Page 39 - IJAMD-1-3

International Journal of AI for
Materials and Design

ORIGINAL RESEARCH ARTICLE
Layer porosity in powder-bed fusion prediction

using regression machine learning models and
time-series features

Vivek Mahato 1,2,3,4 , Suman Chatterjee 1,2,4 *, Anesu Nyabadza 1,2,4 ,
Annalina Caputo 1,3,4 , and Dermot Brabazon 1,2,4 *
1 I-Form Advanced Manufacturing Research Centre, Dublin City University, Dublin, Ireland
2 School of Mechanical and Manufacturing Engineering, Dublin City University, Dublin, Ireland
3 School of Computing, Dublin City University, Dublin, Ireland
4 Advanced Processing Technology Research Centre, Dublin City University, Dublin, Ireland
(This article belongs to the Special Issue: AI Usage in the Analysis of the Additive Manufacturing
Process)

Abstract

Additive manufacturing (AM) using laser powder-bed fusion (L-PBF) has become
a common industrial process for high-end component production. The uptake
of the process has been accelerated through the broad acceptance of the L-PBF
process toward achieving high-quality parts with complex geometry. However, the
*Corresponding authors: L-PBF process faces challenges from the process’s sensitivity to the process build
Dermot Brabazon parameters, which, when incorrectly set, can cause defects such as porosity, which in
(dermot.brabazon@dcu.ie)
Suman Chatterjee turn have a detrimental effect on the produced part properties. On the other hand,
(suman.chatterjee@dcu.ie) the AM processing equipment generates a vast amount of data captured through
in situ sensors such as pyrometers and imaging cameras. Having such an abundance
Citation: Mahato V, Chatterjee S,
Nyabadza A, Caputo A, of process data facilitates the employment of advanced machine learning (ML) tools
Brabazon D. Layer porosity in to understand and extract patterns and information about the underlying AM process
powder-bed fusion prediction using and gain “predictive control.” Driven by this idea, we aimed to employ ML tools over
regression machine learning models
and time-series features. Int J AI pyrometer time-series data from an L-PBF process to predict the porosity percentage of
Mater Design. 2024;1(3): 33-49. layers of an AM-built part. Sensor data are naturally modeled by time series; however,
doi: 10.36922/ijamd.4812 most ML algorithms work with tabular data (i.e., one single vector describes a feature).
Received: September 10, 2024 In the work presented here, feature engineering tools were used to transform the
Accepted: November 13, 2024 time-series data into informative features. These features were fed into the tabular ML
algorithms for evaluation, broadening the selection of ML algorithms available in the
Published Online: December 16, 2024 literature. It was hypothesized that the time-series summary features would capture
Copyright: © 2024 Author(s). the interaction of melt-pool temperature with resulting porosity, from which the
This is an Open-Access article resulting models could better predict porosity occurrence. The dataset contains layer
distributed under the terms of the
Creative Commons Attribution porosity values in the range of 0.00175 – 7.160%, to which we divide the data into
License, permitting distribution, “low” and “high” porous layers using a splitting threshold value of 1%. From evaluating
and reproduction in any medium, these algorithms, it was concluded that classifying “low” versus “high” porosity layers is
provided the original work is
properly cited. relatively easier than predicting the layer’s porosity percentage.
Publisher’s Note: AccScience
Publishing remains neutral with Keywords: Additive manufacturing; Powder-bed fusion laser-beam; Machine learning;
regard to jurisdictional claims in
published maps and institutional TS-Fresh
affiliations.

Volume 1 Issue 3 (2024) 33 doi: 10.36922/ijamd.4812