Materials Technology

Bulk Steel Products with Functionally Graded Properties Produced by Differential Thermo-mechanical Processing Ursula Weidig1), Klaus Hübner2), Kurt Steinhoff1)

1) Institute for Production Technology and Logistics, Chair of Metal Forming Technology, University of Kassel, Kassel/Germany 2) FEMUTEC Engineering GmbH, Kassel/Germany For the production of bulk steel products with tailored properties a new approach of thermo-mechanical processing offers high potential for process innovations. The application of differential thermo-mechanical effects to initially homogeneous workpiece materials combines thermally controlled material flow with functional grading of mechanical properties. The key effect is the control of local microstructural transformation. The detailed investigation and description of all related phenomena, not only for forming processes of bulk steel products but also for light metals and polymers, is the main topic of the recently established Transregional Collaborative Research Centre 30, which is financed by the German Research Foundation. The comprehensive coverage of these phenomena is aimed at developing tools to govern this innovative new type of hybrid production processes. Recent optimisations of the experimental set-up in the field of hybrid metal forming offer new possibilities in the investigation of the main effects. In the following the effects of different cooling strategies on the functional gradation of properties are investigated by temperature and hardness measurements. Finite-element simulations give additional information to experimental observations. The control of these effects offers the possibility of a differential thermo-mechanical driven functional grada-tion of bulk steel products.

Keywords: functionally graded materials (FGM), differential thermo-mechanical treatment, tailored properties, bulk metal forming, tempera-ture point mapping, FEM

DOI: 10.2374/SRI07SP059; submitted on 10 May 2007, accepted on 31 July 2007. Introduction

The ongoing trend towards lightweight components aims at the integration of elevated mechanical properties, ge-ometries adapted to the load profile and reduced joining operations. The answer is the tailoring of geometries and mechanical properties, produced by flexible manufactur-ing processes. However, state-of-the-art technologies are still limited to the utilization of work hardening or adjust-ment of thickness profiles by sheet metal semi-products. The combination of local heating and conventional rolling techniques has led to the development of innovative new hybrid rolling techniques of thermo-mechanically tailored strips [1]. Whilst in general in sheet metal forming new thermo-mechanical techniques are adopted within a short time, there is often a significant delay in comparable ad-vances in bulk metal forming. Up to now, graded proper-ties in bulk components have mainly been achieved by graded material compositions produced by powder metal-lurgy or casting technologies [2, 3]. These methods only enable simple gradients. The interest in forming technolo-gies for the creation of more complex functionally graded materials is increasing significantly. Semi-finished prod-ucts, functionally graded before by powder metallurgy, are forged or extruded to their final shape. Due to their ex-tremely differing flow resistance and formability this ap-proach seems to be of limited success [4, 5]. On the other hand, the commonly used production processes are not cost effective, which is the main reason for the restrictive use of functionally graded components in mass produc-tion, so far.

The aim to improve the efficiency of existing bulk metal forming processes and, at the same time, to establish tech-nological means to increase their flexibility towards a defined adjustment of functional profiles can be achieved by a new thermo-mechanically coupled process. This approach consists of transferring a pre-defined tempera-

ture profile during deformation to a defined material-property profile and/or geometrical shape. The customisa-tion of the local formability to the deformation pattern supports a homogenization of resistance against deforma-tion and, therefore, contributes to a smooth course of the process with the effect of, e.g., an improvement of tool life. This new type of process strategy is accompanied by unconventional phenomena regarding heat transfer mechanisms and shape determining mechanisms. A transi-tion from free heat transfer to heat conduction is observed, as well as from free forming to contact related forming determined geometrically by the given tool shape [1].

