A Bayesian framework for the calibration of cohesive zone models
Autor: | Daniel A. Castello, Emerson B. Albuquerque, Cynthia Guzman, Lavinia Borges |
---|---|
Rok vydání: | 2017 |
Předmět: |
education.field_of_study
Materials science Bayesian probability Population Markov chain Monte Carlo 02 engineering and technology Surfaces and Interfaces General Chemistry Inverse problem 021001 nanoscience & nanotechnology Surfaces Coatings and Films symbols.namesake 020303 mechanical engineering & transports 0203 mechanical engineering Mixing (mathematics) Mechanics of Materials Damage mechanics Materials Chemistry symbols Calibration Applied mathematics Uncertainty quantification 0210 nano-technology education Simulation |
Zdroj: | The Journal of Adhesion. 94:255-277 |
ISSN: | 1545-5823 0021-8464 |
DOI: | 10.1080/00218464.2016.1268055 |
Popis: | Adhesively bonded joints are used in several industrial sectors. Cohezive Zone Modes can be used to predict the adhesive mechanical behaviour. This work presents an approach to calibrate Cohesive Zone Models (CZM) by means of Statistical Inverse Analysis. The Bayesian framework for Inverse Problems is used to infer about the CZM model parameters. The solution corresponds to the exploration of the posterior probability density function of the model parameters. The exploration of the posterior density is performed by means of Markov Chain Monte Carlo (MCMC) methods mixing Population-Based MCMC with Adaptive Metropolis (AD) strategies. The assessment of the approach is performed using measured data from a single-lap shear experimental set-up. Measured data from 5 test-specimens is used for calibration and measured data from five other test-specimens is used for model validation. It is proposed a stochastic effective model for the CZM parameters. The predictions of maximum force and maximum displaceme... |
Databáze: | OpenAIRE |
Externí odkaz: |