Stochastic Metamodel for Material Testing Uncertainty Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for testing materials with non-linear characteristics are time-consuming and costly, requiring hundreds of tests to understand material properties over a range of design parameters, with limited validation due to cost constraints.
Innovation Solution
A simulation-based testing method that selects a first set of points from a design space, generates a stochastic metamodel, determines uncertainty values, and iteratively combines additional points to reduce uncertainty until it meets a predetermined threshold, thereby optimizing the selection of data points for accurate material characterization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If hundreds of tests are performed to understand material properties over a range of design parameters, then measurement precision is improved, but loss of time and cost increase significantly
Solution Approach 1:
The patent performs preliminary testing to generate initial training data before creating the metamodel. This preliminary action establishes a foundation that allows subsequent predictive analysis without requiring complete exhaustive testing, thereby reducing overall testing time while maintaining characterization accuracy.
Solution Approach 2:
The patent creates a metamodel as a computational copy of the physical material system. This metamodel replicates material behavior across the design space, allowing virtual experimentation and prediction of material properties without performing physical tests at every parameter combination, thus dramatically reducing testing duration.
2Measurement precision
If hundreds of tests are performed to understand material properties over a range of design parameters, then measurement precision is improved, but cost increases significantly
Solution Approach 1:
The patent creates a metamodel as a computational copy of the physical material system. This metamodel replicates material behavior across the design space, allowing virtual experimentation and prediction of material properties without performing physical tests at every parameter combination, thus dramatically reducing testing duration.
Solution Approach 2:
The patent replaces the mechanical testing system with a computational metamodeling system. Instead of physically testing materials at hundreds of parameter combinations, the metamodel computationally predicts material response, substituting expensive physical experimentation with lower-cost computational analysis.
3Productivity
If FEA is used to predict material properties, then productivity is improved, but reliability decreases due to limited validation data
Solution Approach 1:
The patent implements a feedback mechanism where the metamodel is iteratively refined using available test data. The model learns from actual experimental results and adjusts its predictions accordingly, creating a closed-loop system that continuously improves prediction accuracy while maintaining computational efficiency.
Solution Approach 2:
The patent transforms the FEA approach by changing the parameters used for prediction. Instead of relying solely on theoretical FEA models with limited validation, the metamodel uses learned parameters from actual test data combined with FEA, creating a hybrid approach that maintains productivity while improving reliability through data-driven parameter adjustment.
Data Source
AI summary
A method implemented using a processor based device for simulation based testing of materials, includes selecting a first set of points from a data generated from a design space and generating a stochastic metamodel based on the first set of points. The method also includes determining an uncertainty value based on the stochastic metamodel. The method also includes identifying a second set of points different from the first set of points, from the data generated from the design space, based on the uncertainty value. The method further includes combining the second set of points with the first set of points to generate a third set of points, assigning the third set of points to the first set of points. The method also includes iteratively generating, determining, identifying, combining, and assigning steps till the uncertainty value is less than or equal to a predetermined threshold value.