Saving Heat Treatment Steps - Means Saving Costs

Residual stresses that evolve in the casting process during solidification and cooling may cause problems during subsequent machining or component use. In one particular case, dimensional changes of a ductile iron main bearing cap for marine applications had been observed after storing of the part. For this reason, the typical practice was therefore to apply an annealing heat treatment to release residual stresses after casting.

The Finnish foundry group Componenta and the marine and energy component manufacturer Wärtsilä jointly investigated whether it would be possible to omit the annealing step for cast iron main bearing caps. The parts are typically exposed to a heat treatment at around 550-600 °C for about a day. Being able to skip this annealing process would advantageously shorten foundry lead times and allow for significant cost saving potentials.

Both partners are used to using information generated from MAGMASOFT® during casting processes. Several simulations and mold cooling rate measurements were performed to find suitable simulation parameters for accurate cooling conditions from solidification up to shake out, which are the basis for reliable residual stress results. The simulated stresses were compared to measurements on the produced castings to fully understand the residual stress state of the main bearing caps. Finally, dimensional measurements of main bearing caps without heat treatment before and after storing them were carried out to verify the robustness of the castings against dimensional changes during storage.

In order to measure the cooling curves, sensors were placed in the mold cavity, and the mold was closed and poured under normal production conditions. The casting was shaken out after 24 hours of cooling, and the temperatures at that point of time were about 290 °C.

 

Figure 1: Main bearing cap

Figure 1: Main bearing cap

Virtual thermocouples were placed at the same positions in the simulated model to compare temperatures with the measured curves. The initially simulated cooling curves differed strongly from the ones measured in the production mold. Based on experience with other simulation projects, it was decided to adjust the sand properties for simulation to attain a better fit of simulated and measured curves. The casting process was then simulated to predict the residual stress formation using the sand properties found using this inverse procedure. After that, real castings were produced and their residual stresses were measured by means of X-ray diffraction at two locations in two perpendicular coordinate directions on the casting surface. The stresses at the casting surfaces are generally affected by the shot peening process, which creates high compressive stresses in the material within about 1 mm of the surface. Therefore, any comparison of measured and simulated residual stresses is only valid for the material below this layer.

Figure 2: Temperature curves – measured and simulated

Figure 2: Temperature curves – measured and simulated

The von Mises stresses were calculated from the measured stresses in the two directions. The values for measurement point 1 can be seen in the figure above, together with the appropriate simulation results.

At position 1 of the casting, the simula- ted residual stress is about 45 MPa, which is confirmed by the measured
values which vary from 16 to 53 MPa (from 1.5 to 5 mm below the casting surface).

The simulated distribution of maximum principal stress in the main bearing cap is shown in the figure below. The highest stress is in the center of the casting, which is not a critically loaded area. The maximum principal stress is about
60 MPa.

Figure 3: Measurement of residual stress in two  directions – marked as phi=0° and phi=90° -  measurement (top) and results (bottom table)

Figure 3: Measurement of residual stress in two directions – marked as phi=0° and phi=90° - measurement (top) and results (bottom table)

The simulation with MAGMASOFT® showed that the residual stresses occurring during casting and cooling were not sufficient to distort the part significantly. From a stress evaluation perspective, no significant dimensional changes were to be expected from storing the casting for some time without prior stress-relief heat treatment.

Figure 4: Maximum principal stress distribution in a main bearing cap

Figure 4: Maximum principal stress distribution in a main bearing cap

In order to check this assumption, 3 main bearing caps were machined without annealing. Their exact dimensions were measured in a 3-D coordinate measurement machine before storing them for one month.

The measurement after storage proved that the dimensions did not significantly deviate from those directly after casting. All dimensions were within the tolerances, particularly the critical ones. Consequently, the stress annealing stage for this type of casting was waived, saving significant time and costs for both parties.

Figure 5: Von Mises stresses calculated from the measurements (left) and indicated by simulation

Figure 5: Von Mises stresses calculated from the measurements (left) and indicated by simulation

Wärtsilä

Wärtsilä from Finland is a global leading supplier of ship engines and power plants.

Componenta

Componenta, a Finnish technology company, specializes in the supply of cast iron and machined components.