In permeameters for rigid materials such as concrete, cement paste, or rock the sample is generally sealed into the test cell with a resin or with a pressurised membrane and the permeability is calculated from the total flow through the apparatus. Clearly any leakage around the edge of the sample or other inhomogeneity in the peripheral region will lead to inaccuracy in the measured permeability. This problem is particularly severe with low permeability materials such as concrete and is further exacerbated by the fact that the true permeability of concrete can vary by many orders of magnitude and there may be little information with which to assess whether particular test results are reasonable.

The divided flow permeameter offers a technique by which edge leakage and other edge effects can be assessed and several permeability results obtained from different areas of a single sample. The basic principle of this permeameter is that either the inflow or the outflow surface, or both, of the sample are divided into separate flow regions. Provided that there is no difference in pressure between the regions there should be no tendency for flow between them and the flow through each should be proportional to its area.

As the technique gives several permeability results for a single sample information about permeability variation within the sample can be obtained; for example data on microcracking, moisture distribution etc. The technique can be used with gas or liquid permeants. The divided flow permeameter system developed at King's College is inexpensive and with a simple system of ‘0’ rings a flow division head can be fitted to many conventional permeameters.

The migration of water vapor through the existing pore system was studied on a series of cementitious, hydrothermally produced and ceramic materials with open porosities ranging between 18 and 79 per cent. Regardless on the nature of the material the water vapor migration rate increased linearly with open porosity. As to the effect of pore size, the apparent diffusion coefficient was moderately greater in pores with radii above 100μm than in smaller pores. As to the effect of relative humidity the water migration rate was greatest at high humidities and exhibited a moderate minimum at about 20–40% r.h.. At equlibrium conditions the amount of water bound in the pore system of samples through which water vapor was alloved to migrate was consistently lower than in identical samples in which just water vapor adsorption took place.

Physical and chemical processes are responsible for the alkali-silica reaction in concrete. Physical processes include (i) migration of hydroxyl and alkali ions (ii) swelling by water imbibition of the alkali-silica gel produced, (iii) development of cracks. Chemical processes involve (i) neutralisation of acidic silanol groups and attack of siloxane bridges, both by OH- ions (ii) formation of an alkali silicate. These reactions result in a breakdown in the silica structure and diffusion of some silicate and impregnate the cement paste filling the capillary pores. The increase in permeability due to the opening and progression of cracks facilitates secondary chemical deteriorations like the formation of ettrinqite, carbonation and corrosion of steel reinforcement. Recent studies have emphasized the role of Ca(OH)2 preventing the dissolution or miqration of freed silica out of the structure. Only the removal of Ca(OH)2 is able to stop the expansion of concrete.

The paper will present some new interpretations of alkali-silica reaction related to the diffusivity of various ions from the pore solution.

Permeability measurements have been conducted for mortars, concrete and hydrated cement pastes. The permeability with water as the penetrating medium has been measured in a low pressure range (˜ 0.5 psi) and at higher pressures up to 400 psi. Samples never exposed to drying and oven dried samples (105 °C) have been investigated. Pore structure of the investigated samples has been characterized using mercury intrusion porosimetry.

The results are discussed with regard to changes in structure during the experiment due to progressing hydration and healing of cracks.

A numerical model has been developed that predicts the freezing process in a cementitious matrix. Freezing point depression due to pore size characteristics is incorporated into the numerical model. An idealized pore size distribution based on a mature paste with a water-cement ratio of 0.6 was adopted. The pore size range was randomly distributed over the test domain. Results are shown for the advancement of the freezing fronts and the progression of the isotherms.

Chlorides are currently the main culprit in the deterioration of concrete structures in most parts of North America and Europe. During the last few years the authors have investigated the following properties: (i) the steady state diffusion of chlorides through cement pastes; (ii) penetration of chlorides into mortar and concrete; (iii) the binding of chlorides in the paste phase; (iv) the threshold value of chlorides below which corrosion of reinforcement is unlikely to occur; and (v)the relationship between the electrical resistivity and diffusivity of paste, mortar, and concrete. The variables investigated include water/cement ratio, cement type, curing conditions, effects of admixtures, and pozzolanic additions and the cover depth.

