Optimize The Design Of Cast Steel Parts

At Eagle Alloy, one of the leading steel foundries in America, cracks occurred during the manufacture of a steel casting. These arose due to stresses occurring in the casting due to the shrinkage inhibition during and after the solidification process. The necessary costly and time-consuming welding repairs meant that the component could not be produced economically in the long term.

To avoid the error pattern, all degrees of freedom that could be selected for optimization were examined. In this case, this affected both the component design and the adaptation of the casting technology. In order to achieve the goal as effectively as possible, costly and time-consuming real tests were dispensed with and the new methodology of virtual test planning and autonomous optimization was chosen using the casting process simulation. In order to make success measurable, the tendency to crack as well as the porosity that occurred were selected as quality criteria and evaluated quantitatively.

Virtual attempts to crack

As part of the conceptual design, 12 different variants were constructed. For this purpose, both the cast part geometry, the cast part position and the feeder configuration were varied. The changes were in an experimental design (DoE) in the Gießprozesssimulationssoftware Magma ⁵ displayed. In the virtual tests, solidification and cooling of the cast part down to room temperature, including the stresses and tendency to crack, were calculated and predicted autonomously. The software delivered the fully automatic evaluation of the respective quality criteria for feeding and hot crack tendency in feed-critical and crack-prone areas.

The engineers used the statistical tools implemented in Magma ^{5 to} evaluate the virtual experiments. This allowed a quantitative evaluation of all designs in order to determine a possible optimal manufacturing technology that leads to more robust solutions. In a direct comparison of the different designs, the opposing trends of porosity and cracking became clear: As expected, the greatest tendency to cracking occurred with optimal feeding with minimal porosity. This was due to the longer solidification times and increased temperature differences between the cast part and feeders. Using the simulation results, however, a solution could be found that provided a good compromise.

Picture gallery

The comparison of the results between the initial situation and the best design showed that only through geometrical changes in the casting as well as in the feed technology could the stresses and porosity be reduced significantly at the same time.

With the new solution, the stresses and plastic deformations in the critical solidification area could be reduced by 61%. The evaluation in Magma ^{5 allowed} a quantitative evaluation of the individual measures: The manufacturing changes led to a reduction of 30%, while the adapted cast part geometry reduced the stresses and elongations that occurred by 44%.

Due to the positive results, both magnetic particle tests and processing time could be drastically reduced. The calculation of the 12 versions was carried out on a workstation and took a total of 22.5 hours. Conducting practical trials for each of these 12 versions would have taken several weeks, causing unacceptable material and labor costs.

The systematic use of the virtual experiment field in Magma ⁵ was the key to the specification and economical production of the cast part at Eagle Alloy. (qui)

Casting process simulation

Current software version automatically optimizes casting processes

Article files and article links

Link: More information about the Magma 5 simulation tool

Link: Link to the MAGMAacademy