SIMULATION OF THE STAMPING PROCESS

Simulation of the stamping process was performed using Autoform software. A thermomechanical analysis is carried out throughout the simulation until a steady state is achieved. The effectiveness of the cooling system is determined by monitoring the temperature of the drawpiece and the temperature on the working surfaces of the tools. The temperature of the drawpieces at the end of the process should not exceed 250°C. This, together with a minimum cooling rate of approximately 27°C/s for MBW-K1900 steel [13], guarantees the desired martensitic structure of the drawpiece.

Computer-aided design models of the tools were prepared for the simulation of the process. Based on a three-dimensional model of the drawpiece, working surfaces of the die, binders, and punches were obtained, as illustrated in Fig.4.

Fig. 4.

Tools working surfaces

Prior to the simulation of the stamping process, tool kinematics and basic process parameters, including tool stroke and stamping speed were selected. Firstly, a correctly cut blank is positioned on the upper surface of the die and the lower binder, and the stamping process is initiated (Fig.5a). The application of the upper binder serves to secure the blank while the punches simultaneously initiate the formation of the initial bottom flange (see Fig.5b). Next, the die and lower binder, mounted on gas springs, move downwards, together with the closed upper binder and punches, forming the top flange (Fig.5c). Afterwards, dies designed for this purpose, attached to sliders with actuators, form the side flanges (Fig.5d). After closing the tool, the press pressure is increased (for improvement of heat exchange between the stamping die and the tool) and the quenching process is initiated. At the end of this stage, the press tool is opened and the tools take their initial position.

Fig. 5.

Stages of drawpieces forming

Once the tool kinematics had been established, the parameters related to the cooling process of the die could be determined. In the actual tool, the cooling channels are drilled into individual sections of the tool. In order to simulate the cooling process, the position of the centre lines of the channels was planned and the appropriate diameters were assigned (Fig.6).

At this point, it is necessary to note a certain aspect related to the modelling of tool cooling channels in the AutoForm package. The latest version of the software does not yet allow the modelling of cooling channels for tools operating in a direction other than the main stamping direction which is a requirement for the designed process (a punch for bending side flanges). The software only permits the assumption of a constant temperature (an initial value of 80°C was assumed for the analyses) on the surface in question. This also serves to illustrate the innovative nature of the proposed solution with a movable punch. Accordingly, simulations of the cooling process of punches for bending side flanges will be carried out after the preliminary design of the tool in the Simcenter FloEFD thermal-fluid analysis module.

The geometric models of the tools and cooling channels were supplemented with the material properties of the drawpiece. In accordance with the design assumptions and with reference to the experience of tool manufacturers, considering suggestions provided by the sheet metal manufacturer, the following was assumed for the process:

It was assumed that the maximum value of the heat transfer coefficient from the blank to the tools would be 3.5 mW/mm²K. This is the value for a tool pressure of 100 MPa. For lower pressures, the heat transfer coefficient decreases. In such cases, the value corrected by the scaling factor (Fig.7) determined by the supplier of the stamping sheet is adopted for the calculations.

Fig. 7.

Distribution of cooling channels a) in the upper part of the tool b) in the lower part of the tool

Fig. 7.

Scaling factor of the pressure

The simulation of tool heating and cooling is performed in an iterative manner. The convergence criterion of the computational process is the state at which the differences in tool temperature in subsequent iterations are no greater than 2°C. In the case under discussion, four complete cycles were required, with the Quick Numerical Cycling option enabled. This method enables a notable reduction in the number of cycles required to stabilise the temperature of the drawpiece and the tool. Instead of starting the analysis with the ambient temperature for the tools, the tool initial temperature is estimated for the first cycle, enhancing the overall efficiency of the analysis.

The calculations provided a number of results describing the course of the stamping process. Primarily, the FLD (Forming Limit Diagram) was subjected to analysis. The results shown in Fig.8 do not indicate any product feasibility issues. No regions susceptible to cracking were noticed.

Fig. 8.

