ESAFORM Benchmark 2025: predicting stainless steel PBF-LB part density using statistical, data-driven, and physics-informed machine learning models derived from process parameters and in-situ monitoring data - 2026
ESAFORM Benchmark 2025: predicting stainless steel PBF-LB part density using statistical, data-driven, and physics-informed machine learning models derived from process parameters and in-situ monitoring data
[en] Abstract
This study benchmarks multiple data-driven methodologies for predicting relative density (RD) of 316 L stainless steel fabricated via Powder Bed Fusion–Laser Beam (PBF-LB), as part of the ESAFORM Benchmark 2025 AMDmodel initiative. Two datasets (DS-01 and DS-02), each with 256 specimens from a 4-factor, 4-level design of experiments, were produced on different PBF-LB systems equipped with equivalent
in-situ
infrared (IR) melt-pool pyrometry. Failed builds (RD = 60%) were retained to allow models to learn from both nominal and catastrophic processing conditions, a scenario rarely addressed in PBF-LB machine learning (ML). Statistical analysis of variance (ANOVA) confirmed that conventional process parameters alone are weak predictors (R² ≈ 0.49). In contrast, sensor-driven supervised ML models using melt-pool thermal descriptors performed substantially better. Recursive feature elimination highlighted the interquartile range and mode of thermal signatures as dominant predictors; an XGBoost model using only these achieved R² = 0.93 on DS-01. Hybrid models combining parameters and IR descriptors performed slightly worse (R² = 0.92), indicating mild redundancy. Cross-system transferability was limited: ML models trained on DS-01 underperformed on DS-02 due to IR input-domain divergence despite RD distributions between both domain sources showing high inter-laboratory consistency. To address this, a physics-informed ML framework (PIML) using symbolic regression (QLattice) embedded dimensionless physical priors. Resulting compact expressions dominated by normalized laser power and volumetric energy density achieved R² = 0.83–0.93 under cross-system validation. Overall, sensor-driven ML models are effective for machine-specific monitoring and layer-wise closed-loop control, whereas PIML provide system-agnostic process parameter-window estimation for design-stage optimization.
Disciplines :
Science des matériaux & ingénierie
Auteur, co-auteur :
Monu, Medad Chiedozie C.
McCarthy, Eanna
Madathil, Abhilash Puthanveettil
Chekotu, Josiah C.
Ilic, Irina
Doğu, Merve Nur
Hughes, Cian
Juan, Rongfei
Amini, Ehsan
Lian, Junhe
Vassiades, Constantinos
Bylya, Olga
Darosa, Kim
Kromer, Robin
Seidou, Abdul Herrim ; Université de Liège - ULiège > Aérospatiale et Mécanique (A&M)
Mohanty, Sankhya
Habraken, Anne ; Université de Liège - ULiège > Département ArGEnCo > Département Argenco : Secteur MS2F
Mertens, Anne ; Université de Liège - ULiège > Aérospatiale et Mécanique (A&M)
ESAFORM Benchmark 2025: predicting stainless steel PBF-LB part density using statistical, data-driven, and physics-informed machine learning models derived from process parameters and in-situ monitoring data
Farrag A Yang Y Cao N et al. Physics-Informed Machine Learning for metal additive manufacturing Prog Addit Manuf 2025 10 171 185 10.1007/s40964-024-00612-1
Kapusuzoglu B Mahadevan S Physics-Informed and Hybrid Machine Learning in Additive Manufacturing: Application to Fused Filament Fabrication JOM 2020 72 4695 4705 10.1007/s11837-020-04438-4
Yang S Peng S Guo J Wang F A Review on Physics-Informed Machine Learning for Monitoring Metal Additive Manufacturing Process Adv Manuf 2024 1 1 2 10.55092/am20240008
Chen K Zhang P Yan H et al. A review of machine learning in additive manufacturing: design and process Int J Adv Manuf Technol 2024 135 1051 1087 10.1007/s00170-024-14543-2
