Material data, powder behavior and laser-material interaction present unique challenges for simulating metal 3D printing.
Simulation is a vital tool in additive manufacturing (AM). Whether it’s thermal simulation, mechanical simulation or simulation of the 3D printing process itself, the potential to optimize both 3D printing time and material usage is substantial. However, the physics of 3D printing – particularly 3D printing metal – creates certain challenges when it comes to applying simulation to this process.
Here are three of the most common challenges for simulation in metal additive manufacturing.
1) Material data
The complex, multi-physics nature of metal AM means that simulating the process requires numerous inputs. For metal powders or wires, these include density, viscosity, thermal conductivity, heat capacity, latent heat, emissivity or absorptivity and, in the case of powders specifically, particle geometry.
The trouble is, this material data typically comes from many different sources based on various experiments, and there’s no guarantee that those experiments match the specific conditions and material properties of the application at hand. For this reason, material data inputs for metal AM simulations should ideally be collected in-situ, on a case-by-case basis. Unfortunately, this increases both the time and cost of metal AM simulations, essentially running counter to the point of running the simulations in the first place.
2) Powder behavior
Metal AM processes that involve powder deposition can be modeled in two different ways with respect to deposition, specifically.
A continuum approach treats the powder bed as a homogenous continuum, with the powder particles as a continuous medium. This is the approach taken by simulations focusing on macroscopic, part-level features. It’s computationally inexpensive, since it doesn’t require capturing the behavior of each individual particle, but it’s also coarse-grained for the same reason.
In contrast, a discrete approach resolves the individual particles in the powder bed. This is obviously better suited to micro- and mesoscale simulations of powder packing density, particle interactions and keyhole formation. However, it’s also much more computationally expensive.
The challenge here lies in determining when to apply which method, a problem that’s compounded by the fact that the best answer can change layer-to-layer over the course of the entire process.
3) Heat source and laser-material interaction
Modeling the heat source (a laser or electron beam) and its interaction with the feedstock material is central to metal AM simulations since this is what defines the thermal boundary condition of the energy balance equation throughout the process. In the context of powder-based metal AM, there are several factors that make modeling the heat source and laser-material interaction challenging for engineers.
To start, there’s the complex physics involved in the process, including absorption, reflection, scattering and re-emission of radiation. Heat transfer is also affected by the size and shape of the powder particles, as well as their effective thermal conductivity and the surrounding environment. Then there’s the stochastic nature of the powder bed itself, with the local packing of the bed having a strong influence on laser penetration depth.
The laser beam characteristics, including power, spot size and shape also impact the thermal profile of the powder bed. In addition, variations in laser intensity over the course of the build, resulting in non-uniform heating of the powder bed.
Finally, the relatively small scale of powder bed fusion and directed energy deposition processes – which typically have a layer thickness of 20-100 micrometers – makes it difficult to measure and validate thermal profiles experimentally.
