Updating a finite element model to the real experimental setup by thermographic measurements and adaptive regression optimization. (December 2015)

• Record Type: Journal Article
• Authors: Peeters, J.; Arroud, G.; Ribbens, B.; Dirckx, J.J.J.; Steenackers, G.
• Is Part Of: Mechanical systems and signal processing. Volume 64/65(2015)
• Page Start: 428
• Page End: 440
• Publication Date: 2015-12
• Subjects: Finite element method -- Image processing -- Infrared thermography -- Optimization -- Inverse problem; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2015.04.010
• Languages: English
• ISSNs: 0888-3270

Abstract

Abstract: In non-destructive evaluation the use of finite element models to evaluate structural behavior and experimental setup optimization can complement with the inspector׳s experience. A new adaptive response surface methodology, especially adapted for thermal problems, is used to update the experimental setup parameters in a finite element model to the state of the test sample measured by pulsed thermography. Poly Vinyl Chloride (PVC) test samples are used to examine the results for thermal insulator models. A comparison of the achieved results is made by changing the target values from experimental pulsed thermography data to a fixed validation model. Several optimizers are compared and discussed with the focus on speed and accuracy. A time efficiency increase of over 20 and an accuracy of over 99.5% are achieved by the choice of the correct parameter sets and optimizer. Proper parameter set selection criteria are defined and the influence of the choice of the optimization algorithm and parameter set on the accuracy and convergence time are investigated. Abstract : Graphical abstract: Build response surface (meta-model) to identify setup and structure dependent parameters. Abstract : Highlights: The use of unique parameter sets is essential to achieve good convergence. The choice of a proper optimizer and algorithm decreases time usage by factor 2. Convergence is independent of algorithm choice. The use of meta-modeling increases time efficiency by factor 20. Accurate FE updating is possible for low conductive thermal measurements.