Review
Progress in Strip Casting Technologies for Steel; Technical Developments
2013 Volume 53 Issue 5 Pages 729-742
Details
2013 Volume 53 Issue 5 Pages 729-742
The present article is a sequel to the previous review on the history of near net shape strip casting facilities. The present review focuses on technical progress made in strip casting over the last three decades. Strip casting is a revolutionary technology that promises the hope for an efficient, economical and environmentally-friendly process to produce hot-rolled, steel sheets. This review provides a summary of the theory, recent research, and progress, in the developments of strip casting operations for steels, along with technical discussions regarding the characteristics and design features of steel strip casting machines. Two strip casting processes are discussed in detail; the Twin-Roll Casting (TRC) process and the Horizontal Single-Belt Casting (HSBC) process. Particular emphasis is placed on topics such as the commercial potential for strip casting technology in the steel industry, and the economic and environmental advantages of direct strip production, versus current continuous casting, fixed mold technologies.
The present article is a sequel to the previous review on the history of near net shape strip casting facilities.1) Near Net-Shape Casting (NNSC) processes have the potential for high volume, friction-free, continuous production of steel sheets. Strip casting processes, given their compact size, low capital costs and low operating costs, together with inherent low carbon dioxide emissions and low energy consumption, could be an ideal replacement for Conventional Continuous Casting (CCC) operations, in the years to come.
In a typical strip casting process, molten metal is cast directly onto a moving substrate (belt or drum), producing metal strips with thicknesses in the order of one to several millimeters. Since the as-cast metal strips are close to the final product thickness, the costly and energy-intensive downstream size-reduction and finishing steps that are needed in the CCC and TSC (Thin-Slab Casting) processes can be minimized, and directly integrated into the casting process. By merging casting and rolling into one single compact and efficient process, strip casting has the potential to provide many economic, environmental and technical benefits to the steelmaking industry.
The development of the Conventional Continuous Casting (CCC) process in the 1970’s, allowed for the continuous production of semi-finished thick slabs (200–400 mm), directly from molten steel. This eliminated the need for the expensive breakdown/roughing mills and soaking pits.1) Similarly, reduced hot mill “crop ends” lead to a decisive improvement in process yield, in the order of +10%.
The next development was Thin Slab Casting (TSC) in 1989, allowing thinner slabs, 50–60 mm thick, to be cast continuously from the melt. TSC uses higher casting speeds vs. conventional continuous casting machines (~5 m/min vs. ~1 m/min) so as to maintain equivalent throughputs in integrated steel plants. Casting and hot-rolling are combined into one continuous process. The as-cast thin slabs pass through a soaking (or muffle) furnace, and are then directly hot-rolled, in-line, into strips.1) As a result, the costs of reheating and rolling can be considerably reduced, as compared to the CCC process,2) but steel quality can be compromised.
Thin strip casting technology picks up where the TSC process leaves off, and aims to produce as-cast steel sheets that are even thinner, directly from the melt. Reheating, or thermal soaking, can be eliminated altogether, and costs associated with rolling can be further minimized. Unfortunately, thermo-mechanical processing is also minimized, severely so for Twin Roll Casting, and this can compromise the final properties of steel sheets produced.
Like other casters, a strip caster comprises a liquid metal metering system and a mould. The molten steel is still transferred from a ladle into a tundish to shield it from re-oxidation and to maintain it at an appropriate temperature. The tundish, or an intermediary holding vessel, then delivers molten steel through a Submerged Entry Nozzle (SEN), or more normally through a more sophisticated flow control system, onto a moving chilled mould(s), where it can solidify to form steel strips.
The defining feature of strip casters is the ‘moving mould’ design itself, where the mould surface moves along with the solidifying melt, thereby reducing the relative speed difference between the two surfaces to zero. Several distinct strip casting technologies exist today: Single- and Twin-Roll Casting (SRC and TRC), as well as Single- and Twin-Belt Casting (SBC and TBC).3)
Figure 1 illustrates a design schematic of the CASTRIP twin-roll caster for the production of low carbon steel strip. The CASTRIP TRC set-up, as one of the few TRC projects initiated in the 1980s that survived to commercialization,1) is chosen as an example to demonstrate the various important features of a typical TRC system. There are other TRC processes that reached industrial-scale operations, including Nippon Steel’s SS strip caster and Eurostrip, both of which were abandoned, as well as other ongoing efforts such as poStrip and Baostrip, as explained in the previous review article.1) The conventional 110-tonne ladle is held in place with an overhead crane. During casting, the molten steel in the ladle is allowed to flow through a nozzle into a tundish, and subsequently through a refractory shroud/tube into a molten metal distributer (transition piece).4,5) The distributer is shaped like a wide box and made of a refractory such as magnesium oxide, and has a series of longitudinally spaced outlet openings for liquid metal on its bottom surface, so as to allow an even flow of melt into the delivery nozzle set below.5,6)
Schematic of the CASTRIP TRC process.7)
The delivery nozzle is made of refractory materials such as alumina graphite and is reportedly an elongated shape with a tapered lower part to direct the flow of metal between the rolls so as to form a casting pool. The delivery nozzle may have a series of longitudinally spaced, vertically extending flow passages, or a continuous slot, to allow the melt to flow downwards at a sufficiently slow velocity towards the “roll nip”, without impinging directly on the casting rolls. The nozzle is usually immersed in a molten pool of steel, which accumulates above the roll nip.4,5,6,8)
To contain the melt pool at the ends of the rolls, a pair of side closure plates made of refractory materials, such as an alumina-boron nitride composite, is used. The plates are scalloped on the side edges to fit the curvature of the stepped ends of the rolls and are held against the roll ends, so as to ensure no leakage occurs. Cooling of the rolls is done internally with water flowing via a series of longitudinally extending, circumferentially spaced, water passages. The rolls used by the CASTRIP process are ~500 mm in diameter, and up to 2000 mm long, so as to cast steel strips up to 2 m wide. All other modern strip casters have used rolls with much larger diameters (~1 m), and smaller widths.5) The CASTRIP process generally produces as-cast strips typically 1.8 mm thick or less, at casting speeds of 60–100 m/min.4,6)
After passing the caster rolls, the strip enters another chamber set below.9) This chamber can be filled with either a non-oxidizing gas such as 99.99% nitrogen or argon, or with a weakly reducing atmosphere such as nitrogen with 2–10% hydrogen. This is to prevent surface oxidation of the hot as-cast strips (1300–1400°C) and scale formation. The casting chamber, itself, is also continuously cooled.6,9) The strip cools in subsequent chambers and its temperature is kept between 950–1200°C on exiting, suitable for in-line hot rolling downstream.9) The CASTRIP twin-roll caster setup is highly effective, maintaining the oxygen contents in the casting chamber to below 100 ppm.9)
After passing through in-line hot rolling mills and typically undergoing 10–50% thickness reduction,4) the hot strip passes onto a run-out table, where it is rapidly cooled by spraying water jets and is subsequently coiled.4,10) The strips may then be uncoiled, cold-rolled, and annealed, so as to obtain more desirable microstructures and better mechanical properties. The final product thicknesses are typically 0.9–1.5 mm.4)
In terms of casting roll design, another TRC process, the MAINSTRIP process, conceived by the Swiss company MTAG Marti-Technologie AG,1) has brought many new features to simplify and potentially improve the TRC process. It utilizes an integrated torque drive system that is combined with the modular, movable casting heads. This design allows the two casting rolls and their motors/gears to be integrated onto a single mobile unit. Such a compact Movable Casting Roll Assembly would enable the rapid changing of casting rolls during casting operations for cleaning and/or maintenance. The laborious process of disconnecting gear box/drive units and the detachment of bearing housings during casting roll maintenance and replacement can thereby be avoided.11,12) Moreover, the cleaning and coating/modification to the casting roll surfaces can be done on the casting head, without having to physically detach the rolls from the assembly.11,13)
In a typical proposed MAINSTRIP casting plant, there would be two movable casting heads and two tundishes working with a twin ladle turret. One casting head would be in operation with a tundish, while the other caster undergoes cleaning, maintenance, or preparation. The second tundish can also be simultaneously preheated. This design minimizes the downtime for maintenance and preparation, since one assembly can be serviced while the other can be brought into operation, quickly and easily.11,12,14) The design schematics of MAINSTRIP’s movable casting head system are shown in Fig. 2.
The design of MAINSTRIP’s integrated, movable casting roll assembly.12)
Two types of surface defects can occur during twin-roll strip casting of steels in CASTRIP operations: “Chatter” and “Crocodile-skin” defects. Chatter defects initiate at the meniscus level of the casting pool, where initial solidification occurs as the melt comes into contact with the water-cooled rolls.
There are two types of chatter defects; low-speed and high-speed chatter defects. Low speed chatter is formed when a low casting speed causes premature freezing at the upper meniscus. This produces a weak shell that is deformed as it is drawn deeper into the casting pool. The positions of the meniscus and solidified shell are shown in Fig. 3.
Simplified schematic of the TRC process, illustrating the position of the meniscus and the solidified shell, adapted from.15)
High speed chatter defects form when the casting speed is excessively high and shells form too far down the casting roll, with liquid accumulating above the forming shell. The liquid feeding the meniscus cannot keep up with the rotating roll surfaces and slippage occurs between the liquid metal and the roll in the upper part of the casting pool. This high speed chatter is characterized by transverse deformation bands across the as-cast strip.5)
CASTRIP have also identified crocodile-skin defects. These occur during solidification, when δ- and γ-iron phases solidify simultaneously, in the forming shells, causing local heat fluxes through the solidifying shell to vary. As the two iron phases have different hot strength characteristics, the variations in heat fluxes will cause localized distortions in the solidifying shell. As these distortions meet at the nip, crocodile-skin defects form on the strip produced.5,16,17)
Finally, the presence of a light oxide deposit with a lower melting point than the metal on the casting rolls has been found to be beneficial, since it ensures a more even heat flux during metal solidification. During casting, the oxide layer melts as the roll surface enters the casting pool, creating a thin interfacial liquid layer between the roll surface and the freezing steel, which promotes enhanced heat fluxes. However, excessive oxide build-up will result in drastic heat flux fluctuations as the oxide first melts to give high heat fluxes, but subsequently re-solidifies and ceases to re-melt as it thickens, decreasing heat fluxes substantially. This problem can be alleviated by uniform cleaning of the rolls, so as to maintain the oxide layer’s thickness to a strict limit. Non-uniform roll cleaning leads to variations in oxide build-up, and heat flux variations across the solidifying shell, leading to localized distortions and creating crocodile-skin defects.5)
The MAINSTRIP team has conceived and patented a unique TRC continuous roll cleaning system for an integrated, movable casting head assembly.14,18) It uses a series of measuring devices to monitor the temperature and profile of the rolls, while using laser beam units, strippers, sand blasters, various suction hoods, and two sets of brushes, so as to continuously clean each of the rolls during casting or maintenance.18)
Both roll surface texture and metal chemistry solutions can also be implemented to correct for chatter and crocodile-skin defects. Research experiments performed at McGill University have revealed how the surface roughness, topography, and substrate material, affect instantaneous heat transfer rates during casting.19,20)
The CASTRIP process uses grit-blasted roll surfaces. The peaks of the discrete projections formed by grit blasting have flat areas of ~100–400 μm2 and a distribution of 5–200 peaks/mm2. Surfaces containing more than 100 peaks/mm2 of surface (average peak density) are used for higher casting speeds. The average height of the projections may be a minimum of 10 μm, and preferably over 20 μm. Their desired surface texture can be produced on the casting rolls, by either grit blasting the metal
substrate and subsequently coating it for protection (e.g., blasting a Cu substrate and plating with a thin Cr coating), or by direct grit blasting of the casting surfaces without applying coatings (e.g., grit blasting on a Ni surface).5)
Another method of creating the required texture is by depositing a coating on a substrate using chemical or electro-deposition. Two suitable alloys for such coatings have been identified; HASTALLOY C (Ni, Cr and Mo) and T800 (Ni, Mo and Co).5)Figure 4 shows a topographical map of a typical roll surface used in the CASTRIP TRC process. Similar types of plasma coatings have been used by Hazelett Strip Casting Corporation for many years21) for belts used for their twin belt non-ferrous casters.
Casting roll surface texture for TRC (CASTRIP).5)
In terms of a steel grade’s chemistry for resolving defective strip surfaces, it has been found that the control of Mn and Si contents of the steel melt within certain limits (Mn > 0.55 wt.% and Si between 0.15–0.35 wt.%), using the aforementioned casting surface texture, makes it possible to cast steel strips at over 60 m/min, without incurring substantial defects.5) This composition was chosen to ensure the deoxidation products of MnO and SiO2 remain as liquids during casting, so as to avoid clogging and to increase the rate of interfacial heat transfer. However, this practice results in a higher volume fraction of non-metallic inclusions in the as-cast strip, as compared to conventionally produced steel.6)
Due to its inherent high cooling rate, especially at the meniscus where the melt first comes into contact with the rolls (over 20000°C/s, and heat fluxes of ~23 MW/m2), the as-cast strip has a higher susceptibility to crack formation, due to stress formation in the solidifying shell.17,22) Furthermore, there are reports of internal defects, such as shrinkage porosity and of cracks forming between the columnar and equiaxed zones within the strip.23)
With regards to internal defects and inclusions, due to the rapid cooling and solidification rate achievable, TRC; as well as other strip casting processes, have the potential of producing a fine and uniform distribution of globular inclusions.6) The size of these inclusions are generally between 0.5–5 μm, and at most, 10 μm. As a result of a smaller degree of downstream rolling (as compared to conventional processes), inclusions in TRC strips are generally not significantly elongated and remain elliptical in the rolling direction.6) This may lead to lower strip formability and shear properties compared to conventionally produced strips.
Many different approaches have been taken to alter the microstructure and improve the mechanical properties of steel strips produced by TRC. The inverse correlation between degree of hot reduction and the size of austenite grains, and hence the strength of the steel strips, is well established.6,24) Moreover, the combined effect of the degree of hot reduction and cooling rate (water sprayed cooling after hot rolling) plays an important role in determining the strength of the steel strip. For small hot reductions of less than 15%, subsequent rapid cooling results in a microstructure that is predominantly acicular ferrite (340–410 MPa yield strength (YS)), while a slower cooling rate results in a mixture of equiaxed ferrite and intragranular acicular ferrite (275–340 MPa YS). On the other hand, after a greater degree of hot reduction in TRC (>25%), slow cooling rates produce a largely equiaxed ferrite microstructure (240–310 MPa YS), whereas rapid cooling will result in a microstructure containing intragranular acicular ferrite and grain boundary ferrite (310–380 MPa YS).24)
Alloying is also used to improve the strength of TRC steel strips. By increasing the Mn content in the typical CASTRIP steel strip product from 0.6% to 0.8%, as well as applying appropriate hot reduction and cooling rates, a 35–70 MPa increase in yield strength was reported.24)
More recent study has focused on microalloying of TRC steel strips with niobium, for its strengthening effect. It was found that in low carbon steel (~0.03% C, ~0.85% Mn and ~0.2% Si), a 0.084% Nb addition would increase the yield strength of hot-rolled strips by 20–30%.17,25) This strengthening is attributed to niobium’s effect in promoting bainite and acicular ferrite formation during austenitic transformation, as well as the formation of Nb-N clusters.17,25)
3.2. Horizontal Single-Belt Casting (HSBC) 3.2.1. Machine LayoutsThe HSBC process was independently conceived by Herbertson and Guthrie, in 198826) and by Reichelt, M. Scheulen, K. Schwerdtfeger, P. Voss-Spilker and E. Feuerstacke,27) also in 1988. The two original patents eventually led to two parallel developments, one primarily in North America (Hazelett, previously BHP, and McGill Metals Processing Center, which is, in turn, supported by an international consortium) and the other in Europe (a consortium of Swedish, Finnish and German companies/research institutes).1) Unlike twin-roll casting, the HSBC technology is currently under commercial development.1) Therefore, specific design features introduced here, apply to pilot-scale casters, and may require modifications for fully-commercialized casters. One such pilot-scale HSBC caster has been operating at the McGill Metals Processing Center (MMPC) since 1999, featuring a 600 lb induction furnace.28,29) During casting operations, steel is first melted and alloyed at the furnace station. The furnace is then transferred to the casting station, once the metallurgy is correct. There, the melt is displaced upwards into the metal delivery system, by the downward motion of a piston into the induction melting furnace, so as to deliver liquid metal at a preselected flow-rate. The displaced molten metal is transferred through a launder/tundish and delivered onto a textured steel belt. The metal subsequently solidifies into as-cast strips, 1–10 mm thick.
An earlier metal delivery and melt distribution system for the MMPC-HSBC caster is shown in Fig. 5. The flow of liquid steel into the metal delivery system is again controlled by the downwards movement of the refractory piston into the melt, while the launder/tundish itself is divided into three chambers, an entry chamber (7), a head control chamber (8), and an output chamber (9). A slag weir (10) is installed at the front of the entry chamber to remove any incoming slag. Liquid metal is passed over a dam (11) before entering the head control chamber, so as to reduce turbulent inflow.29)
Detailed schematic of a liquid metal delivery and melt distribution system for the MMPC-HSBC caster.29)
Prior to casting, the entire metal delivery system, comprising the piston and the launder/tundish system, was, and still is, preheated using resistive heating systems. The induction furnace holding the melt provides direct heating of the melt during casting operations, while located at the casting station. A graphite coating can be applied to the top surface of the belt, which is water-cooled from below. It is also supported by magnetic backup rolls where the melt first contacts the belt. This measure eliminates belt distortion (bulging). The sides of the belt provided by Hazelett are also “cold framed”, so as to compensate for belt expansion during casting operations. During casting, the molten metal flows through a transverse slot onto the water-cooled steel belt moving below, at a pre-selected speed. We practice a “flying start” to the casting operations. After leaving the primary cooling section, the as-cast strip reaches the motorized run-out table fitted with a pinch roll/mini-mill. This rolling stand can act either as a pinch roll to withdraw the strip, or as a mini rolling unit in which the cast strip can undergo up to 20% reduction in thickness.
More recent work on the metal delivery system has focused on modeling the free surface flows as the melt pours from the slot nozzle onto the belt. These models incorporate the effects of surface tension, together with potential back flows of liquid metal at the triple point.30) Similarly, further work is being carried out at the MMPC on the design and implementation of an electro-magnetic flow distribution system to further enhance liquid metal spread and uniformity on the belt.
The pilot-scale HSBC caster at the MMPC has also been used extensively in experimenting new nozzle designs and configurations. Many different nozzle designs have been tested under different operating conditions, in order to evaluate their effects on critical parameters such as melt stability, strip uniformity, nozzle clogging, maximum achievable casting speed, etc. Some of these nozzle designs are shown in Fig. 6.
