Study on Shrinkage Prediction Models and Restraint Crack Formation in Unbonded Post-Tensioned Slabs

Abstract

Most of the time, when developing the structural design of a building, deflection-induced cracking is analyzed and measures to prevent serviceability issues regarding deflection are considered in the design. However, crack formation in concrete structures is also produced by restrain-to-shortening effects induced by volume changes in the members through the passage of time and environmental effects. Therefore, the estimation of creep-and-shrinkage-induced stresses, and the inclusion of restraining crack mitigation details should be included in the slab design process. The first part of this study focuses on the comparison of seven published shrinkage and creep calculation models that aim to predict a concrete members volume-changing behavior. The second part of this study presents a series of architectural configuration prototypes analyzed through a FEM software, where a shrinkage and creep model strain results from part I are used, to determine the combination of shrinkage-and-creep-induced volumetric changes and restraint configuration that can lead to cracking in unbonded post-tensioned slabs and observe the overall behavior of these slabs. From the selected shrinkage models, the GL 2000 model (N. J. Gardner & Lockman, 2001) was the one showing the largest creep and shrinkage results and therefore it was the one selected for the analysis. The analysis results show that restraints tend to increase tension and also compression in the slabs, which suggests that they reduce the efficiency and spread of prestressing forces. The results also show that the increase in length does not exert a significant increase in tensile stresses compared to the effect of restraining walls, and that in case of having perimeter walls a partial wall release would be more effective than a pour strip.