TY - GEN
T1 - Machine learning based Melt Pool Spattering registration and defectcorrelation for LPBF additive manufacturing
AU - Zhang, Haolin
AU - Vallabh, Chaitanya Krishna Prasad
AU - Zhao, Xiayun
N1 - Publisher Copyright:
Copyright© (2022) by American Society for Precision Engineering (ASPE) All rights reserved.
PY - 2022
Y1 - 2022
N2 - Laser powder bed fusion (LPBF) additive manufacturing (AM) process has attracted increasing research and application interest in recent years, due to its potential ability to efficiently produce parts with complicated geometry. Despite the rapid development, LPBF manufactured products still lack the quality required for practical implementation in most industries. In LPBF printing processes, some major defects can be attributed to the melt pool (MP) induced spatters which would cause porosity, powder contamination, and recoater intervention. In this work, we develop a machine learning (ML) based method for automatically and accurately registering MP spattering metrics using in-situ off-axis camera monitoring data. Specifically, each monitored MP image is analyzed for extracting its associated spattering signatures including the MP's center location (coordinates), spatter count, and ejection angle relative to its MP center. Such MP spattering signatures are registered for each monitored MP across each print layer. To evaluate the potential influence of MP spattering on surface defects, a preliminary correlation of each print layer's registered MP spattering signatures to the layer's surface topography measured by a lab-designed in-situ Fringe Projection Profilometry (FPP) system is performed via ML. Overall, this work presents a systematic study on MP spattering monitoring, registration, and correlation to reveal effect on defect formation during LPBF, allowing for online qualification of LPBF-AM processes.
AB - Laser powder bed fusion (LPBF) additive manufacturing (AM) process has attracted increasing research and application interest in recent years, due to its potential ability to efficiently produce parts with complicated geometry. Despite the rapid development, LPBF manufactured products still lack the quality required for practical implementation in most industries. In LPBF printing processes, some major defects can be attributed to the melt pool (MP) induced spatters which would cause porosity, powder contamination, and recoater intervention. In this work, we develop a machine learning (ML) based method for automatically and accurately registering MP spattering metrics using in-situ off-axis camera monitoring data. Specifically, each monitored MP image is analyzed for extracting its associated spattering signatures including the MP's center location (coordinates), spatter count, and ejection angle relative to its MP center. Such MP spattering signatures are registered for each monitored MP across each print layer. To evaluate the potential influence of MP spattering on surface defects, a preliminary correlation of each print layer's registered MP spattering signatures to the layer's surface topography measured by a lab-designed in-situ Fringe Projection Profilometry (FPP) system is performed via ML. Overall, this work presents a systematic study on MP spattering monitoring, registration, and correlation to reveal effect on defect formation during LPBF, allowing for online qualification of LPBF-AM processes.
UR - http://www.scopus.com/inward/record.url?scp=85139776964&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85139776964&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:85139776964
T3 - 2022 ASPE and euspen Summer Topical Meeting on Advancing Precision in Additive Manufacturing
SP - 146
EP - 151
BT - 2022 ASPE and euspen Summer Topical Meeting on Advancing Precision in Additive Manufacturing
T2 - 2022 ASPE and euspen Summer Topical Meeting on Advancing Precision in Additive Manufacturing
Y2 - 11 July 2022 through 14 July 2022
ER -