Some of the various nozzle configurations investigated at the MMPC.31)
In TU Clausthal’s HSBC process, liquid steel from a ladle flows through an opening at the bottom into a tundish/launder system. It is then dispensed onto the water-cooled belt by passing over a weir system.32) The belt is maintained at ~0.7 atmosphere pressure, to reduce belt distortion. While on the belt, the melt/solidified strip is protected by an inert gas atmosphere, so as to prevent ferrous/ferric oxide scale formation. This cooling section is considered to be the ‘primary cooling’ section and the as-cast strip is typically 6–20 mm thick.32)
After leaving the primary cooling section, the now substantially solidified ‘pre-strip’ passes on to the homogenization zone, also under a protective atmosphere, for temperature control and stress relief. It consists of a heat insulated housing and a roller table. In the homogenization zone, the pre-strip can be maintained at a certain temperature, further cooled down, or heated slightly, depending on downstream processing.32) Depending on alloy composition and microstructure requirements, the pre-strip can either pass into a heater, inductive or otherwise, or go directly into the finishing line, where in-line hot rolling is implemented.32)
The finishing line typically consists of three reduction rolls and ends with a set of smoothing rolls. After the desired size reduction (typically from about 50% to over 70%) has been carried out, the strip passes to the final cooling zone where is it cooled down to the coiling temperature. The strip is then cut or coiled.32)
During casting, it is important to maintain a uniform distribution of molten steel over the width of the cooling belt, so as to ensure the uniformity of pre-strip thickness. In the TU Clausthal HSBC system, this is achieved by using a special liquid steel dispenser and feeding system shown in Fig. 7. To enhance an even distribution of the liquid steel and to decelerate the liquid film as it reaches the belt, argon gas ‘rakes’ are used. These are a series of transversely distributed jets of argon gas, blown against the flow of the approximately 10 mm thick liquefied steel.33)
Detailed schematic of the steel feeding system for the TU Clausthal HSBC caster.35)
The relative velocity of the steel with respect to the moving belt has to be minimized in order to ensure undisturbed solidification. For this, a powerful magnetic field, similar to the electromagnetic brakes used in the conventional continuous casting processes, moving synchronously with the cooling belt, can be used. This magnetic field can be generated by a linear inductor called the Electromagnetic Flow Synchronization System (EFSS), installed behind the argon rakes, close to the melt surface. Additionally, the inductor can also improve the uniformity of the melt’s flow profile.33)
For either system, when casting begins and the melt first reaches the cooling belt, it is possible for uncontrolled bulges to form on the cooling belt, caused by the high thermal impact and subsequent non-uniform expansion of the belt. In the McGill-Hazelett caster system, in order to resolve this issue, the first series of support rolls for the belt are fitted with strong neodymium magnetic discs which hold the steel belt down. Similarly, the belt in contact with hot liquid metal can be pre-tensioned by using cold framing around the edges of the belt. This simple yet effective solution produces very good stability in the melt impingement region, and compensates for the initial high heat loads on the water-cooled belt.34)
For the TU Clausthal system, this control of bulging belts is resolved differently. The belt travels across a system of support rollers with staggered support points, while a pre-set negative pressure (~0.3 bar) below the belt, holds it down onto the rollers. In this manner, the inevitable thermal expansion of the belt is only manifested as a large number of smaller deflections, limiting vertical belt movement to below 0.1 mm. This has a negligible effect on the cast surfaces.33) Furthermore, a textured belt surface design is used (either longitudinal grooves or many small burls) to prevent large-scale bulge formation, as well as to reduce overall heat transfer rates. This, in turn, reduces stress formation and distortion in the solidified shell.35) The design of Clausthal’s roll support system is shown in Fig. 8.
Detailed schematic and picture of the belt support system for the Clausthal HSBC caster.35)
Another important feature of HSBC casters is the side containment systems. Lateral containment dams are sometimes required on the cooling belt for thicker sections (>7 mm), so as to prevent melts from flowing sideways off the belt.33)Figure 9 illustrates the moving side dam system implemented on the Clausthal HSBC caster. The system involves two laterally revolving chains of copper blocks, one on each side of the belt. The copper blocks have internal channels for water cooling and are linked via tubes made of high strength, high temperature resistant tubes screwed into the blocks. This brings the blocks into close contact prior to contacting the steel melt, but allows the system to bend in the horizontal direction. During operation, cooling water is supplied from, and removed through flexible tubes, and rotatable links. Rods and bearings are fixed onto each block so they travel in notches along the caster. The block-chains are placed over, and rotate about two horizontal tail discs, while the front tail discs are motorized to drive the block-chain.36) The components of this block-chain side dam system are shown in Fig. 9.
Pictures of the components of the moving side dam system at TU Clausthal.35)
The McGill-Hazelett side-dam system is much simpler. It involves two endless steel strap systems, travelling with the endless steel belt, in which refractory segments come into alignment prior to liquid steel contact, so as to contain any steel flows lateral to the casting direction. Roller guiding supports are installed to assure the stability of the side dams. The McGill-Hazelett moving side dam system is shown in Fig. 10. For the McGill experiments, casting 3–8 mm thick AA6111 and AA5000 series aluminum alloy strips, these side-dam systems were not needed, and were not used. Similarly for the 7 mm thick steel castings, no side dams proved necessary.
The McGill-Hazelett moving side dam system for lateral melt containment.
Steel melt solidification during primary cooling in the Clausthal system takes place under a shrouding protective atmosphere of argon, so as to prevent oxidation from compromising strip surface quality. However, under a pure argon atmosphere, cooling of the exposed, top melt surface is inefficient, due to the relatively low emissivity of the melt, which preserves its liquid metallic reflective shine. As a result, heat removal and solidification is predominantly single-sided, through the intensively water-cooled belt on the bottom side of the strip. Therefore, final solidification will take place close to the top surface of the melt, leaving a porous upper strip surface. As this is undesirable, this effect can be alleviated by adding several percent of carbon dioxide gas into the argon shroud gas. The CO2 gas acts to slightly decarburize the top layer of the melt film, raising its freezing temperature. The resultant supercooling of the liquid steel then leads to heterogeneous nucleation of a thin solid film of steel on the top surface of the melt. This solidified film increases the steel’s surface emissivity, leading to greater heat extraction from the top surface. In this way, solidification rates from the top surface downwards can be increased and the final solidification zone can be shifted down by as much as 40% of the strip thickness from the strip’s upper surface. Any potential micro-porosity is then readily sealed during subsequent in-line hot rolling.33)
The shrouding system used for TU Clausthal’s HSBC caster is designed in such a way that pure argon shrouding is first applied to the steel feed. Carbon dioxide gas is only blended in, after the top surface of the melt has become calmed. A design schematic of the TU Clausthal shrouding system and the effects of the solidification atmosphere on surface quality of as-cast steel strips, are shown in Fig. 11.
