Digital acquisition and analysis of backscattered electron images provide powerful tools for the study of cement-based materials. The techniques can provide useful information on hydration phases, size distributions of unhydrated particles and voids, effects of changes in the watercementitious material ratio and the use of mineral admixtures, and the distribution of microcracks. The results of automated analyses of cement pastes with different water-cement ratios and pastes containing silica fume are presented. The analyses demonstrate that microstructural data vary significantly from image to image, requiring multiple images to limit the effects of scatter. The analyses also indicate that, although the pastes exhibit different degrees of hydration, the size distributions of the unhydrated cement particles are nearly identical. In contrast, the size distribution of larger voids differs significantly as a function of water-cementitious material ratio and with the use of silica fume as a partial replacement for cement. The calcium hydroxide content obtained based on image analysis exceeds but generally parallels that obtained with thermogravimetric analysis. The majority of microcracks in both nonloaded and loaded specimens occur through or adjacent to the lowest density hydration phase.

Backscattered electron imaging of polished cement paste specimens permits a re-evaluation of structural details of hydrated cement paste at the gim level. The primarily microstructural units comprise a highly porous groundmass and large distinct grains (“phenograins”) set in it. The groundmass is composed of several kinds of fine particles, with a significant content of easily detected gross pores. Phenograins are primarily large clinker grains hydrating in-situ, but may be distinct deposits of CH, or may be mineral admixture grains. Detailed EDS analyses indicated that hydrating cement in phenograins has a highly consistent composition, interpreted as C-S-H with a small but regular incorporation of sub-jim CH and calcium monosulfoaluminate. Groundmass particles are highly variable in composition, but appear to consist of C-S-H with variable and occasionally major contents of other hydration products on a sub-μm scale. Incorporation of fly ash does not appear to change the basic microstructure, but silica fume incorporated with superplasticizer drastically modifies the character of the groundmass. Some attempts at quantification of these features by application of image analysis are briefly described.

Cement paste microstructure as revealed in backscatter SEM presents a number of inherent difficulties that interfere with implementing quantitative image analysis. An approach to overcoming these difficulties is presented, involving gray scale segmentation coupled with application of a hole filling algorithm. Using this approach it is possible to isolate the unhydrated and hydrated portions of phenograins separately, and to combine them for analysis of combined phenograins. Pores and coarse calcium hydroxide masses may also be isolated for feature analysis. Results are reported on mature cement pastes prepared at two water:cement ratios (w:c 0.45 and w:c 0.30) and with and without superplasticizer. It was found that superplasticizer greatly reduced the content and the average size of “visible pores” and increased the content and the average size of coarse CH particles compared to corresponding plain pastes. The area per hydrated phenograin was much smaller in the lower w:c ratio pastes and higher in superplasticized pastes. Among the solid features measured, unhydrated cement particles had the smallest circularity values (at about 2.7) and were the most circular features; Hydrated phenograins had the largest circularity values (at 3.5) and were the most elongated features.

Computer modelling of the properties and performance of cement-based materials is complicated by the large range of relevant size scales. Processes occurring in the nanometersized pores ultimately affect the performance of these materials at the structural level of meters and larger. One approach to alleviating this complication is the development of a suite of models, consisting of individual digital-image-based structural models for the calcium silicate hydrate gel at the nanometer level, the hydrated cement paste at the micrometer level, and a mortar or concrete at the millimeter level. Computations performed at one level provide input properties to be used in simulations of performance at the next higher level. This methodology is demonstrated for the property of ionic diffusivity in saturated concrete. The more complicated problem of drying shrinkage is also addressed.

The relationship between microscopical damaging of concrete and its mechanical parameters such as Young's modulus and degree of reversibility is a basic issue. In the same time, transport properties and of course the durability of civil engineering and hydraulic structures are linked to the state of microcracking.

The replica technique has been developed in our laboratory in order to study microcracks with a scanning electron microscope. This non-destructive method allows observation of the imprint of the concrete surface and to follow the evolution of the alterations with time. An objective quantitative analysis requires numerous replicas and many microscopic fields within each replica.

After a brief state of the art of the automatic method of extraction of lines by image processing, we propose a specific algorithm of image analysis which allows the cleaning of the pictures and the extraction of the microcrack skeleton. Today, this procedure has been validated on cement paste samples. It has to be improved to be applied on concrete.

The measured parameters are the microcrack specific lengths and their orientation. The objective values measured by means of this procedure compare favorably to results obtained by hand drawing.

Modern high performance concretes (HPCs) are well characterized by their materials, compositions, and other characteristics, and usually documented by ‘test samples’ from laboratory specimens and trial castings. The characteristics are applied for construction design and estimations of service life time. However, in these young constructions, the internal structure of the HPC often reveals serious differences from that of the ‘test sample concrete’. Accordingly, the characteristics of the constructional concrete may be dramatically different from the expectations, often leading to increased rate of deterioration. A series of examples of such failures in the internal structures from major European HPC civil engineering constructions are presented. A test method is presented for the evaluation of the internal structure of the concrete as it exists in the actual structure. This test method is based on reground vacuum impregnated full sections of concrete core samples, cored from the construction at the moment of stripping the form work. Thus it is made possible to assess the structure during production. The cost of applying the test method approximates 0.1 % of the concrete value.

