By-products from two advanced coal technologies, Fluidized Bed Combustion (FBC) and Limestone Injection Multistage Burner (LIMB), were found to be cementitious when mixed with water and compacted. However, exposure to natural conditions in test cells resulted in losses of strength and increases in permeability over a period of years. Changes in mineralogy and microstructure with time in recovered core samples have been characterized by powder X-ray diffraction (XRD) and scanning electron microscopy (SEM). Up to 80 wt% of the core materials had converted to crystalline hydrate phases by the end of two years. Ettringite, gypsum and portlandite were the initial hydration products. In the LIMB materials, formation of thaumasite in microfractures, channels and voids was detected after one year, and in the FBC material after two years. Thaumasite formation was accompanied by reductions in gypsum and portlandite; it did not appear to be forming at the expense of ettringite. EDS examination of many ettringite and thaumasite crystals showed that the former always contained some Si and the latter some Al, which is evidence for ettringite-thaumasite solid solution. Thaumasite formation accompanied marked losses in strength and increases in permeability.

A grout based on portland cement, Class F fly ash, and bentonite clay was developed as part of the closure system of shallow subsurface structures for disposal of low-activity radioactive wastes. Heat output, volume change, and compressive strength of the sealing grout were monitored with time, at elevated temperature, and in physical models, to determine if this closure grout could maintain adequate volume stability and other required physical properties in the internal environment of the disposal structure.

To determine if contact with an alkaline liquid waste would cause chemical deterioration of the sealing grout, cured specimens were immersed in a liquid waste simulant containing high concentrations of sodium and aluminum salts. Over a period of 21 days at 60 °C, specimens increased in mass without significant changes in volume. X-ray diffraction of reacted specimens revealed crystallization of sodium aluminum silicate hydrate.

The new phase has an X-ray diffraction pattern similar to that of the commercial syntheticzeolite, Losod. Scanning electron microscopy used with X-ray fluorescence showed that clusters of this phase had formed in grout pores, to increase grout density and decrease its effective porosity. The testing was repeated at 100 °C for 5 days using a simulant containing sodium hydroxide and aluminum nitrate and the results were similar. Physical and chemical tests collectively indicate acceptable performance of this grout as a seal-system component.

Wasteforms made by reaction at elevated temperature of a highly alkaline simulated low-level nuclear waste solution, having high sodium ion concentration, with a cementitious blend high in fly ash have been studied. Significant formation of Na-P1 zeolite (gismondine framework) and of a sodalite occurred. The time evolution of the crystalline phases over the first 28 days is reported for both adiabatic and isothermal curing, and the role of these phases in microstructure development is discussed. The level of carbonate ions in solution was found to have a substantial effect on strength evolution and chemistry.

A simulated Low Level nuclear Waste (LLW) solution was tested for long term solidification in a cement-based matrix. The waste is characterized by high pH and high concentrations of sodium, aluminum, nitrate, nitrite, phosphate and carbonate. The effect of diluting the waste with additional water was studied. The cementitious matrix was composed of cement, fly-ash and clay (21°, 68° and 11° respectively) with high solution to solid ratio (1 liter / 1 kg.). Mixes were prepared at 45°C and cured at 90°C for 28 days.

Maximum 28 day compressive strengths and early age heat evolution were achieved by diluting the LLW solution to approximately 67& of its original concentration. More dilution led to a lower heat evolution and compressive strength. No dilution was found to give lower compressive strength, and a heat evolution that was delayed, and lower in intensity. XRD spectra showed formation of zeolites and tobermorite at the higher concentrations (67, 85, and 100& of the concentration of the undiluted simulated LLW), with a change from Na-P1 zeolite for 67& of the undiluted concentration to a sodalite at 100&. SEM observations showed a porous system for the low and high dilution rates but a less porous one for an intermediate level of dilution.

