AI Predicts 3D Porous Metal Structures from 2D Images During Printing
Researchers have developed an AI system that predicts the 3D porous structure of metal 3D prints using 2D images captured during the printing process, achieving an 18x speedup over conventional methods.
A team from the University of Tokyo has created an artificial intelligence system capable of predicting the three-dimensional porous structure of a metal 3D print in real-time. This innovative approach utilizes two-dimensional images captured during the additive manufacturing process to forecast the internal structure.
Traditionally, assessing the internal porosity of 3D printed metal parts requires time-consuming post-processing steps, often involving destructive methods or lengthy simulations. The new AI method bypasses these limitations by analyzing optical data from the printing stage itself. This allows for immediate feedback on the quality and structural integrity of the printed component.
The researchers demonstrated that their AI system can predict the 3D porous structure with high accuracy. Crucially, this predictive capability significantly accelerates the evaluation process. The system achieves an 18-fold reduction in processing time compared to conventional simulation-based techniques, making it a highly efficient tool for quality control in metal additive manufacturing.
This development holds considerable promise for improving the reliability and speed of metal 3D printing. By enabling rapid, non-destructive assessment of internal structures, manufacturers can identify and correct defects early in the printing process, ultimately leading to higher quality parts and reduced waste.
This AI-driven, in-situ monitoring system represents a significant advancement in quality control for metal additive manufacturing. By translating 2D imaging data into real-time 3D structural predictions, it dramatically cuts down evaluation time. This is crucial for high-value applications like aerospace, where rapid defect detection and process optimization are paramount for producing reliable, complex components.
Edited by the news editor with AI and translated into English from the original report — please refer to the original source.