Strip Casting: Anticipating New Routes to Steel Sheet
a proposal to the AISI
by
Alan W. Cramb
Department of Materials Science and Engineering
Carnegie Mellon University
Pittsburgh, Pa 15213
Executive Summary
Strip casting on a commercial scale is now feasible for several grades of stainless steel and the first comercial strip caster has begun operation at the Nippon Steel, Hikari Plant. For carbon steels the economic feasibility has yet to be determined. Despite extreme secrecy, however, it is clear that a carbon steel strip caster will be commercially available within three to five years. The need to obtain a detailed understanding of how this technology is likely to impact future industry competitiveness is now becoming critical.
To obtain a sound grasp of the potential of this breakthrough, in terms of product attributes, we are proposing a high quality, timely, cost effective program, utilizing:
This project will bring together personel from Max Planck, CMU and partner steel companies in a manner which could greatly enrich strategic business planning, for those companies which opt to actively participate.
Introduction
Strip casting of 304 stainless steel was commercialized in September of 1997 at the Hikari Works of Nippon Steel in Japan. The Myosotis Project in France is currently casting 90 tonne heats of 304 stainless on a routine basis at the Isebergue plant of Ugine and 60 tonne heats of carbon steels are routinely cast by the Project "M" team in Port Kembla, Australia, by BHP. Strip casting on a commercial scale for a limited number of grades is technically and perhaps commercially, feasible; however, as yet, it is not clear what the future holds for the strip casting of steels. At the moment there is a great potential for strip casting to reduce energy, process steps, manpower, investment and operating cost while increasing productivity and quality in currently produced grades. There is also a future potential for strip casting in that it may be possible to produce in sheet form (1) new grades that were previously not produced as sheets due to difficulties during rolling, (2) steel chemistries not amenable to current casting technologies, (3) strip with optimized texture for electrical and magnetic applications, (4) steels which have a lower average inclusion size distribution (cleaner steels) and (5) steel grades with higher residual contents than can be tolerated with conventional processing techniques.
Current studies worldwide are aimed mainly at the development of strip casting as a replacement technology for conventional grades where the cast strip should be equivalent in profile, shape and properties to a hot rolled strip; however, a fundamental processing issue that has not been yet been overcome is the control of the product microstructure and texture on grades other than 304 stainless. Initial carbon steel strip cast studies at Bethlehem Steel by a consortium of steel companies in the early 80's indicated that some hot processing might be necessary to develop conventional properties from conventional grades that are strip cast [1]. A recent presentation by BHP has also suggested that either hot rolling, annealing or a chemistry variation would be necessary to develop conventional properties from strip cast carbon steels; however, patent literature by Nippon Steel [2] has indicated that strip cast carbon steel can be cold rolled and annealed to produce a cold rolled steel sheet product with conventional properties and a broad tolerance to residual levels. Thus, at this time, it is not clear if the strip cast material can be used directly without hot reduction from the caster and without significant changes to the grade chemistry. A limitation of strip casting may be that conventional mechanical properties in carbon steels may not be achieved without changes in conventional grade chemistries and significant post caster processing; however, a potential of strip casting may be that significantly improved product properties may be possible with the same changes. Thus it is difficult to determine the future of strip casting for carbon steels as the complete processing route has yet to be determined.
Currently North America does not have a major effort aimed at developing the strip casting process and, based upon the current status of worldwide development, such a project would now be unwise; however, it is clear that all developments up to the present time have been aimed at strip cast process development. Little information is available on the use of strip cast material in conventional applications (other than for 304 stainless steels) and recently the major developers of strip casting technology have carried out projects to study microstructural development in strip cast carbon steels as it has been recognized that casting a defect free steel sheet of appropriate dimensions for further processing is a necessary but not a sufficient criterion for the application of strip cast steels. The development of strip casting is now passing from the realms of the casting engineer to the metallurgist as the strip microstructure must also be controlled to produce an acceptable product.
