Showing papers in "Steel Research in 1996"

Metallurgy and creep properties of new 9-12%Cr steels

Abstract: Current understanding of the creep strengthening mechanisms in newly developed 9-12%Cr creep resistant steels is discussed. It is demonstrated how new computer programs for calculation of equilibrium phases in the steels can contribute to increased understanding of the microstructure of these steels. The calculations are useful for, e.g. estimation of the risk of δ-ferrite formation, and of balancing of the addition of V, Nb and N so M2X formation can be avoided. Combination of the equilibrium calculations with kinetic models and experiment has led to a description of the precipitation of Laves phase in W-alloyed 9%Cr steels. Combination of these results with creep tests indicated that Laves phases have a creep strengthening effect through precipitation hardening, and that any solid solution strengthening effect of W is small.

Modelling of austenite stability in low-alloy triple-phase steels

Abstract: A model for the stability of dispersed austenite in low alloy triple-phase steels has been developed. The model was based on the dislocation dissociation model for classical heterogeneous martensitic nucleation by considering stress effects on the nucleation site potency distribution. The driving force for martensitic transformation has been calculated with the aid of computational thermodynamics. The model allows for the effects of chemical composition of austenite, mean austenite particle size, yield strength of the steel and stress state on austenite stability. Chemical enrichment in C and Mn, as well as size refinement of the austenite particles lead to stabilization. On the contrary, the increase in the yield strength of the steel and triaxiality of the stress state lead to destabilization. The model can be used to determine the microstructural characteristics of the austenite dispersion, i.e. chemical composition and size, for optimum transformation plasticity interactions at the particular stress state of interest and can then be useful in the design of low-alloy triple-phase steels.

Determination of stacking fault energy of austenite in a duplex stainless steel

Abstract: A determination of stacking fault energy (SFE) of the austenite phase of a duplex stainless steel, material no. 1.4462, has been carried out using transmission electron microscopy (TEM). Furthermore, cold rolling tests and microstructural analysis have been realized in order to allow a detailed discussion of the obtained SFE-values. The results of this investigation indicate that the stacking fault energy of the austenite phase within the duplex stainless steel is lower than those of single-phase austenitic stainless steels. This is justified by the chemical composition; mainly by the Cr and Ni alloying contents. Nevertheless, work hardening of the austenite during cold deformation is not as accentuated as expected by the low SFE-values, because at higher deformation levels the deformation mainly occurs within the ferrite phase.

Transformation behaviour of the high temperature martensitic steels with 8 – 14% chromium

Abstract: Comprehensive development work on martensitic steels belonging to the so-called 12% Cr steel group were performed at the Institute for Materials Research (IMF) of Forschungszentrum Karlsruhe on martensitic steels, the so-called 12% Cr steel group, in order to meet the various requirements in nuclear and conventional energy technology. The transformation characteristics of 29 different grades of steel and 38 heats have been determined and continuous cooling transformation (CCT) diagrams have been prepared. The diagrams are first described by groups of subjects in a chronological order because the change in the chemical composition cannot be correlated in all cases with the change in transformation behaviour. The quenching hardness can be satisfactorily described as a function of the C+N content if, taking into account the Nb, Zr, Ta, Hf fractions, a common effectiveness factor is calculated. This effectiveness factor is also integrated in the calculation of the Ms point by modifying accordingly the equation proposed by Steven and Haynes for low-alloy steels and supplemented by the summands for V and W. An equation is introduced for the calculation of the critical cooling rate for pearlite transformation which takes into account the special influence exerted by elements Cr, V, Mo, W, Ge. The comparison between calculated and measured values for Ms and Vcrit is satisfactory, except for some steels.

Dust formation in a BOF converter

Abstract: Minimisation of waste and manipulation of residues to suit succeeding processes today is one of the steel works most important research areas. BOF dust is expensive to depose in landfills and difficult to recycle due to its zinc contents. The simultaneous loss of raw materials and cost increase makes it important to minimise the generation of dust. This can be done by employing better process control systems with regard to the mechanisms governing the generation. Four dust forming mechanisms can be considered in a converter: ejection of metal, ejection of slag, entrainment of charged material and vaporisation. In order to determine the importance of the four dust forming mechanisms and to characterise dust, with special emphasis to zinc, the off-gas from a 100-t BOF was sampled during twenty seven heats. The sampling equipment allows the sampling of both solid dust particles and vaporised elements. Samples from the first and the second part of blowing show significant differences both in chemical composition and in the origin of formation. The major dust formation mechanism is the ejection of metal and slag, respectively. Entrainment of solids plays an important role only during the first part of blowing. Vaporisation of elements from the bath is most important during the end of blowing. The formation of dust is influenced by process operation control, especially lance position, silica contents and time of charging of slag formers. Zinc is mostly found on the rim of ejected particles where a gradual transition of zinc oxide to zinc ferrite to iron oxide is found. At sampling temperatures of between 800 and 1000°C most of the zinc had already condensed.

