Examining of the effect of high temperature on cement mortar properties and modelling by artificial neural network

The changes in physical, chemical and mechanical properties of the materials, which are exposed to high temperature, should be known before in order to assess durability of the constructions. In this study, it was tried to determine the mechanical strength of Portland cement and Portland composite cement in high temperature by artificial neural networks. Experimental study was implemented in six steps as sample production, sample cure, temperature effect, cooling process, and artificial neural networks, physical and mechanical tests. The samples, which are prepared to determine the effect of high temperature on mechanical properties, were heated neutral in eight different temperature degrees as 20, 100, 300, 400, 500, 600, 700 and 900 °C on 7, 28 and 90 th days. Ultrasound velocity, specific gravity weight, moduls of dynamic elasticity, expanse percent, tensile and compressive strength of samples that were cooled to room temperature were determined. Internal structure changes because of the corruption in concretes exposed to high temperature effects were investigated. Experiment results were analyzed statistically and by artificial neural networks, and effect of high temperature on mortars was determined. Consequently, it was seen that compressive strength up to 300 °C has increased than to control samples and the strength has decreased continuously after this temperature. Furthermore, it was determined that there is a difference statistically for mortar samples exposed to high temperature when cooled by air and water; and that tensile and compressive strength should be preticed in an assessment made by artificial neural networks by maximum 8% margin of error.