SP192 In 2000, CANMET, in association with ACI, the Japan Concrete Institute, and several other organizations in Spain and Canada, sponsored a fifth international conference held on June 4-9, 2000, in Barcelona, Spain. More than 120 papers from 35 countries were received and peer reviewed in accordance with the policies of the American Concrete Institute; 73 were accepted for publication. The accepted papers deal with all aspects of concrete durability. In addition, several sessions dealing with sulfate attack, superplasticizers and supplementary cementing materials, and near surface testing for the durability of concrete were organized. In addition to the papers that have been published in the refereed proceedings, more than 30 papers were presented at the conference.

In Portugal, fly ash is sometimes used as a cement replacement material in concrete for road pavements. The percentages of replacement generally range from about 30% in the top pavement layer up to 50% in the bottom layer. A study was carried out to evaluate to what extent the water curing practice could affect some of the properties of concrete with different amounts of cement replaced by fly ash. For this purpose, several properties generally related to the durability of concrete pavements were determined. Abrasion test, as well as determinations of compressive strength, capillary absorption, oxygen permeability and open porosity, were performed on concrete mixtures with different cement and fly ash contents, using limestone coarse aggregate and natural siliceous sand. This paper presents results on the influence of different water curing times on these properties. The results show that fly ash concretes can develop satisfactory resistance to abrasion, even with large amounts of cement replacement. Concretes with lower permeability and higher strength generally tended to exhibit better abrasion resistance. In general, concrete performance was improved with longer water curing times.

A series of nine large concrete blocks (750 mm cube), stored outside on an industrial site at the Appleby Frodingham works at Scunthrope in N.E. England were assessed at 7-8 months, 30 months and 7.5 and 10.5 years of age. The concrete contained a normal total cementitious content of 390kg/m and had different levels of ground granulated blast furnace slag in the range 0, 30, 50 and 70% by weight as replacement material for portland cement. Some of the concretes were prepared using gravel and others with crushed limestone as aggregate. The quality and performance of these concretes were determined at each age by testing drilled cores taken from the side face of the blocks, partly protected form driving rain. The assessment involved early hear release data; measurements of the depth of carbonation; oxygen permeability and compressive strength. Comparisons were made between the different concretes on the basis of their comentitous slag content an aggregate types. The results showed that concretes containing 100% normal portland cement or 50% slag have hardly carbonated, although carbonation had progressed to 5 to 6 mm at 2.5 years where 70% slag was used, but there was little change thereafter. Gas permeability decreased slightly as the slag replacement levels were reduced from 70% to 50% and the coefficient of oxygen permeability (ko) values for the inner concrete were in the range .5 to 9 by 10 to the -18th. M2 by 2.5 years which indicated that all concretes had low permeability. Impressive compressive strength gains were realize with time for the slag cement concretes which at 7.5 years were in the rang 82.5 to 105 MPa compared with the plain portland cement concrete at 67.5 MPa, clearly demonstrating the beneficial long term effect of slag on strength development. Benefits were also derived form the use of water reducing admixtures and from using crushed limestone as aggregate. These are discussed in relation to the longer-term performance of the concretes at 10.5 years.

The work presented details a testing methodology whereby a range of properties of cover-zone concrete can be determined, these properties being ultimately linked to the long-term performance of concrete. Emphasis is placed on the application of a surface-applied, water absorption method in conjunction on the application of a surface-applied, water absorption method in conjunction with electrical measurements taken within the cover zone. In the current work, with electrical measurements taken within the cover zone. In the current work, the influence of cyclic wetting and drying of normal portland cement concretes, with and without blast-furnace slag replacement, were studied. Samples were subjected to both water and chloride solution. The study shows that the depth and rate o water penetration during absorption can be evaluated; when these data are combined with volumetric gain, the effective porosity of concrete in the cover region can be estimated. Conductivity profiles taken over cycles of wetting and drying can be used to provide information on the one of influence of drying action (convective zone), on-going hydration effects, and ionic ingress.

This work estimated experimentally the critical amount of steel corrosion (Xcrit) needed for concrete cover cracking of a reinforced concrete element where only a fraction of the steel bar length is corroding. The amount of corrosion needed to crack the concrete cover (Xcrit) was ~49 um to ~137 um in specimens with localized corrosion, in comparison to ~15 um to 75 um for uniform corrosion reported for other investigations in comparable systems. An empirical equation is proposed for Xcrit as a function of specimen dimensions (concrete clear cover, C; rebar diameter, ; and anodic length, L). In this equation Xcrit is proportional to the first power of C/ and to the higher power of [C/L+1]. Quantitative determinations of the development and magnitude of stresses produced by corroding steel in concrete have been obtained. Estimated pressures at the steel/concrete interface for C/ > 3 reached values comparable to the concrete compressive strength. The potential use of a fracture-energy-based model to predict Xcrit was supported by indications of approximate agreement between estimates of the work of corrosion expansion and the energy required to crack the concrete.