Dynamic model updating of body-in-white structure using sensitivity analysis

• Main Author: Noor Am Zura, Abdullah
• Format: Thesis
• Language: English
• Published: 2017
• Subjects: TJ Mechanical engineering and machinery
• Online Access: http://umpir.ump.edu.my/id/eprint/19775/19/Dynamic%20model%20updating%20of%20body-in-white%20structure%20using%20sensitivity%20analysis.pdf; http://umpir.ump.edu.my/id/eprint/19775/

A large and complex structure such as vehicle body-in-white (BIW) structure consists of several major components made from thin metal sheets that are joined together by several types of mechanical joint such as spot welds and bolted joints. Meanwhile, a finite element model of a structure is built to predict the behaviour of the structure before fabrication of the structure. Experimental work will be performed on the fabricated structure and the measured data are compared with the computational analysis on its finite element model counterparts. However, in case of complex structure such as BIW, the finite element prediction is not always in good agreement with the experimental results due to inaccurate approximation of boundary condition, physical properties of the structure and meshing process. This thesis intends to focus on demonstrating a method called model updating in order to reduce the discrepancies. Initially, the finite element model of BIW structure is constructed in simplified manner. The first model is constructed without including any connecting elements to represent the mechanical joint components. In the next phase of modelling, beam element (CBAR) and spring connector element (CELAS) is included in the model to represent the joint components. The modal properties of the created finite element models of BIW is obtained and compared to the modal properties measured in experimental modal analysis. Experimental work involves the application of impact hammer test for measuring the modal response. Several approaches and set up are used during experimental work which includes different experimental set up and number of sensor used in measuring response, in order to ensure that the quality of the measured response is reliable for correlation purpose. Finally, the procedure of model updating on all created BIW model is performed with the aid of sensitivity analysis in order to identify the potential parameters to be chosen as the updating parameters. Initial percentage of error in all model are below 10%. After model updating procedure is performed on all BIW models, the average percentage of error in every mode are successfully reduced to below 5%, a significant level of improvement. Among all three updating work, the updating procedure performed on BIW model with joint elements shows better error reduction. At the end, the model updating procedure with the aid of sensitivity analysis on the selected parameters has proved to be an effective tool to reduce the discrepancies between prediction and experimental data and able to localize the source of inaccurate assumptions and uncertainties.