Finite element model updating using application programming interface (API) for PARIS (PARameter Identification System).

摘要

This thesis describes a project to improve a structural parameter estimation program developed at Tufts University. The process of structural parameter estimation is a method to update structural parameters in an analysis model, reflecting measured performance of the structure. It can be a valuable component of a Structural Health Monitoring (SHM) system, whereby measured values of a structure under loading are compared to predicted values from an analysis model, and the difference between the two can indicate deterioration in the structure. Typical parameters for static model updating are axial rigidity, bending rigidity, torsional rigidity for frame elements, thickness for shell elements, and modulus of elasticity for solid elements.;A MATLABRTM based computer program developed at Tufts University called PARIS(c), PARameter Identification System, was developed by Professor Masoud Sanayei and former graduate students for structural parameter estimation. The goal of this research is modifying PARIS(c) version 11.0 to the new PARIS(c) API version which is capable of interfacing SAP2000RTM and MATLABRTM in real time to perform multi-response parameter estimation using non-destructive test data of full scale bridges.;The major contribution of this work is developing PARIS(c) API version 12.0. Compared to PARIS(c) version 11.0, the finite element modal responses are evaluated in SAP2000RTM and updated by MATLABRTM functions using API, Application Programming Interface. MATLABRTM activates SAP2000RTM in real time to update the finite element model. This is a feasibility study of interfacing two software packages with graphical user interfaces for creating a more powerful software package for finite element model updating of full-scale structures.;The new version of PARIS(c) developed in this work is applied to several examples including truss, frame, shell and solid elements. These examples use single or several types of elements simultaneously. Finally, a model of Tobin Memorial Bridge is successfully used for finite element model updating using simulated data. The ability to use SAP2000RTM as a slave to PARIS(c) for the parameter estimation of full scale bridges is a major step forward, opening up opportunities for future research.