The influence of shear mixing on selected properties of cement pastes and mortars was investigated by preparing specimens using an ordinary paddle mixer and a high speed shear mixer. The results appear to indicate that shear mixing influences the bond between paste and aggregate, particularily at low water:cement ratios. The properties of hardened cement paste did not change markedly as a result of high speed shear mixing used in this initial study.

The durability of concrete is strongly dependent on the distribution of porosity particularly in the surface layers. To study this distribution, concretes at different water to cement ratios were studied by methanol counter diffusion and by analysis of backscattered electron images. The distribution of porosity in the surface was compared with that in the bulk. Samples of the same concrete were imnersed in sodium sulphate solutions while others were left in lime water as a control. The effects of the immersion on the microstructure was studied with backscattered electrons in the SEM.

In modern society, concrete is one of the materials studied deeply in materials science, and, consequently, it is utilized also in high technology applications. However, the knowledge of gas permeability of concrete seems still to be poor and unreliable. Since 1980, the author has made both theoretical and experimental research to clarify the different aspects of gas permeability of concrete, especially for the applications in nuclear industry. A brief progress report of this work will be presented in this article.

The influence of different admixtures: gypsum, tartaric acid, silica fume and C4A3S upon the HCP developing microstructures has been studied. Investigations of flexural strength, E-modulus, porosity, DT-analysis, X-ray analysis and SEM observations were carried out. The obtained results showed that these samples differ in microstructure and porosity.

The effect of the curing conditions on two properties of concrete, strongly dependent upon the pore structure, namely the strength development and the carbonation rate, was studied in the laboratory. The influence of the cement type, mix proportions and specimen size on the measured properties is also discussed. The importance of the quality of the “covercrete” for good durability is emphasized.

An experimental assessment is presented of the influence of curing regimes on the abrasion resistance of concrete. Specimens have been produced from several different mix compositions and the abrasion resistance has been measured using a rolling-wheel apparatus. To explore these relationships, microstructural features of the cement matrix components of the surface have been studied by mercury intrusion porosimetry. The results indicate the abrasion resistance of concrete is controlled by the pore structure of the surface matrix. The nature of the curing regime having significantly influenced the pore size distribution and total pore volume of the surface matrix of the concrete surface.

The application of transmission electron microscope techniques to the study of cement hydration can reveal the nature of the fine pore structure present in dried cement pastes. Studies of OPC cement pastes and OPC/fly ash or blast-furnace slag blends are presented and compared. Preliminary results of a technique which allows effective imaging of the porosity which is important in permeation, and which is involved in mercury porosimetry measurements, are presented.

Polymer-modified cements were prepared macro-defect free by adding poly(vinyl acetate) to the normal water/cement mixture. The resulting samples, that corresponded to different volume porosity, were subjected to capacitance measurements in the range 5 Hz to 13 Mhz.

Effects of NH4NO3, CH3COOH and HCOOH solutions on the pore structure and other properties of hardened cement pastes made out of portland, slag, and fly ash cements at different water-to-cement ratios are discussed.

Total porosity and the volume of pores detectable by mercury intrusion porosimetry (MIP) increased with increasing water-to-cement ratio, concentration of aggressive solution, and time of immersion. The weight, bulk density, and compressive strenath of the samples decreased at the same time. Use of blended cements did not result in improvement of the long-term durability of the samples. The degradation of the cement pastes increased from NH4NO3 to CH3COOH to HCOOH.

Curing affects sincerely the durability of concrete especially in surface regions. This is reflected in the pore structure and in the permeability. In a systematic research the structural changes due to water cement ratio and curing have been studied by mercury porosimetry, by nitrogen adsorption, by oxygen diffusion and by oxygen permeability. Due to these results diffusion and permeability are connected by a power law. Structural changes alter the permeability by four orders of magnitude. The results can be well explained by pore size distribution both for hcp and for concrete. Carbonation and frost action as well as strength and elastic modulus are closely correlated to permeability and therefore to pore structure. For practical application it is possible to give advice for concreting and curing. Service life can be predicted by permeability measurements in an early stage. The results are comparable with Nyames /20/.