Technical requirements for determining the size distribution of capillary pores in hardened cement paste by SEM image analysis are discussed. Results of such measurements are reported for a set of hardened cement pastes of w:c ratio 0.40 and 0.25, and of ages ranging from 1 to 28 days. Pore size distributions based on conventional mercury intrusion porisimetry are presented for the same pastes. Estimates of pore diameters by mercury intrusion are two orders of magnitude smaller than the sizes revealed by the image analysis. Diameters of air voids are even more drastically underestimated by mercury intrusion. Typical micrographs are provided to illustrate the physical reality of the image analysis results, and the technical reasons underlying the conventional misinterpretation of MIP results for hydrated cements are reviewed.

There is a strong motivation to study the interfacial properties of concrete composites because the interfacial region is often the phase where fracture first develops. The aim of this study is to understand phenomena which are unique at high-performance concrete composite interfaces, and how these influence the bulk properties of a concrete composite. Since processes at interfaces must be considered over a range of scales varying from the atomic to the macroscopic, multidisciplinary research approaches are desirable. Model cement/rock (aggregate) and matrix/fiber interaction experiments were carried out. Morphology and microstructure of interfacial regions among mortar/rock, and fiber/matrix were examined utilizing SEM. Computer image analysis performed along a perpendicular to the interface revealed compositional and physical irregularities. The variations in the volume of pores adjacent to interface zones are documented and supported by microscopic observation. The influences of interfacial properties on concrete composite strength and durability are discussed, and influences of fibers on the fracture and fracture resistance behavior are also discussed. Analyses of debonding along interfaces are used to define the role of debonding in fiber-reinforced concrete composites.

The pore structure characteristics of cementitious materials play an important role in defining their mechanical and durability performance. The mechanical properties of pore reduced cements (PRC) have already been reported [5] as a function of the product density. This paper reports recent data on the durability of PRC and attempts to relate the pore properties of the material to its resistance against deterioration in aggressive environments.

This paper presents results obtained for hardened cement pastes and concretes, ordinary and very-high-performance materials, concerning pore structure and moisture properties.

Because of the inadequacy of classical pore investigation methods (mercury intrusion, nitrogenadsorption) to provide the exact pore network characteristics of cement-based systems, notably in the case of very-high-performance materials, a more pertinent type of experiment was carried out: water vapour desorption and adsorption measurements.

From these experiments, isotherms were determined at hygrometrical equilibrium and at constant temperature (T = 23 °C) for every desorption/adsorption path. The analysis of these curves gives textural parameters in the mesopore range, not accessible by most of other methods. Thus, from the adsorption curve, the specific surface area accessible to water molecules was calculated by the B.E.T. method. The statistical thickness of the water film adsorbed on the solid surface was also determined. And finally, application of the B.J.H. method on the desorption branch provides relevant porosity values and pore size distribution of the hardened materials, and especially intrinsic characteristics of C-S-H hydrates.

This makes it possible to point out the similarities and the differences between the materials, regarding microstructure and moisture properties, and to quantify the influence of mix parameters such as water-cement ratio or silica fume content.

There has been much recent progress on the application of impedance spectroscopy (IS) to the study of microstructure and transport in cement-based materials. The IS spectrum allows for the precise determination of bulk resistance, which is a measure of the pore phase interconnectivity, and calculation of the relative dielectric constant, which is related to the capillary pore size and distribution. High values of the relative dielectric constant (σ105) observed in cement paste at early hydration times are the direct result of the microstructure inducing dielectric amplification. Solvent exchange and freezing experiments, combined with digital-image-based computer modeling, have confirmed the role of large capillary pores in the dielectric amplification in young pastes.

