Lawrence Livermore National Laboratory



This simulation shows laser energy interacting with a layer of individual powder granules. The transformation from blue to red to green corresponds with the powder granules melting and solidifying in consolidated form.

Simulating the critical process of powder melting

The powder model simulates the melting of metal powder, and its resulting densification, to improve the additive manufacturing process.

The powder model also provides data to the effective medium model as part of the internally funded strategic initiative to accelerate the certificaiton of additively manufactured metals. This model resolves individual powder particles in three dimensions and treats the laser–material interaction using ray tracing and a physics-based absorption model. The simulation incorporates not only the melting of the powder but also the flow of the liquid and the behavior of trapped gas. Time scales are on the order of fractions of a second and length scales are fractions of a millimeter.

Treating powder as a system of particles

Recently, a physics-based model that accounts for a number of important phenomena has appeared in the literature. This model is unique in the modeling of additive manufacturing processes in that it treats the powder as a discrete system of particles rather than a homogeneous continuum with effective properties. (Attar: 2011; Körner et al.: 2011) The model is two dimensional, and it uses a lattice Boltzmann approach to investigate melting and re-solidification of the powder particles. This class of approach is seen as being of value in developing additive manufacturing processes for new materials and for use in conjunction with process optimization algorithms, including design-of-experiments.

However, the referenced lattice Boltzmann methods are two dimensional, utilizing a mixed zone treatment and incompressible flow assumption, and compared to similarly advanced methods, are computationally intensive to an extent that may make 3D simulations of additive manufacturing processes impractical. Our approach uses Livermore's three-dimensional, massively parallel, multi-physics codes with thermal-mechanical coupling that includes solid particles, molten liquid, viscosity, and compressibility effects with cooling ambient gas to accurately represent the relevant physical and thermodynamic processes during additive manufacturing.

Mixing molten and solid phases

Modeling intensive mixing of molten phases and solid phase dynamics with sharp material interfaces requires an Eulerian or ALE (arbitrary Lagrange Eulerian) approach. The Laboratory's ALE and Eulerian codes have been successfully used for mesoscale and continuum-scale simulations for manufacturing processes including casting, forming, and extrusion. (Nichols: 2012) These codes have been used on all of the Laboratory's high-performance computing platforms and demonstrated good scaling characteristics up to tens of thousands of processors.

In addition to the existing capabilities, a few new features are being incorporated for representation of the additive manufacturing processes; these may include melt thermal convection enhancements and modification to existing numerical methods for handling element activation or late-time shaping of particles. To track bubble migration and gas escape in the melt, coupling several Laboratory codes may be beneficial. (Lomov et al.: 2009)

Andy Anderson

  • anderson1@llnl.gov

    Topical Area Leader for Powder Modeling
    Computational Mechanics Group -
    Computational Engineering Division (CED

    Andy Anderson

  • anderson1@llnl.gov

    Topical Area Leader for Powder Modeling
    Computational Mechanics Group -
    Computational Engineering Division