Artificial neural nets for prediction of silicon content of blast furnace hot metal

Abstract: Conventional models for prediction of silicon content of blast furnace hot metal are briefly reviewed. Four different artificial neural net (ANN) models, namely, back propagation algorithm (BPA), dynamic learning rate algorithm, functional link network (FLN) and fuzzy neural network (FNN), are trained and tested on operational data from blast furnace (BF1) at Visakhapatnam Steel Plant. FNN can predict silicon mass content of hot metal with a standard error (actual versus predicted) of 0.09% and correlation coefficient of 0.86; standard back propagation predicts with a standard error of 0.08 % and correlation coefficient of 0.79.

Finite element simulation of quench hardening

Abstract: In this study, an efficient finite element model for predicting the temperature field, volume fraction of phases and the evolution of internal stresses up to the residual stress states during quenching of axisymmetrical steel components is developed and implemented. The temperature distribution is determined by considering heat losses to the quenching medium as well as latent heat due to phase transformations. Phase transformations are modelled by discretizing the cooling cuves in a succession of isothermal steps and using the ITdiagrams. For diffusional transformations both Scheil's additivity method and Johnson-Mehl-Avrami equation are used, while KoistinenMarburger equation is employed for martensitic transformation. Internal stresses are determined by a small strain elasto-plastic analysis using Prandtl-Reuss constitutive equations. Considering long cylinders, a generalized plane strain condition is assumed. The computational model is verified by several experimental measurements and by comparison with other known numerical results. Case studies are performed with St50, Ck45 and C60 type of solid and hollow steel components. The complete data and result sets provided for the verification examples establish a basis for benchmark problems in this field.

Evaluation of structural stability and creep resistance of 9-12% Cr steels

Abstract: During creep exposure of modified chromium steels lowering of solid solution strengthening due to precipitation of Laves phase as well as coarsening of all precipitates causes degradation of creep resistance. Two distinct domains of the stress dependence of creep rate and time to rupture have been observed in precipitation strengthened modified chromium steels. The stress characterizing the transition between these domains was found to be closely related to the Orowan stress. This stress consists in these steels of the contribution from large particles on subgrain boundaries (mainly M 23 C 6 and during the limited time also Laves phase) and from small precipitates (Nb(C,N) and VN) inside subgrains. This has to be considered when measuring the interparticle spacing and calculating Orowan stress. Larson-Miller parametric equation is used to elucidate the necessity of long-term creep testing. By means of two heats of CrMoVNbN steel it is shown that reliable extrapolation of creep properties is possible only in a stress and temperature domain in which only one creep creep rupture mechanism operate. In the high stress domain Larson-Miller constant C LM is well above 30 while in the low stress domain this constant does not exceed 25. When the extrapolation is based mainly on short-term creep tests, the C LM constant is close to that valid in the high stress domain and therefore it overestimates long-term creep strength.

A comparative study of artificial neural networks for the prediction of constitutive behaviour of HSLA and carbon steels

Abstract: Backpropagation neural networks are utilized to store and predict the flow stresses of several steels. A convergence algorithm using a varying learning factor is developed which is shown to save one sixth of the learning time when compared with the algorithm in which a constant learning factor is utilized. A performance test shows that the well-trained neural network can interpolate flow stresses very well if the information for interpolation is sufficient in the training pairs. The capability of the network to extrapolate is found not to be impressive. The neural network can handle several groups of data during adaptive learning simultaneously without losing accuracy. The time needed for adaptive learning to reach a reasonable level of accuracy is short. Comparing the predicted results to other models, the output of neural network is shown to have the highest accuracy.