Differential Thermo-Mechanical Processing

In general, conventional - integral - thermo-mechanical

processing consists of simultaneous forming and heat treatment of the entire workpiece. The choice of the effec-tive process temperature level resulting from pre-heating, cooling and deformation-induced heat generation deter-mines whether the microstructure arising will be transfor-mation hardened, multiphase, recrystallized, refined or work hardened. The differential processing, where the temperature profile directs the material flow and thus the final shape, additionally offers the beneficial opportunity to adjust the distribution of various microstructural prop-erties. A wide range of microstructural configurations is a characteristic of workpieces undergoing such special thermally profiled process conditions. It has to be noted that the local microstructural evolution does not only de-pend on the temperature profile itself but also on the strain distribution. In the zones where the maximum temperature locally exceeds the transformation temperature Ac3, a partially hardened structure of bainite/martensite evolves at sufficiently high cooling rates. When cooling down from lower initial temperatures within the two-phase area, a multiphase structure of ferrite/pearlite and bain-

steel research int. 79 (2008) No. 1 59

Materials Technology

ite/martensite develops, exhibiting a hardness profile ac-cording to the hard phase fraction. Zones without any phase transformation show a recrystallized or work hard-ened structure depending on the temperature, strain level and exposure time. Among those effects with significant influence on the final property distribution, not only the aforementioned thermo-mechanical conditions are of par-ticular importance with regard to the resulting gradation, but also the metallurgical potential of the material has to be considered [1]. Therefore, the achievable range of func-tional gradation with respect to material properties and part geometry depends on the ability to control the thermo-mechanical process conditions, and furthermore, to control the dynamic microstructural mechanisms by an appropriate choice of material grade and condition.

Set-up with Controlled Heat Transfer. In order to

eliminate the drawbacks of uncontrolled variation of proc-ess parameters, such as heating time and heat losses during the workpiece positioning, a computerized automation of the initial experimental set-up is integrated. The set-up comprises a hydraulic press with a capacity of 1000 kN, an inductive heating facility and a workpiece transfer unit, Figure 1. The manual/mechanical control systems are complemented by a process-control on the basis of a CAN-Bus-IO-module. This allows for the accurate time dependent control of the machine functions.

Figure 1. Experimental set-up.

The induction facility is especially designed to generate 3-dimensional thermal profiles in workpieces. In the first expansion stage it consists of a high-capacity power sup-ply and 2 independent medium-frequency transformers (10 - 70 kHz), which offer the opportunity to simultane-ously heat two areas with different heating parameters. Heating power, frequency and time schedule can be ad-justed independently according to the required tempera-ture profile. An additional medium-frequency and a high-frequency transformer complement this facility. Specially designed induction coils - single turn, double-turn, three-turn, etc. - are used to enable a great variety of different temperature profiles over the workpiece length and cross-section. The height of the coils, as listed in Table 1, corre-sponds approximately to the height of the area of maxi-mum temperature.

The transfer unit consists of devices for clamping and rotating of the workpiece and for its electro-pneumatically controlled transport. The rotation of the workpiece during the heating guarantees a homogeneous circumferential temperature distribution. All these components are con-trolled by a computer program which executes the overall time control and the data acquisition. The set-up is equipped with several sensors for measuring the die tem-perature and the overall forming forces. The workpiece temperature during heating, transfer, forming and cooling is registered locally by a pyrometer, a camera for thermal imaging records the superficial temperature field (Figure 1). The cooling facility is integrated within the die design.

Modular Die Design. The die design is based on a

modular concept. It consists of a simple container carrying the active components, which can be exchanged in a sim-ple way. This offers the possibility to gradually integrate more complex shapes in the course of the research project going along with an increased understanding of the proc-ess. Furthermore, by a set of components of different die coatings and die materials their influences on the process can be investigated as well as the deterioration of the dies due to mechanical and thermal fatigue and shock after a defined set of cycles.

Figure 2 shows the configuration for workpieces with a single flange consisting of hot forming and cold forming

components. For the cold forming die the steel 90MnCrV8 (1.2842) was chosen, quenched and tem-pered to a hardness of 57-59 HRC, for the hot forming die X38CrMoV5-1 (1.2343) was heat treated to a medium hardness of 45 HRC in order to meet the require-ments for high temperature loading and alternating thermal stress. In the reference layout no active cooling was intended for the hot forming component. In order to vary the cooling rate during the die - workpiece contact, water cooling of the hot forming die insert was integrated and pressure air cooling will follow.