Barrionuevo GO Ramos-Grez JA Walczak M Betancourt CA Comparative evaluation of supervised machine learning algorithms in the prediction of the relative density of 316L stainless steel fabricated by selective laser melting Int J Adv Manuf Technol 2021 113 419 433 10.1007/s00170-021-06596-4
Barrionuevo GO La Fé-Perdomo I Ramos-Grez JA Laser powder bed fusion dataset for relative density prediction of commercial metallic alloys Sci Data 2025 12 1 11 10.1038/s41597-025-04576-x
Gor M Dobriyal A Wankhede V et al. Density Prediction in Powder Bed Fusion Additive Manufacturing: Machine Learning-Based Techniques Appl Sci 2022 12 7271 10.3390/app12147271
Abdulla H, Maalouf M, Barsoum I, An H (2022) Truncated newton kernel ridge regression for prediction of porosity in additive manufactured SS316L. Appl Sci 12. https://doi.org/10.3390/app12094252
Eshkabilov S, Ara I, Azarmi F (2022) A comprehensive investigation on application of machine learning for optimization of process parameters of laser powder bed fusion – processed 316L stainless steel. Int J Adv Manuf Technol 2733–2756. https://doi.org/10.1007/s00170-022-10331-y
Wang J Jeong SG Kim ES et al. Material-agnostic machine learning approach enables high relative density in powder bed fusion products Nat Commun 2023 14 6557 10.1038/s41467-023-42319-x
Mahato V Chatterjee S Nyabadza A et al. Layer porosity in powder-bed fusion prediction using regression machine learning models and time-series features Int J AI Mater Des 2024 1 33 10.36922/ijamd.4812
Zhu Q Liu Z Yan J Machine learning for metal additive manufacturing: predicting temperature and melt pool fluid dynamics using physics-informed neural networks Comput Mech 2021 67 619 635 4222510 10.1007/s00466-020-01952-9
WuY, Sicard B (2024) A review of physics-informed machine learning methods with applications to condition monitoring and anomaly detection. arXivLabs. https://doi.org/10.48550/arXiv.2401.11860
Grose J, Liao A, Foong CS, Cullinan M (2023) Data-efficient surrogate model for rapid prediction of temperature evolution in a microscale selective laser sintering system. J Micro- Nano-Manufacturing 11. https://doi.org/10.1115/1.4064106
Peng B, Panesar A (2023) Predicting temperature field for mental additive manufacturing using PINN. Int Solid Free Fabr Symp 881–896 https://doi.org/10.26153/tsw/50997
Sajadi P, Dehaghani MR, Tang Y, Wang GG (2024) Real-time 2D temperature field prediction in metal additive manufacturing using physics-informed neural networks. arXiv org 1–42 https://doi.org/10.48550/arXiv.2401.02403
Jiang F Xia M Hu Y Physics-Informed Machine Learning for Accurate Prediction of Temperature and Melt Pool Dimension in Metal Additive Manufacturing 3D Print Addit Manuf 2024 11 e1679 e1689 10.1089/3dp.2022.0363
Habraken AM Aksen TA Alves JL et al. Analysis of ESAFORM 2021 cup drawing benchmark of an Al alloy, critical factors for accuracy and efficiency of FE simulations Int J Mater Form 2022 15 61 10.1007/s12289-022-01672-w
Agirre J Bernal D Flipon B et al. The ESAFORM benchmark 2023: interlaboratory comparison benchmark for the characterization of microstructural grain growth and dynamic recrystallization kinetics of a single-phase Ni-base superalloy Int J Mater Form 2025 18 33 10.1007/s12289-025-01893-9
Vanhulst M Lee Y Steinfels D et al. ESAFORM benchmark 2024: study on the geometric accuracy of a complex shape with single point incremental forming Int J Mater Form 2025 18 72 10.1007/s12289-025-01928-1
ASTM B962-23 (2023) Test methods for density of compacted or sintered powder metallurgy (PM) products using archimedes principle. ASTM B962-23 i:1–7
