Localized Defect Detection from Spatially Mapped, In-Situ Process Data With Machine Learning

William Halsey,Luke Scime,Derek Rose,Vincent Paquit,Ryan Dehoff

doi:10.3389/fmech.2021.767444

Abstract

In powder bed fusion additive manufacturing, machines are often equipped with in-situ sensors to monitor the build environment as well as machine actuators and subsystems. The data from these sensors offer rich information about the consistency of the fabrication process within a build and across builds. This information may be used for process monitoring and defect detection; however, little has been done to leverage this data from the machines for more than just coarse-grained process monitoring. In this work we demonstrate how these inherently temporal data may be mapped spatially by leveraging scan path information. We then train a XGBoost machine learning model to predict localized defects—specifically soot–using only the mapped process data of builds from a laser powder bed fusion process as input features. The XGBoost model offers a feature importance metric that will help to elucidate possible relationships between the process data and observed defects. Finally, we analyze the model performance spatially and rationalize areas of greater and lesser performance.

Highlights

Additive manufacturing (AM) is a set of exciting and promising processing methods that are expected to continue to revolutionize manufacturing
We focus on a ConceptLaser M2 Laser Powder Bed Fusion (L-powder bed fusion (PBF)) machine instead of Arcam EBM or directed energy deposition (DED) sensor data
One can even imagine leveraging this method to mine for relationships between other localized properties such as microstructure and mechanical properties by utilizing the feature importance scores produced by XGBoost to discover relationships between the inputs and targets to investigate further with targeted experiments. This analysis represents a first attempt at leveraging in-situ process data for predicting localized build defects

Summary

Introduction

Additive manufacturing (AM) is a set of exciting and promising processing methods that are expected to continue to revolutionize manufacturing. The inability to gain traction comes from the poor repeatability of the process from one build to the stemming from the prevalence of seemingly stochastic defects (Grasso and Colosimo, 2017) and anisotropic microstructure (Kok et al, 2018) yielding inconsistent and heterogeneous properties throughout a component and across nominal duplicates of a component. This is true for all metal AM technologies including the more mature subdomain of powder bed fusion (PBF).

Methods

Results

Discussion

Conclusion