| About this Abstract |
| Meeting |
2023 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2023)
|
| Symposium
|
2023 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2023)
|
| Presentation Title |
Localized Porosity Prediction in Laser Powder Bed Fusion via Deep Learning of Multi-modal Melt Pool Signatures |
| Author(s) |
Haolin Zhang, Chaitanya Krishna Prasad Vallabh, Alexander Caputo, Richard Neu, Xiayun Zhao |
| On-Site Speaker (Planned) |
Haolin Zhang |
| Abstract Scope |
Laser powder bed fusion (LPBF) additive manufacturing utilizes laser to sinter or melt powders for fast production of complex parts. However, due to the complex interplays among laser, powder, printed part, and gas flow, the LPBF process tends to generate severe defects such as pores, which are detrimental to the final part performance. In this work, we develop a deep learning aided porosity prediction framework utilizing in-situ monitored melt pool signatures including multiple thermal, geometrical, and spatter metrics that are derived from our high-speed on-axis single-camera two-wavelength imaging pyrometer and an off-axis camera jointly. Scalograms, transformed from the obtained time-domain MP signatures are used as input to train deep convolutional neural network models for correlating to ex-situ porosity characterization data from X-ray computed tomography. The developed method is shown to be capable of quantifying localized porosity and holds promise to qualify LPBF processes and parts. |
| Proceedings Inclusion? |
Definite: Post-meeting proceedings |