Skip to main navigationSkip to main content
The University of Southampton
Engineering

Stochastic Model Updating in Structural Dynamics and its Application in Structural Joints Seminar

Time:
16:00 - 17:00
Date:
27 November 2018
Venue:
Building 13/3021

Event details

ISVR Engineering Research Seminar

Finite element model updating is a cornerstone of model validation and verification, its purpose being to adjust an mathematical model such that its outputs, in a range of interest, agree with data obtained typically from a vibration test. Deterministic model updating is now a mature technology, widely applied to large scale industrial structures. In the last 10 years, a number of model updating methods have been developed to take account of modelling and test-data variability. In these methods, much emphasis has been placed on what has become known as stochastic model updating.

Joints in engineering structures are one of the most important sources of structural uncertainties. The uncertainties and nonlinearities in joints arise from many complex behaviours such as friction (slip), opening and closing of gaps (slap), relaxation, preloading, fastener bending, tilting.  Joint behaviour also depends on surface roughness and pre-normal stress. It is extremely difficult and expensive to produce identical joints that behave exactly the same, one with another. Consequently, small variability might become significant in the overall dynamics of the structure.

In this presentation, first the problem of deterministic linear model updating is introduced. Stochastic model updating using Bayesian inference is described and its application to the DLR AIRMOD structure will be demonstrated. The presentation demonstrates how the uncertainty may be represented in the contact interfaces and investigates these uncertainty effects in the dynamic responses of the structure, e.g. natural frequencies. The Bayesian model updating method is used for identification of stochastic model of joints.

Speaker information

Hamed Haddad Khodaparast, Swansea University. Dr Hamed Haddad Khodaparast is a Senior Lecturer in College of Engineering at Swansea University. He has extensive research experience in structural dynamics and development of methods for stochastic model identification in this field. In Liverpool, he developed efficient codes for perturbation-based stochastic model updating and interval model updating that were used for determination of irreducible variability in a population of nominally identical test pieces, such as cars off a production line. His codes were used by the German Aerospace Centre (DLR) on stochastic model updating of an aircraft-like structure. He also collaborated with the CFD (Computational Fluid Dynamics) group at Liverpool to examine the problem of uncertainty propagating through large dimensional CFD-based models. In 2015, he was awarded the Royal Academy of Engineering Industrial Secondment (RAEng IS) and Sêr Cymru National Research Network (NRN) awards and worked with Airbus in the area of efficient aircraft design. During the industrial fellowship, he developed numerical tools that are capable of demonstrating the effects of structural uncertainties on the prediction of unsteady aerodynamic loads. He is currently working under EPSRC grant EP/P01271X/1, on the problem of stochastic modelling and identification of nonlinear structural joints and their implementation in realistic aeroelastic models. He is also an active member of the Digital twins for improved dynamic design EPSRC Programme grant EP/R006768/1 and collaborate on nonlinear dynamics.

Privacy Settings