Figure 2. Die concept with cold and hot forming die areas and exchangeable dies.

60 steel research int. 79 (2008) No. 1

Materials Technology

Workpiece Material. The material to be functionally graded is a CrV-alloyed heat-treatable steel (51CrV4) which is generally used for parts in the general mechanical engineering and vehicle construction like gearbox parts, pinions, shafts, gears etc. [6]. The material is delivered as commercially available bar stock in a spheroidized or soft annealed condition with a length of 6 m and a diameter of 35 mm. It is cut to segments with 200 mm length and dressed to size with a diameter of 30 mm. In Table 2 the chemical composition is shown as received from a spark emission spectrometer. The vanadium content could not be analysed because of missing reference of the spectrometer for this element. As the material is commercially available, vanadium is evaluated from data given in literature [6]. A time temperature diagram for continuous cooling of the commercially available steel is shown in Figure 3, as reference for transformation temperatures, cooling rates and resulting microstructure and hardness.

Table 1. Induction coils.

induction coil inner diameter [mm]

height [mm]

single-turn 20 double-turn 40 three-turn

44 60

Table 2. Chemical composition of the workpiece material 51CrV4 in wt.%).

C Si Mn P S Cr V

0.53 0.26 0.88 0.009 0.01 1.03 0.10-0.25

Table 3. Process parameters of reference experiment.

main heating 42 kW (70 % Pmax)

12 s

post heat warming 15 kW (25 % Pmax)

5 s

max. temperature 1350 °C

frequency (approx.) 15 kHz

indu

ctiv

e he

atin

g

configuration single three-turn coil

duration of transfer and positioning 4 s

max. temperature at start of forming 1120 °C

max. temperature at finish of forming < 400 °C

max. forming force 1000 kN

stroke velocity (approx.) 60 mm/s

die-workpiece contact time 28 s

total time forming + cooling 35 s cooling rate of free workpiece surface (1000°C - 800°C) 47 K/s

cooling rate of free workpiece surface (800°C - 500°C) 30 K/s

lubrication graphite / lubricant AS 1400

form

ing/

coo

ling

cooling strategy die contact / passive cooling

Figure 3. Time-temperature-transformation diagram by continuous cooling of commercial heat-treatable steel 51CrV4 [7].

General Experimental Procedure. The general test pro-cedure comprises an inductive heating, followed by an automated workpiece transfer and positioning and finally a forming step with an integrated cooling, see Figure 4. The standard parameters of the reference experiment are given in Table 3.

In order to promote the cold forming of the non-heated areas, the workpieces are lubricated at the bar ends with graphite in an alcoholic suspension, whereas the die is lubricated additionally with AS 1400, a mineral grease with inorganic solid lubricants.

Figure 4. Experimental procedure.

steel research int. 79 (2008) No. 1 61

Materials Technology

At the start of the experiment, the control program reads a script file with the process parameter set. The work-piece is fixed by a clamping device and rotated during the heating sequence.

After a defined heating time, the electro-pneumatically controlled device positions the workpiece into the die. The computer program activates the press at a defined delay after heating, corresponding to the duration of workpiece transfer and positioning. By the downward movement of the upper die, the forming process starts. Different cooling strategies are possible, depending on the set-up configura-tion. During the first steps of transfer, positioning and free forming, heat losses occur due to radiation and convection (Figure 4). Accelerated cooling follows during the inten-sive die-workpiece contact time of 28 s, whereas the die is not actively cooled. After the release, convection heat loss is the dominating mechanism during the remaining cool-ing of the workpiece.

Table 4 summarizes the parameters for the experiments with different die-workpiece contact times and different cooling modes of the hot forming die insert.

Graded Properties by Different Cooling Strategies Die-Workpiece Contact Time. The measurement of the

surface temperature of the workpiece in the zone of the peak temperature reveals the influence of different die-workpiece contact times and die cooling modes on the resulting cooling rates, see Figure 5. The temperature characteristics after the die-workpiece contact are not registered in the given set-up due to the displacement of the workpiece in the final step of the workpiece release.

