Abstract:Early detection of defects, such as keyhole pores and cracks is crucial in laser-directed energy deposition (L-DED) additive manufacturing (AM) to prevent build failures. However, the complex melt pool behaviour cannot be adequately captured by conventional single-modal process monitoring approaches. This study introduces a multisensor fusion-based digital twin (MFDT) for localized quality prediction in the robotic L-DED process. The data used in multisensor fusion includes features extracted from a coaxial melt pool vision camera, a microphone, and an off-axis short wavelength infrared thermal camera. The key novelty of this work is a spatiotemporal data fusion method that synchronizes multisensor features with the real-time robot motion data to achieve localized quality prediction. Optical microscope (OM) images of the printed part are used to locate defect-free and defective regions (i.e., cracks and keyhole pores), which serve as ground truth labels for training supervised machine learning (ML) models for quality prediction. The trained ML model is then used to generate a virtual quality map that registers quality prediction outcomes within the 3D volume of the printed part, thus eliminating the need of physical inspections by destructive methods. Experiments show that the virtual quality map closely matches the actual quality observed by OM. Compared to traditional single-sensor-based quality prediction, the MFDT has achieved a significantly higher quality prediction accuracy (96%), a higher ROC-AUC score (99%), and a lower false alarm rate (4.4%). As a result, the MFDT is a more reliable method for defect prediction. The proposed MFDT also lays the groundwork for our future development of a self-adaptive hybrid processing strategy that combines machining with AM for defect removal and quality improvement.

Highly Adaptive Single-Shot Surface Shape Distortion Measurement Method for High-Power Lasers

Simplifying the detection of optical distortions by machine learning

High robustness single-shot wavefront sensing method using a near-field profile image and fully-connected retrieval neural network for a high power laser facility

Deep learning based wavefront sensor for complex wavefront detection in adaptive optical microscopes.

Distortion Correction in Surface Defects Evaluating System of Large Fine Optics

Deep learning-driven adaptive optics for single-molecule localization microscopy

Wavefront Shaping by A Small-Aperture Deformable Mirror in the Front Stage for High-Power Laser Systems

Deformable mirror resolution-matching-based two-stage wavefront sensorless adaptive optics method

Simulation and Experimental Investigation on the Wavefront Correction Performance of a Dual Magnetic Connection Deformable Mirror under High-Power Laser Irradiation

Single-Shot Multi-Frequency 3D Shape Measurement for Discontinuous Surface Object Based on Deep Learning

Surface Shape Distortion Online Measurement Method for Compact Laser Cavities Based on Phase Measuring Deflectometry

In-situ monitoring laser based directed energy deposition process with deep convolutional neural network

MEMS-based adaptive optics scanning laser ophthalmoscopy

Efficiency improvement of pulse waveform shaping on high power laser facility using deep learning

Surface Error Modeling of Mounted Large Optics in High Power Laser System

Research on the surface shape testing and close-loop performance of a deflectometry system

Atmospheric Turbulence Aberration Correction Based on Deep Learning Wavefront Sensing

Machine learning based adaptive optics for doughnut-shaped beam

Multisensor fusion-based digital twin for localized quality prediction in robotic laser-directed energy deposition

Real-time monitoring of high-power disk laser welding statuses based on deep learning framework

Fast Measurement of Mid-Spatial-frequency Error on Optical Surfaces with Convolutional Neural Networks