Thesis (MScEng (Civil Engineering))--University of Stellenbosch, 2010. / ENGLISH ABSTRACT: Pre-cast sections have the advantages of structural efficiency, better quality control and less construction time, which enable them to be widely used in building structures. The connections of pre-cast buildings play a vital role for the stability and strength of structures. Nowadays, more attention is drawn to the aesthetical appearance of building structures, especially by architects. The Hidden Corbel Connection (HCC) was then developed to make the building structures stable and aesthetically pleasing. A modified HCC was designed and investigated in this study. Amongst all the mechanisms in the connection zone, the mechanism of the end anchorage length of tension reinforcement plays a key role in the economy of the connection and is hence further investigated. In order to investigate whether the end anchorage length of tension reinforcement can be reduced for a simply supported beam, a 2D non-linear finite element model is used to analyze the stress distribution inside the connection zone. Based on the stress distribution in the connection zone, the tensile force was calculated at the face of the support, which directly correlates to the required end anchorage length of tension reinforcement. The confinement in the connection zone increases the bond stress, which in turn reduces the required anchorage length of tension reinforcement. Therefore, a 3D model is used to analyze the region inside the modified HCC to find the position of the best confinement. By comparing the finite element (FE) results with Eurocode 2 (2004), and SABS 0100-1 (2000), it is demonstrated that the FE results require the shortest anchorage length, while the longest anchorage length is specified in SABS 0100-1 (2000). Based on the comparison between the FE results and the design codes, a laboratory experiment was then performed to determine if the end anchorage length of tension reinforcement can be reduced. Four beams with different support conditions and with different end anchorage length of tension reinforcement were tested. The results of the laboratory experiment indicate that the end anchorage length for simply supported beams can be shortened from the specification of SABS 0100-1 (2000).