An alternative simplified method of tensile membrane action (TMA) of thin-flat concrete slabs at ambient temperature was extended to consider composite slabs in fire. The existing simplified method (Bailey-BRE) of estimating the influence of tensile membrane action on the slab strength in fire has some shortcomings in its assumptions and results. The alternative simplified method showed reasonable results, avoiding inherent flaws in the existing method that cause large differences in the results when a marginal change in reinforcement ratio is made.
This thesis takes into consideration the presence of a secondary downstand steel beam which was left unprotected against elevated temperatures. The main objectives of this study were: 1. to examine the mechanics of tensile membrane action of composite slabs in the presence of an unprotected secondary beam; 2. to examine the similarity of the structural behaviour of composite slabs with the results of the alternative method of flat slabs in terms of the benefits of large deflections; 3. to compare the extended method in this study against the existing Bailey-BRE simplified method.
The mechanics behind this study was divided into two main parts: 1. small-deflection analysis; 2. large-deflection analysis. The small-deflection analysis is based on the yield-line theory, where the slab was heated from the bottom side, affecting the unprotected secondary steel beam, until the yield-lines formed in what is known as the “optimum yield-line pattern”. The second part of the study took this yield-line pattern configuration. With further deflection, tensile membrane action started to mobilize, and the slab gained further capacity. The method accounts for the internal work dissipation by the reinforcement and the steel beam as they extend, which contributes to the total internal work of the slab. The small-deflection analysis was coded in Visual Basic in a Microsoft Excel spreadsheet, increasing the temperature to obtain the optimum yield-line pattern. The large-deflection analysis was coded in Matlab in order to conduct a deflection-controlled loop, increasing the temperature until the failure of the slab, giving the temperature enhancement for every deflection level through the analysis.
The study allows the stress pattern around the yield-lines to be monitored. It also monitors the formation of the cracks on the slab, reinforcement fracture in different phases as the temperature increases and the slab deflects, until the structural failure. Comparisons against the Bailey-BRE method confirmed that the latter is conservative when small size reinforcement is used, as previous studies showed.