(top) Gas shrouding system for the TU Clausthal HSBC caster system and (bottom) the effects of shrouding atmosphere on the upper surface quality of as-cast steel strips.35)
As mentioned earlier, the surface texture of the mould can play a significant role on heat transfer and solidified strip quality. Various belt textures, and their effects on interfacial heat fluxes, have been investigated at the MMPC. They include a sandblasted substrate surface comprising many discrete, microscopic projections and valleys, as well as a variety of macroscopically textured substrates. Figure 12 shows the heights and density of the projections, which have a tremendous impact on interfacial heat fluxes. It has been shown that it is possible to achieve a 5–10 fold increase in interfacial heat fluxes by reducing the projection heights from 40 μm to 4.5 μm. By decreasing the roughness of the substrate surface, it is also possible to drastically reduce the effect of contact point density (the tips of these projections that come into direct contact with the solidifying melt) on heat fluxes,37) thus making heat fluxes more consistent over the entire strip.
In an industrial process, the substrate’s surface texture can be modified into one with random patterns of discrete projections by grit blasting. Based on the known relationship between a cooling substrate’s surface texture and interfacial heat fluxes mentioned earlier, it is possible to tailor an ideal casting roll surface texture that can give the desired heat flux that fits the process operating parameters, while simultaneously avoiding defect formation.
Finally, the speed of casting, rolling and coiling have to be all synchronized, if in-line rolling is to be implemented. At TU Clausthal, the as-cast strip thickness is determined by the melt flow. This is controlled by the ladle stopper rod and the moving speed of the cooling belt. At the start of casting, a “dummy-strip” is used to pull the cast into the in-line rolling stand, running at the same speed as the belt. A thermal sensor is used to detect the passing of the head end of the cast strip through the rolls and the gaps are then adjusted to desired levels. Speedometers are installed before and after the rolls in order to measure strip speed, and the rolling speed is then adjusted to match the belt’s moving speed, taking into consideration strip shrinkage during cooling.33)
The MMPC-HSBC caster features an advanced in-line process control system designed and implemented by Hazelett Strip Casting Corporation. Operator control, data entry and visualization of the control system are integrated into a touch-panel user interface. This control system is capable of simultaneously regulating all three major components of the caster: the delivery system, the cooling belt and the pinch roll drive, so as to synchronize casting, solidification and rolling. The flow rate of molten metal is regulated by controlling the speed of motion of the piston displacer, which has to be correlated with strip thickness, the speed of the belt, and the speed of the rolls of the Pinch Roll/Minimill, downstream. The system is also capable of controlling the tension exerted on the cooling belt, the cooling-water pump and the roll-gap distance.
The world’s first commercial-scale HSBC plant is being built by SMS Siemag for Salzgitter Flachstahl GmbH in Peine, Germany. It is scheduled for full commissioning in late 2012. This new plant will be mostly dedicated to the production of AHSS.38,39) The schematics of this new plant are shown in Fig. 13 and it incorporated many of the aforementioned HSBC process features and designs. The entire process line is 60 m long, 11 m of which is the caster. With an 80 t ladle capacity, the caster will be able to produce 1000 mm wide steel strips at up to 30 m/min.
Schematic of the commercial-scale HSBC caster at Peine, Germany, scheduled for commissioning in October, 2012.40)
Like many other continuous casting processes, computational fluid dynamic modeling has been used extensively in strip casting processes, for process simulation, process design, and process optimization.
For example, for the TRC process, the effect of different submerged slot nozzle designs for metal delivery on the fluid flow and solidification behavior of the melt were investigated at the MMPC in the late 1990’s. It was shown that the performance of vertical slot nozzles is far superior (mandatory) to bifurcated delivery nozzles like those used in conventional continuous casting operations. Similarly, wide, slender, vertical slot nozzles (slots as long as the roll width) yield a more uniform solidified shell thickness,41,42) but a much superior method was to deliver the steel across the meniscus to meet the rolls. These findings influenced subsequent development works on metal delivery to TRC machines. Figure 14 illustrates the predicted solid fraction profiles at mid-height of liquid steel in the sump. Clearly, the overall velocity fields generated by different slot nozzle designs greatly affect the uniformity of transverse solidified shell growth in a twin-roll caster system. As can be seen, vertical slot nozzles, especially those extending the full width of the rolls, produce good, uniform, solidification profiles. However, an even better way is to deliver the incoming melt, evenly, across the meniscus, towards the rotating roll surfaces. This eliminates shell erosion deep in the sump, maximizes the thickness and uniformity of the shells being formed, and improves surface quality.43)
Predicted solid fraction profiles (at a horizontal plane 100 mm below melt surface) and fluid velocity fields generated by different slot nozzle designs, a) tubular nozzle with double horizontal ports, b) vertical, tubular nozzle, c) slot nozzle with length = 50% of roll length and d) extended vertical slot nozzle with length = 100% of roll length.41,43)
Another aspect of the TRC process that was investigated using mathematical and physical models is the effect of casting speed on interfacial heat fluxes. Heat fluxes are a very important attribute for any casting process, since they directly affect the degree of solidification, the production capacity of the process and the microstructure of the cast. As such, these fluxes need to be carefully controlled. Work performed by Tavares, Isac and Guthrie had shown that casting conditions of a TRC process has a significant influence on the solidification and as-cast microstructure of the strip,44) as shown in Fig. 15.
