A mechanics‐informed artificial neural network approach in data‐driven constitutive modeling

A mechanics‐informed artificial neural network approach for learning constitutive laws governing complex, nonlinear, elastic materials from strain–stress data is proposed. The approach features a robust and accurate method for training a regression‐based model capable of capturing highly nonlinear s...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:International journal for numerical methods in engineering Ročník 123; číslo 12; s. 2738 - 2759
Hlavní autori: As'ad, Faisal, Avery, Philip, Farhat, Charbel
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Hoboken, USA John Wiley & Sons, Inc 30.06.2022
Wiley Subscription Services, Inc
Predmet:
artificial neural network
Artificial neural networks
Computer architecture
Computer simulation
constitutive modeling
convexity
Dynamic stability
Finite element method
hyperelasticity
Learning
machine learning
Mechanics
Neural networks
Noise sensitivity
Parachute canopies
Regression models
Rigid structures
Robustness (mathematics)
Solid mechanics
stability
Strain
supersonic parachute inflation dynamics
Training
Woven fabrics
ISSN:0029-5981, 1097-0207
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:A mechanics‐informed artificial neural network approach for learning constitutive laws governing complex, nonlinear, elastic materials from strain–stress data is proposed. The approach features a robust and accurate method for training a regression‐based model capable of capturing highly nonlinear strain–stress mappings, while preserving some fundamental principles of solid mechanics. In this sense, it is a structure‐preserving approach for constructing a data‐driven model featuring both the form‐agnostic advantage of purely phenomenological data‐driven regressions and the physical soundness of mechanistic models. The proposed methodology enforces desirable mathematical properties on the network architecture to guarantee the satisfaction of physical constraints such as objectivity, consistency (preservation of rigid body modes), dynamic stability, and material stability, which are important for successfully exploiting the resulting model in numerical simulations. Indeed, embedding such notions in a learning approach reduces a model's sensitivity to noise and promotes its robustness to inputs outside the training domain. The merits of the proposed learning approach are highlighted using several finite element analysis examples. Its potential for ensuring the computational tractability of multi‐scale applications is demonstrated with the acceleration of the nonlinear, dynamic, multi‐scale, fluid‐structure simulation of the supersonic inflation dynamics of a parachute system with a canopy made of a woven fabric.
Bibliografia:Funding information
Air Force Office of Scientific Research, FA9550‐20‐1‐0358; National Aeronautics and Space Administration, 80NSSC21K0228; National Science Foundation, 1937129
ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0029-5981
1097-0207
DOI:10.1002/nme.6957