Distribution of cooling channels a) in the upper part of the tool b) in the lower part of the tool

The next figure (Fig.9) shows the temperature distribution of the drawpiece at the end of the quenching process. It can be seen that the maximum temperature does not exceed 248°C and is lower than the required temperature. It is also worth noting that over almost the entire surface of the finished drawpiece, these temperatures are only slightly higher than 180°C. The two places with significantly higher drawpiece temperatures are where the lifting devices are to be located in the tool. Their function is to lift the drawpiece after the quenching process to a position where it can be picked up by the robot arm. Inside the die tools, during the quenching process, the top surfaces of the lifting devices are positioned approximately 0.5mm below the surface of the die and are therefore not involved in the intense heat exchange between the die and the drawpiece.

Fig. 9.

Temperature distribution in the drawpiece

Fig.10 shows the temperature distribution on the working surfaces of the tools. As can be seen, the highest temperatures only slightly exceed 100°C, which means that the process will be able to run continuously, without the need for additional time to allow the tools to cool down.

Fig. 10.

Temperature distribution in tools

The predicted martensite distribution in the drawpiece (Fig.11) guarantees a product with the required material properties.

Fig. 11.

Share of martensite in the drawpiece

Based on the results of the simulations, a 3D model of the tool was made. It is a two-cavity die, the main components of which are a lower plate (ribbed cast iron) with a die, lower binder and a system of punches working in a direction other than the main stamping direction (Fig.12a), and an upper plate (ribbed cast iron) with an upper binder and forming punches (Fig.12b).

Fig. 12.

Parts of the tool a) lower part of the tool, b) upper part of the tool

Two punch units are built into the tool, for forming flanges in a direction other than the stamping direction (Fig.13).

Fig. 13.

Movable punch

COOLING SIMULATION OF MOVABLE PUNCHES

A special cooling system with drilled channels (Fig.14) was designed for the moving punch system, as for the other parts of the tool, with a separate coolant supply system.

Fig. 14.

Cooling water supply system for stamps

As mentioned above, when analysing the stamping process in the AutoForm package, it is not possible to simulate cooling of the tools operating in a direction other than the stamping direction. For this reason, the cooling system for the moving punches was analysed in the FloEFD thermal-fluid calculation system. Fig.15 shows a model of the system with the computational domain of the cooling fluid highlighted.

Fig. 15.

3D model of the coolant domain

The geometric model thus prepared was discretised. The dimensions of the mesh in the discrete model were selected in an iterative manner. In successive iterations, the size of the finite elements was reduced and the effect on the results obtained (temperatures and velocities) was examined. If the results did not change significantly (more than 2%) in the next iteration, it was assumed that the finite element dimensions from the previous iteration were optimal. Fig. 16 shows, in an example section, a mesh discretising the volume of the punch and coolant adopted for the calculations. The computational model of the die consisted of 1427330 cells, of which 468882 elements represented the coolant volume.

Fig. 16.

FEM mesh of the punch model

No simulation of the stamping process is performed during the heat-flow analysis. It is therefore necessary to determine the heat flux density that the drawpiece gives off to the top and bottom of the tool during the forming and quenching processes. Then, the temperature of the drawpiece decreases from the initial temperature of t₂=810°C (before the forming process), to the final temperature of t₁=180°C (after the quenching process). Knowing the forming and quenching time T=14 s, the mass of the drawpiece m=4 kg, and the specific heat of the steel c=0.46 kJ/(kg°K), it is possible to calculate the heat flux Q˙, transferred to the tool:

For the beam described, the flux is Q˙=83 kW. Assuming that the heat is transferred uniformly to the top and bottom of the tool, a heat flux density of q˙=130 kW/m2 was obtained (Fig.17). The cooling system on the inlet side was supplied with the medium at 20°C and the flow rate of 6 l/min (0.0001 m³/s). Open channels were assumed to be on the outlet side of the medium. Free convection with the ambient temperature of 40°C was taken into account.

Fig. 17.

Analysis boundary conditions

To evaluate the effectiveness of the proposed cooling system, the temperature of the working surfaces of the tools was tested. Fig.18 shows the temperature distribution on the working surface of the moving punches. As can be seen, the temperatures are around 74°C, which is slightly lower than the 80°C initially assumed. It can therefore be concluded that the cooling system for the moving punches was correctly designed.

