Manufactured Part Assessment Using Functional Tolerance Probability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for assessing the quality of manufactured gas turbine engine components, such as integrally bladed rotors, rely solely on geometric tolerances, leading to incorrect rejection of functionally tolerant parts.
Innovation Solution
A dual branch assessment method combining physical testing and machine learning is employed, where data from multiple locations on the component is generated and compared to training data to determine functional tolerance, with a conservative evaluation to accept or reject parts based on a percentage chance of being functionally tolerant.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If manual inspection of dimensions is performed to ensure components are within tolerance range, then manufacturing precision is maintained, but productivity is reduced and functionally tolerant parts may be incorrectly rejected
Solution Approach 1:
The patent replaces manual mechanical inspection with an automated system that uses optical scanning to capture 3D geometry data and machine learning algorithms to assess functional tolerance. This substitution dramatically increases evaluation speed while maintaining or improving accuracy in determining whether parts are functionally tolerant.
Solution Approach 2:
The system creates a digital copy (3D scan) of the physical component's geometry and uses this copy for analysis instead of physically measuring the part. This allows rapid virtual assessment of functional tolerance without handling or physically inspecting the actual component, thereby increasing productivity.
2Manufacturing precision
If strict geometric tolerance limits are enforced, then manufacturing precision is ensured, but reliability is reduced due to incorrect rejection of functionally tolerant components
Solution Approach 1:
The patent applies different assessment criteria to different locations on the component. Instead of uniformly applying geometric tolerance limits across the entire part, the system evaluates each location based on its specific functional requirements. This allows regions that are less critical to function to have more variability, preventing incorrect rejection of functionally tolerant parts.
Solution Approach 2:
The system changes the assessment parameter from strict geometric tolerance to functional tolerance probability. Instead of binary pass/fail based on dimensional limits, the machine learning model outputs a probability percentage indicating how likely the part is to function within tolerance, allowing for more nuanced and accurate acceptance decisions.
3Measurement precision
If manual inspection methods are used to assess component quality, then measurement precision is maintained, but loss of time increases due to slow evaluation process
Solution Approach 1:
The system performs continuous assessment of functional tolerance across multiple locations on the component without interruption. The optical scanner continuously captures geometry data and the machine learning model continuously evaluates each location, eliminating the sequential nature of manual inspection and dramatically reducing total assessment time while maintaining precision.
Solution Approach 2:
The machine learning model is pre-trained on extensive data to quickly assess functional tolerance without requiring time-consuming manual analysis during inspection. The preliminary training phase enables the system to perform rapid evaluations during actual inspection, reducing measurement time while maintaining precision through the trained model's ability to quickly process and interpret geometry data.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method of assessing the quality of a manufactured component includes the steps of manufacturing a component (22) and generating data (24) at each of a plurality of locations on the component. The generated data is passed to a machine learning branch (35). The generated data is compared to training data at each of the locations to determine whether the component (22) is a functionally tolerant dimension at each of the plurality of locations. The manufactured component (22) is accepted (108) or rejected (106) based upon a determined percentage chance the component is functionally tolerant or fails to be functionally tolerant, respectively, at each of the plurality of locations. A system (20) is also disclosed.