Papers

These are the Benjamin Stump papers directly relevant to Condor’s thermal kernel, integration, parallelization, solidification outputs, calibration/CET, and sparse coupling. The list was cross-checked against Benjamin Stump’s Google Scholar link and ORNL publication record; publications on cellular automata, alloys, tribology, and other topics are not part of Condor’s mathematical derivation.

Core numerical method

B. Stump and A. Plotkowski, “An adaptive integration scheme for heat conduction in additive manufacturing,” Applied Mathematical Modelling 75, 787–805 (2019). doi:10.1016/j.apm.2019.07.008. This is the primary source for adaptive quadrature and selective melt-pool evaluation implemented in Condor.
B. Stump and A. Plotkowski, “Spatiotemporal parallelization of an analytical heat conduction model for additive manufacturing via a hybrid OpenMP + MPI approach,” Computational Materials Science 184, 109861 (2020). doi:10.1016/j.commatsci.2020.109861. This provides the parallelization context for independent space-time work.

Model fidelity and process use

B. Stump, A. Plotkowski, and J. Coleman, “Solidification dynamics in metal additive manufacturing: analysis of model assumptions,” Modelling and Simulation in Materials Science and Engineering 29, 035001 (2021). doi:10.1088/1361-651X/abca19. It compares conduction-only predictions with added physics and motivates explicit model limitations and calibration.
B. Stump, “An algorithm for physics informed scan path optimization in additive manufacturing,” Computational Materials Science 212, 111566 (2022). doi:10.1016/j.commatsci.2022.111566. This supplies the CET convention $\Delta T_c=(aV)^{1/n}$ used by Condor’s equiaxed-fraction expression and provides scan-optimization context.
S. Bi, B. Stump, J. Zhang, Y. Lee, J. Coleman, M. Bement, and G. Zhang, “Blackbox optimization for approximating high-fidelity heat transfer calculations in metal additive manufacturing,” Results in Materials 13, 100258 (2022). doi:10.1016/j.rinma.2022.100258. This provides the model-calibration context. Condor’s current Calibrate mode uses its own grid or gradient-descent search over two effective material factors rather than the paper’s Bayesian or directional-smoothing methods.