za računalniško vodenje EOP-VOD tehnologije izdelave nerjavnih jekel, Žel. Zbornik, 1986, 3.
4 N. Smajič: Študij kinetike vakuumskega razdušičenja su-perferitnih talin, Poročilo Metalurškega inštituta v Ljubljani, N 89-007. 1989.
N. Smajič: Superferitna nerjavna jekla. Žel. zbornik, 1988.2.
6 N. Smajič: Vakuumsko razdušičenje nerjavnih jekel, Žel. zbornik. 1990, 1.
' N. Smajič: Production D'Acier Ferritique Inox Avec Car-bone et Azote Tres Bas Par La Technologie EAF-VOD, Bulletin du Cercle de Metaux, 29ems Journees 'Etudes de Metaux, Avril 1990, Saint Etienne.
8	N. Smajič: Vakuumsko razdušičenje nerjavnih jekel, Žel. zbornik, 24, 1990, št. 1.
9	N. Smajič: CAPSS, Demonstration on Technova 92 International, Graz 1992.
10	N. Smajič: Integralni matematični model EP-VOD tehnologije, Poročilo IMT N. 92-09, Ljubljana, 1992.
11	Mc Coy et al., Electr. Furn. Conf. Proc. 21, 1963, 17-26.
12	Hilty. D.C. cit. v Handbook of Stainless Steel, 1977, 3-32.
13	Peckner, D.: Handbook of Stainless Steel, Mc Gravv Hill Book Inc., N.Y. 1977, 3-26.
14	Otto J. et al., Stahl und Eisen 96, 1976, 1939-45.
15	Otto J.. Disertation, Aachen 1975.
16	M. Schmidt et al., Stahl und Eisen, 88, 1968, 153.
17	H. Bauer et al, ibid. 90, 1970. 725.
18	C. Wagner: Thermodynamics of Alloys, Addison-Wesley, Cambridge, 1952.
19	J.F. Elliot et al., Thermochemistry for Steelmaking, vol. 2, Addison-Wesley, London 1963.
20	Ban-ya et al.. Tetsu-to-Hagane, 60, 1974, 1443.
21	F. Tsukamoto, Transactions of ISIJ, 26, 1986, 273-81.
22	B.I. Leonovič et al., Metalli, 1980, 4.
23	Fujio Is hi i et al., Tetsu-to-Hagane 68, 1982, 946-55.
24	K. Mori, Transactions of ISIJ, 28, 1988, 246-261.
25	N.P. Vladimirov et al., Metalli, 5, 1973, 89-95.
26	M. Murakami et al., Transactions ISIJ, 27, 1987.
27	K. Ishihara et al., Proceedings of 100111 ISIJ Meeting, October, 1980, 836.
28	K. Mori et al., Tetsu-to-Hagane, 52. 1966, 1443-45.
29	V.P. Luzgin et al.. Izvestija VUZ, Čem. Metal. 9. 1963, 50-54.
30	J.V. Javojskij, Izvestija VUZ, Černaja Met. 1977. 7.
31	E.T. Turkdogan et al.. Trans. Metali. Soc. A IME, 1963, 227.