Simulation of diffusion controlled phase transformations in steels

Abstract: Phase transformations in steels are currently treated with empirical approaches which make use of the vast amount of existing knowledge and experience. The aim of this paper is to show that more fundamental approaches are complementary, and mandatory in order to arrive at predictive capabilities. Such approaches have to take into account the local conditions on a microscopic level such as inhomogeneities, element distributions, driving forces, thermodynamic properties of the phases. With this sophistication it becomes possible to treat complex situations such as the long term stability of microstructures under service conditions, giving room for new optimizations.

Effect of boundary conditions and workpiece geometry on residual stresses and microstructure in quenching process

Abstract: In this study, the internal and residual stress states in quenched C60 steel cylinders are analyzed both numerically and experimentally in order to investigate the effects of boundary conditions (such as quench severity and temperature of quench bath) and specimen geometry. Specimen geometry has been analyzed by introducing a hole in a cylinder and varying hole diameter and its eccentricity. In the numerical analysis, the finite element method is applied and both temperature gradients and phase transformations are considered. Experiments include microstructural examination and X-ray measurements of residual stresses of the first kind. It has been found that the value of the convective heat transfer coefficient is very critical to obtain simulation results close to real ones. For instance, when a constant value obtained as the mean of a temperature dependent distribution is used for this parameter, residual stresses are seriously underestimated (up to 40%). The temperature of the quench bath affects directly the convective heat transfer coefficient. The lower the bath temperature, the higher are the resulting residual stresses. Under the same quenching conditions, if the diameter of the hole is greater than a critical value, a transition occurs from the shallow hardening case to the through hardening case, i.e., the residual stress distribution is reversed. On the other hand, for a constant hole diameter, if the eccentricity ratio reaches a critial value, a complex residual stress state results, i.e., compressive/tensile stress transition regions along the circumference are observed.

Influence of gases dissolved in cooling water on heat transfer during stable film boiling

Abstract: Results of investigations concerning the determination of the influence of the water quality on the heat transfer of quenched hot metal surfaces are presented. With an immersion quenching device, tests have been carried out for various cooling water qualities. Apparently there was no correlation between the water quality and the cooling process (Leidenfrost temperature). High concentrations of dissolved gases yield stable vapour films (low Leidenfrost temperatures are measured). Salt displaces the dissolved gases from the water; the Leidenfrost temperature rises: the higher the salt concentration, the lower the gas concentration and, therefore, the higher the Leidenfrost temperature. If salt concentrations are high or the cooling water is degassed, the vapour film falls apart just after immersion. An extended model considering the influence of the gases allows the correlation of the water quality/salt concentration and the heat transfer/Leidenfrost temperature.

Precipitation behaviour of 25% Cr - 5.5% Ni austenitic stainless steel containing 0.87% nitrogen

Abstract: In this paper, a 25Cr - 5.5Ni steel (mass contents in %), which originally had a ferritic-austenitic duplex microstructure (DIN material no. 1.4460), and to which mass contents of 0.87% nitrogen were added to become fully austenitic was studied. The phase transformations that occurred between 700 and 1350° C. were studied using several complementary techniques. Firstly, various treatments were carried out in the range 1050-1350° C for 1h to select the best temperature for solution heat treatment. Chromium nitride dissolution and absence of ferrite were was observed at 1250° C for 1h. Subsequently, the specimens were aged in the range 700-1100°C, for 90s up to 100h. The first phase to precipitate was chromium nitride. Ferrite precipitation occurred after long aging times, followed by sigma phase precipitation. A TTT diagram has been proposed.

Creep rupture strength of tungsten-alloyed 9-12% Cr steels for piping in power plants

Abstract: The creep rupture strengths of the three tungsten-alloyed 9-12% Cr-steels E 911, NF 616 and HCM 12A are discussed. It is shown that Larson-Miller assessments may lead to an overestimation of the long-term creep rupture strength because of microstructural changes. The presented assessments indicate that all three alloys will have similar creep rupture strengths in the range of 110-120 MPa for 10 5 h and 600 °C. To obtain reliable estimates for creep rupture strength, however, further long-term tests are needed. In any case, the tungsten-alloyed steels will enable the main steam temperature to be raised by about 20 °C over that possible with P 91.

Carbon or nitrogen alloyed quenched and tempered stainless steels : a comparative study

Abstract: Some conventional stainless quenched and tempered steel grades were modified by substituting nitrogen for carbon and by variing the contents of chromium, molybdenum and nickel. Results of a comparative study of carbon and nitrogen steels are presented in a wide range of tempering temperatures. In nickelmartensitic steels nickel can be partially substituted by nitrogen without harming the properties. Due to their finer microstructure nitrogen steels with 15-17 %Cr, 2 %Ni and 0.2 %N, at the same strength level, offer higher toughness and better corrosion resistance than the respective conventional steel X20 CrNi 17 2.

