Session: 04-01 Emerging Technologies: Smart AM

Paper Number: 100301

100301 - Engineering-Informed Machine Learning for Additive Manufacturing Accuracy Control 

Rapid technology advances in additive manufacturing (AM) have been significantly expanding manufacturing capability and utility.  One key technological challenge is how to ensure AM product quality and reduce the number of iterative runs. Quality control (QC) in AM currently relies heavily on human expertise and intervention due to product and process complexities. Physical modeling and simulation of AM is still computationally prohibitive for timely QC. Applying popular AI techniques to automate QC not only demands large amounts of costly AM data, but also falls short of gaining engineering insights for knowledge generalization and adaptation. This talk presents domain-informed convolution modeling and learning of layer-by-layer fabrication for shape accuracy prediction for a wide range of 3D products including both smooth and non-smooth surfaces (polyhedral shapes). The key idea of learning heterogeneous deviation surface data under a unified model is to establish the association between the deviation profiles of smooth base shapes and those of non-smooth polyhedral shapes. The association, which is characterized by a novel 3D cookie-cutter function, views polyhedral shapes as being carved out from smooth base shapes. In essence, the AM process of building non-smooth shapes is mathematically decomposed into two steps: additively fabricate smooth base shapes using a convolution framework, and then subtract extra materials using a cookie-cutter function. The proposed joint learning framework of shape deviation data reflects this decomposition by adopting a sequential model estimation procedure. The model learning procedure first establishes the convolution model to capture the effects of layer-wise fabrication and sizes, and then estimates the 3D cookie-cutter function to realize geometric differences between smooth and non-smooth shapes. A new Gaussian process model is proposed to consider the spatial correlation among neighboring regions within a 3D shape and across different shapes. The case study demonstrates the feasibility and prospects of prescriptive learning of complex 3D shape deviations in AM and extension to broader engineering surface data.

Presenting Author: Qiang Huang University of Southern California

Presenting Author Biography: Dr. Qiang Huang is currently a Professor at the Daniel J. Epstein Department of Industrial and Systems Engineering, University of Southern California (USC), Los Angeles. His research focuses on AI and Machine Learning for Manufacturing, in particular, Machine Learning for Additive Manufacturing (ML4AM). He was the holder of Gordon S. Marshall Early Career Chair in Engineering at USC from 2012 to 2016. He received IISE Fellow Award, NSF CAREER award, and 2021 IEEE CASE Best Conference Paper Award, 2013 IEEE Transactions on Automation Science and Engineering Best Paper Award, among others. He has five patents on ML4AM. He is a Department Editor for IISE Transactions and an Associate Editor for ASME Transactions, Journal of Manufacturing Science and Engineering.

Authors: