Bayesian Optimization-Based Parameter Identification for Discrete Element Method Simulation of Consolidation and Its Application to Powder Spreading Analysis
Round 1
Reviewer 1 Report
Comments and Suggestions for Authors- Originality and Scientific Contribution
This manuscript presents a Bayesian optimization-based framework for parameter identification in discrete element method (DEM) simulations of powder consolidation, followed by its transferability assessment to powder spreading. The study combines data-driven optimization with physically grounded modeling, which represents an innovative and timely contribution to additive manufacturing research, particularly in Laser Powder Bed Fusion (PBF-LM). The originality lies in integrating Bayesian optimization as an adaptive calibration tool to efficiently identify DEM parameters while maintaining physical consistency. The work provides meaningful insights into the limits of parameter transfer between distinct powder-handling processes.
- Methodology and Research Design
The methodological design is clear, systematic, and technically rigorous. The two-parameter optimization (surface energy density k and rolling resistance coefficient μr) is well motivated, and the experimental reference (FT4 powder rheometer) is appropriate. The application of the GPyOpt library for Bayesian optimization is well justified. However, further clarification would strengthen the study—specifically, how hyperparameters such as the acquisition weight (wa) were tuned, and how robust the optimization is against initial sampling or noise. Including a brief sensitivity or convergence analysis would enhance methodological transparency. Moreover, a short comparative note on other optimization strategies (e.g., genetic algorithms, Latin hypercube sampling, gradient-based methods) would situate the work within a broader computational context.
- Results and Interpretation
The results are logically structured and supported by clear visualizations (parameter maps, consolidation snapshots, and powder-spreading simulations). The optimization successfully identified near-optimal parameters within 12 trials, demonstrating efficiency. Importantly, the finding that parameters calibrated for consolidation are not directly transferable to spreading is physically well interpreted — the transition from normal-force-dominated to tangential-sliding-dominated mechanics is convincingly discussed. Nevertheless, the discussion would benefit from deeper scientific reflection and comparison with related literature which address DEM calibration and powder-bed behavior in PBF-LM. A quantitative or conceptual comparison to these studies would significantly strengthen the contextual interpretation and highlight the advancement of the present work.
- Formal and Language Aspects
The manuscript is well written, technically sound, and adheres to the journal’s formal standards. Figures and tables are of high quality, and terminology is consistent throughout. Some methodological sections are overly descriptive; these could be condensed or moved to supplementary material to improve readability. The graphical illustrations could include scale bars and more consistent referencing to figure numbers and equations. The English language quality is very good and only requires minor stylistic adjustments for conciseness.
- Recommendation to the Editor
The manuscript presents a well-conceived and competently executed study with significant methodological and practical value for the simulation-based analysis of additive manufacturing processes. The work deserves publication after minor revision.
Recommendation: Minor Revision
Suggested Revisions:
- Expand the discussion by comparing the findings with other recent studies on DEM parameter calibration in PBF-LM.
- Clarify the tuning procedure and robustness of the Bayesian optimization framework (hyperparameters, acquisition function, and sensitivity).
- Shorten some descriptive methodological sections and emphasize the interpretation and implications of the results.
Once these revisions are made, the paper will be well suited for publication in JMMP.
Author Response
We appreciated for reviewing our manuscript. Taking into account for the comments from the reviewer1, we have modified our manuscript. The modifications made are listed in the enclosed reviewer response documents and are highlighted by YELLOW in the revised manuscript. We hope that our modifications will meet the acceptance criteria of your journal.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsDear Authors
The manuscript addresses an important and relevant topic and has the potential to make a meaningful contribution to the field. However, the validation of the powder spreading stage and the methodological justification of scaling require further clarification to fully support the conclusions.
Abstract: I suggest rewriting the summary... here's a suggestion: “This study develops a Bayesian optimization framework to calibrate two discrete element method (DEM) parameters—the cohesion-related surface energy coefficient (𝑘) and the rolling resistance coefficient (𝜇𝑟)—based on experimental void ratio data obtained from powder consolidation tests. The optimized parameter set reproduces the experimental consolidation curve with a relative error below X%, demonstrating an efficient and physically plausible calibration under confined loading. However, when these parameters are applied to powder spreading simulations, the resulting powder beds become excessively cohesive, leading to poor layer uniformity. This discrepancy is attributed to (i) the mismatch in mechanical similarity (σ/E) between the experimental and simulated conditions, and (ii) the shift in dominant particle-scale mechanics from normal-force-controlled consolidation to shear-dominated spreading. The results indicate that DEM parameter calibration for powder-bed-based additive manufacturing should incorporate shear-related experimental metrics and scaling considerations rather than rely solely on consolidation-based fitting.”
Introduction: The reference list could be expanded to include recent DEM calibration studies involving multi-mechanism fitting.
Materials and methods: The statement “Data not available” limits reproducibility. Even if experimental data cannot be released, providing optimization scripts and baseline parameter files would enhance transparency. Since the optimized parameter space is relatively narrow and based on a limited number of trials, the robustness of the identified parameters should be demonstrated by analyzing how small variations in 𝑘 and 𝜇𝑟 impact the predicted behavior.
Results: Explicitly reporting the deviation metric used for model comparison (Figure 5). The comparison presented in Figures 6 and 7 is predominantly qualitative. The study would be significantly strengthened by including measurable performance indicators, such as local bulk density, packing fraction distribution, or surface roughness of the powder bed layer.
Discussion: The manuscript mentions the σ/E scaling issue, but does not provide a systematic procedure to ensure physical similarity between experiments and simulations. A more explicit dimensional similarity framework or scaling guideline would improve the rigor of Section 4.1. The manuscript identifies that consolidation is dominated by normal contact forces, whereas spreading is dominated by shear-induced sliding. This implies that DEM calibration must include shear-based experimental metrics. The authors should explicitly state this and discuss how such experiments could be incorporated in future calibration.
Conclusion: The conclusion should explicitly state the main contribution of the work. It is suggested to briefly indicate how the results may guide researchers or engineers in the use of DEM for PBF-LM. The conclusion can be strengthened by clearly stating that future calibrations should include experimentally obtained shear-related metrics.
Comments on the Quality of English LanguageA light English proofreading is recommended to improve fluency and reduce redundant phrasing.
Author Response
We appreciated for reviewing our manuscript. Considering the comments from the reviewer2, we have modified our manuscript. The modifications made are listed in the enclosed reviewer response documents and are highlighted by YELLOW in the revised manuscript. We hope that our modifications will meet the acceptance criteria of your journal.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsDear Authors,
The revised version has been substantially improved.
In my opinion, the manuscript can now be accepted in its current form.
Comments on the Quality of English LanguageThe English could be improved to more clearly express the research.