Cold rolling texture development of α/γ duplex stainless steels

Abstract: The cold rolling texture development of two α/γ duplex stainless steels (DSS) with similar volume fractions of both phases but with totally different microstructures were investigated. Due to the limited number of available pole figures using X-rays, for the calculation of the ODFs both a direct method and a recent iterative series expansion method were used. The results were checked by neutron diffraction measurements. The austenitic phases of both DSS behave similarly to single phase materials with a low stacking fault energy which develop a brass-type rolling texture. In contrast, the texture development of the ferritic phases strongly differs from those of single phase ferrites. Instead of a fibre type texture the α-phase in both DSS exhibits a peak dominated texture regardless of whether it is the matrix phase or not. These differences, as well as the sharpness of both phases, are explained by the presence of the second phase.

Numerical results achieved with an inverse computational method for determining the constitutive parameters usina the hot torsion test results

Abstract: It was demonstrated in a previous paper that to obtain a unique set of parameters for a constitutive relationship from the hot torsion test, two separate objective or error functions must be considered. They are based on the load-displacement and load-displacement rate data respectively. The algorithm described earlier, has been implemented in a rigid viscoplastic FE code. The best set of the constitutive parameters that minimizes the objective functions is determined by using an optimization scheme. To demonstrate the predictive capabilities of the analytical approximation described in the previous part and to simulate the problem of identification of material parameters, a set of pseudo-experimental load-displacement/displacement rate data has been used for calculating the objective functions. The results confirm the inadequacy of using only the load-displacement results for the identification of the constitutive parameters from the hot torsion results. This is in agreement with the theoretical results in the previous paper. It is shown subsequently that the newly proposed inverse computational method can be used successfully to determine a unique set of the constitutive parameters.

Characterization of medium- and low-temperature carbides in a low-carbon steel by internal friction

Abstract: The formation of carbides during the aging of a quenched steel with mass contents of 0.04% carbon was investigated by means of internal friction by measuring the carbon Snoek peak height. The aging was carried out at different temperatures ranging from 40 to 350 °C so that different types of carbides were formed. Low-temperature carbides develop after aging below 100 °C. The Johnson-Mehl equation is applied to evaluate the kinetics of precipitation and re-dissolution of the different carbides. The solubility of carbon in ferrite is influenced by the pre-existing carbides, which is, in part, the reason for the variation of thermodynamic data in previous work.

Influence of roll geometry and strip width on flattening in flat rolling

Abstract: The size of local roll flattening and its distribution along the direction of the roll axis in flat rolling were calculated by means of 3-dimensional finite element method. For analysis of elastic flattening deformation of the roll stack the well-known classical and analytical solutions are usually employed which were derived from elastic half space theory or two dimensional contact theory. By comparison of results from both the different methods the validity of the classical formulae was examined. Some of the formulae are more appropriate for calculating local flattening along the work roll/strip interface, however, they may result in a great deviation in calculating the flattening along the work roll/ back-up roll interface, especially for the back-up roll. Using the Fe model the influence of roll geometry and strip width under a specific rolling force on the flattening was taken into account, which is difficult to be treated with classical models.

Effects of tramp elements Cu, P, Pb, Sb and Sn on the kinetics of carburization of case hardening steels

Abstract: Investigations on the effect of tramp elements and impurities on the gas carburization of case hardening steels in CO-H 2 -H 2 O gas mixtures at 930 °C in a flow apparatus by thermogravimetric experiments were carried out. The influence of surface segregation on the kinetics of the surface reaction and the influence of decreasing carbon transfer coefficient β on the carbon concentration profile were estimated. Numerical calculations were developed to determine the carbon transfer coefficient β.

An inverse computational method for determining the constitutive parameters using the hot torsion test results

Abstract: Recently a number of inverse computational methods were presented for determining constitutive parameters from the results of mechanical tests. These are based on minimization of the least square errors of the measured and calculated load-displacement results of the given test. Uniqueness of the minimum obtained from these methods has not been investigated properly. This paper presents an approximate solution for the minimum error in the inverse computational determination of constitutive parameters using the hot torsion test. The approximation is based on an analytical solution for the torsion test. It shows that for the hot torsion test, the existing inverse computational methods based on a load-displacement objective function fail to predict a unique set of constitutive parameters for an assumed rigid-viscoplastic constitutive relationship. A new objective function was developed to ensure the uniqueness of the solution for the constitutive parameters. An algorithm is presented for applying the new method to the hot torsion test results.

