Abstract:Laser powder bed fusion (PBF-LB) is a promising metal additive manufacturing technology. However, ensuring consistent quality remains a pressing challenge due to defects that arise during the PBF-LB process. These defects significantly impede subsequent molding procedures, emphasizing the urgent need for real-time monitoring technology to steer part production effectively. Present monitoring techniques are limited, typically addressing specific defect types or contour deformations individually and lacking a simple and effective method to achieve multi-scale defects detection for PBF-LB parts. This study aims to pioneer a method that can detect defects and part contour deformations during the PBF-LB process using only an industrial camera as the monitoring device. The proposed method consists of two key components: a semantic segmentation model based on deep transfer learning (SS-DTF model) and a contour deformation detection module. To begin with, an off-axis in-situ monitoring system is deployed with an industrial camera to capture powder bed images for each layer after laser scanning and powder spreading. This enables the production of datasets for post-scanning and post-spreading anomalies. The SS-DTF model is trained on the above two datasets respectively to achieve segmentation for multi-category defects. The contour deformation detection module provides automatic early warning by comparing the true contour of the part with the nominal contour layer-by-layer, using statistical process monitoring method. The proposed method is then used to monitor parts with various geometries under real processing conditions, aiming to clarify the potential correlation between monitoring outcomes and the parts' multi-scale defects in PBF-LB. The results show that the proposed method provides a promising solution for in-situ monitoring of the PBF-LB process.

Qualitative and quantitative characterization of powder bed quality in laser powder-bed fusion additive manufacturing by multi-task learning

Powder Bed Defect Extraction of Laser Powder Bed Fusion Additive Manufacturing with Tensor Robust Principal Component Analysis

Low-rank Adaptive Transfer Learning Based for Multi-Label Defect Detection in Laser Powder Bed Fusion

Online Quality Monitoring in Material Extrusion Additive Manufacturing Processes Based on Laser Scanning Technology

Deepening the scientific understanding of different phenomenology in laser powder bed fusion by an integrated framework

Predicting Laser Powder Bed Fusion Defects Through In-Process Monitoring Data and Machine Learning

Research on an Online Monitoring Device for the Powder Laying Process of Laser Powder Bed Fusion

A Novel Approach of Online Monitoring for Laser Powder Bed Fusion Defects: Air-borne Acoustic Emission and Deep Transfer Learning

Multi-scale feature pyramid approach for melt track classification in laser powder bed fusion via coaxial high-speed imaging

Layer-wise multi-defect detection for laser powder bed fusion using deep learning algorithm with visual explanation

In-situ visual monitoring for multi-scale defects detection in laser powder bed fusion

Uncertainty-driven Trustworthy Defect Detection for High-Resolution Powder Bed Images in Selective Laser Melting

A deep learning-based model for defect detection in laser-powder bed fusion using in-situ thermographic monitoring

A hybrid deep learning model of process-build interactions in additive manufacturing

A lightweight high-resolution algorithm based on deep learning for layer-wise defect detection in laser powder bed fusion

Quality Monitoring in Additive Manufacturing Using Emission Spectroscopy and Unsupervised Deep Learning

Efficient and Lightweight Layer-Wise In-Situ Defect Detection in Laser Powder Bed Fusion Via Knowledge Distillation and Structural Re-Parameterization

Transfer Learning-Based Quality Monitoring of Laser Powder Bed Fusion Across Materials

Process modeling in laser powder bed fusion towards defect detection and quality control via machine learning: The state-of-the-art and research challenges

A machine learning guided investigation of quality repeatability in metal laser powder bed fusion additive manufacturing

Multi-source Information Fusion for Enhanced In-Process Quality Monitoring of Laser Powder Bed Fusion Additive Manufacturing