Additive Manufacturing Layer Monitoring for Internal Defect Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing additive manufacturing systems cannot predict or prevent defects inside the manufactured object, requiring post-manufacturing inspection which is limited in scope and accuracy.
Innovation Solution
An additive manufacturing system that includes a measurement unit to assess layer formation, a control unit to store reference information from defect-prone samples, and an estimation unit to predict defects using machine learning, allowing for real-time defect prevention during the manufacturing process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If post-manufacturing inspection is performed using X-ray CT scanning, then defects inside the object can be detected, but defects cannot be predicted or prevented in advance and material waste increases
Solution Approach 1:
The system performs measurement of layer formation states during the additive manufacturing process itself, before the object is complete. The estimation unit predicts potential defects in advance by analyzing measurement information from intermediate layers, allowing preventive action to be taken before material waste occurs.
Solution Approach 2:
The system establishes a feedback loop where measurement information from each layer is fed back to the estimation unit, which compares it against reference information from sample objects. This continuous feedback enables real-time prediction and prevention of defects, reducing the need for post-manufacturing inspection and material waste.
2Reliability
If post-manufacturing inspection is performed, then defects can be detected, but the inspection is limited in scope and accuracy and cannot prevent defects
Solution Approach 1:
The estimation unit acts as an intermediary between measurement information and defect prediction. It uses machine learning algorithms to process measurement data and compare it with reference information from sample objects, enabling accurate defect prediction without requiring complex post-manufacturing inspection equipment like X-ray CT scanners.
Solution Approach 2:
The system replaces physical inspection methods (X-ray CT scanning) with an information-processing approach using machine learning. The estimation unit analyzes measurement information and predicts defects through computational methods, substituting mechanical/physical inspection systems with a software-based prediction system that is less complex and more accurate.
3Loss of time
If traditional inspection methods are used after manufacturing, then defects are detected late, but manufacturing time is increased due to rework
Solution Approach 1:
The measurement unit continuously measures the formation state of each layer during the additive manufacturing process, and the estimation unit continuously predicts potential defects in real-time. This continuous monitoring and prediction eliminates the need for separate post-manufacturing inspection stages, maintaining continuous productive action throughout the manufacturing process.
Solution Approach 2:
By predicting defects in advance during the manufacturing process itself, the system allows for corrective actions to be taken before the object is completed. This preliminary defect prevention eliminates the need for time-consuming post-manufacturing inspection and rework, significantly improving manufacturing efficiency and reducing overall production time.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables accurate and timely prediction of defects, reducing material waste and manufacturing time by preventing defects before they occur, and enhancing the reliability of the manufacturing process.
Implementation Method 1
reference information based on internal defect information indicating a defect existing inside a sample object which is shaped by the additive manufacturing unit and which includes the plurality of layers, based on an electromagnetic wave which has passed through the sample object
Data Source
AI summary
An additive manufacturing system includes an additive manufacturing unit configured to shape an object including a plurality of layers, a measurement unit configured to measure a state of each of the plurality of layers, and a control unit. The control unit includes a storage unit configured to store reference information based on internal defect information indicating a defect existing inside a sample object shaped by the additive manufacturing unit and including the plurality of layers, based on an electromagnetic wave which has passed through the sample object, and sample measurement information indicating a measurement result of the plurality of layers of the sample object measured by the measurement unit, and an estimation unit configured to estimate whether a defect occurs inside the object, based on measurement information indicating a measurement result of the plurality of layers of the object measured by the measurement unit and the reference information.