Carter LN Martin C Withers PJ Attallah MM The influence of the laser scan strategy on grain structure and cracking behaviour in SLM powder-bed fabricated nickel superalloy J Alloys Compd 2014 615 338 347 10.1016/j.jallcom.2014.06.172
Monu MCC Ekoi EJ Hughes C et al. Resultant physical properties of as-built nitinol processed at specific volumetric energy densities and correlation with in-situ melt pool temperatures J Mater Res Technol 2022 21 2757 2777 10.1016/j.jmrt.2022.10.073
Awad M Fraihat S Recursive Feature Elimination with Cross-Validation with Decision Tree: Feature Selection Method for Machine Learning-Based Intrusion Detection Systems J Sens Actuator Networks 2023 12 67 10.3390/jsan12050067
Fabian P Gael V Alexandre G et al. Scikit-learn: Machine Learning in Python J Mach Learn Res 2011 12 2825 2830 2854348
Buitinck L, Louppe G, Blondel M et al (2013) API design for machine learning software: experiences from the scikit-learn project. 1–15 https://doi.org/10.48550/arXiv.1309.0238
BÜYÜKKEÇECİ M OKUR MC A Comprehensive Review of Feature Selection and Feature Selection Stability in Machine Learning Gazi Univ J Sci 2023 36 1506 1520 10.35378/gujs.993763
Chicco D Warrens MJ Jurman G The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation PeerJ Comput Sci 2021 7 e623 10.7717/peerj-cs.623
Lundberg S Lee S-I A Unified Approach to Interpreting Model Predictions Adv Neural Inf Process Syst 2017 2017–Decem 4766 4775
Thomas M Baxter GJ Todd I Normalised Model-Based Processing Diagrams for Additive Layer Manufacture of Engineering Alloys Acta Mater 2016 108 26 35 10.1016/j.actamat.2016.02.025
Manikandan P Venkatesan K The Influence of Linear Energy Density on Density, Defect Formation, Residual Stress, Microstructure, and Texture in 310 Austenitic Stainless Steel by Laser Powder Bed Fusion J Manuf Process 2024 131 2191 2207 10.1016/j.jmapro.2024.10.022
Wang Z Zhang J Zhang F Qi C Impact of laser energy density on the structure and properties of laser-deposited Fe–Ni–Ti composite coatings Front Mater 2024 11 1408333 10.3389/fmats.2024.1408333
Buhairi MA Foudzi FM Jamhari FI et al. Review on volumetric energy density: influence on morphology and mechanical properties of Ti6Al4V manufactured via laser powder bed fusion Prog Addit Manuf 2023 8 265 283 10.1007/s40964-022-00328-0
Minervini I, Yang P, Higgins H, Ziade E (2023) Thermal conductivity of am fabricated stainless steel 316L Part. SANDIA-REPORT: 2023-00333. https://doi.org/10.2172/2432179
Demir P Kacar E Akman E Demir A Single Pulse Laser Ablation of AISI 316L Stainless Steel Surface Using Nd: YAG Laser Irradiation ACTA Phys Pol A 2014 125 439 441 10.12693/APhysPolA.125.439
Aldeia GSI Zhang H Bomarito G et al. Call for Action: towards the next generation of symbolic regression benchmark Proc Genet Evol Comput Conf Companion 2025 1 2529 2538 10.1145/3712255.3734309
Broløs KR, Machado MV, Cave C et al (2021) An approach to symbolic regression using Feyn. https://arXiv.org/abs/210405417
González-Estrada E Cosmes W Shapiro–Wilk test for skew normal distributions based on data transformations J Stat Comput Simul 2019 89 3258 3272 4008824 10.1080/00949655.2019.1658763
Haferkamp L Spierings A Rusch M et al. Effect of Particle size of monomodal 316L powder on powder layer density in powder bed fusion Prog Addit Manuf 2021 6 367 374 10.1007/s40964-020-00152-4
Choi J-P Shin G-H Lee H-S et al. Evaluation of Powder Layer Density for the Selective Laser Melting (SLM) Process Mater Trans 2017 58 294 297 10.2320/matertrans.M2016364
Huang G Chen H Ma Z et al. Microstructure and mechanical properties of stainless steel addictively manufactured via laser powder bed fusion in high-dense process parameter window Mater Sci Eng A 2025 928 148033 10.1016/j.msea.2025.148033
