The paper describes research to evaluate permeability and the microstructure of mortars containing fly ash as a partial replacement for cement. The replacement levels of cement by fly ash were 35 and 50% by volume and 50% by weight; effectiveness of mortars made with four types of fly ash at two water/cement ratios, 0.47 and 0.65, were assessed. Three types of permeability measurements were made: (a) water permeability, (b) oxygen permeability and (c) permeability to chloride ion by the 6-hour rapid permeability test. Measurements were performed at 7, 28 and 160 days age. Strength gain in the mortar mixes was also monitored. In an attempt to explain trends observed in permeability, the pore structures of oven-dried and solventexchanged specimens were examined by mercury porosimetry. Mortars manufactured with fly ash displayed superior engineering properties when compared to the control mortars. Significant correlations were found among measurements of permeability by water, oxygen and chloride ion. Relationships were also observed between permeability and some microstructural indicators, especially the volume of mercury intruded up to a pressure of 41 MPa (6000 psi).

The transport of ions through cement pastes and mortars with variable contents of fly ashes and granulated blast-furnace slag from different sources and with variable composition has been investigated. The research included the determination of chloride diffusion rate and chloride permeability in relation to microstructure development. The median pore size generally was much diminished in mature blended material compared with Portland cement (PC) pastes and mortars. It appears that, at the same age, a finer microstructure is generally developed in blended specimens compared to PC specimens. Also, it was found that the microstructure approaches a limiting value at longer ages of hydration. That limiting value may be reached at earlier ages with the blends. The chloride diffusion rates and permeabilities in the blends were significantly lower than PC mixes. A comparison between the blends containing fly ashes and those containing slag was made.

The effects of curing conditions on the chloride permeability of concrete with various replacements of Portland cement by a flyash and a blastfurnace slag was investigated. In order to relate the porosity and pore size distribution of concretes to their chloride permeability, mercury intrusion porosimetry measurements were also conducted. The results showed that exposure of concretes to a relatively low humidity at early ages increased their chloride permeability. It was also found that the chloride permeability of concrete increased proportionally with increasing volume of pores larger than 0.1 μm in diameter.

Well-cured slag cement blends are known to give rise to paste matrices of low permeability. Electron microscopy has been used to follow microstructural development in slag cement pastes. An initially dense layer of hydrated product develops around the slag grains. Its chemistry and microstructure are not constant but evolve with time. These variations have been explored by analytical electron microscopy, X-ray diffraction, etc. Mg is concentrated in this zone and serves as a useful chemical marker. However an outer matrix of mixed hydration products also develops, to which slag and cement both contribute. The additional contribution of Ca, Al and Si derived from slag to this matrix is considered to explain its gradual densification and resultant low permeability.

The objective of this summary is to report on work in progress that is examining parameters, measurable through chemical and XRD analyses, that could indicate whether a fly ash will enhance, degrade or have no effect on the sulfate resistance of fly ash concrete.

Five compositions with 10% to 50% metakaolin for cement substitution were studied. The rate of hydration was studied from the compressive strength after up to 6 months of curing and from the hydrates formed (DTA-XRD). The metakaolin addition considerably reduced portlandite content in the hydrated cement and contributed to the formation of hydrated gehlenite which is not present in OPC paste. The microstructure study (SEM) shows that pozzolanic cement pastes were less crystallized than plain pastes. Mercury intrusion was used to measure porosity of hydrated cement pastes. The porosity with blended cements was higher than that with OPC, except for 10 and 20% metakaolin substitution. Evolution of the pore size distribution was studied: the pozzolanic pastes enhance small diameters.

Strengths developed in fly ash concretes usually equal or exceed that of similar plain concrete after a few months, with much of the response usually attributed to “pozzolanic” reaction between ash and secondary calcium hydroxide (CH). The CH contents of pastes made with five different fly ashes were determined by DTA for periods up to six months. The CH contents found did not decrease notably over the period, and were substantially identical to that expected for plain cement pastes diluted with the same amount of inert material as the amount of fly ash used. Scanning electron microscope examination of the pastes showed only minimal evidence of reaction even up to 1 year of age, although many fly ash grains were in intimate contact with CH. Non-evaporable water contents of the fly ash pastes were substantially higher than expected at each age, suggesting that the fly ash promoted more complete cement hydration or that the hydration products formed bound substantially greater amounts of water than plain cement paste ordinarily does.

The pore structure of low porosity cement pastes containing varying quantities of silica fume has been examined using mercury intrusion porosimetry (MIP) and nitrogen adsorption (NA) measurements. The water:solid ratio for all DSP pastes studied was 0.18. It was observed that, as for conventional pastes, removal of water by solvent replacement with methanol minimizes changes to the pore structure. Vacuum or oven drying severely reduces specific surface area and obscures important trends.

It was found that capillary porosity in excess of 10 nm (100 Å) is essentially absent and that the pore volume measured can be considered an intrinsic part of the binding phase. The influence of silica fume and curing times on pore structure has been measured and the implications of the data will be discussed. Comparisons will be made with conventional cement pastes.

This study is aimed at the quantitative evaluation of the impermeability of soil-cement in relation with soil conditions, mix proportions and curing ages. Considering the permeability coefficient as the macro property and the pore size distribution as the micro property, this paper presents the relationship between the permeability coefficient and the pore size distribution of the soil-cement. From this study, a dense soilcement with dispersed fine voids is recommended to improve the impermeability of the soil-cement.