Fig. 18.

Temperature distribution in moving punches

PROBLEM OF PRODUCT JAMMING. ANALYSIS OF THERMAL SHRINKAGE OF THE DRAWPIECE AND TOOLS

The problem associated with large-size drawpieces jamming in the tool after quenching is relatively common. Shrinkage and quenching stresses can cause problems when removing the drawpiece from the die tools. The value of the force exerted by the drawpiece on the die after cooling to the final process temperature is significant and exceeds 40 kN (Fig.19).

Fig. 19.

The force of the drawpiece on the die

In the example described, jamming is eliminated in two ways. Firstly, based on the simulation results (Fig.20), a tool is made with a dimension that compensates for the shrinkage of the drawpieces. It is worth noting that the magnitude of this shrinkage along the beam is approximately 0.8 mm, which is half the thickness of the drawpiece. During the beam production, stamping conditions change and, within the limits agreed with the supplier, the material changes. This can result in the tool being insufficiently adjusted and the die continuing to jam. This phenomenon is going to be eliminated by the use of movable punches (retracted after hardening).

Fig. 20.

Drawpiece deformation after hardening (in relation to nominal shape)

Thermal expansion of dies and punches can also be the cause of a drawpiece jamming and problems with correct tool operation. For this reason, a thermo-mechanical analysis was carried out for the lower and upper part of the tool and their deformations were determined. Simplified 3D models of the tools were developed for this analysis. These simplifications consisted in removing from the model all parts whose influence on the load and strength of the die is negligible. For reasons of symmetry, half of the top and half of the bottom of the tool were calculated (Fig.21). The models prepared this way were analysed using the finite element method in the Simcenter package.

Fig. 21.

FEM model for calculating the expansion of a) the lower part of the tools b) the upper part of the tools

The upper part of the die and lower part of the die were loaded with a heat flux q˙=130 kW/m2 determined by analysing the flow through the moving punch system. The temperature in the cooling channels was assumed to be 20°C. The model was supported on the lower and upper surfaces of the dies where the die is in contact with the press bed. In addition, symmetry conditions were assumed in the model (Fig. 22). Material data was added. (base material is GG25 cast iron, working parts are CR7V steel).

Fig. 22.

Boundary conditions for the analysis of (a) the lower part of the tool (b) the upper part of the tool

The problem was completed in two stages. First, the thermal analysis problem was solved and the temperature distribution in the tool was determined. This was followed by a static analysis under the temperature load obtained from the previous analysis. As a result of the calculations, deformation values were determined for the lower and upper parts of the tool (Fig.23 and 24) due to their thermal expansion. These values do not exceed 0.07 mm for the lower part of the tool and a maximum of 0.024 mm for the upper part of the tool. These deformations are very small and do not affect the stamping process, in particular the jamming and the dimensional accuracy of the drawpieces. It is worth mentioning that the minimum value of the clearance between the moving parts of the stamping tool (dies, punches and binders) is assumed to be 0.4 mm, i.e. many times greater than the deformation values determined.

Fig. 23.

Deformation in the lower part of the tool

Fig. 24.

Deformation in the upper part of the tool

TESTING OF MANUFACTURED PARTS

In accordance with the described project, a stamping die was made (Fig.25) and mounted on an experimental stamping line. In order to assess the correct performance of the stamping die, a test batch of 150 drawpieces was prepared (Fig.26). Mechanical parameters were then measured using both non-destructive and destructive testing methods. Non-destructive testing was carried out on 20 randomly selected drawpieces and 5 drawpieces were chosen for destructive testing.

Fig. 25.

Stamping die (lower part)

Fig. 26.

Photo of a sample of a drawpiece

Non-destructive testing - measurement of mechanical properties of the drawpiece

The 3MA (Micromagnetic Multiparameter Microstructure and Stress Analysis) instrument from the Fraunhofer Institute [14] was used to measure basic mechanical parameters of the drawpiece, such as yield strength Rm and hardness in Vickers degrees. Measurements were taken at 10 control points as shown in Fig. 27.

Fig. 27.

