Quality problems of forged steel products may originate from the ingot casting process. Simulation tools for both casting and forming processes are available to analyze and optimize the quality and productivity of each. There is a clear need for a through-process simulation of both to predict possible defects and to optimize the entire process chain.

The final quality of forged steel products is the result of their sequential production steps. After steelmaking, molten metal is tapped from a ladle and poured into a mold, where it solidifies. The solidified ingot is then brought to its semi-finished shape through a series of reheating and forging steps, each of which influences final product quality. Many defects in forged or hot-rolled products originate from the casting process.

    Computer simulation is a tool used to investigate, understand and predict the effects of production processes on product quality. For the two main parts of the process – ingot casting and forging – dedicated simulation solutions are available. Simulation is applied in many plants to predict and optimize the properties of an as-cast product as well as a forged one.

    Using casting simulation models it is possible to predict shrinkage, centerline porosity, segregation, inclusions, residual stresses and cracks that originate during casting. With forging simulation it is possible to conduct virtual hot, cold, bulk, sheet and incremental forming processes to predict the shape of the part, the process forces and the resulting material properties. These are typically temperatures, strains and stresses, but material damage, phase constitution and grain size are increasingly simulated. Even die wear and die stresses can be analyzed.

Ensuring the Best Quality

Until recently, the worlds of casting and forging have only rarely intersected. There is significant reason that they should, however, since the origin of defects in forged or hot-rolled products may be found in the casting process. As-cast ingot quality is the starting point for all the subsequent reheating and deformation steps and plays a decisive role in final product quality. For example, the severity of porosities may be decreased by closing voids during forging if favorable forging conditions are achieved. Also, the position of local defects in the semi-finished product is affected by material flow during deformation. Inclusions or residual stresses remaining in the as-cast ingot negatively impact product quality when the ingot is intensively deformed.

    This article illustrates how casting and forging simulations can be combined to predict the influence of as-cast defects on forged steel products. The simulated properties of the as-cast ingot are transferred to a subsequent forging simulation. The shape of the shrinkage cavity at the top of the cast ingot is predicted by the casting simulation. This shape is used as the starting point for the forging simulation. Also, parts of these linked simulations are the position and extent of defects in the semi-finished product and the effect of the deformation on the severity of these defects.

Casting Simulation

In the past 30 years, casting process simulation has developed considerably. Virtual casting processes are done to determine the potential risks for defects and to predict material properties. Temperatures, metal-flow velocities, flowing particles, potential defects and also material properties can be analyzed. With MAGMA simulation software, parameters affecting ingot quality –
such as shrinkage, porosities, macrosegregation, cleanliness and cracks – are predictable.  These can be modified to help limit or eliminate quality problems. Figure 1, for example, shows the temperatures at one particular point of time during teeming of a 90-ton ingot.

Linking Casting and Forging Simulations

To simulate casting and forging processes, specialized software packages are applied sequentially for each process step. The transfer of simulations from one software package to the next is an often-discussed bottleneck of closed process-chain simulations. The problem lies not with different file formats so much as with the elements, meshes and symmetry conditions tailored to the specific processes.

    The main challenge for the transfer of results is transferring the meaning of the results themselves. Certain results are only meaningful in combination with a specific material model, which may be only valid for the specific application. For example, porosity may be a geometric result in one simulation but a material property in a subsequent simulation. A clear concept of how to deal with such transitions is needed when the transfer of results is to be designed.

    In this example, the casting simulation (MAGMA) and forging simulation packages (Simufact.forming) both support the transfer of results from one package to the other. For a quick transfer of data from MAGMA to Simufact.forming, a file is produced from the casting simulation using the I-DEAS universal format. This information is imported as “geometry with results” into the forging simulation, where it can be used with the existing mesh from the casting simulation or with any new mesh tailored to the forging process. In the present case, the porosity and the local concentration of specified elements (macrosegregation) are carried over from casting to forging as results. During the import, porosity is interpreted as a relative density, which will change during the forging process. 

Details of Open-Die Forging Simulation

A breakdown (cogging) process was simulated using an ingot (Figure 1) as the input condition. Simufact.forming was used for this, utilizing the implicit finite-element method. Because the solver allows for rigid-body movements, all movements of the workpiece can be simulated. Elastic/plastic material behavior was used in a fully mechanical-thermal coupled analysis with hexahedral elements known to deliver the most accurate predictions.

    The software contains a kinematics module that enables the simulation of open-die forging and ring-rolling processes. The standard cogging kinematics control used for this simulation allows the setup of complex simulations using input parameters that are close to factory language. Several heats with several passes each can be defined. For each pass, the final height and translational and rotational movements must be specified.