Two micrographs of the internal transverse microstructure of carbon steel strips produced by TRC under difference casting conditions. 6 mm thick strip cast at 4 m/min (top) and 3.5 mm thick strip cast at 7.5 m/min (bottom).43)
Figure 15 shows the solidified dendritic structure in two different as-cast TRC strips produced under different casting conditions (“slow” and “fast” casting speeds), as revealed using Oberhoffer’s reagent. This reagent is capable of showing the degree of micro-segregation, and in this context, it is able to distinguish the iron-rich dendritic region (as the black region) and the impurity-rich inter-dendritic spaces (as the white region).44)
As shown in the figure, the slower cast, thicker strip, exhibits a columnar dendritic structure, with the dendrites meeting in the centerline of the strip and the absence of an equiaxed region. However, as marked by the arrows, various broken and bent dendrites were observed, showing signs of interaction between columnar dendrites growing from the two sides. Moreover, the arrow on the right side of the micrograph highlights a columnar dendrite able to grow across the centerline, since it was not blocked by other dendrites growing inwards from the opposite surface. This usually occurs due to relatively large inter-dendritic spaces commonly associated with low cooling rates, which was exactly the case for the slower casting speed.44)
The second micrograph in Fig. 15 was obtained from a thinner as-cast strip at higher casting speed. Because of the higher heat fluxes and higher cooling rate, the primary dendritic spacing is much finer. Furthermore, the columnar dendrites do not reach the centerline, and an equiaxed zone can be seen in the middle of the strip.44)
Earlier studies have observed the presence of an equiaxed zone in the center of some twin-roll cast steel strips, and have attributed it to the incomplete solidification of the strip at the roll gap.45,46) Tavares, Isac and Guthrie proposed a different mechanism for the formation of the central equiaxed region in twin-roll cast strips – one of melt-recirculation and deformation in the semi-solid state. Since Oberhoffer’s reagent makes impurity-rich regions appear white, a darker colored equiaxed zone (in Fig. 15, bottom) means that the non-columnar zone is depleted of impurities. The interdendritic liquid, rich in impurities, was believed to have been ‘squeezed’ back up into the sump, by the action of the rolls during the semi-solid state reduction, resulting in the observed impurity distribution. The double heat flux peaks obtained at higher speeds and cooling rates, suggest that the equiaxed zone was formed owing to a certain degree of hot-deformation of the strip before it left the rolls. These equiaxed grains are formed as a result of heterogeneous nucleation on broken off ‘tips’ of the columnar dendrites.44)
Further physical modeling work has also shown that the prior austenite grains, observed in twin-roll cast strips, display much larger average grain sizes (~400 μm vs. 20–50 μm) than conventionally processed strips of the same thickness. Thicker strips (produced at lower casting speeds) forms larger austenite grains, since the lower heat fluxes allow the strip to remain at higher temperature for a longer period of time, favoring the growth of γ grains.44)
Experiments performed at IMI, Boucherville under Project Bessemer, suggested that both the magnitude and trend in interfacial heat fluxes are heavily influenced by the casting speed, as shown in Fig. 16. Heat fluxes were determined from temperature-time data measured by thermocouples embedded in the rolls, using the Inverse Heat Conduction Problem (IHCP) method.47) At slow casting speed (~4 m/min), there was only a single peak in the heat flux vs. contact time plot, whereas at higher casting speeds (6–7.5 m/min), two peaks could be identified, both of which were significantly greater than the peak from the slower casting speed. This was explained by the mechanism proposed by Guthrie, Isac, Kim and Tavares, which took into account various phenomena such as thermal resistance from an air-film entrapped at the melt/mould interface, the presence of a thin insulating coating on the rolls causing poor initial melt/mould contact, thermal expansion of casting rolls, increasing metallostatic pressure further down the sump, diminishing air-film thickness due to increasing contact pressure, and the effect of rolling and solid shell shrinkage47) (See Fig. 17).
Variation of interfacial heat fluxes against contact time, for slow speed casting, thicker strip (4 m/min and 6–7.5 mm, top) and fast casting, thinner strip (7–8 m/min and 4 mm, bottom).47)
Mechanisms proposed by Guthrie, Isac, Kim and Tavares, to explain variations in interfacial heat fluxes at different casting speeds and strip thicknesses.47)
From the interfacial heat fluxes shown earlier, it can be seen that the maximum interfacial heat flux did not appear at the first moment of contact between the melt and the rolls, when the temperature gradient is the greatest. This would seem counter-intuitive. However, the copper rolls were coated with a thin protective layer of unspecified chemistry, thermal properties and thickness.
As mentioned earlier, the inherent high cooling rate of the TRC process means that surface crack initiation is easier, due to stronger stresses during casting. The residual stress field right after solidification is difficult to be measured experimentally, since cracking releases these surface stresses.22) Numerical simulation of the stress field for a SS304 austenitic stainless steel during TRC casting has revealed that a) tensile stress on the strip surface is at a maximum at a position 5–10° above the roll nip, b) the stress is tensile on the strip surface and compressive in the strip centerline, and c) the surface tensile stress on the strip increases with larger roll diameter and strip thickness.22) This numerical modeling work provides a theoretical groundwork on which solutions to the surface crack problem in actual production can be developed.
Another potential issue for the TRC process is the thermal deformation of the rolls (thermal crown).48) This can lead to uneven strip profiles which can cause enormous problems during subsequent rolling steps. Finite element analysis, incorporating shrink-fit effect and plastic deformation, has shown that the thermal crown for a typical TRC roll is ‘M’ shaped.48) It has also shown that the size of the thermal crown increases with increasing interfacial heat fluxes, casting speed, sleeve thickness and roll width and decreases with increasing roll diameter.48)
One feature that exists in all strip casting operations is the presence of a triple point – the point where a three-phase interface forms as the liquid melt comes into contact with the solid cooling substrate in a gaseous shrouding environment. This gas can be adjacent, or not, to the refractory wall. As mentioned earlier, the successful control of this meniscus is absolutely critical to the successful operation of the strip casting process in terms of surface quality of the strip.49) Numerical modeling has been heavily used, coupled with experiments, to understand the correlation between different operating parameters and the shape and stability of the meniscus.47,50) Figure 18 shows the result from one such study carried out for the MMPC-HSBC delivery system, where the effects of a back-wall gap of 0.9 mm, and various casting speeds, on meniscus shape was investigated.
The comparison between predicted (top) and experimental (bottom) fluid flow and meniscus behaviour in a HSBC process, adapted from.30)
It is now understood that entrapped air within the crevices on the surface of a cooling substrate, when coming into contact with a liquid metal, will play a significant role in determining interfacial heat fluxes between the melt and a substrate. Recent research work has revealed a mechanism, as shown in Fig. 19, for the formation of air pockets that can explain the “dimpled” bottom surface characteristics of HSBC as-cast strip.37) The understanding of this phenomenon has allowed for the accurate prediction of instantaneous interfacial heat fluxes from first principles, and this has been verified with experimental data.
Proposed mechanism for air pocket formation, a) expansion of entrapped air, b) melt loses contact with substrate at weak points and c) air pocket formation and growth.37)
Thus, a considerable amount of numerical modeling work has been done at the MMPC for the HSBC process. In addition to aspects similar to those investigated for the TRC process, such as fluid flow, triple point meniscus, nozzle design and applications of HSBC to bulk amorphous alloy production,41,51) substrate surface texture is also an important field where modeling has been used extensively in the development of the HSBC process.
