Journal of Materials in Civil Engineering

This paper presents experimental results of full‐scale fire tests with loading on concrete‐filled steel‐tube (CFST) columns using fire‐resistant (FR) steel for building structural use. By covering the tube surface with protection having lower insulation properties, CFST columns will achieve a superior fire resistance, taking advantage of the heat capacity of inner concrete effectively, and they could resist the bending moments caused by eccentric loading without inner reinforcing bars by keeping steel temperature below 600°C. Reinforcing bars are difficult to arrange because inner diaphragms are provided at the beam‐column connections against earthquakes in Japan. Fire tests with loading of CFST columns using ordinary steel were also performed for comparison. Aluminum hydroxide calcium carbonate cement‐type protection (ceramic protection) and foaming‐type intumescent coating were used for the protection material. The fire tests with loading were carried out at the furnace in the General Building Research Corporation, Osaka, Japan.

The principal in‐plane compressive strengths of pultruded E‐glass fiber‐reinforced flat sheet material were measured using a nonstandard test procedure. Over 100 specimens, each having nominal dimensions of 7.0 by 2.0 by 0.635 cm, were tested with the longest side either parallel or perpendicular to the pull direction in the pultrusion process. The experimental test procedure and the test rig for thick (up to 1.6‐cm) specimens are described. The strength data are used in a comparison between statistical methods that may be used to determine design allowables for the purpose of structural design. These strength allowables are compared with the corresponding values (in accordance with standard ASTM D695‐90) provided by the manufacturer. Results show that compression strengths of the material have coefficients of variation of 9% in a parallel direction and 6% perpendicular to this. This large variation indicates that the ASTM D695‐90 specimen batches of five are too low. Design allowable strengths determined from the log‐normal distribution are 25% and 12% higher than manufacturer's allowables.

This investigation is carried out: (1) To study the effect of magnesium‐sodium sulfate environment on the performance of two plain and three blended cements; and (2) to elucidate the sulfate attack mechanisms on these cements in the mixed magnesium and sodium sulfate environment. After two years of exposure, deterioration is observed in all cements. However, the deterioration is more pronounced in blast‐furnace‐slag (BFS) and silica‐fume (SF) cements. Deterioration in these cements significantly exceeds that observed in plain and fly‐ash‐(FA‐) blended cements. XRD and SEM analyses indicate that the greater deterioration in BFS‐ and SF‐blended cements may be attributable to the depletion of the hydrated calcium hydroxide as a result of pozzolanic reaction. In the absence of $\mathrm{Ca}(\mathrm{OH}{)}_2$, magnesium ions react more directly and extensively with the cementitious calcium silicate hydrate to generate gypsum and noncementitious magnesium silicate hydrate resulting in aggravated deterioration.

This study was designed to evaluate the relative corrosion and sulfate resistance of concrete made with portland cements containing 2%–14% $C_{3}A$ without and with 50%, 60%, 70%, and 80% cement replacement by blast‐furnace slag (BFS). The results show that BFS blended‐cement concretes had a significantly superior corrosion‐resistance performance. The best corrosion protection was obtained with 50% BFS cement, which, depending on the $C_{3}A$ content of the parent cement, showed a 3.82‐3.16 times better performance in terms of corrosion‐initiation time compared to the parent plain‐cement concrete. BFS blending was specially beneficial in improving the corrosion‐resistance performance of Type V low $C_{3}A$ cements. Performance on exposure to sodium‐sulfate (N$\overline{S}$) solution, replacement level only at 70% and above, showed sulfate resistance to be better than that of the Type V sulfate‐resistant cement. BFS blending, even with high $C_{3}A$ cement (9%, 11%, and 14%) at 70% and above‐replacement level, imparted a high degree of sulfate resistance. The cement with high $C_{3}S/C_{2}S$ ratio has a perceptible adverse‐interactive effect and causes sulfate deterioration even with $\mathrm{low}‐C_{3}A$ sulfate‐resistant cements. In M$\overline{S}$‐N$\overline{S}$ environment, due to the magnesium‐gypsum type of attack, the 60% BFS cement deteriorated even more severely than the plain Type V and Type I cements.