Transformation of austenite into bainitic ferrite and martensite

Abstract: The stress induced martensitic transformation in the upper metastable intermediate state of γ-α-transformation in ferrous materials, structured as ferritic bainite, is discussed. The fibrous structured ferritic bainite consists of retained austenite and ferrite platelets growing in the [111]α//[101]γ direction. The ferrite growth induces carbon enrichment of the adjacent austenite at the phase boundaries. Strengthening at high stress levels up to the yield point causes dislocation tangles in the ferrite fibre and the formation of shear bands crossing each other in the retained austenite. At lower carbon contents of the austenite, lath martensite precipitates at the shear band intersections and at high shear band densities martensite blocks are observed. In carbon enriched austenite martensite lenses formed by shear processes have been observed. At alternating loading conditions, exceeding the stress level for athermic martensite formation, various shear planes are activated forming characteristic patterns of plate martensite.

The effect of coiling temperature and continuous annealing on the properties of bake hardenable IF steels.

Abstract: Five bake hardenable IF steels were investigated. The transformation and recrystallization behaviour, the precipitation and the mechanical properties were studied. The coiling temperature was varied between 580 and 720 °C. Soaking was carried out in a temperature range from 700 to 900 °C and the soaking time ranged from 60 to 240 s. It was found that boron significantly retards transformation and recrystallization. The bake hardening is increased but deep drawability is lowered by boron additions. In the Ti-containing steels, TiS. Ti4C2S2 and MnS precipitate competitively during the hot rolling process after reheating to 1250 °C. To obtain a reasonable bake hardening effect, the amount of carbide forming elements should be substoichiometrical related to C.

Modification of a classical formula for determination of roll flattening in flat rolling

Abstract: Local elastic roll flattening in direction of the roll axis in flat rolling is one of the important displacement components related to the roll gap shape. Up to now some classical formulae to solve roll flattening have been proposed. These formulae were derived from the elastic half space theory or the 2-dimensional contact theory. On calculating the flattening of the back-up roll they may result in a great deviation. In the present effort a comparison was made between the flattening results obtained by a 3-dimensional finite element method and by a classical formula. The relation between the deviation and the back-up roll geometry was also examined and therefore the classical formula was modified. Calculated results indicate that the modified formula can more accurately predict the back-up roll flattening.

Austenite stabilization from direct cementite conversion in low-alloy steels

Abstract: Transformation induced plasticity (TRIP) effects associated with austenite dispersions in low alloy Fe-Mn-Sl steels can be enhanced by austenite stabilisation. Austenite which forms during conventional intercritical annealing does not possess the required stability in order to exhibit TRIP effects. In this work, thermodynamic calculations indicated that it is feasible to form austenite by a cementite to austenite conversion which occurs under paraequilibrium conditions, i.e with partition of carbon but with no partition of substitutional alloying elements. In this way the austenite inherits the manganese content of cementite and is chemically stabilised. A treatment consisting of a two-step annealing has been examined. In the first step, soft annealing, an Mn-enriched cementite dispersion in ferrite is formed. In the second step, intercritical annealing, austenite nucleates on the cementite particles, which are consumed to form austenite. It was experimentally determined that this austenite has been enriched in manganese and carbon and, therefore, is stabilised. The conversion reaction is followed by the conventional austenite nucleation at ferrite grain boundaries. This austenite is lean in manganese and is not stable. The net effect of the two-step annealing treatment is a significant austenite stabilisation relative to simple intercritical annealing, indicating a potential for enhanced TRIP effects in this class of steels.

Thermal expansion and crown evaluations of rolls in rolling processes

Abstract: An understanding of the thermal expansion of work rolls is essential to the proper control of the flatness of a rolled strip in operating a modern, high-speed rolling mill. In this paper, an analytical solution has been developed to provide the thermal displacements and expansions of the roll. The thermal expansion data, in turn, can be utilized to estimate the thermal crown for shape control and contributes to alleviating many shape problems of the rolled product. The thermal displacement potential function has been used to solve the governing partial differential equations for the thermal displacements. The general stress function is then obtained to modify the displacement potential solution to satisfy the boundary conditions of the free roll surface. To demonstrate the accuracy and reliability of the solution developed, the results of the present solution have been compared with previously published numerical and experimental results at typical rolling conditions; good agreement has been found between them. A parametric study is then performed to provide the required information for proper control of the roll and strip shape in operating a high-speed rolling mill.