Understanding the mechanism of air-pocket formation and growth was also deduced from the numerical modeling of melt solidification on a copper substrate. In the work by Guthrie, Isac and Li,49) a 3D multi-phase ab-initio thermal model of the HSBC process, taking into account the effect of air-gap evolution and substrate surface texture, was developed. The model was able to accurately predict transient melt/substrate interfacial heat fluxes, as well as the corresponding melt solidification behavior. Some details of the said model are shown in Fig. 20.
Details of the 3D ab-initio HSBC model developed by MMPC, showing the 3D mesh design of the substrate (top left), the predicted vs. measured interfacial heat flux (top right) and the predicted transient melt solidification behavior (bottom).49)
Recent studies on the effect of belt surface characteristics on initial heat fluxes during HSBC operations, offer opportunities for process optimization. It was found that the density and heights of the microscopic pyramidal projections generated by sandblasting the belt have a strong impact on the initial heat fluxes to the cooling substrate. HSBC simulator experiments, using a sandblasted copper mold and AA6111 aluminum alloy melts, have shown that initial heat fluxes can be increased from ~1.7–3.5 MW/m2 to ~12.8–13.3 MW/m2, by reducing the average pyramidal projection height from 40 μm to 4.5 μm. Furthermore, for 40 μm projection height, it is possible to increase initial heat fluxes from 1.7 to 3.5 MW/m2 by increasing contact point density from 100 to 2500 points/mm2, thereby doubling initial heat fluxes.37)
Consequently, it should be possible to create cooling belts of different textures with different cooling capabilities, optimized for the production of specific grades of steel strips, requiring specific microstructures.
Various types of macroscopically textured and treated substrates have also been investigated on the HSBC simulator at the MMPC, as shown in Fig. 21. Through experimentation and mathematical modeling, two textures, d) and e), for the casting of aluminum AA6111 alloy were identified, which produce casts with improved (reduced) grain sizes. These results are shown in Fig. 22.52) Similar belt texture optimization for HSBC casting of steel appears to be very practical, and further research is required.
3D models (top) and actual textures (bottom) of various macroscopic substrates textured investigated at the MMPC.52)
Effect of substrate texture on grain size of HSBC cast AA6111 aluminum strips.52)
Finally, the use of HSBC strip casting processes for the production of bulk amorphous alloy strips has also been developed with the help of mathematical and physical modeling as well. These numerical models are useful in predicting the interfacial heat fluxes and cooling rates of strip casting processes, which are crucial in the continuous production of bulk amorphous alloys. The effects of various operating parameters such as casting speed, superheat, gas shrouding, etc. can be evaluated using well-constructed numerical models.51,53)
One of the most prominent characteristics of strip casting technology that separates it from conventional and thin-slab continuous casting operations is the drastic simplification of processing operations, and downstream stages, owing to the significant reduction in as-cast thickness. Typical process characteristics of CCC, TSC, TRC and HSBC are shown in Table 1. This section will cover the most significant advantages that strip casting technologies can offer the steel industry.
| Process variable | CCC | TSC | TRC | HSBC |
|---|---|---|---|---|
| Product thickness (mm) | 150 – 300 | 20 – 60 | 0.7 – 5 | 5 – 20 |
| Total solidification time (s) | 600 – 1100 | 40 – 60 | 0.15 – <1.0 | ~6 |
| Casting speed (m/min) | 1.0 – 2.5 | 4 – 6 | 30 – 150 | 12 – 60 |
| Avg. mould heat flux (MW/m2) | 1 – 3 | 2 – 3 | 6 – 15 | ~ 11 – 13* |
| Weight of melt in caster (kg) | > 5000 | ~ 900 | < 400 | ~ 120 |
| Avg. shell cooling rate (°C/s) | ~ 12 | ~ 50 | ~ 1700 | 400 – 500 |
| Scale loss (kg/m2) | – | 7.8 | < 0.2 | – |
| Plant capacity (Mt/yr) | 4 – 10 | 2 | 0.4 – 0.6 | Up to 3 |
The production costs of steel strips rises sharply with increasing initial as-cast thickness and with thinner final product thicknesses. In conventional processes (CCC and TSC) using oscillating moulds, the dramatic increase in production costs with decreasing product thickness is primarily attributed to the intense rolling and reheating required.57) However, it is difficult to cast thinner steel slabs using oscillating moulds, since thinner strips solidify faster and require proportionately faster casting speeds. Faster casting speeds, in turn, result in greater friction between the mould and the cast, which ultimately causes surface defects such as cracks.58,59)
So as to reduce mould friction in conventional continuous casting, mould oscillation and slag lubrication at the mould-slab interfaces are used. But these techniques cannot be applied to the casting of thin strips, because mould oscillations would leave marks on the cast surface that would induce defects with sizes similar to the strip thicknesses being produced at high casting speeds.
The defining feature of strip casting technology is that it uses moving moulds. The moulds (rolls or belts) travel at the same velocity as the forming strip, so that mould-slab friction is eliminated. This allows for the production of thin strips directly from casts at high speed, with good surface quality, while reducing operating costs and uncertainties by eliminating the use of mold fluxes.
5.1. Economic BenefitsStrip casting enables the direct production of thin strips from the melt and eliminates most of the subsequent stages of hot-rolling required in conventional continuous casting processes.48) Only a few rolling steps, a finishing step and conventional coiling, are required. Figure 23 provides a comparison of the typical equipment and processing stages involved in various continuous casting processes. This shortening of the process route can reduce investment costs by a factor of 4 to 10 times. The specific investment costs can be reduced by as much as 40% per ton of steel.57) The Eurostrip caster reported up to a 68% reduction in capital costs, and a 40% reduction in space usage, compared to the TSC process.7) Of the strip casting technologies, the HSBC process offers the lowest specific capital cost, due to its five (or many more) fold higher production capacity compared to TRC.60) Figure 24 shows the approximate investment costs for a continuous casting plant with different processing routes.
Typical process layout and equipment for various forms of continuous casting.61)
Comparison of typical capital costs for various casting process routes.3)
The operating costs of strip casters to produce steel strips have been found to be relatively independent of the thickness of the final product – in sharp contrast to conventional casting technology, in which the operating costs rise drastically as final gauge thickness decreases.62) However, an accurate calculation of operating cost is difficult, since uncertainties still exist in the life of key components of the caster (such as rolls/belts, side-dams, refractories, etc.).
