Crack control in base-restrained reinforced concrete walls

Abstract

Following casting, concrete undergoes early-age thermal (EAT) and long-term (LT) shrinkage volumetric changes. If restrained to move, concrete invariably cracks due to its low tensile strength. Crack control is of particular concern in structures like retaining walls, liquid-retaining tanks, and cut-and-cover tunnels, where through-cracks can lead to water leakage unless their width is adequately controlled with steel reinforcement. The aim of this thesis is to increase the confidence with which engineers can predict and control crack widths in reinforced concrete (RC) walls with edge restraint and in walls with combined edge and end restraint. The research compares reinforcement areas required to control crack widths to Eurocode 2 (EN 1992) and the previous UK code (BS 8007). EN 1992 can require very different areas of reinforcement to BS 8007 to control crack widths – more in some situations (e.g. in walls with end restraint and in thick sections) and less in other (e.g. in thin sections) – both being of equal concern to the construction industry. In addition, there is no guidance for reinforcement design to control cracking in walls with combined edge and end restraint, which is very common in practice (e.g. RC wall cast on a stiff base and between adjacent pours). In such situations, the engineer very often uses end restraint design equations leading to onerous designs. Experimental data on edge-restrained walls are limited, and data for walls with combined edge and end restraint are not available in literature. For these reasons, an experimental methodology has been designed and developed by the author to investigate the influence of different reinforcement arrangements on early-age (EA) and LT crack widths in RC walls restrained at their bases and in walls restrained at their bases and ends. The tested walls measured 3.5 m long by 180 mm thick with heights of 500 mm or 750 mm and were monitored over a period of several months to allow for both EA and LT shrinkage cracks to develop. Temperatures, wall displacements, surface strains, crack widths and crack spacings were carefully monitored over this period. Because of time, cost and laboratory space constraints, it was not possible to systematically vary all parameters believed to influence cracking in the tests. A non-linear finite element analysis (NLFEA) program, ADAPTIC, was thus used as an important tool to extend the laboratory study using time-dependent and time-independent models. Initially, test results were used to validate the NLFEA. Once verified, parametric studies were conducted and the influence of various parameters not investigated in the tests were carried out, including the effects of the ratio of bar diameter to reinforcement ratio, wall aspect ratio and wall height on crack widths in edge-restrained walls and in walls with combined edge and end restraint. The main findings from the experimental and numerical investigations are highlighted in this thesis. In particular, the thesis highlights the importance of the wall geometry (i.e. wall aspect ratio and wall height) in the prediction of crack widths in edge-restrained walls. This thesis concludes by comparing these findings to available design code rules and by suggesting an improved method to design reinforcement against cracking in edge-restrained walls. Test and NLFEA results suggest that code end restraint equations do not give sensible crack width predictions in walls with combined edge and end restraint. This thesis suggests to design reinforcement against EA and LT cracking in walls with combined edge and end restraint based on the design method used to design reinforcement in edge-restrained walls.