Parts Matching via Graded Clustering for Tolerance Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In manufacturing settings, the assembly of products from multiple components with varying tolerances and performance measures often results in compounded tolerance issues that affect the overall quality and performance of the final product, where each component may be on either the upper or lower limit of allowable tolerance, leading to suboptimal assembly outcomes.
Innovation Solution
A method and system for parts matching in manufacturing, utilizing a parts processing engine that analyzes training data to create a graded product cluster template, which is then used by automated manufacturing devices to select components for assembly, aiming to produce the highest-grade product with minimal tolerance variations and maximum performance measures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If components are sourced from multiple manufacturers with varying tolerances, then manufacturing flexibility and component availability are improved, but assembly quality and performance consistency deteriorate due to compounded tolerance issues
Solution Approach 1:
The system performs preliminary grading and clustering of components based on their tolerance characteristics before assembly. By analyzing component measurements and creating graded clusters in advance, the system pre-sorts components into compatibility groups, ensuring that only components with compatible tolerance profiles are assembled together. This preliminary action prevents tolerance compounding issues before they occur in final assembly.
Solution Approach 2:
The system applies different quality grading criteria to different components based on their specific tolerance characteristics and impact on final assembly performance. Rather than applying a uniform tolerance standard to all components, the system identifies which components are critical to assembly quality and applies stricter grading to those specific components, while allowing more flexibility in non-critical components.
2Ease of manufacture
If traditional assembly methods are used without parts matching, then assembly process simplicity is maintained, but product performance and quality are reduced due to tolerance accumulation
Solution Approach 1:
The system replaces traditional mechanical tolerance management methods with an automated information-processing system. Instead of relying on physical measurement and manual sorting, the system uses digital component databases, automated measurement data analysis, and algorithmic clustering to match components. This substitution maintains ease of manufacture through automation while significantly improving product performance through precise parts matching.
Solution Approach 2:
The system enables components to effectively select their own compatible mates through automated grading and clustering. Each component is characterized and placed in appropriate clusters based on its tolerance profile, allowing the system to self-organize compatible component sets without human intervention. This self-service approach improves reliability while maintaining process simplicity.
3Manufacturing precision
If graded product cluster templates are created and used for parts matching, then assembly quality and tolerance optimization are improved, but system complexity and data processing requirements increase
Solution Approach 1:
The system segments the component population into distinct graded clusters based on tolerance characteristics. By dividing components into discrete groups with similar tolerance profiles, the system simplifies the matching process while achieving precise tolerance optimization. Each cluster represents a segment of the overall component population with defined compatibility boundaries, making the system manageable despite the complexity of handling multiple tolerance dimensions.
Data Source
AI summary
A method includes receiving a training data set for a first plurality of assemblies, wherein the training data set includes a plurality of components for each of two or more types of components of the first plurality of assemblies. The method analyzes the training data set for the first plurality of assemblies. Responsive to receiving a set of product component information for a second plurality of assemblies, the method creates a graded product cluster template for the second plurality of assemblies based on the analyzed training data for the first plurality of assemblies. The method sends the graded product cluster template to an automated manufacturing device, wherein the automated manufacturing device manufactures a first assembly from the second plurality of assemblies based on the graded product cluster template.