Jaskari M Hamada A Allam T et al. Effects of volumetric energy density on defect structure and fatigue behaviour of powder bed fusion manufactured 316L stainless steel Mater Sci Eng A 2025 925 147868 10.1016/j.msea.2025.147868
Tan HT Lopez E Selbmann A et al. Parameter analysis to correlate density with surface roughness and productivity in Powder Bed Fusion – Laser Beam (PBF-LB) of AlSi10Mg Procedia CIRP 2024 124 157 162 10.1016/j.procir.2024.08.090
Scipioni Bertoli U Wolfer AJ Matthews MJ et al. On the limitations of Volumetric Energy Density as a design parameter for Selective Laser Melting Mater Des 2017 113 331 340 10.1016/j.matdes.2016.10.037
Smith C Hommer G Keeler M et al. Assessing Volumetric Energy Density as a Predictor of Defects in Laser Powder Bed Fusion 316L Stainless Steel Jom 2025 77 737 748 10.1007/s11837-024-06946-z
Buhairi MA, Foudzi FM, Jamhari FI et al (2022) Review on volumetric energy density: influence on morphology and mechanical properties of Ti6Al4V manufactured via laser powder bed fusion. Prog Addit Manuf 2022 1–19. https://doi.org/10.1007/S40964-022-00328-0
Forien JB Calta NP DePond PJ et al. Detecting keyhole pore defects and monitoring process signatures during laser powder bed fusion: A correlation between in situ pyrometry and ex situ X-ray radiography Addit Manuf 2020 35 101336 10.1016/j.addma.2020.101336
Yadav P Rigo O Arvieu C et al. In Situ Monitoring Systems of The SLM Process: On the Need to Develop Machine Learning Models for Data Processing Crystals 2020 10 524 10.3390/cryst10060524
McCann R Obeidi MA Hughes C et al. In-situ sensing, process monitoring and machine control in Laser Powder Bed Fusion: A review Addit Manuf 2021 45 102058 10.1016/j.addma.2021.102058
Yu L Liu H Efficient Feature Selection via Analysis of Relevance and Redundancy J ofMachine Learn Res 2004 5 1205 1224 2248015
Manvatkar V De A DebRoy T Spatial variation of melt pool geometry, peak temperature and solidification parameters during laser assisted additive manufacturing process Mater Sci Technol (United Kingdom) 2015 31 924 930 10.1179/1743284714Y.0000000701
Mohr G, Altenburg SJ, Ulbricht A et al (2020) In-situ defect detection in laser powder bed fusion by using thermography and optical tomography—comparison to computed tomography. Met (Basel) 10. https://doi.org/10.3390/met10010103
Yan Z Liu W Tang Z et al. Review on thermal analysis in laser-based additive manufacturing Opt Laser Technol 2018 106 427 441 10.1016/j.optlastec.2018.04.034
Riensche A Bevans BD Smoqi Z et al. Feedforward control of thermal history in laser powder bed fusion: Toward physics-based optimization of processing parameters Mater Des 2022 224 111351 10.1016/j.matdes.2022.111351
Hu L Li Y Luo G et al. Constructing solidification microstructure selection map for AlSi10Mg alloy processed by laser powder bed fusion J Alloys Compd 2024 976 173308 10.1016/j.jallcom.2023.173308
Huang Y Fleming TG Clark SJ et al. Keyhole fluctuation and pore formation mechanisms during laser powder bed fusion additive manufacturing Nat Commun 2022 13 1 11 10.1038/s41467-022-28694-x
Forien J, Calta N, DePond P, Guss G, Roehling T (2019) MM Detecting keyhole porosity and monitoring process signatures in additive manufacturing: an in situ pyrometry and ex situ X-ray radiography correlation. 1–3. https://doi.org/LLNL-JRNL-782115
Kotadia HR Gibbons G Das A Howes PD A review of Laser Powder Bed Fusion Additive Manufacturing of aluminium alloys: Microstructure and properties Addit Manuf 2021 46 102155 10.1016/j.addma.2021.102155
Kurz W, Fisher DJ, Rappaz M (2023) Fundamentals of Solidification, 5th edn. Trans Tech Publications Ltd https://digital.casalini.it/9783036410159
