Laser powder bed fusion of multilayer thin-walled structures based on data-driven model

An Chen Lee, Ruei Yu Huang, Trong Doan Nguyen, Chung Wei Cheng, Mi Ching Tsai

Research output: Contribution to journalArticlepeer-review

Abstract

The laser powder bed fusion (LPBF) process has the advantage of directly building metal parts with complex geometries and lattice structures. The lattice structures are usually composed of different thin-walled structures. Therefore, precise control over the shape of structures on the LPBF fabricated lattice structures is important. In this study, a symbolic regression solution that describes the relationship between scan track width and process parameters (laser power and scanning velocity) is generated by empirical data. This regression model is then implemented into the control scheme that stabilizes the structure width for single-layer and multilayer thin-walled structures. The experiment results showed that the average errors for the single-layer track with desired width from 120 μm to 180 μm are all under 5%, and the average error for the ten-layered structures with desired width 180 μm is about 2.2%.

Original languageEnglish
Pages (from-to)38-44
Number of pages7
JournalJournal of Laser Micro Nanoengineering
Volume15
Issue number1
DOIs
StatePublished - 1 Jun 2020

Keywords

  • Data-driven model
  • Selective laser melting
  • SLM
  • Symbolic regressed solution

Fingerprint Dive into the research topics of 'Laser powder bed fusion of multilayer thin-walled structures based on data-driven model'. Together they form a unique fingerprint.

Cite this