Nitrogen adsorption is quick, inexpensive method for measuring the surface area of mesoporous materials. The use of this method for studying cement-based materials has generated much controversy in part because the values obtained are quite variable and differ from those measured using water adsorption, SAXS, SANS, or NMR techniques. No two measure the same surface area of cement paste. Although nitrogen may not measure all of the surfaces in a cement paste, a quantitative measurement of the surfaces accessible to nitrogen is till a valuable number. When variation due to sample preparations minimized, the effects of variables on the nitrogen B.E.T. surface area become clear. Evidence from the effects of w/c and age on surface area shows that there are two types of C-S-H can be manipulated, with potential effects on physical properties such as drying shrinkage and creep.

Concrete structures that are expected to last a long time in a severe environment must be built with proper attention to design, materials selection and proportioning, and construction practices. This paper addresses the construction issues, including placement and curing of concrete. In placement, the importance of cover depth, effects of pumping on air content, need for properly functioning vibrators and screed, and the consequences of improper consolidation are described. The necessity of proper curing is addressed by explaining cracking that result from loss of moisture. The variation in strength between that of the test samples and that of the member is described for concrete subjected to steam curing.

Concrete may be considered a three phase material, consisting of cement matrix, aggregate and the interfacial transition zone (ITZ). The ITZ is studied usually by collecting backscattered electron (BSE) images of polished samples of concrete and by carrying out a quantitative analysis of these images. This technique makes use of the contrast between pores and cement hydrates to analyse the microstructure. Once the images are collected, grey level thresholding is used to segment regions associated with capillary porosity. The size of the ITZ is determined using this data. Different techniques have been suggested to minimise errors introduced at this stage of the analysis. The authors have carried out an investigation on 17 different concrete mixtures in an attempt to assess the role of threshold values. The results indicated that the size of the transition zone is not affected by the range of threshold values used. However, the porosity of both the ITZ and the bulk cement paste varied with different threshold values. The paper proposes a method which can be used to determine a reliable threshold value.

Concrete structures are deteriorating at an alarming rate. While a substantial body of research exists to describe the corrosion process of pristine concrete systems, this paper describes a recent study in which the corrosion process of pristine concrete systems, this paper describes a recent study in which the corrosion and concrete systems, this paper describes a recent study in which the corrosion and mechanical response were assessed for beams exposed to various loading histories. Specifically, different levels of preloading were applied to generate damage while substained loading was also used to investigate the interaction between load level and corrosion rate. Results illustrate that loading history can significantly influence the corrosion and mechanical response of reinforced concrete elements. Corrosion is initiated in cracked beams much aster than in uncracked beams, presumably due to the cracks which facilitate the ingress of aggressive agents to the surface of reinforcing steel. Specimens with high levels of loading illustrated higher corrosion rates. Corrosion of the beams under sustained loading illustrated a similar load deflection history until the point at which significant corrosion was initiated. After significant corrosion occurred, the creep/corrosion behavior resulted I increased deformations which ultimately resulted I a creep/corrosion failure of high load level beams. Beams with higher loading levels were observed to have corrosion imitation sooner and undergo larger deflection. These results indicate that corrosion is accelerated in cracked structures and even further accelerated I structures where the load is maintained over a long-period of time. This suggests the need for models which assess the impact of the loading history in addition to corrosion driving forces, environmental conditions, and material proportions.

This paper describes the measured behavior of a bridge made with precast, prestressed, high-performance concrete (HPC) girders. The concrete was considered high-performance, because it was specified to have a compressive strength of 51 MPa (7400 psi) at release and 69 MPa (10,000 psi) at 56 days. By using HPC instead of normal-strength concrete, the bridge designer was able to reduce the number of girder lines from seven to five. These girders were the first to be constructed in Washington State using HPC. To monitor the girders, vibrating-wire strain gages with thermistors were installed in five girders, and camber was monitored by various means, including a stretched-wire system that could be monitored by the data-aquisition system. Temperatures measured during fabrication indicate the presence of a large and unexpected temperature gradient over the height of the girder. As a result the concrete strength at release may have been lower at the bottom of the girder than at the top. The long-span girders were stressed to an initial bottom stress of approximately 28 MPa (4000 psi). The strength of the concrete was higher than usual and permitted the high initial stress. However elastic modulus rises only with the square root of strength, so elastic shortening strains, and creep strains that are assumed to depend on them, are higher for high-strength concrete. In these girders, elastic shortening and creep dominated the loses. The measured losses were compared with the predicted losses form two standard methods, but neither method was able to provide a universally superior match. In general, the measured losses exceeded the calculated losses initially, but with time, the discrepancies decreased.