The use of certain aggregate in harden concrete may cause in a particular chemical process in whichvarious silica forms of aggregate react with alkali hydroxides dissolved in the pore solution of concrete,attributing to the alkali silica reaction (ASR). The ASR can produce hydrous calcium-alkali silicate andalkali-silicate gels. This so-called ASR gel adsorbs water and the resulting in swelling expansion, causescracks in the aggregate grains and in the surrounding cement paste matrix leading to loss of strength andreductions in the elastic modulus and durability of the concrete. Therefore, ASR is a major liability for thedurability of concrete structures. Generally, the expansion process occurs due to the formation of gel bythe ASR reaction. It may be deduced that the rate of expansion due to ASR depends on both the contentsof ASR as well as on the capacity available for swelling of the gel.The main objective of the research in this dissertation is to clearly understanding the ASR mechanism as acause of damage. One way to approach the mechanism is to analyze the chemical compositions andstructure of ASR products. A chemical model is presented here to simulate the ASR formation volume indeteriorated concrete due to ASR. The ASR formation was studied with experiments involving siliceousmaterials (Yoro-chert, Seto- chert, Silica sand, and Pyrex glass) and Ca(OH)2 with alkalinity underaccelerated conditions at different temperatures (60, 70, and 80°C). The investigation employedInductively Coupled Plasma Atomic Emission Spectrometry (ICP-AES), X-Ray Diffraction (XRD), 29SiNuclear Magnetic Resonance (29Si-NMR), Scanning Electron Microscopy with Energy Dispersive X-RayAnalysis (SEM/EDX), and Electron Prove Micro-Analyzer (EPMA) analysis. The ICP-AES, XRD, 29Si-NMR, SEM/EDX, and EPMA results showed that the formation of ASR occurred in model system iscompatible to C-S-H product. Considering these results, the prediction of ASR products was made usingPHREEQC program, in which the results were in line with the experiment data. Additionally, the reactivepyrex glass was assumed to be the most reactive material in this work. The expansion measurements wereconducted by mean of ASTM 227 standard (mortar testing), indicating the conclusion that PG is a highlyreactive compound inducing expansion due to ASR. The dissertation is organized into seven chapterswith the references. The contents of each chapter are presented as follows.In Chapter 1, the study context and motivation, the significance for practice, the objectives of the studyand the research methodology are viewed. Attention in Chapter 2 had been paid to the basic chemicalreaction, the model description of ASR expansion, the identification of ASR in deteriorated structures, thelaboratory identification of ASR products and the examination of quality of aggregates. In Chapter 3, thematerials and experiments (including model system, simulation of phase assemblage and ASR-inducedexpansion of mortar) carried out in this investigation are revealed. The results and discussion of modelsystem are described in Chapter 4. It is well-known that ASR is a chemical reaction which is highlysensitive to temperature. With the dissolution rate of soluble silica determined by ICP-AES analysis, theincreasing temperature increases in the contents of soluble silica. It has been pointed out here that the Caions have an effect on the content of free soluble silica in pore solution of cement matrix. This is becausewhen Ca ion is almost consumed, the contents of soluble silica remarkably increased in the solution. Forinvestigation of insoluble products, the C-S-H formed in model system can be attributed to the ASRoccurring between the available SiO2 and Ca ions. In XRD results, the greater calcium hydroxideconsumption supports the idea that Ca ions play an important role in accelerating the ASR reaction, withASR formation of chemical compositions similar to C-S-H. It was noticeable that by the addition of smallcontent of calcium hydroxide, the XRD peaks of C-S-H peaks located 29.0° 2 shift to 30.0° 2 andbecame broader. The spectra of 29Si Nuclear Magnetic Resonance (29Si-NMR) indicated a morereasonable relationship between ASR gel and C-S-H when Ca ions are present in particular. It is believedthat alkali silicate hydrate (Na/K-S-H) may be supposed to be the first product in the sequence of the ASR.IIIPrior to the completed C-S-H formation, C-Na/K-S-H (the Q1 site dominantly) could be formed by theincorporation of Ca ions into Na/K-S-H. Later, excess Ca ions are closely involved in the precipitation ofASR gel, likely to form a more polymerized structure of C-S-H (dominated by the Q2 site). With loweramounts of Ca ions, the Q3 site becomes detectable and finally, the presence of C-Na/K-S-H and Na/K-SHmay be attributed to the complete consumption of CH. According to SEM/EDX observation, the Ca/Siratios of solid samples can be identified. For this result, we also assume the precipitation of C–S–H withan Ca/Si ratio of 0.83 for chert sample and that of 1.66 for Pyrex glass sample. Consequently, thechemical sequence of ASR gel formation for Yoro-chert and Pyrex glass was simulated by PHREEQCprogram (more detailed in Chapter 5). The sequence in the simulated process of ASR can be divided into4 steps and the simulation model tends to give predictions of C-S-H, C–Na–S–H, and Na-S-H that are inagreement with the experimental data from the corresponding tests, as mentioned previously in the XRDand 29Si-NMR results. Hence our simulation strongly confirms that sequence of the ASR of Pyrex glass isthe same as those of both cherts (Yoro and Seto). This also suggests that the ASR examination usingPyrex glass can be used for and would be effective to understand the chemical and physical behavior ofthe ASR induced expansion of mortar and concrete. For this reason, the expansion measurement made atconstant temperatures of 40°C, using two alkali constants (0.6 and 1.2%), allow the ASR reactivity ofPyrex glass to be investigated in this research (shown in Chapter 6). Supporting the evidence by EPMA,the formation of ASR existed in the cracked in the Pyrex glass or in the rim near the reactive site. ForASR product, it was found that at inside the cracks of Pyrex glass, high intensities of Si, Na and K withlow Ca were investigated, in which the Ca/Si ratios of ASR gel were less than 1.0.One of the important constituents of this study is to associate the dissolution rate of soluble silica withmortar expansion due to ASR. Based on Arrhenius law, three parameters including the constant rate ofdissolution at 40°C, the dissolution rate of soluble silica at 40°C and the activation energy (Ea) could bedetermined by considering the dissolution rate of soluble silica at different temperatures (60, 70, and80°C) in model system. The results show a strong relationship between the rate of dissolution of Si fromaggregate in the model and its contribution to ASR-induced expansion of mortar under temperature of40 °C. A quantitative consideration was given to the damage of mortar from a point of view of therelationships among the expanded mortar, the dissolution rate of soluble silica, and the gel composition ofeach reactive-siliceous materials. Ultimately, the simulation confirmed that the ASR gel formation isrelevant to the expansion of mortar. The summary of results is revealed and contributed to therecommendations for future works in Chapter 7.
(主査) 教授 名和 豊春, 教授 廣吉 直樹, 教授 佐藤 努, 准教授 胡桃澤 清文
工学院（環境循環システム専攻）