Investigation of methods to analyse global stability of a reinforced concrete stabilising core

Sammanfattning: Ensuring the structural stability of a building is an evident requirement for the building to be safe to use. This becomes a qualified task for a structural engineer. Ensuring stability has become especially important nowadays while stabilising elements often are kept to a minimum and tall buildings are common. A common approach of securing structural stability is to use stabilising shear walls of reinforced concrete. The walls are often placed around staircases and elevator shafts and together they make a stabilising core. Important to keep in mind regarding stabilising elements of reinforced concrete is the effect of material non-linearity, the risk of cracking, creep and second order effects. Regarding global stability, i.e., stability concerning an entire structure, the horizontal forces arising are of great interest while they have to be stabilised by the stabilising elements. This thesis treats methods of controlling global stability of a building stabilised by reinforced concrete core walls. Geometries and loads acting on the analysed structure are based on an already existing building, called Sergelhuset, located in Stockholm City. The analyses are performed using different software applying the finite element method, FEM. Comparisons are made between methods considering linear-elastic material behaviour and a method that are taking the non-linearity that is typical for reinforced concrete into account. A non-linear analysis is generated to represent areal-life behaviour of the structure where material non-linearity, risk of cracking and second order effects are considered. Linear analyses are generated where manual stiffness reduction is applied to take the material behaviour of a loaded concrete structure in to account. Second order effects are applied using the nominal stiffness method. The manual stiffness reduction is implemented using several approaches and are compared to the non-linear analysis. The concrete structure was modelled using two different software, FEM-Design 2021 and ATENA3D. Limitations were made in order to simplify the modelling and to achieve the requested behaviour of the structure. A global buckling instability mode was emerged since boundary conditions preventing twisting or sideways deformation was setup. The models were made so that full interaction between wall elements as well as free rotation in the connection between the floor slabs and the wall was implemented. A low global critical buckling ratio implied that second order effects were to be accounted for. Besides the generation of the models, a literature study regarding global stability of buildings including relevant loads and combinations of loads were performed. Also, relevant properties of reinforced concrete such as creep, shrinkage, cracking, stiffness and reinforcement were discussed. A comparison between the non-linear analysis and the different approaches of stiffness reduction made in the linear analyses was made by observation of the displacements. It was concluded that a stiffness reduction according to nominal stiffness values of the entire stabilising shaft, overestimated the final displacements of the structure. When stiffness reduction was applied on the shaft, only on floors that included elements exceeding the tensile strength of concrete, the displacements were also overestimated compared to the non-linear analysis. However, stiffness reduction of separate elements or of walls that were showing tensile stresses, displayed a displacement behaviour more like the non-linear case. Reduction of elements in the upper part of the shaft that did not show tensile stresses above the tensile strength did not have any significant impact.  The effect of cracking did strongly affect the results while the effect from creep turned out to be quite small. Other factors that strongly affected the displacements were the stiffening behaviour of the connecting floor slabs as well as the connection to the ground. This implies that these parameters should be carefully investigated while performing a stability analysis. A thorough understanding of the software used were shown to be especially important in order to generate time-efficient and stable models.