It is the purpose of this proposal to outline a plan to determine the potential of strip casting and to determine the post casting processing that will be necessary to ensure that strip cast material meets or exceeds current requirements. Our first assumption is that all technological problems in the production of a steel strip directly from liquid steel are close to a solution and that the production of a steel strip of required dimensional tolerance will be possible and available commercially within 3 years from a number of vendors. Given this assumption, this project is aimed at determining the future potential of strip casting within five years. The proposed study is critical for the steel industry as the proposed research will provide the impetus for understanding the technology on a firm scientific basis.
Technical Issues
While the technological issues associated with strip casting are currently being addressed by the steel industry worldwide, a fundamental issue (hurdle) that has not been overcome is the control of the product microstructure and texture. This is particularly significant for strip cast products since there is very little scope for tailoring the microstructure by significant downstream thermomechanical processing (TMP) as in conventional sheet production. Since the solidification and deformation conditions during strip casting are significantly different from those in conventional processing, a commercial steel of nominal composition may exhibit a completely different microstructure and texture than a conventionally processed steel sheet of the same composition. Hence, it may be required to either modify the steel composition or develop an alternative processing route to obtain the same (or better) properties as found in a conventionally processed steel.
The key parameters which influence the microstructure and texture in the cast strip include the steel composition, the superheat, the casting speed, the heat transfer characteristics between the solidifying strip and the rolls, and the flow characteristics of the incoming liquid metal. These parameters essentially determine the endpoint of solidification during strip casting. The texture is determined by the relative widths of the columnar and the equiaxed zones in the cast microstructure. While the columnar zone consists of grains which are generally oriented with <100> parallel to the heat flow directions, the equiaxed grains both on the chill surface and possibly at the mid-thickness have a random texture. The grain size depends essentially on the nucleation density on the chill surface and the density of homogeneous nucleation in the bulk of the metal. The end of solidification also determines the magnitude of yet another important process variable, namely, the extent of hot reduction that occurs between the rolls. While it is desirable to minimize the extent of hot rolling in order to maximize the roll life in the strip casting of steels, a certain amount of reduction may be necessary to break down the cast structure and the texture.
Many opportunities exist to exploit the structure and properties of strip cast steels through control of their physical metallurgy. The motivation for this is that the microstructure of strip cast material differs markedly from that of the comparable material manufactured by conventional methods, i.e. hot band. For some applications, there are opportunities to take advantage of the differences. Electrical steels, for example, are sensitive to the crystallographic preferred orientation (texture) - the typical weak cube texture formed in solidification may be advantageous for certain magnetic applications. The high solidification rates for strip cast steels will also permit us to explore new avenues for control of microalloying additions and precipitated inclusion size distribution.
Hutchinson recently reviewed the relationship between steel making practice and mechanical anisotropy and noted that high purity steels (interstitial-free steels) offer the best thinning resistance, as measured by r-value [3,4] . This observation is correlated with the strong <111>||ND fiber that can be obtained in the interstitial free steels by cold rolling and recrystallization. The first stage of hot rolling is now known to yield significant variations in texture, despite the expectation that the g-a transformation (for plain carbon steels) will tend to randomize the texture[5]. This variability in "hot band" texture often has an effect on the subsequent cold rolling texture. The presence of carbide formers (e.g. Ti, Nb) can lead to precipitation during hot rolling which restrains recrystallization and therefore tends to promote stronger hot band textures. Cold rolling typically generates textures which contain strong g ({111}<uvw>) and a ({hkl}<110>) partial fibers [6] . Thus any study of strip casting must contain a program to determine the factors controlling texture formation during further processing as the role of hot rolling on texture formation will be minimized in strip casting.