Santos Macías JG Douillard T Zhao L et al. Influence on microstructure, strength and ductility of build platform temperature during laser powder bed fusion of AlSi10Mg Acta Mater 2020 201 231 243 10.1016/j.actamat.2020.10.001
Khorasani M Ghasemi AH Leary M et al. A comprehensive study on meltpool depth in laser-based powder bed fusion of Inconel 718 Int J Adv Manuf Technol 2022 120 2345 2362 10.1007/s00170-021-08618-7
Leung CLA Marussi S Atwood RC et al. In situ X-ray imaging of defect and molten pool dynamics in laser additive manufacturing Nat Commun 2018 9 1355 10.1038/s41467-018-03734-7
Scipioni Bertoli U Guss G Wu S et al. In-situ characterization of laser-powder interaction and cooling rates through high-speed imaging of powder bed fusion additive manufacturing Mater Des 2017 135 385 396 10.1016/j.matdes.2017.09.044
Abrami MB Tocci M Obeidi MA et al. Prediction of Microstructure for AISI316L Steel from Numerical Simulation of Laser Powder Bed Fusion Met Mater Int 2022 28 2735 2746 10.1007/s12540-022-01168-x
Mohr G Altenburg SJ Hilgenberg K Effects of inter layer time and build height on resulting properties of 316L stainless steel processed by laser powder bed fusion Addit Manuf 2020 32 101080 10.1016/j.addma.2020.101080
Vallabh CKP, Hinnebusch S, To AC, Zhao X (2022) Layer-wise melt pool temperature evolution in laser powder bed fusion: an experimental study using a single camera based two-wavelength imaging pyrometry, vol 2. Manufacturing processes; manufacturing systems. American Society of Mechanical Engineers https://doi.org/10.1115/MSEC2022-85387
Krishna C, Vallabh P, Zhang H et al (2024) Melt pool width measurement in a multi – track, multi – layer laser powder bed fusion print using single – camera two – wavelength imaging pyrometry. Int J Adv Manuf Technol 2575–2585. https://doi.org/10.1007/s00170-024-13486-y
Kusano M Takata Y Yumoto A Watanabe M Effects of time per layer and part geometry on thermal history and microcracking in the fabrication of nickel superalloy samples by laser powder bed fusion Addit Manuf 2024 80 103987 10.1016/j.addma.2024.103987
Medvedev AE Brudler S Piegert S et al. Interlayer time as a robust, geometry-agnostic predictor of microstructural and mechanical properties evolution in PBF-LB/M Ti6Al4V alloy J Mater Process Technol 2025 340 118858 10.1016/j.jmatprotec.2025.118858
Zhang W Tong M Harrison NM Multipart Build Effects on Temperature and Residual Stress by Laser Beam Powder Bed Fusion Additive Manufacturing 3D Print Addit Manuf 2023 10 749 761 10.1089/3dp.2021.0143
Karson M (1968) Handbook of methods of applied statistics. Volume I: techniques of computation descriptive methods, and statistical inference. Volume II: Planning of surveys and experiments. I. M. Chakravarti, R. G. Laha, and J. Roy, New York, John Wiley; 1967, $9.00. J Am Stat Assoc 63:1047–1049. https://doi.org/10.1080/01621459.1968.11009335
Calixto E Calixto E Lifetime Data Analysis Gas and Oil Reliability Engineering, Second Edi 2016 Boston Elsevier 1 92
Vegas Sánchez-Ferrero G, Alberola-López C (2024) Estimation and inference. In: Frangi AF, Prince JL, Sonka M (eds) Medical Image Analysis. Elsevier, pp 85–111 https://doi.org/10.1016/B978-0-12-813657-7.00016-9
Ramdas A Trillos N Cuturi M On Wasserstein Two-Sample Testing and Related Families of Nonparametric Tests Entropy 2017 19 47 3608466 10.3390/e19020047
Panaretos VM Zemel Y Statistical Aspects of Wasserstein Distances Annu Rev Stat Its Appl 2019 6 405 431 3939527 10.1146/annurev-statistics-030718-104938
Sun B, Feng J, Saenko K (2016) Return of frustratingly easy domain adaptation. Proc AAAI. Conf Artif Intell 30. https://doi.org/10.1609/aaai.v30i1.10306
Rahman MM Fookes C Baktashmotlagh M Sridharan S Correlation-aware adversarial domain adaptation and generalization Pattern Recognit 2020 100 107124 10.1016/j.patcog.2019.107124
