Sensitivity of L-DED melt-pool descriptors to point-cloud representation and gradient reconstruction

Laser directed energy deposition (L-DED) simulations often export melt-pool fields as unstructured point clouds. These fields contain rich transport information, but they do not automatically provide reproducible evidence across process conditions. Coordinate duplication, irregular sampling, local gradient reconstruction, threshold selection, and region masking can each alter the apparent flow topology. We analysed 30 exports spanning six powers (200-450 W) and five late-time snapshots to determine which thermal and flow descriptors were preserved across coordinate aggregation, time, and weighted least-squares design. Direct temperature, temperature-gradient, and velocity descriptors were separated from reconstructed Q, λ2, and ΩN fields. At 0.70 s, the sampled powers showed different thermal and velocity patterns, but no power met all predefined temporal criteria. The 350 W maximum-velocity contrast arose from three coordinates within central distributions that overlapped the 400 W distribution. Equal-weight reconstruction recovered affine manufactured fields to numerical precision, whereas nonlinear recovery remained scale- and weighting-dependent. The reconstructed tensor descriptors therefore yielded no comparative topology conclusion. These findings show that point-cloud representation and gradient reconstruction materially constrain the interpretation of solver-exported melt-pool fields; physical fidelity against experiment, a continuous power response, and a physical mechanism remain outside the scope of the present data.

Authors

Institutions

Publication Details

Journal
PLoS ONE
Published
2026-11-20
DOI
https://doi.org/10.1371/journal.pone.0362947
Primary Topic
Additive Manufacturing Materials and Processes
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Sensitivity of L-DED melt-pool descriptors to point-cloud representation and gradient reconstruction

Xiaoli Lin, Zhiyong Yu, Boxue Song, Hanghuang Xu
2 links
PLoS ONE
Additive Manufacturing Materials and Processes
article

Sensitivity of L-DED melt-pool descriptors to point-cloud representation and gradient reconstruction

Xiaoli Lin, Zhiyong Yu, Boxue Song, Hanghuang Xu
article en

Abstract

Laser directed energy deposition (L-DED) simulations often export melt-pool fields as unstructured point clouds. These fields contain rich transport information, but they do not automatically provide reproducible evidence across process conditions. Coordinate duplication, irregular sampling, local gradient reconstruction, threshold selection, and region masking can each alter the apparent flow topology. We analysed 30 exports spanning six powers (200-450 W) and five late-time snapshots to determine which thermal and flow descriptors were preserved across coordinate aggregation, time, and weighted least-squares design. Direct temperature, temperature-gradient, and velocity descriptors were separated from reconstructed Q, λ2, and ΩN fields. At 0.70 s, the sampled powers showed different thermal and velocity patterns, but no power met all predefined temporal criteria. The 350 W maximum-velocity contrast arose from three coordinates within central distributions that overlapped the 400 W distribution. Equal-weight reconstruction recovered affine manufactured fields to numerical precision, whereas nonlinear recovery remained scale- and weighting-dependent. The reconstructed tensor descriptors therefore yielded no comparative topology conclusion. These findings show that point-cloud representation and gradient reconstruction materially constrain the interpretation of solver-exported melt-pool fields; physical fidelity against experiment, a continuous power response, and a physical mechanism remain outside the scope of the present data.

PLoS ONEVol. 21(11)
Putian University (CN)
Putian University
Additive Manufacturing Materials and Processes
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon
Sensitivity of L-DED melt-pool descriptors to point-cloud representation and gradient reconstruction — Xiaoli Lin, Zhiyong Yu, et al. · PLoS ONE (2026) | TGRS Research Map | TGRS