AM Simulation: a Guide to the Best Part Orientation and Support Configuration

The process of producing CMF implants is challenging. If not heat-treated, the implants tend to suffer from a so-called “spring-back” effect in certain build configurations due to residual stresses. This spring-back effect causes significant deformations in the final design, which is a problem since it may not fit the individual patient anymore. How severe this spring-back effect will be depends on your part orientation and support configuration. However, developing a proper configuration is complex and time-consuming.

In this case study, we investigate different part-support configurations of a personalized CMF implant, and determine which configuration has the least spring-back and, therefore, is most suitable for the additive manufacturing process. We do this by simulating deformations via a well-calibrated inherent strain method* and comparing the final deformations of the parts after support removal. We performed an additional comparison with the manufactured design configurations to confirm the simulation results.

As we explained in our previous simulation post, simulation is a powerful tool to minimize failed builds. It provides you with valuable feedback on how (not) to support or position critical regions. Let’s have a look at the simulation workflow.

1. Obtain the CAD geometry of the part and support configurations
2. Voxelize part geometries and consider support structures via boundary conditions
3. Simulate the layer-by-layer build-up via the inherent strain method
4. Interpolate the voxel results onto the original CAD geometry

We decided to focus on simulation speed and, therefore, the voxelization of the parts is fairly coarse. Our goal here is not to simulate the exact distortion of every configuration but to determine which configuration qualitatively deforms the least. In only three minutes, we simulated the entire voxelized platform.

The Magics Simulation software was used to simulate and view the results. A function to selectively load the most relevant simulation data inside Magics further enhanced the workflow to find the best-fit orientation. The image below shows three different part-support configurations (transparent) and the simulated deformations of the CMF implants after removal from the support structure. As you can see, the part-support configuration with the least deformation is the configuration in the middle. This one does not contain any red or yellow zones.

To confirm the results from the simulation, we performed a design-deviation comparison of the deformed parts with the original CAD geometry for both the simulation and printed parts. As you can see below, the simulated and the printed designs show the same deviation patterns. The top configuration has the biggest geometrical deviation from the original CAD geometry, and the middle one has the smallest deviation.

In this case study, we used finite element simulation to quickly predict the global deformations of three different part-support configurations of a patient-specific cranio-maxillofacial implant. The coarse voxelization approach allowed for fast simulation and quick feedback about the qualitative deformation tendencies of the configuration. Actual test prints confirmed that out of the three proposed configurations in this study, the configuration in the middle showed the least deformation after support removal.

Therefore, this second simulation case also confirms that the predictive power of simulation is a valuable tool for AM engineers. With the use of the Magics Simulation module, they can evaluate their designs prior to manufacturing, allowing them to quickly find the optimal part orientation and support configuration. Materialise leverages over 30 years of AM experience to develop AM-specific simulation approaches that will make “AM lives” easier.