The punching shear behavior of steel fiber reinforced concrete slabs was investigated in the present study. Each of 14 square slabs was simply supported along four edges and loaded to failure under a concentrated load over a square area at the center. The test parameters were the effective span to depth $a/d$ ratio, volume fraction of steel fibers ${\rho }_f$, slab thickness $h$, concrete strength $f_c^{\prime }$, and size of load‐bearing plate $r$. Test results indicate that the load‐deflection curve of slabs exhibits four distinct regions that may be characterized by first cracking, steel yielding, and ultimate load. Within the scope of the test program, an increase in the values of ${\rho }_f$, $h$, or $r$ was found to lead to an increase in both the punching shear strength and the ductility of the slab. The ultimate punching shear strength of the slabs was compared with the predictions of equations available in the literature and code equations for reinforced concrete. The British Standard CP110's equation was found to estimate the punching shear strength of the test specimens reasonably well.

This research was conducted to investigate the performance of fresh and hardened concrete containing discarded foundry sands in place of fine aggregate. A control concrete mix was proportioned to achieve a 28‐day compressive strength of 38 MPa (5,500 psi). Other concrete mixes were proportioned to replace 25% and 35% by weight of regular concrete sand with clean/new foundry sand and used foundry sand. Concrete performance was evaluated with respect to compressive strength, tensile strength, and modulus of elasticity. At 28‐day age, concrete containing used foundry sand showed about 20–30% lower values than concrete without used foundry sand. But concrete containing 25% and 35% clean/new foundry sand gave almost the same compressive strength as that of the control mix.

The paper describes an experimental program in which four deformed commercial fibers with widely different geometries were investigated in steel‐fiber reinforced concrete. Three matrices with compressive strengths of 42, 52, and 85 MPa were reinforced with fibers at a dosage rate of 40 kg/m³. Compressive and flexural strengths were measured along with the elastic moduli. The focus of the study, however, was to measure and characterize the toughness improvements in the basic matrices due to the addition of various fibers. To this end, flexural load‐deflection curves were analyzed in accordance with the ASTM and Japan Society of Civil Engineers (JSCE) standard methods and also using a proposed‐analysis scheme. The paper points out the limitations of the current techniques of toughness characterization and identifies this as an area with immediate research needs. For the fibers and the matrices investigated, a strong influence of both fiber geometry and matrix strength on the toughness characteristics of fiber‐reinforced concrete was observed. End‐deformed fibers were, in general, found to perform superior to those deformed throughout the length.

A two‐stage freezing model is proposed for concrete at a near‐saturated state. The first‐stage involves a mechanism considered by Powers, while the second‐stage is based on a theory due to Everett. The volumetric consequences of this model checked by dilatometry do not perfectly materialize, since the pore air is not compressible enough as assumed in Powers' theory. Experiments qualitatively confirm the expected abrupt dependence of frost damage on degree of saturation. The role of air entrainment in minimizing the frost damage in the field is explained primarily by lowering the rate of saturation and partly by its retardation of the second‐stage freezing by increasing the compressibility of the gas phase. Experiments also confirm that the dilatometric expansion on freezing is a measure of significant damage as predicted by the proposed model.

Both unconfined‐ and confined‐creep tests have been widely used to evaluate and characterize rutting potential in hot mix asphalt (HMA). However, no reference has been made suggesting that either the unconfined‐creep test or the confined‐creep test is better than the other. The unconfined and confined creep of a dense graded laboratory mix were measured at the conditions at which rutting is most likely to occur in HMA pavements. The relationship between sample air voids and the unconfined, as well as the confined, creep were evaluated in the laboratory and compared to the relationship between in‐place air voids and HMA rutting under traffic. In addition, the effect of mix preparation and specimen height on confined and unconfined creep were also evaluated. The creep test (unconfined‐or confined‐) that produced the test results most consistent to HMA rutting has the higher probability of evaluating the rutting performance of HMA, and is therefore recommended as the preferred method for conducting the creep test. The confined‐creep test was found to be more representative of in‐place performance.