During the die-workpiece contact the cooling rate of the free surface is nearly identical regardless of the total con-tact time. Table 4 reveals the same cooling rate for the first interval from 1000 - 800 °C.

After the contact time all work-pieces undergo a free heat transfer by radiation and convection but with starting temperatures de-pending on the total contact time, see Figure 6. This difference in the starting temperature is the main parameter to determine whether a fully martensitic, par-tially hardened or non-hardened microstructure will be obtained, independent of the cooling mode of the die insert. Figure 6, for example, shows the resulting hardness profiles in the flange area, depending on the die-workpiece contact time. As can be seen in Figure 6a, a steep hard-ness increase from the centre to the surface is only obtained with a contact time of 28 s with a non-cooled die insert. The hardness values near the surface corre-spond to the maximum hardness values of a fully martensitic struc-

ture of 750- 760 HV, cf. Figure 3. A 50 % reduction of the total contact time (14 s) results in a low increase of hard-ness values comparable to the results obtained with a short contact time of 4 s, corresponding to a ferrite-pearlite-bainite structure (390 - 420 HV), cf. Figure 3. With a water cooled die this change in hardness evolution behav-iour is shifted to shorter contact times, i.e. 4 s.

experi-ment

cooling rates of free workpiece surface

die insert

heating parameters

contact time tdie-wp [s]

Cwp °−

•

8001000,T [K/s]

Cwp °−

•

500800,T

[K/s]

max. temp.

max,Tdie

[°C]

cooling rate

dieT•

[K/s]

modus

S 42 kW/ 12s 15 kW/ 5s

28 47 30 375 50

SZ-14 14 52 32

SZ-4 4 47 -

die contact/ passive cooling

WS 28 46.5 32 325 80

WZ-14 14 50 33

WZ-4 4 52 -

die contact/ active cooling: water flow rate: 26 l/min water temp.: 20 °C ±5 °C

Table 4. Test series cooling parameters, other parameters according to reference experi-ment.

Role of Die Cooling Mode. The influence of an acceler-

ated cooling of the hot forming die by water circulation in coolant boreholes (water flow rate: 26 l/min; water temp.:20 °C ±5 °C) is shown in Figure 7. It reduces the peak temperature of the die by about 50 K (375 to 325°C). The cooling rate of the insert increases from 50 to 80 K/s.

Although the water cooling increases the cooling rate of the die insert, only slight differences are observed in the temperature characteristics of the free workpiece surface compared to the behaviour with a non-cooled die, cf. Fig-ure 5. The maximum workpiece surface temperature is reduced by about 30 K and in the interval between 500 - 400 °C the cooling rate is increased by about 50 % (20 K/s resp. 13 K/s).

The conductive heat transfer workpiece - die depends on the temperature gradient and the thermal diffusivity and conductivity of both die and workpiece. The die water cooling has an influence only on the temperature gradient, because it primarily lowers the maximum die temperature. The experiments show a reduction of the die temperature in a range of about 50 K, as explained above. On the other hand, the maximum workpiece temperature is lowered by about 30 K. Consequently, there is only a slight increase of the temperature gradient, not sufficient to have a seri-ous effect on the cooling rate, measured at the free work-piece surface during the contact cooling. Thus, in order to significantly vary the cooling rate during the die contact, it is essential to change the thermal conductivity, i.e. the hot forming die insert material.

62 steel research int. 79 (2008) No. 1

Materials Technology

Figure 5. Cooling rates of free surface of workpiece during the die-workpiece contact, depending on cooling mode.

Figure 7. Die temperature evolution during the die-workpiececontact, depending on cooling mode, contact time 28 s.

Figure 6. Hardness distribution in radial direction, depending on die-workpiece contact time, a) non-cooled die insert, b) water cooled die insert.

