الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 314 |  |  |
| | | from | 314 | | |
Abstract
Nowadays, climate change is one of the most critical issues and became a major global concern all over the world. The primary contributor to climate change is carbon dioxide (CO2) emissions. To tackle the issue of climate change, there have been increasing efforts to reduce CO2 emissions and transition towards a low-carbon future. As Portland cement (PC) industry is a major cause of CO2 emissions, geopolymer concrete (GPC) is being an alternative to PC concrete (PCC) due to its environmental friendliness. On the other hand, concrete buildings are exposed to several natural and manmade hazards, such as fire, which may cause partial or complete collapse of the building. Exposure to fire or elevated temperatures causes several chemical and physical changes in concrete matrix. Concrete slabs are significantly affected by fire when their large surface, relative to its depth, is being exposed to fire, especially from one side, a gradation in temperature over slab thickness is expected. This gradation of temperature inside the concrete slab may cause more cracks and accelerates its collapse. Although previous studies found many advantages of GPC such as early compressive strength, high mechanical properties, low permeability, excellent durability, and fire resistance behavior, the performance of GPC elements under exposure to elevated temperatures needs to be well studied. Moreover, as fire or elevated temperatures exposure tests consume a lot of time, effort, and high cost, the using of machine learning techniques in studying the behavior of GPC under exposure to fire or elevated temperatures has become a pressing necessity. This investigation consists of an experimental part and an analytical part. The experimental part consists of three phases. The first phase was designed to investigate the physical (mass loss) and mechanical properties (compressive, flexural and indirect tensile strengths) of different geopolymer (GP) mortars after exposure to elevated temperatures (200 ℃, 400 ℃, 600 ℃, and 800 ℃). Different parameters including molarity of hydroxide solution (12M, 14M and 16M) silicate solution to hydroxide solution ratio (2,2.5 and 3), sand to binder ratio (1.5, 2, 2.5, and 3), alkaline solution to binder ratio (0.4, 0.5 and 0.6), alkaline solution type (sodium silicate, potassium silicate, sodium hydroxide and potassium hydroxide), binder type (fly ash (FA) and ground granulated blast furnace slag (GGBS)) and curing condition ( heat and ambient curing) were considered. The second phase was designed to investigate the physical (mass loss), mechanical (compressive, flexural, indirect tensile strengths and elastic modulus), and thermal properties (thermal conductivity, thermal diffusivity and specific heat) of different GPC mixes after exposure to elevated temperatures (200 ℃, 400 ℃, 600 ℃, and 800 ℃). Main parameters were binder type (PC, FA and GGBS), coarse aggregate type (basalt, crushed dolomite and gravel), and coarse aggregate size (14 mm, 20 mm and 37.5 mm). The third phase included studying the behavior of reinforced concrete slabs (RC) and reinforced GPC slabs (RGPC) after exposure to 800 ℃ from one side for 2 hours. The RC and GPC slabs were divided into three main groups including non-exposed slabs, slabs exposed to 800 ℃ from tension side and slabs exposed to 800 ℃ from compression side. Flexural behavior and heat transfer throughout the slab thickness were studied. The analytical part includes using artificial neural networks (ANN) to predict the compressive strength of ambient and heat cured GPC mixes after exposure to elevated temperatures. The results of phase one showed that using molarity of 16M, Na2SiO3 to NaOH ratio of 2.5, sand to FA ratio of 2 and alkaline solution to FA ratio of 0.5 gave higher residual mechanical properties of heat cured GP mortar than other ratios. Using GGBS as a partial replacement of FA by 25% enhanced the mechanical properties of ambient cured GP mortars. The results of phase two showed that the type and size of coarse aggregate remarkably affects the mechanical properties of GPC mixes after exposure to elevated temperatures. Residual compressive, flexural, indirect tensile strengths and modulus of elasticity after exposure to 800 ℃ for GPC mix containing 75% FA, 25% GGBS and basalt aggregate were 61.5%, 28.24%, 39.1%, and 39.47%, respectively. Thermal conductivity and thermal diffusivity of GPC mixes were higher than PCC mixes which indicated that heat transfer faster in GPC than PCC. The results of phase three showed that exposure to high temperatures at tension side of RC or RGPC slabs caused more deterioration on the structural behavior. Moreover, RGPC slabs showed higher cracking, yielding and ultimate loads than RC slabs after exposure to 800 ℃. Ultimate load of RGPC slab cast with basalt was higher than ultimate load of RC slab cast with basalt by 102.27% and 41% after exposure to 800 ℃ from tension side and compression side, respectively For RGPC slab cast with basalt, the cracking load, yielding load, ultimate load, stiffness, and toughness after exposure to 800 ℃ from tension side was higher than RGPC slab cast with dolomite by 28%, 26%, 39%, 35.5% and 52.5% and higher than RGPC slab cast with gravel by 876%, 740%, 247%, 942% and 373%, respectively. Higher heat transfer through the thickness of RGPC slabs than RC slabs was observed. Furthermore, predicting the compressive strength of the ambient and heat cured GPC mixes after exposure to elevated temperatures using ANN models were applicable with high accuracy. R2 value reached 0.94 and 0.887 for ambient and heat cured GPC mixes, respectively.