To determine strip cast potential it is necessary to produce strip cast material under defined conditions and then process it under closely controlled conditions to understand microstructural evolution. Clearly this means that a strip caster must be available for the program so that defined grade chemistries can be cast. The caster must be able to produce material which replicates the solidification and recrystallized microstructure of a commercial twin roll caster, but it does not need to produce material on a commercial scale. The pilot caster only needs to produce material for testing which will then be fully characterized with respect to mechanical properties and the effect of post casting processing on mechanical properties. Thus a pilot caster which can produce material adequate for testing is necessary.
Direction of Research
To overcome these technical hurdles it is necessary to institute a program aimed at (1) the determination of the post caster processing steps necessary to develop appropriate mechanical properties and (2) the determination of the feasibility of the process for production of novel grades that are tailored to take advantage of the uniqueness of the process. This project will have three major thrust areas: casting, secondary processing and characterization of processed materials. In order to accomplish these three tasks it is proposed that two centers of steelmaking research ( the Center for Iron and Steelmaking Research at Carnegie Mellon University and The Max Planck Institute in Germany) combine with a number of steel and associated companies in a three year long program to allow the potential of strip casting for steels to be developed.
Casting will be carried out in the initial stage (18 months) of the project at the Max Planck Institute in Dusseldorf, Germany using their pilot facility and at the British Steel Research Laboratories in Middlesborough. Grade selection and processing parameters will be determined jointly with the steel industry, MPI and CMU and secondary processing and characterization of the strip cast material will be carried out at CMU, MPI and at selected steel industry labs. Within this joint project, MPI will concentrate on the features of the as-cast steel strip. The material will be produced and characterized; the phenomena during casting and the properties of the cast strip will be studied; ways for improvement will be investigated. These research steps will be carried out in collaboration with the other project partners. Casting of carbon steel grades at British Steel will also be pursued, to allow different process parameter ranges to be studied. Work at Carnegie Mellon and Max Planck will be aimed at developing a fundamental understanding of the effect of heat transfer in strip casting structure development, determining strip cast structures, developing and characterizing microstructural development and determining the relationship between microstructural and mechanical property development. Various laboratories will secondary process the strip cast material and aid in property determination.
Team members will be A. W. Cramb (project leader, CMU), A. D. Rollet and two graduate students from Carnegie Mellon University, Dr. K. H. Tacke and Dr. Buechner at MPI and members of the technical staffs of LTV, USS, Dofasco, National Steel and AK Steel. A project management team lead by AISI will be set up with a representative from each member organization involved in the program.
Benefits of Program
Strip casting is a process which could either become a small niche market technology or an outright replacement for thick and thin slab casters. Thus there is the potential for either small or large changes to the steel industry based upon the extent of implementation of the technology. This program is aimed at determining the potential of strip casting and will lead to a better understanding of the future of the technology within the steel industry. If the program is successful new grades may be identified and the processing route necessary for commercial grades developed. Strip casting has great potential to reduce both the operating and investment cost for steel casting.
Goals
To determine the potential for strip casting in the steel industry
To develop the fundamental knowledge necessary to allow the role of strip casting in the modern steel industry to be understood
Objectives
Phase 1
To determine the operating window and types of solidification structures produced by the pilot machines and benchmark the solidification structures versus known literature values from other operating machines
Phase 2
To strip cast various low and ultra-low carbon steels and determine their cast structure, mechanical properties and the effect of post casting processing on mechanical properties
To determine if a modification in steel grade composition will allow variations in processed strip properties to be developed
To test the response of strip cast commercial steel grades to varying residual element contents
Phase 3
To optimize the post processing of strip cast material
To determine the potential of strip casting for novel or unconventional applications based upon the information gathered during phases 1 and 2
Project Plan
The project on strip casting of steels will be carried out using the pilot casters at the Max Planck Institute in Germany and at British Steel to supply cast strip of commercially significant grades for metallurgical analysis. Phase 1 of the project will last six months and will be aimed at developing a knowledge of the solidification structures developed by the pilot machines to determine the operating window for the machines and its relationship to the material produced by larger machines. Phase 2 will constitute the next 18 months of the project and will be aimed at determining a baseline characterization of the material to determine appropriate processing conditions for strip cast material. Phase 3 will encompass the last year of the program and will be aimed at exploiting the information gained during the first two years of the program and determining future potential of strip casting for more novel and unconventional applications.