Figure 8. Hardness distribution in radial direction, depending oncooling mode, a) contact time 28 s, b) contact time 14 s.

steel research int. 79 (2008) No. 1 63

Materials Technology

The influence of the die cooling mode on the hardness profile is shown in Figure 8. A contact time of 28 s, the parameter of the reference experiment, leads to identical maximum surface hardness of 750 HV. At 50 % reduction of the contact time (14 s) a complete change in the result-ing hardness profile (Figure 8b) occurs. Whilst a water- cooled die results in high surface hardness of 750 HV, comparable to a fully martensitic structure and lower core hardness values (450 HV), a non-cooled die induces only a slight increase from a core hardness of about 310 HV to 410 HV at the surface.

Figure 9. Temperature mapping of the measured experimental temperature distribution to the FEM-model. Experimental data correspondto specimen/experiment SZ14 (cooling mode non cooled die, contact time 14 s).

The use of water cooling of the die guarantees a fast re-initialisation of the experimental process parameter “die temperature” but it does not increase the general work-piece cooling rate, as stated above. Considering this fact and the difference in die temperature of 50 K, this discrep-ancy in the surface hardness between 750 HV and 450 HV at an intermediate contact time of 14 s depending on the die cooling mode is rather surprising. Although the start-ing temperature for the free heat transfer after the forming and cooling is identical (430°C) and the course of the free cooling rate after the die-workpiece contact is supposed to be independent of the die cooling mode, it seems to be decisive in the given case. Differing local temperatures at the invisible die-workpiece interface may be the underly-ing reason for a different cooling behaviour after the die contact and thus for the resulting higher amount of hard phases.

Assuming the interface temperature to be lower with a water-cooled die, the inner parts of the flange may be of lower temperature than with a non-cooled die after a con-tact time of 14 s. After the release of the workpiece at a temperature above the martensite start temperature, the heat transfer of the core material is responsible for the final cooling rate, leading to a decelerated cooling by an experimental configuration with a non-cooled die with the consequence of a transformation in the ferrite/bainite phase area. The reference contact time of 28 s does not show this cooling mode depending behaviour because the die-workpiece contact ends at a temperature below the martensite start temperature.

Figure 10. Temperature characteristics of distinct points in the simulated process after a temperature point mapping of the initial workpiece temperature distribution prior to the simulated forming and cooling process.

With regard to the fact that the temperature characteris-tics of the workpiece during and after the forming process are only measured in visible areas, the lack of reliable data for non-visible areas, especially the die-workpiece inter-face, constrains a detailed study of the influence of an active die cooling on the phase transformation behaviour. Local cooling rates at distinct workpiece positions may explain the contradictory results. Due to the present con-figuration of the experimental set-up and the resulting restricted space, miniaturised thermal and pressure sensors are needed for the measurement of local temperatures and forces in the die-workpiece interface. A complemented solution may be offered by finite element simulations, e.g.

64 steel research int. 79 (2008) No. 1

Materials Technology

by a tracking of the temperature evolution at distinct points.

Temperature Point Mapping and Tracking by Finite

Element Simulations. In order to establish a method for determining the volume temperature distribution, a map-ping of experimental temperature data to the FE-model is introduced. For a given experimental parameter set the temperature was measured at distinct positions on the surface and across the workpiece by thermocouples. The temperature data were transferred to a finite element model of the initial workpiece prior to the simulation of the forming and cooling process. For all simulations the software MSC.SuperForm was used. At distinct nodes of the workpiece model the experimental temperature data prior to the forming process were mapped manually. An extended compensation time with the mapped temperature data as fixed boundary condition led to an interpolation of missing nodal temperatures. The resulting temperature distribution is shown in Figure 9. With this experimen-tally obtained temperature distribution the forming process is simulated including the contact cooling step as well. An overall temperature of 50°C and Coulomb friction with constant µ=0.3 are assumed for the rigid dies.

Figure 10 shows the temperature characteristics during the forming process for the experiment SZ 14 with a die-workpiece contact time of 14 s visualized by finite-element simulations.