Modelling on melting of sponge iron particles in iron‐bath

Abstract: In this work the melting process of sponge iron particles in iron-bath is mathematically modelled. Based on the fact that the heat conductivity of sponge iron is much smaller than that of dense iron, a simplified analytical solution for the system has been derived. This solution agrees very well with the exact one from numerical calculation. The melting time and melting course of sponge iron particles are calculated. The melting time depends mostly on the particle radius. The amount of un-molten substance in the iron-bath at continuous feeding of sponge iron is estimated. The results show that under smelting reduction conditions the melting runs fast and must not be the rate determining step of the whole process.

Measurement of aluminum or silicon activity in Fe-Ni (< 30 %) alloys using mullite and ZrO2-based solid electrolyte galvanic cell

Abstract: Using mullite (3Al 2 O 3 .2SlO 2 )-tube type and ZrO 2 (MgO mole fractions of 9 %)-plug type solid electrolyte, the activities of Al in the Fe-15, 29% Ni-0.002 to 0.84% Al (mass contents in%) alloys were measured at 1873 K in an Al 2 O 3 crucible, and those of Si in the Fe-13, 28% Ni-0.13 to 2.3% Si alloys were measured at 1823 K in a mullite crucible with an excess of SiO 2 . The interaction coefficients of e Al Nl and r Al Ni were determined as -0.0173 ± 0.0052 and 0.00021 ± 0.00007, respectively, and those of e Si Ni and r Si Ni were determined as 0.0094 ± 0.0049 and -0.00008 ± 0.00004, respectively. The oxygen activities obtained by the EMF measurements were compared with those obtained from the contents of analyzed oxygen. The oxygen activities in the Fe-Ni-Al alloys were found to be supersaturated. The interaction coefficients of e O Ni and r O Ni were determined as 0.0027 ± 0.0012 and 0.00012 ± 0.00004, respectively, from the contents of analyzed oxygen in Fe-Ni-Sl alloys.

Austenite in the process of thermomechanical treatment of microalloyed spring steels

Abstract: On a classical CrV-bearing spring steel 50CrV4 (1.8159) the effects of thermomechanical treatment (TMT) and additional microalloying with titanium and niobium on the processes in austenite were investigated. The aim of this study was to achieve an austenite state, that promises -after subsequent quench and tempering- mechanical properties of spring steels superior to those after conventional treatment. For laboratory tests the hot deformation simulator Wumsi was employed. It was found that after reheating to usual austenitization temperatures the austenite microstructure of steel 50CrV4 is widely uninfluenced by the initial as-delivered microstructure. Despite the finer austenite grain size occurring after austenitization, additional microalloying with Ti and Nb raises the recrystallization temperature considerably. The finding that the incubation time of austenite recrystallization can be prolonged up to several minutes by microalloying is decisive for the practical applicability of TMT. In this way the substructured state of polygonized austenite beneficial for martensite formation can be preserved over the time needed for additional production steps between hot rolling and hardening of leaf springs. Moreover, the deformation of a not recrystallizing austenite may favourably influence the distribution of undesirable tramp elements in spring steels.

Effects of interface resistance on heat transfer in steel cold rolling

Abstract: The thermal contact resistance created in the bite region during rolling depends on surface roughness, contact pressure, and the coolant, lubricant or oxide scale between the roll and strip. Therefore, to estimate temperatures of the roll and the strip accurately, the interface resistance in the contact region should be considered when modelling. The purpose of this study is to more effectively analyze the thermal behaviour of the steel rolling process by considering the interface resistance of the roll and strip in the contact region. Since the interface is very thin in comparison with its length, it is modelled as one-dimensional heat transfer with friction heat generated along the interface. For the estimation of the thermal contact resistance, different surface situations of cold rolling are considered. The finite element method is adopted to evaluate the deformation and friction power dissipated in the rolls and strip during rolling process. Roll and strip thermal properties, such as thermal conductivity and specific heat, are considered to be temperature dependent during the calculation.