Journal of Materials in Civil Engineering

The following fracture models are assessed on the basis of an extensive testing program of large wedge‐splitting specimens: Bažant's size effect law; Hil‐lerborg's fictitious crack model; Carpinteri's brittleness model; and Shah and Jenq's two‐parameter model. Experimental data reinforce the validity of the size effect law with respect to specimen sizes. Fracture toughness values determined through the compliance method are compared with the model predictions. Results are discussed regarding specimen size requirement for linear elastic fracture mechanics (LEFM) applicability. On the basis of the size effect law and the fictitious crack model, the wedge‐splitting specimen sizes used are still too small for a classical LEFM analysis to be valid. The fracture models and their corresponding brittleness number are compared and found to be related to each other. They can be traced to Griffith's linear elastic fracture mechanics theory. The experimental results are then extrapolated to large structures, such as dams, and the limits of validity of both linear elastic and nonlinear fracture mechanics structural analyses are discussed.

The effects of curing in mild, hot, and dry conditions on strength and quality of concrete skin were studied. The skin quality was determined by absorption tests and depth of carbonation (natural and accelerated). It has been found that a deficient water treatment in a hot and dry environment could have a mild negative effect on strength, but a marked detrimental influence on the properties of the skin, particularly in high‐fly‐ash content concretes. Thus, compressive strength can be a misleading parameter if it is used to characterize the quality of concrete skin. Accelerated carbonation and rate‐of‐absorption tests are more valid for the latter purpose, and can provide information that is at least of semiquantitative significance.

This paper examines the uncertainty in the mathematical models used for predicting the ultimate and first‐crack moment capacities of ferrocement structural elements. Uncertainty studies are performed on five ultimate‐moment models and two first‐crack moment models using experimental data. The studies include the uncertainty in each model arising from the basic variables as well as the prediction error in each model due to the simplifications and assumptions made in deriving the model. The results obtained provide a basis for comparing the accuracy of each model as well as for use in reliability‐based design. The mechanism approach and the equilibrium method are comparatively more reliable for predicting the ultimate and first‐crack moment capacities, respectively, among the models studied.

A two‐year aging study was presented for four sealants: a low‐modulus silicone, a medium‐modulus silicone, a one‐part polyurethane, and a two‐part polyurethane. A specimen, with red granite substrates, is used. The four aging exposures are: ordinary room conditions; outdoors in Houghton, Michigan; outdoors near Hilo, Hawaii; and aging in a Q‐Panel (QUV) weathering unit. Following an initial six‐week cure, room‐ and outdoor‐aged specimens are tested at zero‐, six‐, 12‐, and 24‐month intervals. The QUV‐aged specimens are tested after 0, 500, 1,000, and 5,000 hours. In the postaging tests, specimens are given 20 displacement cycles at the sealant's rated movement capacity and elongated to failure. Reported data include nominal stresses at 25%, 50%, and 100% elongation, the ultimate tensile strength, and the corresponding elongation. Additionally, observations are given for surface cracking and dirt pickup. Adhesion to the substrates is good for all four sealants. Results of a pilot study combining QUV exposure with a displacement cycle are also reported. The overall findings indicate that the silicones' performance is superior to that of the urethanes.

This paper deals with glass‐fiber‐reinforced plastic (GFRP) beams produced by the pultrusion process. Pultruded composite members are being used extensively as beams for structural applications. Widespread use is motivated primarily due to the light weight and corrosion resistance of composite materials. Low‐cost, mass‐produced pultruded beams are becoming competitive with conventional materials like steel and reinforced concrete. Common structural shapes have open or closed sections of thin composite walls. The ultimate bending strength of pultruded composite beams is limited by various failure mechanisms. Most failure modes are precipitated by local buckling of the thin walls. Analytical models for several local buckling modes are used in this work to model observed behavior in commercially available composite beams. Experimental data for composite beams are presented for comparison. Local buckling initiates a failure mode that eventually results in material degradation and total failure of the beam. Due to the large elongation to failure of the composite material, only postbuckling deformations can subject the material to deformations large enough to produce failure. Experimental results are shown to sustain these arguments. Analytical models for local buckling are developed and correlated with observed behavior.

A new method for measuring the pore size distributions (PSDs) of saturated porous materials has been developed. With this new methodology, referred to as conductometric phase transition porosimetry (CPTP), the changes in electrical conductance and temperature of saturated porous materials that have been subjected to a cycle of capillary freezing and melting are measured and used to calculate PSD. Based on this methodology, a porosimeter was constructed. Methods and procedures for determining the PSD of 4‐in. (10.6‐cm) diameter by 4‐in. (10.6‐cm) high portland cement mortar and concrete cylinders are introduced. The CPTP method has been found to be free of the problems of entrapment of mercury inherent to mercury porosimetry. Also, the pressures a sample is subjected to during testing are much less than those necessary with mercury porosimetry. Therefore, deformation or destruction of pores is less. It is well known that for many geomaterials, pore structure is dependent on moisture content. With CPTP, samples do not have to be dried prior to testing as with other techniques.

Mortar sheet products can be made stiff and brittle by reinforcing with stiff fibers having strong chemical bonding to cement paste, such as asbestos. They can be made tough and pliable by reinforcing with softer fibers having poor bonding to cement, such as polypropylene, where the fiber stress is transferred over longer distances from developing cracks. It is also known that the stiffness and fracture energy of mortar can be altered by surface treatments of the fiber to either increase its bonding with the paste, making the composite stiffer, or to decrease its bonding, making the composite more pliable. Less understood are the increases in both stiffness and toughness of mortar by reinforcing it with fiber hybrids—one stiff and/or well‐bonded fiber combined with one soft and/or poorly bonded fiber. This study experimentally evaluates the synergism of combining mineral wool to improve stiffness and polyethylene pulp to improve toughness for mortar sheet products. A design example illustrates the determination of fiber weight ratios for specified bending stiffness and bending toughness at minimum material cost.