Abstract:This paper presents an in situ monitoring method for numerical controlled manufacturing of large conceptual prototype based on multi-view stitching fusion (MSF). With the extensive application of numerical controlled manufacturing, like additive and subtractive manufacturing (ASM), the forming size tends to increase. Nevertheless, limited by optical lens and affiliated machine vision parameters, it is difficult to avoid the existence of blind areas when detecting defects including scratch, crack, pore, and thermal deformation by machine vision. To overcome the obstacles, multi-view information, including visible and depth sequences, were captured during the manufacturing progresses to expand monitoring scope. The overlapped regions of various frames were confirmed by stitching fusion of multi-views with homography matrix. The flaws were detected by superposing anomalous regions of multi-views which were extracted via multi-dimension synthetical anomalous degree map. The finite element analysis (FEA) was conducted to extract susceptible regions and locally adjust threshold. Inspired from digital twin visualization of Boundary representation (B-Rep), the numerical controlled manufacturing process of large textureless conceptual prototype can be quantized and visualized synchronously. The numerical outcomes were evaluated and verified that the MSF was effective in distinguishing wide-range defect regions of in situ large parts, thereby generating feedbacks. The physical experiment was implemented on W12 engine upper block and thin-walled shell of engineering machinery with laser stereo lithography appearance (SLA). The physical experiment prove that it is of great significance for fabricating large-scale high reflective objects with low defect and high efficiency, further for prototype verification of conceptual design.

2-D Transformer-Based Approach for Process Monitoring of Metal 3-D Printing via Coaxial High-Speed Imaging

In situ monitoring for numerical controlled manufacturing of large conceptual prototype based on multi-view stitching fusion

A New Approach for Online Monitoring of Additive Manufacturing Based on Acoustic Emission

Multi-view Online Vision Detection Based on Robot Fused Deposit Modeling 3D Printing Technology

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

A CNN-Based Adaptive Surface Monitoring System for Fused Deposition Modeling

Sketch-Based Tensor Decomposition for Non-Parametric Monitoring of Electrospinning Processes

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

Visual Detection of Surface Defects Based on Self-Feature Comparison in Robot 3-D Printing

In-situ monitoring additive manufacturing process with AI edge computing

A Real-Time Defect Detection Strategy for Additive Manufacturing Processes Based on Deep Learning and Machine Vision Technologies

Accelerating Process Development for 3D Printing of New Metal Alloys

Dense Robust 3D Reconstruction and Measurement for 3D Printing Process Based on Vision

In-Situ Melt Pool Characterization via Thermal Imaging for Defect Detection in Directed Energy Deposition Using Vision Transformers

Investigation on domain adaptation of additive manufacturing monitoring systems to enhance digital twin reusability

TDIP: Tunable Deep Image Processing, a Real Time Melt Pool Monitoring Solution

Transformer and cross-attention-based multi-sensor in-situ monitoring of molten pool stability and part quality in laser powder bed fusion

A New Method for Automatic Detection of Defects in Selective Laser Melting Based on Machine Vision

Design of a Multi-sensor Monitoring System for Additive Manufacturing Process

MLCPM: A process monitoring framework for 3D metal printing in industrial scale