Beam T-junction Model Accuracy Improvement Based on Experimental Modal Analysis

For structural analysis of bus and coach tubular structures, finite element analysis with beam type elements is widely approved and utilized. These elements have an infinitely rigid behavior at the joint level, which represents an unrealistic stiffer estimation of the modeled structures. This cannot be neglected since it introduces stiffness variations when analyzing complex structures. To overcome this limitation, in this article an alternative beam T-junction model, in which 6 elastic elements are included at the joint level in order to improve the local stiffness behavior, is utilized. A new methodology to determine these 6 elastic constants is proposed, based on the analysis of the modal response differences between the beam model and the real prototype T-junctions. In this way, a completely new methodology for the determination of the rigidity of the elastic elements based on the analysis of the modal characteristics of these structures is presented. The methodology is validated against experimental modal analysis. The results show that using this methodology with the alternative beam T-junction model is possible to obtain a more accurate T-junction model with an average deviation from real models of less than 1.5%, and a maximum deviation among the studied modes of less than 4.