In order to know the main cause of the cracking of concrete ties we tried to reproduce the deterioration process in the laboratory, using the same thermal cycle and the same materials. Model concretes were first steam cured then examined for ASR using the CSA-A-23-2-14A accelerated test. Linear expansion of concrete prisms were measured and fracture surfaces of concrete after treatment were observed under scanning electron microscope. They showed ASR gels and secondary ettringite. A simultaneous thermal mechanical computation gave the global microcracking of concrete induced by steam curing. A second computation showed the accelerating role of temperature on the local development of ASR products.

This paper reports some preliminary results from a study of the effect of elevated temperature curing on mortars and the phenomenon of delayed ettringite formation (DEF). Mortars made from cements with sulphate levels of 3%, 4%, and 5% and with 5% sulphate and added alkali were cured at 20 and 90° C and subsequently stored in water. Expansion measurements showed a pessimum effect with increasing S03 content. Mortars which expanded showed a corresponding decrease in strength. X-ray diffraction (XRD) studies indicated that no ettringite is present after heat treatment but re-forms over time within the material. However, the ultimate levels of ettringite reached do not correspond to the magnitude of expansion observed. X-ray microanalysis shows that immediately after the heat treatment the aluminate species and most of the sulphate species are incorporated within the C-S-H gel. The concentrations of these species decrease during expansion, such that at the end of expansion the amounts remaining correspond to the presence of AFm phase mixed with C-S-H.

X-ray microtomography can be used to generate three-dimensional 5123 images of random materials at a resolution of a few micrometers per voxel. This technique has been used to obtain an image of an ASTM C109 mortar sample that had been exposed to a sodium sulfate solution. The three-dimensional image clearly shows sand grains, cement paste, air voids, cracks, and needle-like crystals growing in the air voids. Volume fractions of sand and cement paste determined from the image agree well with the known quantities. Implications for the study of microstructure and proposed uses of X-ray microtomography on cement-based composites are discussed.

Microstructural features associated with freeze - thaw deterioration have been studied using scanning electron microscopy and mathematical modelling. Concrete subjected to freeze—thaw cycles in the laboratory and concrete deteriorated in the field were examined in the SEM (BEI). The microcracks produced after a few freeze—thaw cycles are largely vertical and extend deep (several millimeters) into the specimen. With additional cycles horizontal cracks are produced, connecting the vertical cracks to produce spalling. This crack development probably results from localized expansive stresses produced by hydraulic pressure, as described by Powers, plus localized shear stresses produced by differential volume changes.

Granulated condensed silica fume is not easily dispersed. In contact with the pore fluids of the cement paste the granulates rapidly turn into alkali-silica gel nodules. With time this gel absorbs Ca and at least the rim of the AS-nodules transforms into a C-S-H product. The gel by itself does not cause expansion as it forms before the cement paste hardens. However, if exposed to alkali salts the AS-nodules may become further enriched in alkalies, which in turn may trigger an expansive reaction and cracking.

Mechanical properties of any material, including hardened cement paste, are assumed to be controlled by its microstructure. An attempt has been made here to establish a link between bulk fracture parameters of hardened cement paste and its microstructure. Paste microstructure has been varied by changing the initial w/c ratio, curing time and curing temperature, and by addition of chemicals to change the calcium hydroxide morphology. It has been found that, like compressive strength, fracture parameters depend directly on porosity. Contrary to our initial hypothesis, CH morphology was found to have no effect on the fracture parameters.

Analysis of images from SEM, optical and confocal microscopes provides insight into the roughness of fracture surfaces of cement-based materials. Several techniques including 3D surface reconstruction and 2-D vertical contour analysis are compared. Roughness parameters and fractal dimensions are computed to describe the degree of tortuosity of the fracture surface. The tortuosity of the surface (i.e. crack path) is an aspect that has not been adequately integrated into fracture mechanics models of cement-based materials. Topographic maps from confocal microscopy reveal orientation of the crack front throughout the fracture process, allowing analysis of mixed-mode character of nominally mode I fracture. These techniques allow quantification of microstructural parameters to be linked to material properties.

The goal of this research is to improve the physical properties, mechanical properties and durability of cement-based materials by controlling chemistry and processing to produce the desired microstructures and properties. An accompanying initiative is to establish the basic scientific understanding relating processing and microstructure with physical and mechanical behavior, for the future development of new, special purpose, high performance cement-based materials. The research involves both experimentation and theory elucidating fundamental strengthening mechanisms for materials such as warm-pressed, MDF, DSP, polymer, and fiber composites. SEM and computer image analysis were used. Of interest are creating process models that serve as a basis for real applications such as in the following areas: High-performance highway concrete, and the immobilization of radioactive waste.