The caster at the Max Planck Institute für Eisenforschung is fed by a 75 kg furnace and can produce material that is 120 mm wide and a thickness varying from 1 to 3 mm. The rolls are water cooled copper. This machine will be used to cast material that can be used for initial analysis of strip casting. The MPI machine will allow microstructural characterization and the necessary strategy in post processing of strip cast carbon steels to be developed. This machine has supplied material to the ECSC program on strip casting. MPI will characterize the as-cast strip with respect to castability, with respect to the mechanical properties, and, in conjunction with the project partners, with respect to microstructural properties.
According to previous experience, to cast conventional grade specifications on a twin roll caster can sometimes lead to satisfactory results, in other cases, it may not be optimal. MPI will therefore study the effect of casting conditions and of alloying composition both on castability and on the properties of the cast material. Ways to improve the castability of the steel grades in the twin roller casting will be investigated. This may include modified chemistries of the grades to be cast, but also improvements of the casting process and its parameters.
Since this project is concerned with the potential and problems of the full production line from casting to the properties of cold rolled strip, a feedback loop will be established with the other project partners. They will receive material from MPI and carry out characterization, tests and post processing steps with the samples. According to their results, it may become necessary to further modify the composition of the steel, which will be carried out at MPI.
The MPI machine will supply test material but will not be able to control nitrogen and oxygen levels to that normally found in current steelplants. British Steel have committed in principle to producing cast strip for analysis for the program and the British Steel machine supplies a larger coil with better temperature and chemistry control than Max Planck; however, heat transfer conditions will be significantly different at British Steel than Max Planck.
The rolls at British Steel are 750 mm in diameter and 400 mm wide. The bodies of the rolls are steel and the casting surface is a 16 mm thick martensitic stainless steel sleeve. Water cooling is from one end of the roll to the other along the axis of the roll with multiple channels around the periphery of the roll. Both rolls are connected in series to the water supply source. The water channels are 16 mm below the surface of the roll and appear to be at the interface between the roll surface layer and body. The machine was built by Davy before they got involved with POSCO. The design is almost identical to that of POSCO. The roll arbors are very large over twelve inches in diameter indicating that they have a very rigid system. They indicated that they can take reductions at the nip using their machine.
The machine is driven by hydraulic motors that are mechanically coupled. They can currently cast between 16-18 m/min. and are capable of casting up to 36 m/min with their current drives. They cast about 1 tonne of metal and produce strip that is up to 90 m in length. Normal cast thickness is between 2.5 to 3.5 mm and the design range is 1 to 5 mm. They machine a concave surface in the rolls to accommodate any thermal expansion that occurs in an attempt to obtain a flat cast strip profile. The machine has a fixed roll gap. A scraper made of some type of plastic sealing material is located on the top of the roll at the 11 o'clock position, if the nip is at three o'clock, away from the casting surface to scrape any splashes from the rolls and air knives are located at nine o'clock to clean rolls and cool if necessary. British steel was one of the companies producing strip for the ECSC project and has the capability to produce stainless and carbon steels. Both casters allow cooling after casting.
