Using Self-Organizing Maps for In situ Monitoring of Melt Pool Thermal Profiles for Porosity Prediction in Laser Based Additive Manufacturing Processes

Abstract

The objective of this technical note is to use unsupervised machine learning to characterize the underlying thermophysical dynamics of laser-based additive manufacturing (LBAM) captured by melt pool signals to predict porosity during the build. Herein, a novel porosity detection method is proposed based on morphological features and the temperature distribution of the top surface of the melt pool as the LBAM part is being built. Self-organizing maps (SOMs) are then used to further analyze the2D melt pool dataset to identify similar and dissimilar melt pools. The significance of the proposed methodology based on melt pool profile is that this may lead the way toward in-situ monitoring to minimize or eliminate pores within LBAM parts.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jun 01, 2019
Accession Number
AD1078016

Entities

People

  • Linkan Bian
  • Mark Tschopp
  • Mohammad Marufuzzaman
  • Mojtaba Khanzadeh
  • Sudipta Chowdhury

Organizations

  • United States Army Research Laboratory

Tags

Communities of Interest

  • Autonomy

DTIC Thesaurus Topics

  • Abstracts
  • Additive Manufacturing
  • Additives (Chemicals)
  • Data Acquisition
  • Data Analysis
  • Fabrication
  • Fused Deposition Modeling
  • Learning
  • Machine Learning
  • Manufacturing
  • Melting Point
  • Military Research
  • Monitoring
  • Porosity
  • Systems Engineering
  • Two Dimensional
  • Unsupervised Machine Learning

Readers

  • Nanocomposite Materials Science
  • Naval Personnel Management
  • Polymer Science and Engineering.

Technology Areas

  • AI & ML
  • Directed Energy