    To simulate the closing of centerline porosity from casting, a macroscopic approach was used. The porosity is described by the relative density of the material, which is increased during the forming process based on a material law using the following equation:

                                                   (1)

where r_r is the relative density, r₀ is the initial relative density, s_m is the hydrostatic pressure obtained from the analysis and p_max is the maximum hydrostatic pressure needed to close all voids.

    Once the relative density reaches 1.0, all voids are closed and the material is fully consolidated. The implemented relative-density model assumes that the initial porosity allows treating the material as homogeneous and that the effect of voids on the material properties during hot forging can be neglected.

    The simulated process consists of three heats with a total of 23 passes with different rotations, leading to a total of 459 blows. The maximum diameter of the ingot is reduced from 100 to 55 inches and the length simultaneously increases from 177 to 492 inches. The simulation uses about 50,000 hexahedral elements with automated re-meshing. To precisely include the mechanical and thermal interactions of the workpiece not only with the saddles but also with the manipulators, a single, uniform mesh with elastic-plastic material was used for the whole workpiece. Figure 2 shows the simulation model.

Shrinkage and Porosities

The solidification pattern of ingots leads to a characteristic shrinkage in the as-cast ingot (Figure 3). There is always a shrinkage cavity in the hot top, and it must not extend into the final workpiece. During the forging process, the remaining shrinkage cavity is deformed together with the workpiece, and the forging simulation tracks this deformation together with the workpiece’s overall shape. Figure 3 shows that even a small shrinkage cavity can lead to a considerable affected area at the end of the cogging process. With the simulation, different cogging strategies can be evaluated easily to minimize this effect.

    In many cases, problems with centerline porosity are reported. This porosity is small in comparison to the hot-top shrinkage cavity and is typically found along a line in the center of the ingot, as shown in Figure 4 (left). Depending on the size and location of porosities, it is possible to close them in subsequent forging processes if the hydrostatic pressure in the area of the voids is sufficiently high. The process design should be optimized to provide the required hydrostatic stresses.

    Casting simulation can be applied to optimize the process to prevent porosity from forming in the first place. If its presence is inevitable, it is important to transfer information about the size and position of the porosity to the forging simulation, where it is possible to determine the process parameters required to close the porosity. Figure 5 shows how centerline porosity is removed during a simulated breakdown (cogging) process.

Macrosegregation

Segregation is an inhomogeneity of the concentrations of alloying elements and impurities in the steel. Most alloying elements are more soluble in the liquid phase than in the solid phase. Thus, as the metal solidifies, alloying elements in the “mushy” zone are rejected by the growing solid dendrites into the neighboring interdendritic liquid, which becomes increasingly enriched with alloying elements. This is termed microsegregation.

    Macrosegregation can result in an ingot with regions having a composition quite different from the nominal value. Segregation can lead to locally lower material properties and to variations in thermochemical behavior, such as the formation of precipitates or local hot spots that induce shrinkage porosities.

    The casting process simulation shows local concentrations of all relevant elements in the steel chemistry as they may be expected in the cast ingot in Figure 4 (right). If this information is transferred to the forging simulation, changes of the distribution of the concentrations due to the material flow during deformation can be analyzed and the expected local chemistry of the forged workpiece can be predicted (Figure 6).

    During forging, cooling and reheating operations, local chemistry concentrations are not only influenced by the material flow but also by several diffusion effects. The simulation of these effects is an area of ongoing research. Simufact.forming supports this by its flexible data structure for material data. The material database considers the chemical composition and provides the infrastructure for phase-dependent material properties, which are used in the industry today for phase-transformation simulations.

Conclusions

The existence and the locations of defects in a semi-finished product have been predicted by simulations of the casting process. These play a significant role in the quality of the final, deformed product. The coupled through-process simulation gives valuable information that could not otherwise be obtained. This information can be used to optimize both casting and forging processes to target the best properties of the final forged product.

Co-author Dr. Marc Schneider is president and CEO of MAGMA Giessereitechnologie GmbH in Aachen, Germany. He may be reached M.Schneider@magmasoft.de. Co-author Dr. Ingo Hahn is product manager. He may be reached I.Hahn@magmasoft.de. For additional information please visit http://www.magmasoft.de. Co-author Dr. Hendrik Schafstall is managing director & CTO of Simufact Engineering GmbH, Hamburg, Germany. He may be reached at hendrik.schafstall@simufact.de. Co-author Dr. Christian Barth is manager of Engineering Services of Simufact Engineering GmbH, Hamburg, Germany. He may be reached at Christian.barth@simufact.de. For additional information please visit www.simufact.com.