A cement-based grout (90% Type 50, 10% silica fume, 0.4 < water-to-cement ratio, w/c < 0.6) has been used in field trials at AECL's Underground Research Laboratory in Canada and at the OECD/NEA Stripa Mine in Sweden, to evaluate suitable grouts and grouting techniques that could be used for sealing a nuclear fuel waste disposal vault mined deep in granite. Laboratory studies have been carried out to determine the following grout properties: hydraulic conductivity (k); resistance to piping and erosion during setting; influence of grout on the pH and chemical composition of water permeating grouted rock; and the ability of the grout to self-seal after fracturing.

Laboratory tests have confirmed the low intrinsic k of these cement mixtures (10−14 m/s). Using a specially developed cone-in-cone apparatus, we have studied the effect of fracture dilation and temperature changes on the k of thin films of cement. If fractured, the grout has an ability to self-seal and the rate of self-sealing increases with increasing temperature.

The pH and ionic composition of the water permeating grouted fractured granite rock were found to vary with grouted fracture aperture and grout/rock volume ratio. The field tests demonstrated that grout can penetrate and seal very fine fissures (apertures less than 50 μm).

Cement-based grouts are being considered for use as sealing materials in the Canadian concept for nuclear fuel waste disposal. This paper describes laboratory studies of the longevity of these materials, with special emphasis on the effect of hardened grout porosity and cement type. The longevity properties determined for reference grout (90% Type 50 cement, 10% silica fume and superplasticizer) are compared with those of a slag cement grout.

The fractional factorial statistical method of Box-Behnken was used to design a series of leach tests, which covered a wide range of conditions that could occur in a nuclear waste disposal vault. The leach tests have been carried out to determine the effect of temperature, ionic strength of groundwater, and cation exchange capacity (CEC) of clay (which may be in contact with the grout) on the leach resistance of the cement. The temperature ranged from 25 to 150°C and the ionic strength of the groundwaters from 0.0015 to 1.37 mol. Leach rates of Ca and Si were taken as the major indicators of the long-term chemical stability of the grouts.

Preliminary analysis suggests that the reference grout would be more stable than slag cement in the high-temperature environment of a nuclear fuel waste disposal vault. Within the ranges investigated, decreasing the porosity appears not to significantly decrease leach rates.

The seawater leachability of portland cement solidified cadmium and lead wastes is investigated. The synthetic seawater leachates were analyzed for metals content using atomic absorption spectrophotometry. The pH and alkalinity of the leachate was also measured. The cumulative cadmium release after 46 days of leaching was approximately 1.0 percent of the initial total amount added to the portland cement mixture. Lead was not detected in the leachate. The pH was initially higher and the alkalinity lower and subsequently approached the pH and alkalinity values for seawater after 30 days of leaching.

The microstructure of the solidified waste was investigated using the SEM, XRD, MIP and helium pycnometry. Cadmium was detected as cadmium hydroxide while lead was not detected in crystalline form. The total porosity of cadmium and lead wastes were similar however the cadmium waste had a higher volume of pores larger than 0.2 microns.

During the leaching process the surficial microstructure of the solidified waste exhibited a dynamic layer of calcite, aragonite and brucite while the internal structure showed large amounts of ettringite crystals in the cadmium waste only which caused excessive expansion and cracking. A proposed leaching mechanism experienced by the solidified waste is related to the microstructural characteristics of the matrix.

The changes in porosity of OPC and an OPC-fly ash blended cement during hydration have been studied at water/solids ratios of 0.35, 0.47 and 0.59, cured for times of up to 1 year at 25°C. The porosity was measured indirectly by methanol exchange and methanol adsorption techniques and, directly, by quantitative image analysis using backscattered electron imaging in the scanning electron microscope. Measurements of porosity and of remaining anhydrous material by image analysis showed good correlation with indirect methods. Measurement of the diffusion of methanol and of the compressive strength were made in parallel with the determination of the porosity during hydration and attempts were made to relate the properties to the microstructure. For both binders the reduction of total porosity with increased reaction was small. The major change in pore structure was the subdivision of coarse pores by gel to form finer pores. Compressive strength and diffusion properties were dominated by the relative volume of coarse pores.