The conductivities (σ) and relative dielectric constants (εr) of ordinary portland cement (OPC) pastes were monitored during cooling and solvent exchange with isopropanol and methanol. Dramatic decreases in σ and εr, in some cases over two orders of magnitude, occurred at the initial freezing point of the aqueous phase in the macropores and large capillary pores. The same dramatic decreases in a and er were observed at the onset of solvent exchange. Both effects provide experimental support for the dielectric amplification mechanism within the microstructure on the μm-scale. A secondary dielectric amplification was observed in the frozen and solvent exchanged pastes, which produced dielectric constants on the order of 103. This effect is attributed to amplification on the nm-scale within the layered calcium silicate hydrate (C-S-H) gel microstructure. Additional insight into the variable nature of the C-S-H microstructure was obtained by comparing the dielectric behavior of methanol-exchanged OPC pastes to isopropanolexchanged OPC pastes.

The electromechanical behavior of hardened portland cement paste has been investigated as a function of measurement frequency and DC electrical bias using an inductance technique. Large field-induced shape changes on the order of 104 Å have been found in millimeter thick specimens exposed to moisture. Both reversible and irreversible contributions to the total electrically-induced strain were observed. In addition, strong hysteresis effects were found on cycling a large AC field. Dry samples were found to exhibit no field-induced deformations. The mechanism underlying this anomalous behavior is believed to be an electro-osmotically induced swelling of the gel-pore structures.

NMR is used to study the evolution of the discrete distribution of proton magnetization fractions in hydrating synthetic white cement. The results at 100 hours hydration time are modelled using a trimodel pore size distribution. A good correspondence is found between the NMR analysis and SANS results reported in the literature.

The chemistry of cement, its hydration and the development of microstructure in cement-based systems is extremely complex, and it becomes even more complex in the presence of additives. The elucidation of the mechanisms of these processes is a challenging problem and requires the applications of multiple techniques including the latest microscopic methods. The applications of molecular spectroscopies, surface spectroscopies and microscopies have helpeddevelop models and mechanisms for the retardation of cement setting by Zn, Cd and Pb, the chemical and structural effects of superplasticizers, and the interaction of hydrating cement with aggregates, selective sorbents and fillers. The results of these studies indicated that the inhibition of hydration is controlled by dispersion of various charges present in hyperalkaline solution in cement paste. According to this charge dispersal model, the Ca2+ ions from initial hydration form a tightly-bound bilayer with the negatively charged C-S-H surface. Consequent to this intrinsic process, the metalhydroxy or superplasticizer anions immediately surround the bilayer to constitute a trilayer which inhibits further hydration.

Compressive yield stresses have been measured for pastes (0.35 ≤ w/c ≤ 0.50) of portland cement, calcium aluminate cement, and weakly and strongly flocculated alumina (Φ0 = 0.20) using the centrifuge sediment height technique. Equilibrium sediment heights are reached quickly, allowing all measurements to be taken during the cement's induction period. The compressive behavior showed little dependence on the compressive history. Compressive yield stress was, however, dependent upon initial volume fraction, decreasing as the initial volume fraction increases. This behavior was observed in both the cements and alumina suspensions, implying that strong dependencies on initial structure may be a general property of the compressive behavior of flocculated suspensions.

Stereo pair imaging in a scanning electron microscope was used to document the details ofthe topography of fracture surfaces at various degrees of magnification. The fracture surfaces had been produced by fracture testing of chevron-notched beam specimens according to ISRM Method I. The surfaces examined included the fracture surfaces of dolomite and andesite rocks, cement pastes with and without silica fume, the fracture surfaces from specimen prepared by casting cement paste on previously fractured rock specimen faces, and the fracture surfaces of mortars prepared using the same rock types and cement pastes. The principal topographic features of the various fracture surfaces are described.

There is a strong motivation to study the interfacial properties of concrete composites because the interfacial region is often the phase where fracture first develops. The aim of this study is to understand phenomena which are unique at high-performance concrete composite interfaces, and how these influence the bulk properties of a concrete composite. Since processes at interfaces must be considered over a range of scales varying from the atomic to the macroscopic, multidisciplinary research approaches are desirable. Model cement/rock (aggregate) and matrix/fiber interaction experiments were carried out. Morphology and microstructure of interfacial regions among mortar/rock, and fiber/matrix were examined utilizing SEM. Computer image analysis performed along a perpendicular to the interface revealed compositional and physical irregularities. The variations in the volume of pores adjacent to interface zones are documented and supported by microscopic observation. The influences of interfacial properties on concrete composite strength and durability are discussed, and influences of fibers on the fracture and fracture resistance behavior are also discussed. Analyses of debonding along interfaces are used to define the role of debonding in fiber-reinforced concrete composites.

