To stay competitive, forgers must continuously integrate advanced manufacturing technologies into their production. Long gone are the days when forging was simply a means of shaping metal. For decades now, hot deformation has been performed under controlled process conditions in consideration of the subsequent cooling. This combination of the effects of plastic deformation and the controlled cooling during or directly after forging to achieve the required mechanical properties is referred to as controlled thermomechanical processing (TMP).

The results of TMP rely on the proper designation of the chemical composition of the forged material, as well as the correct utilization of process conditions both during forging and subsequent cooling. Proper process design of the hot forging and controlled cooling as a continuous sequence of interdependent thermal and/or mechanical processing stages can be complex and difficult. Simulation by the finite-element method (FEM) has been used for many years for the prevention of metal-flow defects, material-usage optimization, die stress analysis and selection of appropriate equipment. Commercial codes have been evolving rapidly and are incorporating a wealth of new features to simplify and extend the usability of numerical methods in comprehensive analysis of complex processes.

The implementation of more-complex material models such as thermomechanical interaction between forged material and elastic tools at anisothermal heat-transfer conditions, combined with improved data storage and increased data processing speed, have allowed a wider range of phenomena to be taken into consideration. This allows comprehensive analyses of microstructure evolution, failure prediction and, of course, heat-treatment modeling.

Case Study – Simulation of Forging Process

The first task of the simulation was the optimization of the forging process itself. In this example, ruptures occurred in the flash of the actual forging, so one of the basic goals of simulation was to determine the required material volume and the range of workability of the material under forging conditions.

Besides basic thermomechanical parameters such as metal displacement, strain, strain rate, temperature and other parameters of importance, the simulation allowed investigation of the grain-flow pattern that was tracked with flow lines starting as a rectangular grid in the billet (Fig. 1).

An intricate flash outline is a good opportunity to verify the correctness of metal flow in a simulation. Figure 2 shows the impressive consistency of the simulation to the actual dimensions of the flash. This also speaks to the accuracy of the boundary conditions, such as the friction factor.

One important aspect of the forging temperature is how it affects plasticity. Some separation was observed in the flash at the original higher temperature of 1220˚C (Fig. 2b). An analysis of the damage hazard was also carried out for a lower temperature of 1180˚C (2156˚F). The results are shown in Fig. 2.

The analysis used ductile-fracture criteria, which determines the risk of cracking based on the appropriate configuration of tensor components of stress and strain. These criteria are presented in the form of an integral that calculates the risk of fracture from the beginning of the process to failure strain values. The representative criteria for two ductile-fracture models were selected: Cockcroft & Latham and Rice & Tracey. The cracks observed in the flash of the forged part determined the critical values of the criteria. It was also possible to determine the moment and place of the initiation of the crack. (Fig. 3).

The results of fracture risk show the compliance of two accepted criteria with the location of cracks observed in the actual forging (Fig. 4). The crack is caused by complex material flow and an unfavorable state of stress, leading to a failure strain. Essentially, the rupture results from excessive material being pushed out of the impression. Areas with an increased risk of cracking are in the immediate vicinity of the final product, which carries the risk of damage to the product. This fracture is eliminated by reducing the billet size and lowering the forging temperature.

The simulation confirmed that the billet volume could be reduced by 6.5%, which is a 50% reduction of material waste. It also confirmed the temperature could be reduced from 1220°C to 1180°C and confirmed the reduced forging load, proving it could be forged on a smaller 1,600-ton mechanical press (Fig. 5).

Case Study – Analysis of Continuous Cooling

The second portion of this simulation was the thermal analysis to investigate the cooling curves tracked at points during continuous cooling in varying environments. The cooling simulation started from the end of the forging simulation, including inherited temperature and deformation fields from the forging and taking into consideration the cooling time provided for trimming operation. The temperature-time plots obtained were compared for consistency with experimental results to estimate the reliability of the model and defined characteristics of heat transfer.

This analysis was based on the definition of moving domains to simulate atomizers/vents, with special boundary conditions corresponding to the action of the cooling medium inside the domain (Fig. 6). Heat exchange depends on the distance between the cooled surface and the spray-unit nozzle (the intensity decreases with the distance from the nozzle) and is inhomogeneous through the jet section perpendicular to its axis (decreases from the center to the periphery). The assumption made in the software is that the cooling intensity is proportional to the relative velocity of the fluid at the observed point of the spray cooling-unit jet.

Parts of the workpiece may “shade” the areas lying behind them. We used the same domains for simulation for both atomized spray cooling and ordinary vents. This approach was based on the occurrence of the shadowed areas during vent cooling that were verified by means of flow simulation in Solid Works (Fig. 7b), which was the basis for the definition of geometry parameters of the sprayer domain. The action of convection was included in the heat-transfer coefficient defined inside the domain and the environment.

The simulation of direct cooling relies on an estimation of the temperature fields at the moment cooling starts. The temperature distribution in the forged part was calculated simultaneously with metal-flow analysis. Calculated fields of temperature at the end of three consecutive forging stages are shown in Fig. 8 and their changes in time in Fig. 9. Using these as initial data for subsequent cooling, modeling of continuous cooling was carried out, passing through four domains (Fig. 10d) with defined spray or vent heat-transfer coefficients.

The movement of the part through a domain and the results of cooling are shown in Fig. 10 for selected time intervals. The result of simulation was prediction of structural components in the volume of the finished part. Intermediate results of austenite transformation are shown in Figs. 11 a-c, illustrating content of untransformed austenite, pearlite and ferrite. In the end, these amounts in the points of interest were established at 41% ferrite, 58% pearlite and about 1% bainite. The calculated volume fractions of the structural components in steel are qualitatively in agreement with the microstructure revealed in point P2 (Fig. 12b). However, statistical methods will be needed for quantitative verification.

Evaluation of the simulation’s credibility can be done by temperature-plot comparisons. As seen in Fig. 12a, the calculated plot fits nicely to the experimental plots, which is in favor of the models despite significant sophistication of the process, complexity of the part, and relying on different sources of data and characteristics.

Conclusions

Evaluation and verification of these results show that emerging advances in modeling can be reliably used as a design aid in the thermomechanical processing of forged parts. Advanced models were used to simulate metal flow and thermal problems, taking into consideration cracking hazard diagnostics and predicting structural behavior in heat-treated steel.

The accuracy of metal flow was shown by flash dimensions, damage prediction and accurate indication of crack initiation. The simulation reduced scrap by 50%, eliminated potential defects and allowed the forging to be produced on a smaller press. The accuracy of thermal and heat-treatment problems was also very good, where (despite sophisticated processes and complex geometry) a close fit of temperature plots and credible microstructure fractions were observed.

Co-authors Piotr Skubisz, Lukasz Lisiecki and Magdalena Wolk are faculty members of AGH University of Science and Technology, Department of Materials Engineering and Industrial Computer Science, Kraków, Poland. Co-author Andrey Shitikov is part of the Metal Forming Department of Bauman Moscow State Technical University, Moscow, Russia. Co-author Tom Ellinghausen is with Forge Technology Inc., Woodstock Ill. He can be reached at 815-337-7555 or tom.ellinghausen@gmail.com. For additional information visit www.ForgeTechnology.com and www.qform3d.com.