High strength cement pastes include hot pressed, autoclaved, impregnated low water/cement ratio, macrodefect free, ultrafine particles arrangement systems. The densification of the microstrucure is mainly related to a low porosity and to the formation of poorly crystalline hydrates. In composite systems like mortars and concretes, the interfacial bond between the cement paste and aggregates is moreover less porous and more finely crystallized than the normal “auréole de transition”.

It has been observed that the heat of hydration for concrete under practical conditions not only depends on the type of cement and the content of cement and silica fume in the concrete but also depends on the w/c-ratio of the concrete. On.microscopic examination of concrete having low w/c-ratio it has been observed that the cement compounds are left partly unhydrated. However, the unhydrated cement compounds exist for all values of the w/c-ratio, but appear to increase with decreasing value of the w/c-ratio, particularly when w/c < 0.35.

This fact wIll influence the present way of estimating the heat of hydration when a thermal stress analysis is to be carried out for an ultra high strength concrete in order to prevent the concrete from crackling due to temperature differences across a massive cross-section.

The tests are carried out with concrete having a compression strength greater than 100 MPa (∼ 14,000 psi). In order to obtain samples drilled from the same casting for compression tests, microscopic investigatlon, penetration tests and freezing/thawing-investigation rather large specimens are used. A very effective insulation of the concrete is provided, and the temperature of the concrete is recorded by thermo-couples.

The structure of the cement paste is studied using petrographical examination of thin-sections and polished samples. The tests on penetration and freezing/thawing are not dealt with in this paper.

The microstructures of high strength pastes of OPC and high alumina cements prepared by the high shear mixing of a low water/cement ratio paste with water soluble polymers have been studied by transmission electron microscopy. In the case of high strength OPC, the usual hydration products are present, however, the CSH gel lacks the fibrillar morphology often observed in conventional cements. Pastes based on high alumina cement do not contain the normal crystalline calcium aluminate hydrates but a small quantity of gel containing the organic polymer forms a continuous network structure bonding clinker grains. Microanalysis of the polymer phase in high alumina cement revealed the presence of Ca and Al while Ca-rich interstitial gel material was found in OPC pastes.

Ultimately the strength of cement is determined by the properties, amounts and distributions of the phases it contains and by the'way these are bonded together. Here an example is presented of the determination of the amount and distribution of calcium hydroxide, by BEI and quantitative image analysis, in a cement paste and cement/fly ash mixture which had previously been studied by TGA. Examination of crack paths on BEIs of polished cement surfaces gives some indication as to the importance of calcium hydroxide in determining the properties of the cement.

Materials of very high flexural strengths (> IOOMPa) can be made by high shear mixing of cement pastes with small amounts of water soluble polymer. Two such systems - high alumina cement/partially hydrolysed polyvinyl acetate and OPC/polyacrylamide - have been examined using a variety of techniques including isothermal calorimetry, infra-red spectroscopy and X-ray diffraction. In both systems the polymer addition appears to become cross-linked by ions released by the cement powder leading to an increase in stiffness of the material. In particular the reactions involved in the OPC system are analogous to those occurring in certain dental cements. The importance of removing excess water from the pastes is demonstrated.

The mechanical properties of cement-based materials must be controlled by the microstructure, pore geometry and chemical composition of the cement, by the properties of the aggregate, and by the nature of the cement-aggregate bond. While the precise form of the strength vs. microstructure relationship is as yet only imperfectly understood, enough is known to permit us to predict what alterations in the microstructure are required for the production of materials with very high strengths. There are also techniques available for reducing the brittleness that is often a characteristic of high-strength materials. The present paper presents an overview of the strength vs. microstructure relationships that can be used to predict the properties of high strength cement-based materials, and a brief review of some of the methods for achieving high strengths.

Autoclaved cement-silica mortars having flexural strengths of up to 57 MPa were prepared by pressing and casting methods of fabrication. It was found that the Griffith dependence of strength upon the size of the largest flaw, the modulus of elasticity and the surface energy was applicable. The relative contributions of these three parameters were evaluated and it was established that although the maximum flaw was the dominant factor in a specific material, surface energy changes were responsible for major strength differences associated with water-solid ratios or variations in the preparation method. Preliminary studies identified the various microstructural features that operated as Griffith's flaws.

A model has been developed for the flexural strength of Portland cement pastes, based upon observed fracture behaviour of both normal and high strength pastes. Fibrillar or foil-like elements pull apart at a yield stress which is characteristic of the number of elements and interfacial shear strength. The former can be maximised by using a low water/cement ratio and the latter by inclusion of water-soluble polymer, followed by suitable drying. It is proposed that this is the mechanism by which high strength may be attained in Portland cement.

