A mixture of experts approach to handle ambiguities in parameter identification problems in material modeling

Morand, Lukas; Helm, Dirk

doi:10.1016/j.commatsci.2019.04.003

Cited by 23 publications

(21 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Like mentioned before, Equation ( 2 ) is classically solved by gradient-based optimization methods or genetic algorithms. Disadvantages of this method are: The success of optimization algorithms is highly depending on the chosen starting point, which is usually not known Many iterations are needed to find appropriate parameters for complex material models, which leads to high computational costs A high number of error function evaluations are needed Even for small changes in the experimental setup or the examined material the computationally and time-consuming search must be repeated [ 6 , 10 ] …”

Section: Materials and Methodsmentioning

confidence: 99%

“…Even for small changes in the experimental setup or the examined material the computationally and time-consuming search must be repeated [ 6 , 10 ]…”

Section: Materials and Methodsmentioning

confidence: 99%

“…Due to the complexity caused by structural load cases on industrial parts, FE simulations are usually used for this, which require respective material cards to describe the material-mechanical properties. The calibration of the material models (material parameter identification) is typically performed by a trial-and-error approach or by a semi-automatic iterative optimization process and requires a high degree of expert knowledge [ 4 , 5 , 6 ]. Furthermore, this commonly used process is computational- and time-expensive and thus cost-intensive.…”

Section: Introductionmentioning

confidence: 99%

“…In [ 11 ] improvements for this direct inverse approach have been proposed by Unger and Könke using a Bayesian neural network [ 27 ] to additionally verify the accuracy of the identified parameters and their correlation. Morand and Helm investigated in [ 6 ] problems occurring if the PI problem is non-unique and developed an approach using a mixture of expert model to partition the non-unique problem in subtasks. In [ 28 ] different strategies for calibrating non-linear mechanical models with NN are reviewed.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Artificial Neural Networks-Based Material Parameter Identification for Numerical Simulations of Additively Manufactured Parts by Material Extrusion

et al. 2020

View full text Add to dashboard Cite

To be able to use finite element (FE) simulations in structural component development, experimental investigations for the characterization of the material properties are required to subsequently calibrate suitable material cards. In contrast to the commonly used computational and time-consuming method of parameter identification (PI) by using analytical and numerical optimizations with internal or commercial software, a more time-efficient method based on machine learning (ML) is presented. This method is applied to simulate the material behavior of additively manufactured specimens made of acrylonitrile butadiene styrene (ABS) under uniaxial stress in a structural simulation. By using feedforward artificial neural networks (FFANN) for the ML-based direct inverse PI process, various investigations were carried out on the influence of sampling strategies, data quantity and data preparation on the prediction accuracy of the NN. Furthermore, the results of hyperparameter (HP) search methods are presented and discussed and their influence on the prediction quality of the FFANN are critically evaluated. The investigations show that the NN-based method is applicable to the present use case and results in material parameters that lead to a lower error between experimental and calculated force-displacement curves than the commonly used optimization-based method.

show abstract

Section: Materials and Methodsmentioning

confidence: 99%

“…Even for small changes in the experimental setup or the examined material the computationally and time-consuming search must be repeated [ 6 , 10 ]…”

Section: Materials and Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Artificial Neural Networks-Based Material Parameter Identification for Numerical Simulations of Additively Manufactured Parts by Material Extrusion

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Furthermore, perhaps it will be determined whether there exists a minimum required concentration for the serrated flow to occur. Machine-learning methods could also be employed to elucidate the relationship among the additive concentration, microstructure (such as dislocation dynamics), and serrated-flow behavior [306][307][308][309][310]. Consequently, a greater understanding of solute atom-dislocation pinning dynamics in HEAs would be achieved.…”

mentioning

confidence: 99%

A Review of the Serrated-Flow Phenomenon and Its Role in the Deformation Behavior of High-Entropy Alloys

Brechtl

Chen

Lee

et al. 2020

Metals

View full text Add to dashboard Cite

High-entropy alloys (HEAs) are a novel class of alloys that have many desirable properties. The serrated flow that occurs in high-entropy alloys during mechanical deformation is an important phenomenon since it can lead to significant changes in the microstructure of the alloy. In this article, we review the recent findings on the serration behavior in a variety of high-entropy alloys. Relationships among the serrated flow behavior, composition, microstructure, and testing condition are explored. Importantly, the mechanical-testing type (compression/tension), testing temperature, applied strain rate, and serration type for certain high-entropy alloys are summarized. The literature reveals that the serrated flow can be affected by experimental conditions such as the strain rate and test temperature. Furthermore, this type of phenomenon has been successfully modeled and analyzed, using several different types of analytical methods, including the mean-field theory formalism and the complexity-analysis technique. Importantly, the results of the analyses show that the serrated flow in HEAs consists of complex dynamical behavior. It is anticipated that this review will provide some useful and clarifying information regarding the serrated-flow mechanisms in this material system. Finally, suggestions for future research directions in this field are proposed, such as the effects of irradiation, additives (such as C and Al), the presence of nanoparticles, and twinning on the serrated flow behavior in HEAs.

show abstract

Overview of Machine Learning in Geomechanics

STEFANOU

2024

Machine Learning in Geomechanics 1

View full text Add to dashboard Cite

A mixture of experts approach to handle ambiguities in parameter identification problems in material modeling

Cited by 23 publications

References 24 publications

Artificial Neural Networks-Based Material Parameter Identification for Numerical Simulations of Additively Manufactured Parts by Material Extrusion

Artificial Neural Networks-Based Material Parameter Identification for Numerical Simulations of Additively Manufactured Parts by Material Extrusion

A Review of the Serrated-Flow Phenomenon and Its Role in the Deformation Behavior of High-Entropy Alloys

Overview of Machine Learning in Geomechanics

Contact Info

Product

Resources

About