Numerical Modelling of the Structural Behaviour of Joints

Abstract

It is the sign of the progress and it is the competition of the steel with concrete to become more and more efficient construction which needs less material and less effort in manufactoring leading to an optimal cost-benefit-ratio. Therefore new construction technologies are developed. They have to be verified and the ultimate load as well as the deformation characteristic have to be calculated. In former days it was out of question to carry out a lot of experiments to get the significant attributes. Even if a construction should be generalized for a broad range for each parameter combination a couple of experiments have to be performed because of the statistical deviation. The safety factor in the determined design criteria has to be high in order to cover uncertainties (reasons of deviation, lack of tested combinations). Nowadays the characteristic of a construction can be determined by finite element calculations. The devolopment of a new construction can be done numerically. Experiments are only needed to verify numerical results. This procedure is not as time- and cost-consuming as the pure experimental method. Numerical parametric studies are easy to conduct, the failure mechanisms are visible and therefore easy to study. So, the construction’s safety factor perhaps can be reduced. But the finite element calculations have to be used carefully and their results have to be analysed critically to achieve correct qualitative as well as quantitative answers about the tested construction. In this context the main question is: How reliable are the numerical results?