A Comparison of Effects of Welding Parameters in T-Joints Using Experimental Modal Analysis and Finite Element Modelling

Accurate determinations in the frequency domain are crucial in finite element analysis (FEA), particularly for ensuring the structural integrity of welded joints. FEA investigations are conducted on structures exposed to dynamic forcing conditions in various applications. When analysing dynamic conditions, employing modelling techniques in FEA is imperative to prevent failures occurring at natural frequencies. The natural frequency is influenced by welding parameters, which may vary based on different modelling approaches. This study compares the result of a modal analysis of 13 samples with the FEA results obtained through five distinct model methods of solid and shell model. Each sample in the experimental studies is prepared with different welding parameters, allowing for an examination of the effects of both welding and geometric parameters on the natural frequency. Additionally, the verification is performed by analysing the results of the samples scanned with a coordinate measuring machine (CMM) laser scanner. The obtained experimental data are compared with the finite element analyses and error rates are determined for each of the five model types. CMM model exhibits the highest accuracy with an error rate of 1.86%. Additionally, it is observed that solid model yields more precise results in instances where a CMM is unavailable, displaying an average error rate of 2.58%. As a result of the study, possible error rates in other model types in finite element analyses, either solid or shell model, are presented.