At the free surface (point A) a cooling rate of approxi-mately 20 K/s is observed in the temperature interval be-tween 1000-800°C. This is slightly lower than the experi-mentally determined cooling rate. With respect to the given simplifications it shows a quite good agreement between experiment and simulation. At point B, in the middle of the cross-section, 10 mm from point A, the cooling rate is determined in the same temperature interval with 30 K/s. The higher cooling rate at this point may be explained by the cooling effect of the die-workpiece con-tact during the die controlled forming, cf. Figure 10, which does not happen at point A. At a lower temperature interval (600-500°C) during the die-workpiece contact cooling different cooling rates are determined, 15-16 K/s at point A and B, but 34 K/s at point D, which corresponds to the die-workpiece contact area at a distance of 10 mm from point C. This striking difference in the cooling rate may be caused by the contact cooling effect of the die. It confirms the interpretations of the experimental results concerning the elevated cooling rate - compared to the free surface - of the invisible and thus non-measurable contact areas of die and workpiece.

Conclusions and Prospects By a differential thermo-mechanical processing func-

tionally graded properties are feasible with bulk steel products. The increase in reproducibility by an automation of the experimental procedure allows for an in-depth in-vestigation of isolated main process parameters. For ex-ample, the effect of cooling rates on the functional grada-tion of properties is limited by the workpiece material characterisitcs, especially the thermal conductivity. As a

consequence, these material properties constrain the adap-tation of the press hardening approach to bulk metal form-ing due to cooling rates and transformation kinetics of the chosen steel. Active water cooling of the die has only a slight impact on the general workpiece cooling rate but has an effect on local cooling rates at the die-workpiece interface, resulting in a different hardness distribution. Die-workpiece contact times influence the workpiece hardness depending on the temperature at the end of the die-controlled cooling and the transition to air cooling. High die contact end temperatures lead to a dominantly air cooling effect with the result of a low amount of hard phases, low die contact end temperatures, near or below martensite start temperature, yield in a high amount of hard phases up to a fully martensitic structure with maxi-mum hardness at the flange edge.

By governing these effects even a differential thermo-mechanically driven functional gradation of bulk steel products with several lateral extruded geometrical ele-ments is possible. A first successful approach is shown in Figure 11. This shaft with a double flange is produced by differential thermo-mechanical processing within a single press stroke and two heated areas.

Figure 11. Shaft with a double flange produced by differential thermo-mechanical processing.

Acknowledgments This paper is based on studies carried out by the Trans-

regional Collaborative Research Centre SFB/TR TRR 30, which is kindly supported by the German Research Foun-dation DFG. References [1] Steinhoff, K.; Weidig, U.; Scholtes, B.; Zinn, W.: Steel Research Int.,

76 (2005) No. 2/3, 154-159 [2] Joenssen, M.; Kieback, B.: Formgebungsverfahren für Gradienten-

werkstoffe. Pulvermetallurgische Formgebung im Wandel – Pulver-metallurgie in Wissenschaft und Praxis, 15 (1999), 23-42.

[3] Kieback, B.; Neubrand, A.; Riedel, H.: Materials Science and Engi-neering, Part A, 362 (2003) No. 1-2, 81-105.

[4] Raßbach, S.: Grundlegende Untersuchungen zum Umformverhalten von Gradientenwerkstoffen unter Anwendung von Druckumformver-fahren. Freiberger Forschungshefte B 323, 2002, Freiberg.

[5] Raßbach, S.; Szczepanik, S.; Lehnert, W.; Lehmann, G.: Untersu-chungen zur Herstellung von Bauteilen mit makroskopisch gradierter Werkstoffzusammensetzung durch Umformprozesse. In: Verbund-werkstoffe und Werkstoffverbunde. Whiley-VCH Verlag, 2001, p. 223-228.

[6] Stahlschlüssel, Verlag Stahlschlüssel Wegst, Marbach, 1998 [7] Data sheet 1.8159, Dörrenberg Edelstahl, Engelskirchen, Germany.

steel research int. 79 (2008) No. 1 65