304 stainless will be cast in phase 1 as a benchmark material for solidification structures. Phase 1 casting of 304 stainless and the low carbon steel will allow benchmarking of the casters. Further grades to be cast at the Max Planck and British Steel will include: (1) a low carbon (0.04%) , aluminum killed steel grade typical of that used in deep drawing and automotive applications and (2) an ultra low carbon titanium stabilized grade. Aim cast thickness will be 2 mm. The goal of this study will be to determine the potential application of strip casting for these common grades. These grades will be especially interesting as their properties are well known within the industry and will act as an excellent reference group. Phase 2 will include casting of both of the above grades and a third set of experiments focused upon the effect of residuals on cast properties. Strip casting may be insensitive to residual levels as hot cracking due to high copper contents will be eliminated in this technology. In Phase 3 the knowledge gained in Phase 2 will be used to develop a test plan to define the potential of strip casting in current and future grades. All studies will include a study of the statistical variation inherent in the samples.
Casting Plan
Melt stock will be supplied by member companies. The program is aimed at (a) developing a baseline knowledge of the material (phase 1), (b) determining the structure and potential properties of low carbon and ultra-low carbon titanium or titanium-columbium stabilized grades and (c) determining the effect of residual and solute elements on strip cast properties (phase 2) and on optimized processing based on developed knowledge (phase 3).
Phase 1
Max Planck: 2 heats - 304 stainless and a low carbon (1006 grade). British Steel: 304 stainless and a low carbon steel (1006).
Phase 2
Max Planck and British Steel - low carbon, ultra-low carbon titanium stabilized or dual stabilized (Ti + Cb) heats. Low carbon heats with (1) copper content of 0.5%, (2) sulfur content of 0.050%, (3) phosphorus content of 0.050%.
Phase 3
Max Planck: Optimized low carbon and ultra-low carbon grades and potentially casting of novel grades based upon the information gathered during phases 1 and 2.
Processing and Characterization
After casting there will be a program aimed at characterization of the strip cast material. CMU will investigate both the strip casting solidification structure, suface roughness, surface quality and the recrystallized microstructure to determine the types of microstructure and inclusion type and size distribution that are developed during the strip casting process. Another objective will be to define whether and/or how the strip cast material produced at MPI and BSC will differ in this respect from material produced on larger scale strip casters. CMU will investigate the microstructure and heat treatment of strip cast sheet steel with a particular focus on texture and mechanical properties, both of which will be measured at CMU. Thermomechanical treatments will include cycling through the phase transformation for microstructural refinement and texture randomization, as well as conventional rolling and annealing cycles. Mechanical properties such as strength, elongation, ultimate strength and r-value will be measured. These properties will be obtained from instrumented tensile tests using extensometers, for example, to measure lateral strain. Texture will be measured with a combination of X-ray pole figures and the new analytical tool of Orientation Imaging Microscopy (OIM). OIM permits microstructure to be mapped based on crystallographic orientation. This characterization technique is yielding new insights into basic processes such as recrystallization (1). CMU is leading the field in this characterization technique, and has an NSF-supported Materials Research Science & Engineering Center that is geared towards accelerating OIM and developing a systematic approach to the determination of grain boundary properties. This and other modern characterization techniques will be used to understanding microstructure-property relationships. Based on the insights gained, processing methods will be developed for the optimization of properties exhibited by strip cast steels. This fundamental understanding will also lead to grade specification for strip cast material as the chemistry and processing conditions are tailored to allow strip cast material to be processed to the same property criterion as conventionally processed material.
Details of the proposed processing and characterization are as follows:
(1) Characterization of As-cast Material
This study will define the starting cast strip properties. Solidification structure, microstructure and mechanical properties will be determined.
Responsibility: CMU, Max Planck and partner companies
(2) Cold Reduction of As-cast Material
This study will determine if it is possible to simply cold reduce strip cast material and develop commercial properties.
(a) Cold reduction: This study will be carried out at two levels of reduction: 50% and 75%.
Responsibility: Partner companies
(b) Recrystallization kinetic studies will be carried out with cold rolled sheet using salt baths and tube furnaces to determine time and temperature combinations for subsequent annealing process simulation.
Responsibility: Partner companies and CMU
(c) Annealing simulations of both box and continuous annealing will then be carried out using simulators available at partner companies. Mechanical properties and texture measurements will then beperformed on the anealed sheet.
