Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Data-driven representation and clustering discretization method and system for design optimization and/or performance prediction of material systems and applications of same

a data-driven representation and clustering technology, applied in the field of materials, can solve the problems of large computational resources, high cost, and long execution time of finite element software, and achieve the effect of validating the efficiency and accuracy of multi-scale modeling

Pending Publication Date: 2021-11-18
NORTHWESTERN UNIV
View PDF0 Cites 21 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Benefits of technology

The patent describes a method for reducing the number of degrees of freedom needed to represent a material system by generating clusters from the material system. These clusters are then used to train a machine learning model that connects processes / structures to responses / properties of the material system directly. The method also assigns some response predictions as a training set and validation set to verify the efficiency and accuracy of the material system model. The technical effect of this patent is to improve the efficiency and accuracy of modeling material systems by using fewer degrees of freedom and more accurate data.

Problems solved by technology

However, resolving a fine mesh requires significant computational resource, such as High Performance Cluster computing or GPU computing, and a long execution time for the finite element software.
This means cost is very high.
Because of the cost, there is limited practical application of such schemes, and they have seen little use.
Therefore, a heretofore unaddressed need exists in the art to address the aforementioned deficiencies and inadequacies.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Data-driven representation and clustering discretization method and system for design optimization and/or performance prediction of material systems and applications of same
  • Data-driven representation and clustering discretization method and system for design optimization and/or performance prediction of material systems and applications of same
  • Data-driven representation and clustering discretization method and system for design optimization and/or performance prediction of material systems and applications of same

Examples

Experimental program
Comparison scheme
Effect test

example 1

Clustering Discretization Methods for Generation of Material Performance Databases in Machine Learning and Design Optimization

[0356]Mechanical science and engineering can use machine learning. However, data sets have remained relatively scarce; fortunately, known governing equations can supplement these data. This exemplary study summarizes and generalizes three reduced order methods: self-consistent clustering analysis, virtual clustering analysis, and FEM-clustering analysis. These approaches have two-stage structures: unsupervised learning facilitates model complexity reduction and mechanistic equations provide predictions. These predictions define databases appropriate for training neural networks. The feed forward neural network solves forward problems, e.g., replacing constitutive laws or homogenization routines. The convolutional neural network solves inverse problems or is a classifier, e.g., extracting boundary conditions or determining if damage occurs. In this example, we...

example 2

Data Science for Finite Strain Mechanical Science of Ductile Materials

[0476]In this example, a mechanical science of materials, based on data science, is formulated to predict process-structure-property-performance relationships. Sampling techniques are used to build a training database, which is then compressed using unsupervised learning methods, and finally used to generate predictions by means of mechanistic equations. The method presented in this example relies on an a priori deterministic sampling of the solution space, a K-means clustering method, and a mechanistic Lippmann-Schwinger equation solved using a self-consistent scheme. This method is formulated in a finite strain setting in order to model the large plastic strains that develop during metal forming processes. An efficient implementation of an inclusion fragmentation model is introduced in order to model this micromechanism in a clustered discretization. With the addition of a fatigue strength prediction method also...

example 3

Predictive Multiscale Modeling for Unidirectional Carbon Fiber Reinforced Polymers

[0537]This exemplary study presents a predictive multiscale modeling scheme for Unidirectional (UD) Carbon Fiber Reinforced Polymers (CFRP). A bottom-up modeling procedure is discussed for predicting the performance of UD structures. UD material responses are computed from high fidelity Representative Volume Elements (RVEs). A data-driven Reduced Order Modeling (ROM) approach compresses RVEs into Reduced Order Models so material responses can be computed in a concurrent fashion along with the structural level simulation. The approach presented in this example is validated against experimental data and is believed to provide design guidance for future fiber reinforced polymers development process.

[0538]In modern engineering applications, composite materials are receiving growing attention for their extraordinary lightweight and strength. To understand the mechanical performance of various CFRP designs, ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

A method and system for design optimization and / or performance prediction of a material system includes generating a representation of the material system at a number of scales, the representation at a scale comprising microstructure volume elements (MVE) of building blocks of the material system at said scale; providing inputs to the MVEs; collecting data of response fields of the MVE computed from a material model of the material system over a predefined set of material properties and boundary conditions; applying machine learning to the collected data to generate clusters; computing an interaction tensor of interactions of each cluster with each of the other clusters; and solving an governing partial differential equation using the generated clusters and the computed interactions to result in a response prediction usable in an iterative scheme in a multiscale model for the material system. The performance of each scale can be predicted for design optimization.

Description

CROSS-REFERENCE TO RELATED PATENT APPLICATION[0001]This application claims priority to and the benefit of, pursuant to 35 U.S.C. § 119(e), of U.S. provisional patent application Ser. No. 62 / 731,381, filed Sep. 14, 2018, entitled “MULTISCALE MODELING PLATFORM AND APPLICATIONS OF SAME”, by Wing Kam Liu, Jiaying Gao, Cheng Yu and Orion L. Kafka, which is incorporated herein by reference in its entirety.[0002]This application is related to a co-pending U.S. patent application, entitled “INTEGRATED PROCESS-STRUCTURE-PROPERTY MODELING FRAMEWORKS AND METHODS FOR DESIGN OPTIMIZATION AND / OR PERFORMANCE PREDICTION OF MATERIAL SYSTEMS AND APPLICATIONS OF SAME”, by Wing Kam Liu, Jiaying Gao, Cheng Yu, and Orion L. Kafka, with Attorney Docket No. 0116936.152US2, filed on the same day that this application is filed, and with the same assignee as that of this application, which is incorporated herein by reference in its entirety.FIELD OF THE INVENTION[0003]The invention relates generally to materi...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(United States)
IPC IPC(8): G06F30/23G06F30/27
CPCG06F30/23G06F2113/10G06F30/27B33Y50/00G06T17/20G06F2111/10G06F2119/08
Inventor LIU, WING KAMGAO, JIAYINGYU, CHENGKAFKA, ORION L.
Owner NORTHWESTERN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products