Fly ash from power stations is used as concrete additive to improve strength and durability. Surprisingly, studies of ashes of identical mineralogical composition from two different places have reported different results in terms of the rheological properties of the fresh material. The viscosity of the pastes made from these different fly ashes seems to be linked to the proportion of spherical and smooth-shaped grains found in them. A quantitative image analysis was carried out to characterize the shape of the grains of these two ashes from different geographical origins. The main result proves that the higher the glassy particle content of the fly ash, the more the hydraulic matrix is fluid.

The morphology of early hydrates of a cement clinker was studied by SEM and EDS. The hydrates were formed on thin plates cut from the clinker and immersed in a solution containing Ca(OH)2, CaSO4·2H20, and varying amounts of sucrose and KOH. The results show the different influences of sucrose and of KOH on the formation of the hydrates, in particular on the crystallization of AFt. The sucrose plays a leading role, and altered the time of primary formation of AFt and its morphology. KOH did not. In CaSO4·2H20 saturated solution without sucrose, regardless of the presence or absence of Ca(OH)2, the foil -like hydrates, CxAHy are formed as the first result of reaction of C3A with water. Then CxAHy reacts with CaSO4·2H20 to form AFt. However, in the presence of sucrose the C3A reacts directly with CaSO4·2H20 to form AFt. It is supposed that sucrose acts as a catalyst for the formation of AFt crystals. During the early stage of cement hydration, AFt crystals appear to form by both topochemical and by through-solution reaction mechanisms.

Calcium sulfoaluminate, 4CaO • 3Al2O3 • SO3, forms the basis of a new class of structural cements. Microstructural examination of commercial clinker is not easy because it is quite friable. However, backscattered imaging of impregnated clinker is successful and reveals an unusual pattern. The clinker has a sponge-like texture connecting better-sintered, denser regions. The microstructure is, however, very similar between regions of differing density. Irregular to equant, well-shaped crystals of 4CaO • 3Al2O3 • SO3 are enclosed in a matrix composed mainly of Ca2SiO4 together with a little titanium-rich, ferrite-like phase. The minor phases occluded within ferrite include oxide phases one of which concentrates Pb and Bi; the other, Cr, Fe and Ni. Other minor oxide phases in the matrix include lime, periclase, gehlenite and calcium sulfosilicate and probably bredigite. The crystal chemistry and elemental analyses of the individual phases are presented.

Gypsum-free portland cement is a low porosity hydraulic binder based on finely groundportland cement clinker with addition of synergetic system containing an anion-active surface active agent (usually a sulphonated polyelectrolyte) and an inorganic salt (usually sodium carbonate) for regulation of the hardening process. The properties of GF cement are different from ordinary portland cement; they display, for example, higher strength, better corrosion resistance and thermal stability. These positive differences arise from the different mineralogy and microstructure of the hydration products, for example the absence of portlandite crystals. The main component of the binder product in hardened GF cement pastes is C-S-H (mean C/S ratio 2.7, based on EDAX analysis) intergrown with very fine Ca(OH)2 and highly dispersed C-A-H phases (hexagonal and cubic). The absence of crystalline formations in the GF hardened pastes is responsible for highermechanical strength. In the Czech Republic, GF cement is produced in the cement works of CEVA Prachovice Inc. ( Holderbank group) and is used for special works in the building industry.

The use of cementitious materials dates back to the beginning of the Epipaleolithic period. Examples for ancient cementitious materials from Israel, Egypt, Turkey, and Italy are numerous.

Prior to Aspdin's patent of portland cement at the first half of the 19th century, cementitious materials were composed of earth, mixture of earth and limestone, calcium sulfates, and slaked lime with and without pozzolans. The latter comprises pozzolanic materials from volcanic and sedimentary origin, crushed burnt clay brick, and dust brick. Frequently, organic fibers were incorporated for reinforcement. This paper describes the evolution of the cementitious materials through time and highlights the durability of ancient cementitious materials as compares to that of portland cement concrete.

Although modem concrete is characterized by its high strength and low permeability, it often faces durability problems. In turn, ancient concretes examined exhibit low strength but have proved to be durable materials. Microstructural examination reveals that the groundmass of the latter has been carbonated and is highly porous. Nevertheless, no specific cracking pattern could be observed. The outstanding performance of ancient concrete structures implies that thermodynamic stability rather that mechanical strength is a key point for a long-term durability.

By mixing various industrial wastes, as garbage combustion ashes, bottom ashes, fluidized bed ashes, lignite power station ashes, fume purification sulfates and sulfites and lime it is possible to produce cements and binders on the basis of alinite, calcium sulfoaluminate and belite depending on the chemical variety of used wastes.

The fabrication process for these cements was studied by laboratory experiments and the different phases and properties were studied in detail. Alinite cement was already produced on a larger scale in a rotary kiln of 10m length. These cements can be used for application purposes in mining mortars, expansive cements, rapid hardening binders and in landfill technologies. The hydration process and workability can be controlled by using various additives. hus industrial wastes can be a secondary resource for special cement production.