While the concept of 3D-printing – or Additive Manufacturing (AM) – is not necessarily new, it has certainly gained much more traction in the last decade. This is because digitization is continually drawing the design and manufacturing domains closer together. The thought of creating the rightsized component, just when it’s needed, with minimal waste and post-build machining is very appealing. Adopting AM theoretically gives way to more complex geometries, tighter process control and lower part-to-part variability. This is not the reality, however. Many sources of variability persist in AM, just as they do in traditional manufacturing processes, and have substantial impact on the durability performance of the final components.
Each type of AM method requires a different machine, and the multitude of settings on each machine must be dialed-in through exacting, repeated trials in order to create viable parts. To add complexity to the issue, the machines themselves are highly sensitive to a host of factors, including location, time of day, and raw material feed powder stock characteristics. For example, an entire sub-industry is centered around AM metal powder reuse: how many times can it be recycled before substantial degradation in properties is apparent; best practices for mixing and homogenizing virgin powder with recycled powder; etc. Depending on the “speed and feed” parameters of the machine, there will be defects built into the parts (voids, non-fused powder, layer melting/re-melting zones). Corner, edge, and overhang effects are magnified by these process variabilities. The challenges that face international standardization groups like ASTM F42 are put succinctly: how can AM “processes” be certified when AM “machines” demonstrate such wide sources of variability? Engineers at Oak Ridge National Laboratory (ORNL) and Aerojet Rocketdyne tackle these issues daily, and each have recently partnered with VEXTEC to help quantify and manage the risks to performance that are associated with Additive Manufacturing.
VEXTEC used its experience in material modeling and our predictive VPS-MICRO® software to evaluate the effects of AM on simple geometry specimens in fatigue testing (ORNL) and complex geometry parts in overload burst testing (Aerojet). As with any manufacturing, the signature of a component’s processing history lies within its material microstructure. Grain size, orientation, defect size and population distributions, strengthening mechanisms…they are all products of the specific AM technique and raw materials used. VPS-MICRO is a mechanical behavior modeling technology (not a process modeling technology), and it uses these quantifiable microstructural parameters to build a virtual representation of the material in a 3D digital space. The software computationally marries this material information with established design/structural analysis packages like Abaqus, to provide durability predictions that capture the true physics of failure.
At ORNL’s Manufacturing Demonstration Facility, test blocks of Ti-6Al-4V alloy were created via electron-beam melting (EBM). The goal was to predict the fatigue performance of specimens harvested from these blocks. Previous VEXTEC work included modeling of conventionally-forged/heat treated Ti-6Al-4V material. By understanding the microstructural differences between the EBM and the forged material, VEXTEC adapted the modeling parameters of the conventionally-produced alloy in order to capture the nature of the variabilities introduced by the AM technique. VEXTEC then used VPS-MICRO with this adapted material model to predict fatigue durability.
Aerojet Rocketdyne was evaluating the effects of “off-nominal” SLM machine settings on the resulting microstructure for a rocket nozzle. The component was made from Mondaloy, a nickel-based superalloy which exhibits high strength in oxygen-rich environments and is therefore a prime target for rocket engine manufacturers. Aerojet determined there were a multitude of distinct microstructural states arising from the transitions between the “nominal settings” and the “off-nominal settings”. At each setting, the microstructure varied throughout the part. VEXTEC accurately predicted both the burst test location and the range of static burst test pressure for various AM machine settings.
Studies like these allow us to build on current understanding of materials engineering, to help create computational methods for AM certification. Please contact us about your specific needs, and to see if VEXTEC’s VPS-MICRO software and engineering services are the missing tools for your AM toolbox!
