On the influence of masonry infills in concrete buildings

The stresses induced in a structure by a seismic action have a low probability of occurrence and their magnitude is such that the structure is forced beyond the ultimate limit state. The effective standards in Portugal, and the ones that will be effective shortly, consider the possibility of a seismic action in all the territory, fact that has been neglect by designers and authorities in the last years. This new standard, Eurocode 8, Design of structures for earthquake resistance, defines new standards for infill walls, imposing the use of reinforcement but failing to give detailed information besides the type of reinforcement. Furthermore, the structural designer is the responsible for these non-structural elements. Therefore, with the goal of contributing to the creation of simple design rules for these infills, a shaking table test program of reinforced concrete frame buildings with infill walls, reinforced and unreinforced, will be carried out. Predeceasing these tests, different numerical simulations of the buildings were carried out, with two main objectives: i) understand the behavior of the structure when subjected to a seismic action, in order to more accurately define the experimental program; ii) using different computational programs and numerical elements to perform static non-linear analyses, assess the capability of a commercial Finite Element Method (FEM) software, SAP2000, to simulate infill walls by means of a diagonal strut, comparing it to a more generic and powerful tool, DIANA. With SAP2000, the analyses were done with and without infills, and these were simulated using diagonal strut bar elements. In all models, hinges with concentrated non linear relations (axial force, bending moment or both, depending on the element) were assigned to the bar elements. With DIANA, the same static non-linear analysis were made, simulating the infill walls by means of a regular mesh, as an alternative to the diagonal truss elements of the previous analysis. Non-linearity was implemented using a total strain crack model. After analysing the obtained results, using bar elements on SAP2000 and DIANA led to similar results, regardless to the presence of infill walls simulated as diagonal struts. The failure modes were not affected by the use of different elements to simulate the infill although, to better analyse this phenomenon, a FEM mesh with a constitutive law that more accurately considers the shear failure should be used