It is frequently reported that the higher the strength of cement based materials, the more brittle is their behavior. It could he useful to quantitatively express the degree of brittleness. Many attempts [1–13] have been made to use linear elastic fracture mechanis (LEFM) to quantitatively express the degree of brittleness. For example, by testing notched beams one can calculate, using the formulas developed from LEFM, a quantity called fracture toughness and termed KIC from the measured maximum load and the initial notch-length. Unfortunalely, it has been observed that K thus calculated is dependent on the dimension of the beams. Many researchers have attempted to analyze this size dependency. Such approaches are usually quite cumbersome and are often based on expensive nonlinear finite element programs. In this paper a direct method is suggested to calculate two size-independent fracture toughness parameters from the experimental results. The method was developed based on tests on notched-beams of different mix proportions and different sizes.

Characterization of cementitious composites by their fracture properties has been difficult due to controversial results reported in the technical literature. Existing studies on the fracture behavior of plain concrete reveal some fracture characteristics that differ from those normally observed in metallic materials. Among these characteristics is the existence of a micro-cracking zone or process zone at the tip of an advancing crack. The determination of the fracture process zone in concrete is a difficult experimental problem, because the resulting deformation is strongly localized.

In the present study, in-plane displacements in front of notched high strength concrete have been monitored. Laser speckle metrology, which is a special technique utilizing the speckle patterns of laser light for measurement of in-plane displacements, is employed. Experimental results indicate that a precise description of the fracture process zone is possible by speckle metrology.

Partial replacement of cement by fly-ash allows the production of concretes with high strength and low permeability. The correct analysis and prediction of engineering behaviour requires a knowledge of the development of pore-structure of these materials. However, a study of the relationships between engineering and microstructural characteristics has revealed problems in the accurate assessment of pore-structure.

Porosities of plain and blended pastes were analysed by both helium pycnometry and mercury porosimetry. Pastes showing pozzolanic reaction gave values of helium porosity whidh were different from the mercury porosity (measured on the same sample); pastes showing no pozzolanic reaction gave equal values for mercury and helium porosity. Also, significant differences in porosity and pore-size distribution were found for identical specimens when prepared by different techniques, namely direct oven-drying and solventreplacement; these differences occurred whether there was pozzolanic reaction or not.

It is suggested that as well as experimental difficulties, there may be fundamental differences in the way pore-structure develops in plain and blended cement pastes.

The pore-structure strongly influences the carpressive strength of hardened cement paste (hcp) and other porous materials, as well as other mechanical properties. The simplest but most currently used expression representing the relationship between the pore-structure and compressive strength is fram Balshin: σ = σ0 (l-P)A, in which only the total porosity P is involved as a single parameter and σ0 and A are empirical constants. The influence of pore size distribution and pore shapes etc. are not considered.

The authors introduce second parameter w - the factor of relative specific surface area of the pores other than the total porosity P into consideration and a new expression is proposed:σc=K11-p/1+2p(K2(1-p))K3w+K4 all the constants K1 - K4 can be determined experimentally. By using of this expression the new information relating the influence of pore-structure on the caopressive strength of hcp can be predicted.

Substantial increases of the strength of cement paste and mortars may be obtained in conventional processing by optimizing the materials components, the rheology and the curing, and thereby improving the microstructures. Cementitious materials with high proportions of granulated blast-furnace slag have been investigated. Higher strengths of ASTM C 109 mortars were obtained with 40 to 65% substitution of portland cement by slag, than with ordinary mix compositions and processing.

For one set of mixtures, 28 day strengths ≥ 100 MPa (some as high as 240 MPa) were consistently attained after curing at temperatures ranging from 27 to 250°C. The slag substitutions developed finer pore structures as revealed by intrusion porosimetry measurements, than those with pure portland cement. This is believed to be a major reason for their enhanced durability. At each stage from 3 to 28 days, increase of curing temperatures from 27 to 90°C decreased porosity and increased the strength, reflecting an increased maturity.

Implications for practice and suggestions for further work are discussed.