Position of the measurement points on the drawpiece

Analysing the obtained results (Tab. 1, Tab. 2), it can be concluded that for each of the drawpieces, the required mechanical parameters were achieved at each measurement point. Hardness exceeded the required value of 400 HV and the strength limit was within the expected range of 1400÷1900 MPa.

Destructive testing

Tensile test

Paddle samples with a gauge length of 50 mm were cut from the flat areas of the test drawpieces using a plasma cutter. These samples were then elongated on WDW-300S testing machine from TEST Lab. During the test, the yield point and ultimate strength were determined. Fig.28 shows a photograph of a selected sample after the test, and Fig.29 shows the results of the measurements.

Fig. 28.

Photo of a sample

Fig. 29.

The results of the tensile tests

The report of these measurements (Tab. 3) confirms the production of drawpieces with the required mechanical parameters.

Tab. 3.

Tensile test results

Sample number	Yield strength Rp02, MPa	Tensile strength Rm, MPa
1	1355	1851
2	1316	1873
3	1262	1830
4	1284	1838
5	1336	1804
6	1279	1858

Vickers hardness measurement and microstructure analysis

Vickers hardness tests and microstructure analysis after the hardening process require special preparation. Samples for microstructure analysis were cut from selected areas on the drawpiece using a water-abrasive cutter (Fig.30).

Fig. 30.

Locations where samples were taken for testing

Next, the prepared material was positioned with a metallographic pin and then embedded in phenolic resin. In the next step, the samples were ground with diamond discs and polished using diamond suspensions with grain sizes of 9 μm and 3 μm and oxide suspensions with grain sizes of 0.05 μm (Fig.31).

Fig. 31.

Locations where samples were taken for testing

Examples of Vickers hardness results are shown in Tab. 4. As can be seen, the obtained values for the tested samples exceed the 400HV value which is considered to be required.

Tab. 4.

Results of HV hardness measurements

Measurement number	Imprint/value
1		579.1
2		538.7
3		533.4
4		561.9
5		573.3
6		583.6

In the subsequent phase of the experiment, the samples that had been previously prepared were etched in a solution of 4% Nital-4 acid. The structure was then examined at 50x magnification using a microscope of the AmScope ME1200TC-5M Inverted Trinocular Metallurgical Microscope + 5MP Camera type. Fig.32 shows an example of a photograph taken using this microscope.

Fig. 32.

Microstructure photograph of the sample

The resulting structure is a martensitic structure, which is a required outcome of the hot stamping process. It is distinguished by high mechanical properties and was observed in all samples obtained from each of the measurement areas.

Hydrogen embrittlement test

Hydrogen embrittlement occurring during the hot stamping process can lead to uncontrolled cracking of body drawpieces during normal automotive operation. A simple workshop test of a material resistance to hydrogen embrittlement is four-point bending of samples taken from the drawpiece. Such a sample, rectangular with the dimensions of 110 x 25, is placed in a special tool [14]. The punch of the tool is tightened with a screw so that the deflection of the sample induces a stress equal to 80% of the tensile strength limit. The sample is left in the tool for 300 hours and next visual inspection is performed. Table 5 presents the photographs of the tests performed. None of the tested samples showed cracks or scratches on the surface, which could mean the onset of cracking. Therefore, it can be concluded that the samples passed the four-point bending test.

Tab. 5.

Results of four-point bending test

Sample	During study	After completion of the study
1
2
3
4

Measuring the shape-dimensional accuracy of a drawpiece

To check the shape-dimensional accuracy of the manufactured drawpieces, we first analysed the requirements defined in the technical documentation. For the drawpiece in question, the shape and dimensional accuracy was to be within ±0.5mm at the adjacent surfaces and ±1.2mm in other areas. Based on this analysis, a measurement plan was drawn up and five parts were randomly selected and measured on the CMM (Coordinate Measuring Machine). Fig.33 shows an example of the measurement report.

Fig. 33.

Example measurement report

Fig. 33.

The image from the thermal camera

It was found that at none of the measurement points, the assumed tolerances of the drawpiece production were exceeded, particularly on the adjacent surfaces significant for the assembly of the drawpieces.