Responsibility: simulations of annealing by partner companies and texture characterization and mechanical properties by CMU and partner companies.
(d) Temper Rolling on Selected Panels
This will be carried out to determine the influence of temper extension on the strip cast material.
Responsibility: Partner companies for rolling, CMU for property determination
(e) Characterization of annealed (recrystallized) microstructure, texture and mechanical properties.
Responsibility: Partner companies and CMU.
(3) Hot Band Annealing
This study will determine the effect of thermal cycling on strip microstructure and on the subsequent development of properties during cold reduction, using the same schedule as above.
In this study an initial coupon study will be used to determine the effect of temperature and time on structure and to develop recommended practices. Samples will then be treated based on an optimal practice and cold reduced.
Responsibility: coupon study and characterization - CMU; cold reduction - partner companies.
(4) Hot Rolling
This study will determine the effect of thermal cycling and hot reduction on the subsequent development of properties upon cold rolling. This would simulate the likely configuration of a production unit which may include in-line hot eduction.
Responsibility: Partner companies and CMU
Schedule, Milestones and Deliverables
The project is budgeted and scheduled to last three years.
The project schedule includes the following milestones:
Project Start: April 2, 1998
Complete Phase 1 casting Sept 1 , 1998
Complete Phase 1 analysis Sept 1, 1998
Begin Phase 2 Sept 1, 1998
Begin cold reduction studies Sept1, 1998
Complete Phase 2 casting of carbon and
ultralow carbon steel April 1, 1999
Complete cold reduction studies
of low carbon steels April 1, 1999
Year 1 Report April 1, 1999
Initiate Hot Band Annealing
and Hot Rolling April 1, 1999
Complete Casting characterization
of ultralow carbon and carbon steels Sept 1 1999
Complete cold reduction studies of
Ultralow carbon steels Sept 1 1999
Complete casting of residual content steels Sept 1 1999
Complete Annealing and hot rolling
studies of carbon steels Dec 1, 1999
Complete study of effect of residuals
on cold roll properties April 1 1999
Year 2 Report April 1, 2000
Initial optimized casting trials April 1, 2000
Completion of study of Hot Rolling and
Hot Band Annealing on ultra low
carbon steels July 11, 2000
Complete casting trials Octber 1, 2000
Complete study of effect of residuals
on Annealing and hot rolling October 1, 2000
Complete analysis of optimized
processing February 1, 2001
Final Report April 1, 2001
Project deliverables for year 1 are:
Project deliverables for year 2 are:
Project deliverables for year 3 are:
References
1. L. T. Shang and P. J. Wray: "The Microstructure of Strip Cast Low Carbon Steels and Their Response to Thermal Processing", Met Trans A, Vol. 20, July, 1989, p 1191-1198.
2. T. Mizoguchi, Y. Ueshima, T. Moroboshi and K. Shio: "Thin Cast Strip and Thin Sheet Steel of Common Carbon Steel Containing Large Amounts of Copper and Tin and Process for producing the Same", Us patent 5,662,748
3. Hutchinson, B. (1994), Practical aspects of texture control in low carbon steels. Mater. Sci. Forum, 157-162, 1917-1928.
4. Hutchinson, W. B. and L. Ryde (1995). Mechanisms, kinetics and crystallography of recrystallization in cold rolled steels. In Ris~-16, Microstructural and crystallographic aspects of recrystallization, (N.
H. e. al., ed.) Ris~ National Laboratory, Roskilde, Denmark ) p. 105.
5. Raabe, D. and K. L~cke (1994), Rolling and Annealing Textures of BCC Metals. Mater. Sci. Forum, 157-162, 597-610.
6. Ray, R. K., J. J. Jonas and R. E. Hook (1993), Cold Rolling and Annealing Textures in Low Carbon and extra Low Carbon Steels. Int. Mat. Revs., 39, 129-172.