The understanding of the interface between aggregate and cement is crucial in determining the properties of concrete. Recently, a pushout experimental technique and a theoretical model have been developed to determine the stiffness, strength and surface energy of the interface layer. The validity of these material parameters was verified in the present study by examining the effect of diameter and embedment length of aggregate. In addition, the effect of the pretreatment of aggregate surface and various admixtures was investigated. All pushout tests were performed in a closed-loop manner to obtain the load vs. displacement relationship.

The interfacial zone was further investigated by using backscattered electron imaging and energy dispersive analysis of x-rays (EDAX) to characterize the microstructure of the interface. The relationship between mechanical properties and microstructure of interfacialzone was studied. It was shown that the microstructure of the interface plays a substantial role in the mechanical behavior of the aggregate/cement bond.

The chemical nature and microstructure of the interfacial transition zone (ITZ) are influenced by the interfacial mechanisms of hydration which are primarily controlled by surface and nearsurface phenomena. Recent works from our laboratory involving superplasticizers, metal anions that retard hydration and interaction with silicate based aggregates have established that the interfacial chemistry is of fundamental importance in understanding the mechanisms of CSH formation, surface chemical compound formation and mineralization of the interfacial transition zone and its interaction with the hydrated cement particles and the aggregate. A surface charge control reaction model that accounts for :he importance of calcium and other cations and anions will be outlined and used to discuss the interfacial physical chemistry of the ITZ and its ultimate chemical nature and structure.

A new testing method is presented in which the properties of the bond zone between aggregate and matrix in concrete are investigated. The reason for starting the investigation is the improvement of the lattice model, which is developed at the Stevin Laboratory. In the lattice model the material is schematized as a lattice of beam elements. The fracture properties of the individual beam elements are purely brittle, and the use of fracture energy parameters as a material property can be avoided for simulating the non-linear fracture process. In spite of the fact that a locally brittle law is used, the global behaviourshows softening of the material. An improved optical microscopy technique in combination with image processing is adopted for investigating the fracture process during loading.

The presence of poly(vinyl alcohol) (PVA) in portland cement during its hydration was found to induce changes in the microstructure and behavior of the paste that forms around aggregate. The microstructure was studied with scanning electron microscopy in secondary electron, backscattered electron, and EDX mapping modes and with Fourier-transform infraredspectroscopy. The adhesion between cement paste and aggregate was examined with planar aggregate surfaces. With concentrations of the order of 1% by weight of PVA based onthe weight of the cement, the deposition of calcium hydroxide on the aggregate surface was found to be diminished and the porous layer that usually surrounds the aggregate to be reduced in thickness. Also, the bond between the aggregate and cement paste was increasedenough for the bond to approach the strength of the cement.

Concrete specimens of constant water-to-cement ratio and varying amounts of gravel aggregate were tested under 3-point bend. The fracture toughness of the composite and surface roughness of the paste are determined. Fracture parameters obtained from the Two ParameterFracture Model (TPFM) such as the critical stress intensity factor (KIc) and critical effective crack extension (Δao) are found to be related to the average surface roughness of the paste. The Interfacial Transition Zone (ITZ) and the bulk paste are distinguished by different values of roughness. The surface roughness of the area adjacent to the aggregate particle is evaluated as a function of the distance from the aggregate surface. In particular, the surface roughness of the paste near the aggregate is greater than that of the paste far from the aggregate, and it decreases with distance from the aggregate. The higher roughness of the paste near the aggregate indicates that aggregate particles, and more specifically the ITZ associated with them, act to toughen the paste in concrete.