Additive Manufacturing Defect Detection Using Correlated Process Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Additive manufacturing processes face challenges in accurately determining defects such as pores, cracks, and surface deviations due to thermal effects and material properties, which can lead to mechanical property issues in the final product.
Innovation Solution
A computer-implemented method that uses locally resolved process data from additive manufacturing to train an adaptive algorithm for defect detection, correlating measurement data with process data to identify and prevent defects through nondestructive imaging and machine learning techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If additive manufacturing processes are used to produce objects, then manufacturing flexibility and complexity are improved, but defect detection accuracy deteriorates due to thermal effects and material properties
Solution Approach 1:
The system performs preliminary defect detection during the additive manufacturing process by analyzing process data (laser power, temperature, layer formation) in real-time before the object is complete. This allows early identification of defects such as pores, cracks, or incomplete melting, enabling intervention before the entire manufacturing process completes and resources are wasted.
Solution Approach 2:
The system implements feedback by continuously monitoring process data during additive manufacturing and using machine learning models to predict defects. The detected defects feed back into the control system, which can adjust manufacturing parameters (laser power, scanning speed, layer thickness) in real-time to prevent defect formation or alert operators to modify the process.
2Reliability
If comprehensive process data is collected and analyzed using adaptive algorithms, then defect detection capability is improved, but computational complexity and processing time worsen
Solution Approach 1:
The system segments the defect detection task by dividing process data analysis into distinct phases: data collection during manufacturing, preliminary processing of raw sensor data, feature extraction (identifying relevant patterns), and final defect classification. This segmentation allows each stage to be optimized independently, reducing overall computational complexity while maintaining high detection capability.
Solution Approach 2:
The system applies partial action by focusing computational resources on analyzing only the most critical process parameters and regions of the object most prone to defects. Rather than processing all possible data equally, the adaptive algorithm identifies and prioritizes analysis of key features (laser power fluctuations, temperature anomalies, layer adhesion issues) that have the highest predictive value for defect formation.
Data Source
AI summary
Described is determining defects of an object produced using an additive manufacturing process, including: determining spatially resolved first data relating to n objects, the first data defines a process coordinate system for each of the n objects, determining measurement data relating to the n objects by imaging the n objects, the measurement data defines, for each of the n objects, an object representation in a measurement coordinate system, determining which coordinates of at least one section of the measurement coordinate system are defect coordinates assigned to a defect in the object representation; correlating the at least one section with a corresponding section of the process coordinate system in order to collect training data, training an adaptive algorithm for determining defect coordinates in spatially resolved data, by means of the training data, determining spatially resolved second data, and analysing the second data for defects by means of the adaptive algorithm.