The relationships between the autogenous shrinkage, and the hydration reaction of cement and the structural change of hardened cement paste have been investigated by hermetically curing the cement pastes of normal portland cement, type B - blast furnace slag cement and 10% silica-fume blended cement prepared at W/C of 0.5 and 0.25 and continuously measuring the humidity changes and the shrinkage strains of hardened cement pastes to obtain the basic data for elucidating the mechanism of autogenous shrinkage. Although there was a time lag between the autogenous shrinkage and the humidity change of hardened cement paste, no autogenous shrinkage took place in the cement paste prepared at W/C of 0.5 showing no humidity reduction. The autogenous shrinkage observed in the cement paste prepared at W/C of 0.25, therefore, is considered to be caused by the self dessication at a relative humidity (RH) from 100 to 80%. The autogenous shrinkage occured mainly during a period from 8 hours to 4 days and slightly increased after that. It is considered that the autogenous shrinkage takes place because the free water contained in pores particularly in fine gel pores formed by producing a large quantity of C-S-H is consumed by the hydration reaction and the humidity in the hardened cement paste is reduced. The autogenous shrinkage of type B - blast furnace slag cement paste was approximately 1,800~ which was about 1.8 times that of normal portland cement paste. Since type B - blast furnace slag cement paste produces more C-S-H than normal portland cement paste, thereby causing remarkable autogenous shrinkage. The autogenous shrinkage of 10% silica-fume blended cement paste is about 1,000u almost same as that of normal portland cement paste, since the pozzolanic reaction is suppressed in its paste prepared at W/C of 0.25 up to 4 days.

At the early stages of development, RCD concrete often tended to be inferior in strength and durability to any other conventional dam concrete because RCD concrete had the properties of stiff consistency and lean mixture. In this research, a laboratory test was conducted to determine its optimum sand percentage, sand ratio and optimum mixture proportions in order to improve durability of RCD concrete. In addition, a field test was conducted to determine the optimum lift height and optimum compacting method. Furthermore, a new construction method was developed to improve durability of RCD concrete that any existing RCD concrete can be strengthened by casting one of the conventional types of dam concrete on the upstream or downstream slope of the dam. Hence the proposed method in combination with the recommended mixture proportions and construction method has enabled one to rebuild many of the dams and to keep then in serviceable condition for a long time.

The paper presents a brief review of the literature pertaining to the effect of lithium salts on alkali-silica reaction (ASR) and some recent data from a laboratory study of the efficacy of lithium hydroxide monohydrate (LiOH.H2O) in suppressing ASR expansion in concrete containing reactive U.K. aggregates. A synthesis of the published data indicates that the effect of lithium on ASR is not strongly influenced by the form of lithium used and that mortar-bar expansion may be effectively eliminated if sufficient lithium is added such that [Li/(Na+K)] > 0.70 to 0.80. -Expansion tests were carried out on concrete prisms cast with two different reactive U.K. aggregates and a range of cement and alkali levels. Some mixtures mcorporated lithium hydroxide monohydrate (LiOH.H2O) such that the lithium-cement alkalies molar ratio was 0.74 (,i.e. [Li/(Na+K)] = 0.74). Lithium was effective in reducing the expansion and preventing cracking in all the concretes; the maximum expansion observed in lithium-bearing concretes was 0.023% after 1 year. Higher expansions were observed in lithium-bearing concrete containing a highly-reactive siltstone compared with similar concrete containing flint sand, although expansions were still low. Although little significance can be attached to these differences, the possibility of a relationship between the safe level of lithium and the level of reactivity of the aggregate is discussed (i.e. more reactive aggregates require higher dosages of Iithium). Lithium additions resulted in shrinkage of fly ash concrete containing reactive aggregate. It is probable that lower levels of lithium are required when used in combination with pozzolans, if the level of such material is not sufficient to completely eliminate expansion by itself.

Cores from five concrete pavement sites in eastern South Dakota, U.S.A., have been examined for microstructural and microchemical evidence of alkali-carbonate reactivity (ACR) using optical microscopy, scanning electron microscopy, and energy-dispersive X-ray microanalysis (EDX). Dolomitic aggregate used at each of the five sites shows evidence of reaction with the cement paste resulting in dedolomitization products and textures. In altered aggregate, dolomite (CaMg(CO3)2) is replaced by an irregular, fine-grained (l-10 um) intergrowth of crystals that have either Ca or Mg as the only peaks in their EDX spectra. These reaction products are interpreted to be calcite (CaC03) and brucite (Mg(OH)2), respectively, as would be expected from dedolomitization. Dedolomitization is most prevalent in fine aggregate, and is generally concentrated just inside the aggregate rim or along cracks penetrating into the particle. An unreacted rim of dolomite, 5-10 um thick, is present, however, even in the most altered particles. Unreacted dolomite and dedolomitized zones have nearly indistinguishable Ca/Mg, indicating that Ca and Mg are conserved in the reaction, and Si02+A1203 is typically < 1 % by mass, so it is unlikely that clay mineral impurities play a role in the deterioration of these aggregates.

Nine bridges under the control of the Roads and Traffic Authority (RTA), in New South Wales (NSW), were inspected in two different regions of the State, and seven with various degrees of cracking and/or repair were selected for examination, particularly with respect to alkali-aggregate reaction (AAR). Twenty-eight cores were drilled from four bridges in the north-east coastal area and 14 cores from three bridges in the same area but 250 km inland. Detailed petrographic examination and scanning electron microscopy of the cores has shown that of the three inland bridges, definite signs of AAR are present in two bridges but the third bridge has a less clear diagnosis, although the aggregate is the same. For the bridges in the coastal area the glassy basalt may have reacted, although typical visual signs of AAR products were not evident. Clay alteration products in the basalt used for two of the bridges may have caused some dimensional instability. For one of the bridges, signs of alkali-carbonate reaction (ACR) were seen in some cores, but the contribution of this reaction to cracking of the bridge is not clear. This is the first report of ACR in Australia. The expansive behavior of cores and alkali levels are also discussed. subjected to elevated temperature, humidity,