Studies have estimated that in TRC’s, refractories have to be changed no less than every three consecutive 110-ton casts. The time required for changing the rolls is estimated at no longer than 10 minutes.
Unlike CCC and TSC processes, in which slow cooling results in coarser grains, and multiple solidification zones (chill, columnar and central-equiaxed), together with segregation of impurities, Strip Casting, thanks to its rapid solidification rates and high cooling speeds, is capable of producing strips with much lower levels of macro-segregation. In the TSC of carbon steel, the steel scrap used must not contain more than 0.15 wt% Cu and 0.015 wt% Sn, so as to avoid the formation of surface cracks. This threshold can be increased by 3.5 and 10 times respectively to 0.55 wt% Cu and 0.16 wt% Sn, in the case of TRC strip casting.57) The HSBC process is also likely to have similar effects in improving scrap recyclability and higher impurity tolerance.
5.2. New Steel GradesAdditional economic benefits can be obtained from the production of special metallic alloys now made possible by strip casting. In certain forms of strip casting, such as the single-roll melt-quenching or melt spinning process, with very high cooling rates (104–107°C/s), it is quite possible to produce ultra-thin steel strips with far-from-equilibrium crystalline, or even amorphous, structures. This allows for the production of special alloys, metallic glasses or bulk amorphous sheet material.3,63)
The extended recyclability of steel scraps offers new opportunities of alloying in carbon steels. Studies have shown that HSBC is capable of alloying copper into steel as an efficient corrosion-inhibiting element, without incurring segregation and affecting strip quality. By alloying medium carbon steel produced by strip casting with 1 wt% Cu, the corrosion rate can be reduced to 10% of normal in a 3.5% NaCl solution.33)
High manganese-containing Advanced High Strength Steels (AHSS, sometimes called High Strength Ductility, or HSD steels), such as the TRansformation-Induced Plasticity (TRIP) and the TWinning-Induced Plasticity (TWIP) steels, are ideal for manufacturing some automotive components, due to their high strength and high ductility. These parts cannot be produced by CCC or TSC on a commercial scale for two reasons; high hot-cracking susceptibility means that these steel casts cannot be bent after being cast from vertical continuous casters, and high strength means that they work harden very rapidly, making them very difficult, and very costly, to be reduced to strips for commercial applications.
The remarkable mechanical properties of AHSS/HSD steels (such as TWIP and TRIP steels) as compared to conventional steels and stainless steels, are shown in Fig. 25 below.
Mechanical properties of AHSS (TRIP & TWIP) steels compared to conventional and stainless steels, adapted from.35)
The HSBC strip casting technology offers a new, cheap and efficient alternative processing route for the production of AHSS. Research has shown that AHSS strips produced by pilot-scale HSBC casters exhibit satisfactory strip qualities, with room for further improvements.33)
5.3. Energy and Environmental BenefitsBesides considerable economic benefits, strip casting also offers steelmaking companies with a significant savings in energy/ton, with reduced emissions and environmental impact. It is important to note that although energy saving capacity of strip casting is significant, its contribution to overall cost savings is modest, since energy costs are only a small part of the operating cost. It was estimated that energy cost savings is one order of magnitude lower than the total cost saving from strip casting.2) Studies have estimated that twin-roll strip casting is capable of reducing energy consumption by up to 90% and greenhouse gas emission by 80% as compared to CCC.3) The industrial-scale CASTRIP operation in Crawfordsville, Indiana, reports that the overall energy consumption, including the energy requirement for a 50% size reduction in hot-rolling reduction, of the TRC process were 81–89% lower, compared to conventionally-produced steel strips. Greenhouse gas emissions were also reduced by 71–80%.64,65)
As for the HSBC process, due to the need for a greater degree of in-line hot rolling, energy savings achievable are somewhat lower, at approximately 75% primary energy consumption reduction of the CCC process. However, since the HSBC process is yet to be industrialized,38,39,40) it is reasonable to believe its energy consumption and emission levels will both decrease further, to a level comparable to current TRC technologies, through process optimization in more advanced stages of its development.
Emissions reduction of an industrialized HSBC plant is expected to be ~60% over CCC process, corresponding to 170 kg CO2 per ton of hot strip.38) The comparative energy savings and greenhouse gas emissions of HSBC and TRC process against TSC and CCC processes are shown in Fig. 26.
Higher tolerances of tramp element will generate environmental benefits in addition to financial savings. Recycled steel from electric arc furnaces (EAF) can be used instead of virgin steel made via the blast furnace-basic oxygen furnace (BOF) route. Production of EAF steels consumes over 60% less primary energy than BOF steels (7.5 GJ versus 19.2 GJ per ton).33)
Furthermore, as-cast products from strip casters have a much higher surface-area-to-volume ratio due to their small as-cast thicknesses. This allows for more efficient heat transfer to, or from, the environment, which, in turn, enables faster cooling/heating. This permits the casts to be re-heated, if required, in downstream processing, to the target temperatures rapidly, reducing energy consumption and processing times, and lowering steel losses due to scale formation.57)
Although the current environmental benefits and energy saving of strip casting technology do not convert proportionately into monetary benefits, with rising oil prices, extensive implementation of carbon taxes worldwide and cap-and-trade regulations, we believe the cost-saving potentials of strip casting will undoubtedly improve further. As an example, in Canada, the carbon dioxide emission tax charged in British Columbia on industries was increased to 30$ per ton CO2 in July 2012. The emission reduction capability of strip casting (calculation using data from CASTRIP, 0.16 ton CO2 per ton steel), as compared to the CCC process, would translate into approximately 5 US$ of cost-saving per ton of steel, from the provincial carbon tax alone.66)
Thanks to the intensive research and development efforts undertaken throughout the world since the 1980’s, strip casting technologies for steel have finally come of age. Thus, the commercialization of TRC concept developed by Bessemer in 1856 has finally been successfully realized. Similarly, commercialization of the much more recent HSBC concept is expected very soon. These processes may well represent a step change for the steel industry.
The HSBC process presents many advantages over the TRC process for integrated steelworks, such as lower operating costs, higher productivity, and a simpler operation. Similarly, strip casting processes have the potential for producing advanced steel grades not easily possible via conventional hot rolling mill processes.
In the light of these findings, it appears highly plausible that strip casting technologies will be extensively implemented globally in the coming decades, revolutionizing the way steel strips are made, and delivering savings for the global steel industry no less significant than the inventions CCC and TSC technologies decades ago.
