A computational framework for Bayesian inference in plasticity models characterisation

Asaadi, Erfan; Heyns, P.S.

doi:10.1016/j.cma.2017.04.017

Cited by 15 publications

(21 citation statements)

References 55 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In other words, the response must contain sufficient information about the sought-after parameters. The Bayesian inverse modeling relies on the expectation that after observing the material response, the posterior knowledge, described by p (θ|D, D), reduces the initial uncertainty about the parameters, described by p (θ); see [10,28] for more details.…”

Section: Bayesian Model Identification and The Ranking Of Materials Rementioning

confidence: 99%

“…Other sampling techniques have been adopted to approximate the terms needed in Eq. (10) or Eq. 5, to estimate the mutual information and entropy, see [31].…”

Section: Information Quantification: Computational Aspectsmentioning

confidence: 99%

“…Referring to [34], we opt for Nested Sampling (NS) developed by Skilling [25] to approximately either sample from the posterior, required in the entropy estimation, or estimate the marginalized likelihood, required in the mutual information estimation. As it was demonstrated in [10], compared with other sampling algorithms, the advantage of nested sampling is that it allows to estimate the posterior samples and the marginalized likelihood simultaneously, tuning the algorithm's parameters does not involve a cumbersome calibrating process, and the acceptance rate can easily be kept close to 1 (see [34][35][36]).…”

Section: Information Quantification: Computational Aspectsmentioning

confidence: 99%

“…A more informative, yet demanding approach to solve an inverse problem is to take the influence of the uncertainty sources into account. This is done through devising a probabilistic inverse problem for which the solution provides the uncertainty of the identified model parameters as well [10][11][12][13][14][15][16]. In the probabilistic formulation of the material identification problem, the material parameters are not point estimates; rather they are random variables described by their probability distributions.…”

Section: Introductionmentioning

confidence: 99%

“…A close look at attempts aimed at formulating and solving probabilistic material model identification problems (see e.g. [10][11][12][13][14][15]) reveals the absence of a formal strategy for measuring the value of different material responses. The important question to ask is: from the set of candidate material responses, D = {D 1 , D 2 , .…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

On the value of test data for reducing uncertainty in material models: Computational framework and application to spherical indentation

Asaadi

Heyns

Haftka

et al. 2019

Computer Methods in Applied Mechanics and Engineering

Self Cite

View full text Add to dashboard Cite

Section: Bayesian Model Identification and The Ranking Of Materials Rementioning

confidence: 99%

“…Other sampling techniques have been adopted to approximate the terms needed in Eq. (10) or Eq. 5, to estimate the mutual information and entropy, see [31].…”

Section: Information Quantification: Computational Aspectsmentioning

confidence: 99%

Section: Information Quantification: Computational Aspectsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

On the value of test data for reducing uncertainty in material models: Computational framework and application to spherical indentation

Asaadi

Heyns

Haftka

et al. 2019

Computer Methods in Applied Mechanics and Engineering

Self Cite

View full text Add to dashboard Cite

Considerations on the identifiability of fracture and bond properties of reinforced concrete

Dobrilla

Matthies

Ibrahimbegović

2023

Numerical Meth Engineering

View full text Add to dashboard Cite

This work tackles the issue of identifiability of fracture and bond properties in reinforced concrete. The basis for modeling of fracture is a computational model capable of describing damage and failure mechanisms in concrete, as well as bond‐slip which is a result of degradation of the concrete‐steel interface. The discrete approximation combines ED‐FEM for concrete crack representation in each element and X‐FEM representation of bond‐slip along a particular reinforcement bar. The uncertain model parameters are modeled as random variables and identified via Bayesian inference with the help of observations from tensile tests on concrete tie beams with a single embedded reinforcement bar. We discuss how the choice of observation type affects the parameter identifiability and propose combinations which improve the estimation capabilities and reduce the discrepancy between the computed and observed quantities of interest.

show abstract

Exploiting data in smart factories: real-time state estimation and model improvement in metal forming mass production

2019

View full text Add to dashboard Cite

Modern production systems have numerous sensors that produce large amounts of data. This data can be exploited in many ways, from providing insight into the manufacturing process to facilitating automated decision making. These opportunities are still underexploited in the metal forming industry, due to the complexity of these processes. In this work, a probabilistic framework is proposed for simultaneous model improvement and state estimation in metal forming mass production. Recursive Bayesian estimation is used to simultaneously track the evolution of process state and to estimate the deviation between the physics-based model and the real process. A sheet bending mass production process is used to test the proposed framework. A metamodel of the process is built using proper orthogonal decomposition and radial basis function interpolation. The model is extended with a deviation model in order to account for the difference between model and real process. Particle filtering is used to track the state evolution and to estimate the deviation model parameters simultaneously. The approach is tested and analysed using a large number of simulations, based on pseudo-data obtained from a numerical sheet bending model.

show abstract

A computational framework for Bayesian inference in plasticity models characterisation

Cited by 15 publications

References 55 publications

On the value of test data for reducing uncertainty in material models: Computational framework and application to spherical indentation

On the value of test data for reducing uncertainty in material models: Computational framework and application to spherical indentation

Considerations on the identifiability of fracture and bond properties of reinforced concrete

Exploiting data in smart factories: real-time state estimation and model improvement in metal forming mass production

Contact Info

Product

Resources

About