Design for AM

May 15, 2017


What is it?

exaSIM™ is a predictive simulation tool that enables Additive Manufacturing (AM) designers to rapidly optimize their designs for AM fabrication without the need to perform multiple trial and error iterations of a physical build. In short, exaSIM saves designers time and money when generating high quality, accurate AM parts.

What is the hurdle?

Figure 1: Component fabricated in 304SS, courtesy of Croft Filters

Figure 1: Component fabricated in 304SS, courtesy of Croft Filters

The increase in AM adoption has resulted in many designers finding themselves needing to design a part for metal powder bed AM processes such as metal laser sintering to save time and money for their company and customers. Designers who are new to the AM industry, and even those with many years’ experience, can find themselves in situations where intuition and experience do not give them the insight needed to determine how the AM layer-by‐layer build process will affect their component’s accuracy, aesthetics and performance. The most common method for determining the optimal design is to physically build several versions of the part to see which version will best result in a conforming piece of hardware. This trial and error design modification process can take weeks or months and cost tens of thousands of dollars.

Figure 2: (Left) STL File of Basic Component. (Right) exaSIM prediction of distortion showing areas at greater than 0.2mm displacement in dark red.

Figure 2: (Left) STL File of Basic Component. (Right) exaSIM prediction of distortion showing areas at greater than 0.2mm displacement in dark red.

This Tech Brief reviews a scenario where a part was designed and refined using exaSIM Beta during the design phase. The challenge was to find a configuration of the component that would conform to the requirement of less than 0.2mm profile tolerance. The as‐built part is shown in Figure 1. Since the part lacks symmetry, designer intuition could not be used to anticipate the significant deformation
seen in the as‐fabricated part as originally designed. Figure 2 shows the exaSIM predicted asymmetric distortion for the original design

How was simulation used to clear the hurdle?

Through the use of exaSIM the designer was able to visually see both the magnitude and location of distortion for the original design without needing to waste time and money on a physical prototype. The designer could immediately start to consider strategies for improving the accuracy of the part without delays.

For this part, some designers may try to support the asymmetric arm, where displacement is highest, with a beam or reinforcing rib to restrain the bar. This would work if the deformation were due to mechanical loading. However, due to the thermal nature of the AM process, part shrinkage in a beam or reinforcing rib can lead to greater asymmetry, and in many cases even greater distortion, as shown in Figure 3.

 

Figure 3: (Left) Design iteration using a square beam to support the asymmetric arm and predicted displacement. (Right) Design iteration using a triangular beam to support the asymmetric arm and predicted displacement. (displacement > 0.2mm is shown in dark red).

Figure 3: (Left) Design iteration using a square beam to support the asymmetric arm and predicted displacement. (Right) Design iteration using a triangular beam to support the asymmetric arm and predicted displacement. (displacement > 0.2mm is shown in dark red).

 

The designer utilized exaSIM to quickly see effects of various design changes on final part distortion and
determined that uniform stiffening was necessary. A single rib was added up the entire length of the asymmetric
arm. This led to a nearly conforming predicted displacement of 0.2001mm, as seen in Figure 4. The designer
then modified the design to explore slightly more complex reinforcement geometries, including an I‐beam
support. This resulted in a predicted displacement less than 0.2mm throughout the part, as shown in Figure 4.

Figure 4: (Left) Design iteration using a uniform rib with predicted displacement ~ 0.2mm. (Right) Design

Figure 4: (Left) Design iteration using a uniform rib with predicted displacement ~ 0.2mm. (Right) Design

Were the exaSIM predictions accurate?

The uniform rib part was built and measured to compare the accuracy of exaSIM predictions with physical part measurements. Figure 5 and Figure 6 show the measured results compared to simulated results of the
asymmetric arm of the uniform ribbed version.

Figure 5: Overlay plot of measured data (red) and predicted data (green) for the uniform rib design iteration

Figure 5: Overlay plot of measured data (red) and predicted data (green) for the uniform rib design iteration

Figure 6: (Left of image) Displacement map comparing the measured geometry with the STL file for the uniform rib iteration of design. (Right of image) exaSIM predicted displacement trend compared to the same STL file.

Conclusion?

exaSIM’s distortion prediction capabilities provide Design for AM value. Using exaSIM, designers significantly reduce the need for costly and time‐consuming trial and error iterations to achieve part conformance.

Download this TechBrief

Categorised in: