Immediate Deflection Calculations Of Transition And Cracked Prestressed Concrete Sections

Abstract

Prestressed concrete is commonly used in floor, roof, and bridge members. In prestressed concrete, the steel strands are tensioned/pulled and then concrete is cast around the strands in the form. Once the concrete has set, the strands are cut, and the member is put into compression from the pre-tensioned strands. Concrete is classified as uncracked, transition, and cracked based on the stress in a critical section (ACI, 2019). In practice, the member is designed to remain uncracked during its service life. When this is not possible, the code requires additional calculations that can increase the complexity of the design, one of these being deflection calculations. Transition and cracked section deflection calculations utilize an effective or cracked moment of inertia that adjusts the stiffness according to the relationship between the applied moment and cracking moment ratio. Though these methods are commonly used, current equations and models are unable to estimate the immediate deflections of transition and cracked prestressed concrete sections in flexure reliably and accurately. The goal of this research is to determine if a solution with current methods exists for transition and cracked sections used in practice, particularly double-tee members. A sample of rectangular and double tees (n = 26) with prestressed reinforcement ratios varying from 0.06% to 0.45% was selected. Each specimen was modeled with five different equations to predict the deflection and loaded incrementally, to provide a plot of applied moment to deflection, and loads at specific stress limits and service moments. As the section progressively cracks under increasing loads, the neutral axis gets smaller as the cracks propagate towards the compressive face. This reduces the available area in the section to provide stiffness and resist the loads. Without calculating the cracked section properties, the reinforcement ratio is used to predict the cracked moment of inertia. The methods consider different ways to account for the changes in the member as it cracked, when to begin softening of the member, and alternative moments to shift the behavior closer to the experimental data. Some limitations to this study include a lack of full-scale double tee deflection specimens and specimens with low reinforcement ratios tested to failure. The methods considered in this study did not provide any greater performance than those already in use. As expected, there was general agreement between the estimated and tested data during the uncracked phase. But as the specimens were further stressed with loading past the cracking moment, the results became less consistent between methods. The equations that had the best results were the PCI and Auburn methods. These methods trended conservatively in their predictions and were computationally straightforward. The other alternative methods diverged from the experimental results as the stress in the member increased, making them a poorer choice for deflection prediction in cracked sections. Due to the overall lack of confidence in the deflection methods for transition and cracked sections, the recommended method continues to be the PCI for its simplicity in computation and performance in this group of test specimens. The Auburn method shows promise and can be revisited with the inclusion of additional data from deflections in full sized specimens.