This tutorial demonstrates a simple two-beam extension of the AdditiveFOAM AMB2018-02-B single-track tutorial.

This tutorial is intended to show how neighboring beams can alter the thermal field and solidification conditions relative to a single-beam setup.

This tutorial uses the IN625 material configuration from $ADDITIVEFOAM_ETC/materials/IN625.cfg.

The original AMB2018-02-B tutorial contains one moving heat source. This tutorial uses two moving heat sources:

sources (beam1 beam2);

The two beams follow the same scan direction, power, speed, and heat source parameters of the original AMB2018-02-B scan path, but are translated +/- 50 microns in the hatch direction.

Because each beam uses the original AMB2018-02-B laser power, the total applied laser power is twice that of the single-beam baseline.

Build AdditiveFOAM against OpenFOAM-14, source both environments, and run:

source /path/to/OpenFOAM-14/etc/bashrc
source /path/to/AdditiveFOAM/etc/bashrc
cd "$ADDITIVEFOAM_TUTORIALS/multiBeam"
./Allrun

Use ./Allclean to remove generated mesh, decomposition, and result files.

The baseline case is based on the AMBenchmark 2018 AMB2018-02 description:

https://www.nist.gov/ambench/amb2018-02-description

The model coefficients used by the AMB2018-02-B tutorial, including absorption and heat source dimensions, were calibrated in:

G.L. Knapp, J. Coleman, M. Rolchigo, M. Stoyanov, A. Plotkowski,
Calibrating uncertain parameters in melt pool simulations of additive
manufacturing (2023), https://doi.org/10.1016/j.commatsci.2022.111904.

The important files for this tutorial are:

constant/heatSourceDict

Defines the two moving heat sources, their absorption models, heat source models, and mesh refinement model.

constant/scanPath_1

Defines the first beam path.

constant/scanPath_2

Defines the second beam path.

constant/dynamicMeshDict

Defines the dynamic mesh/refinement settings.

system/blockMeshDict

Defines the base computational mesh.

This tutorial uses two heat sources:

sources (beam1 beam2);

Both beams use the same heat source parameters as the calibrated AMB2018-02-B single-beam case.

beam1
{
    pathName            scanPath_1;

    absorptionModel     constant;

    constantCoeffs
    {
        eta             0.33;
    }

    heatSourceModel     superGaussian;

    superGaussianCoeffs
    {
        k               2.0;
        dimensions      (85.0e-6 85.0e-6 30e-6);
        nPoints         (10 10 10);
    }
}

beam2
{
    pathName            scanPath_2;

    absorptionModel     constant;

    constantCoeffs
    {
        eta             0.33;
    }

    heatSourceModel     superGaussian;

    superGaussianCoeffs
    {
        k               2.0;
        dimensions      (85.0e-6 85.0e-6 30e-6);
        nPoints         (10 10 10);
    }
}

eta

Constant absorptivity applied to the laser power from each scan path.

k

Super-Gaussian shape exponent. In this tutorial, k = 2.0, giving a Gaussian-like source.

dimensions

Sets the heat source dimensions used by the moving heat source integration, taken as 2sigma.

nPoints

Controls the sub-cell sampling resolution used when integrating each heat source over mesh cells.

This tutorial can use the same targetCellLoad refinement model as the single-beam AMB2018-02-B tutorial. The buffers entries are keyed by source name and are applied relative to each path interval for each beam.

A representative refinement setup is:

refinementModel
{
    refinementModel         none;

    //refinementModel         targetCellLoad;

    refinementTemperature   1000;

    buffers
    {
        beam1               (85.0e-6 85.0e-6 100e-6);
        beam2               (85.0e-6 85.0e-6 100e-6);
    }

    targetCellLoadCoeffs
    {
        targetCellsPerProc  5000;
        nBufferVolumes      4;
        maxSearchIter       10;
        timeTolerance       1e-4;
    }
}

refinementModel

Selects the refinement model. targetCellLoad projects refinement ahead along the scan paths and adjusts the projected volume to target a cell count per processor.

refinementTemperature

Temperature threshold used to mark hot cells for refinement during AMR updates.

buffers

Source-specific scan-path projection buffers. Entries are keyed by heat source name and are applied to the corresponding beam path. The vector components define the buffer size in the scan direction, transverse direction, and build direction.

targetCellsPerProc

Target cell count per MPI processor. The targetCellLoad model adjusts the projected refinement volume to keep the total mesh size near this load.

nBufferVolumes

Minimum projected scan-path volume expressed as a multiple of the combined source-buffer volume. This keeps the projected refinement region from shrinking below the local scan-path coverage.

maxSearchIter

Maximum number of bisection iterations used when searching for the scan-path time interval that gives the target refinement volume.

timeTolerance

Stopping tolerance, in seconds, for the scan-path time-interval search.

Optional AdditiveFOAM function objects are listed in system/controlDict and are controlled by their enabled entries. To write additional data, set the selected function object entry to:

enabled true;

To disable a function object, set:

enabled false;

meltPoolDimensions writes melt-pool length, width, and depth data. solidificationData writes solidification events for CET analysis. ExaCA writes temperature history data for ExaCA input files.

The Allrun script calls the reconstruction helpers after the solver finishes:

reconstructExaCAData

reconstructSolidificationData

These commands exit quietly when no matching function object data were written.

After reconstructing the temperature data, run ExaCA from the case directory so that the relative paths in ExaCA/input.json resolve correctly:

mpirun -np <nProcs> <path-to-ExaCA> ExaCA/input.json

Plot absorbed power from the solver log with:

plotPower

Plot melt-pool dimensions:

plotDimensions

Plot CET data:

plotCET