Patch loading of stainless steel girders : experiments and finite element analyses

Sammanfattning: This thesis deals with the behaviour of stainless steel girders subjected to patch loading. The current loading situation arises for example at beam- column connections or at crane girders where the load is travelling along a girder. At bridge launching it is also important to consider patch loading. Most of the work on patch loading has been focused on carbon steel and therefore the codes for stainless steel are based on results from carbon steel studies without any accurate validation. The work within this thesis consists of both tests and numerical simulations to form a basis for conclusions concerning the utilization of the existing codes or if a new has to be developed. The study is based on nine experimental tests performed on welded I-shaped girders made of austenitic stainless steel of grade EN 1.4301 subjected to a concentrated force applied at one flange. Two different situations are studied. The first with the load applied far from the girder end and the second with the load applied at an unstiffened girder end. The investigation also consists of a numerical study, which includes models of the patch load tests as well as a parametric study. Results from the tests and the finite element models are compared to the existing design models presented in ENV 1993-1-1:1992 (Part 1-1) and ENV 1993-1-5:1997 (Part 1-5). In these comparisons the yield strength is substituted by the 0.2% proof stress of the stainless steel. A part of the study is also devoted to the influence of in-plane bending moment and results from numerical analyses is compared to results obtained with the interaction formula in Part 1-5. Finally a statistical evaluation of the design guidance in Part 1-5, concerning the resistance to concentrated forces, is presented. It is shown that the model presented in Part 1-5 gives a good prediction of the resistance to concentrated forces and that this design procedure gives results that are in better agreement with the tests results than the one presented in Part 1-1. Together with the numerical simulations, the experimental results verify that Part 1-5, instead of Part 1-1, shall be used to predict the resistance to patch loading. The results from this study fall in the same scatter range as the results for structural steel, which formed the basis of the rules in Part 1-5.