A Physics Informed Machine Learning Framework towards Density Prediction of Additively Manufactured Components

Abstract:

Laser powder bed fusion (LPBF) has become a key manufacturing route for high-value components, yet accurate prediction of part quality, particularly relative density, remains challenging due to complex interactions among process parameters, material properties, melt-pool physics and shielding environment. Traditional physics-based simulations offer mechanistic insight but suffer from model-form uncertainty and computational cost while purely data-driven machine learning (ML) models often lack interpretability, physical consistency and transferability across build conditions. To address these gaps, we propose a physics-informed machine learning (PIML) framework that integrates structured domain knowledge with symbolic regression to derive compact, interpretable analytical expressions for LPBF density. The framework constructs a knowledge base (KB) comprising dimensionless and normalized physics-informed process descriptors (PIPDs) that encode energy input, thermal diffusion and melt-track geometry; these descriptors form a physically consistent feature space for learning. The framework also serves as a foundation for the proposed future community-driven, knowledge-graph-based modelling of LPBF processes. The capability of the framework is demonstrated by modelling the relative density of additively manufactured maraging steel and evaluating cross-atmosphere transferability from Argon-shielded to Nitrogen-shielded builds. The resulting symbolic models provide a transparent, extensible and physically meaningful alternative to black-box ML, achieving high predictive performance under the training regime (R = 0.964) and strong generalization across printing atmospheres (R = 0.896).

• Periodical: Key Engineering Materials (Volume 1047)
• Pages: 73-82
• Online since: April 2026
• Authors: Abhilash Puthanveettil Madathil*, Olga Bylya, Andrew Sherlock
• Keywords: Additive Manufacturing, Density Modelling, Laser Powder Bed Fusion, Physics-Informed Ml

References

[1] DebRoy, T., Wei, H. L., Zuback, J. S., Mukherjee, T., Elmer, J. W., Milewski, J. O., … Zhang, W. (2018). Additive manufacturing of metallic components: Process, structure and properties. Progress in Materials Science, 92, 112–224.

DOI: 10.1016/j.pmatsci.2017.10.001

[2] Kan, W. H., Chiu, L. N. S., Lim, C. V. S., … Lim, C. V. S. (2022). A critical review on the effects of process-induced porosity on the mechanical properties of alloys fabricated by laser powder bed fusion. Journal of Materials Science, 57, 9818–9865.

DOI: 10.1007/s10853-022-06990-7

[3] Bruna-Rosso, C., Demir, A. G., & Previtali, B. (2018). Selective laser melting finite element modeling: Validation with high-speed imaging and lack of fusion defects prediction. Materials & Design, 156, 143–153.

DOI: 10.1016/j.matdes.2018.06.037

[4] Tang, C., Tan, J. L., & Wong, C. H. (2018). A numerical investigation on the physical mechanisms of single-track defects in selective laser melting. International Journal of Heat and Mass Transfer, 126, 957–968.

DOI: 10.1016/j.ijheatmasstransfer.2018.06.073

[5] Zakirov, A., Belousov, S., Bogdanova, M., … Shakhov, F. (2020). Predictive modeling of laser and electron beam powder bed fusion additive manufacturing of metals at the mesoscale. Additive Manufacturing, 35, 101236.

DOI: 10.1016/j.addma.2020.101236

[6] Chen, K., Zhang, P., Yan, H., Chen, G., Sun, T., Lu, Q., ... & Shi, H. (2024). A review of machine learning in additive manufacturing: Design and process. The International Journal of Advanced Manufacturing Technology, 135(3), 1051-1087.

DOI: 10.1007/s00170-024-14543-2

[7] Tapia, G., & Elwany, A. H. (2015). Prediction of porosity in slm parts using a mars statistical model and bayesian inference, 2015 International Solid Freeform Fabrication Symposium.

[8] Tapia, G., Elwany, A. H., & Sang, H. (2016). Prediction of porosity in metal-based additive manufacturing using spatial Gaussian process models. Additive Manufacturing, 12, 282-290.

DOI: 10.1016/j.addma.2016.05.009

[9] García-Moreno, A. I., Alvarado-Orozco, J. M., Ibarra-Medina, J., & Martínez-Franco, E. (2020). Image-based porosity classification in Al-alloys by laser metal deposition using random forests. The International Journal of Advanced Manufacturing Technology, 110(9), 2827-2845.

DOI: 10.1007/s00170-020-05887-6

[10] Khalad, A., Telasang, G., Kadali, K., Zhang, P. N., Xu, W., & Chinthapenta, V. (2024). A generalized machine learning framework for data-driven prediction of relative density in laser powder bed fusion parts. International Journal of Advanced Manufacturing Technology, 135, 4147–4167.

DOI: 10.1007/s00170-024-14735-w

[11] Chen, Y.-M., Lu, J.-L., Yu, D., … Wang, J.-C. (2025). Accurate identification of high relative density in laser-powder bed fusion across materials using a machine learning model with dimensionless parameters. Acta Metallurgica Sinica (English Letters), 38, 1645–1656.

DOI: 10.1007/s40195-025-01895-1

[12] Maitra, V., Arrasmith, C., & Shi, J. (2024). Introducing explainable artificial intelligence to property prediction in metal additive manufacturing. Manufacturing Letters, 41, 1125-1135.

DOI: 10.1016/j.mfglet.2024.09.138

[13] Kapusuzoglu, B., & Mahadevan, S. (2020). Physics-informed and hybrid machine learning in additive manufacturing: Application to fused filament fabrication. JOM, 72(12), 4695–4705.

DOI: 10.1007/s11837-020-04438-4

[14] Liu, R., Liu, S., & Zhang, X. (2021). A physics-informed machine learning model for porosity analysis in laser powder bed fusion additive manufacturing. International Journal of Advanced Manufacturing Technology, 113(7), 1943–1958.

DOI: 10.1007/s00170-021-06640-3

[15] Staszewska, A., Patil, D.P., Dixith, A.C. et al. A machine learning methodology for porosity classification and process map prediction in laser powder bed fusion. Prog Addit Manuf 9, 1901–1911 (2024).

DOI: 10.1007/s40964-023-00544-2

[16] Atwya, M., & Panoutsos, G. (2024). In-situ porosity prediction in metal powder bed fusion additive manufacturing using spectral emissions: A prior-guided machine learning approach. Journal of Intelligent Manufacturing, 35(6), 2719–2742.

DOI: 10.1007/s10845-023-02170-9

[17] Barrionuevo, G.O., La Fé-Perdomo, I. & Ramos-Grez, J.A. Laser powder bed fusion dataset for relative density prediction of commercial metallic alloys. Sci Data 12, 375 (2025).

DOI: 10.1038/s41597-025-04576-x

[18] Thomas, M., Baxter, G. J., & Todd, I. (2016). Normalised model-based processing diagrams for additive layer manufacture of engineering alloys. Acta Materialia, 108, 26-35.

DOI: 10.1016/j.actamat.2016.02.025