The concept of ‘intrinsic strength’ of cement was addressed by testing the bending strength propertes of MDF cement pastes[1–5], containing no organic material, over a range of pore volumes and flaw lengths. The MDF pastes, made only of cement and water, exhibited interesting behaviour, particularly the well packed compositions which gave remarkably low porosity, about 1%, and high modulus, 75 GPa. It was demonstrated that the bending strength of these materials depended on flaw size in a brittle fashion, and that the toughness increased as the volume porosity fell. Fracture mechanics described the results satisfactorily. Therefore, theories based on intrinsic strength, or on the need for polymers to provide high strength [7, 10] are unnecessary.

For concrete, a grossly heterogeneous material, guidelines for selection of materials for its production, especially high strength concrete, should consider the interactions of its different components under load. Materials selection and production techniques for achieving concretes, mortars and cement pastes with strengths in excess of 10,000 psi are discussed. Emphasis is placed on the production of concretes having a compressive strength in the range of 10,000 to 15,000 psi using readily available materials and conventional production techniques. Emphasis is placed on the practical and technical significance of the factors involved in the selection of the materials and their proportions to achieve uniform, economical, high quality concrete. Selection of the concrete ingredients and their proportions is discussed in terms of their relative contribution to the compressive and flexural strength, elastic properties and observed failure mode of the paste, mortar and concrete. Production of concretes having compressive strengths in excess of 15,000 psi using exotic materials and special production techniques is also discussed.

The current ultimate strength design practice is based on experimental information obtained from concretes of compressive strengths in the range of 21 to 42 Mpa. For developing a satisfactory procedure for design of structures using higher strength concretes, additional considerations, and validation or modification of existing strength design methods may be necessary.

Lightweight aggregate concrete under triaxial compressive load shows a different failure behavior compared with normalweight concrete. The increase of strength due to confining pressure is not as large as it is with normalweight concrete. It depends on the type of aggregate. The larger the stiffness of the aggregate the larger the increase of strenoth. Under hydrostatic pressure the low strength lightweight aggregate concrete (Bc ∼ 15 N/mm2 ) showed a confining strength which was about 2 to 2,5 times higher than the urfiaxial strength. Concrete with aggregates with high stiffness (E ∼ 14,000 N/mm2 ) and a uniaxial strength between 33 and 48 N/mm2 reached at a confining factor of 25% a strength increasing factor of about 4. The volumetric change curves show two remarkable facts. With confining factors of 0,05 and 0,10 lightweight aggregate concrete has at a stress level of 80% to 90% of the confining strength a volumetric increase which changes into a volumetric decrease shortly before reaching the ultimate load. With a confininn factor of 0.25 the volume decreases constantly until the fracture happens. The temporary volumetric increase is caused by a widening of the cracks. The fracture of lightweight aggregate concrete under confining pressure is mainly caused by an internal collapse of the aggregate grains.

Until relatively recently, concrete mix materials were selected and proportioned on the basis of certain general principles established through empirical tests, past experience with similar materials, and published tables. Increasingly, however, wc. need to produce high-strength concretes with what has been considered sub-optimal or even marginal quality materials. And, whereas in the past we could afford to achieve high strengths at age 28 or 56 days, there is much greater emphasis now on achieving strengths in excess of 8,000 psi in 24 hours or less.

In response to these new demands for high-strength concretes, the concrete industry is setting aside many of the traditional methods for selecting and proportioning materials. Now, mix materials are selected deliberately for the characterisitics of their microstructure (e.g., the mineralogy and particularly crystal structure of the aggregates), and are proportioned to optimize the microstructure of the concrete.

In this paper the author illustrates how the characteristics of a material's microstructure are used as tools in selecting and proportioning mix materials. The significance of the microstructure of the aggregate particles, mortar-aggregate interface, and mortar are discussed in detail.

During 1984 summer, an experimental concrete column was cast in a 26-story high-building using a 100 MPa silica fume concrete. Only the coarse aggregate was selected specifically for this concrete. The cement used was an ordinary type 1 Portland cement. The superplasticizer was a naphthalene base. The field conditions were among the most unfavourable, the transportation was about 3/4 of an hour. The placing lasted 15 minutes. However, after 1 h 00 the slump of this concrete having a water/cementitious ratio of 0.25 was still higher than 100 mm so that the concrete was placed without any problems.

The strength development of mortars made with a high strength paste containing microsilica was investigated as a function of aggregate type and content. Strength could be modelled using composite theory; the predicted strengths of the paste and aggregate component reflected the influence of the cement-aggregate bond. Scanning electron microscopic examinations gave a qualitative idea of the extent of bond failure, which correlated satisfactorily with the predicted behavior. The modulus of elasticity for SiC mortars could be predicted well by Hirsch's model or the aggregate model.