Journal of Performance of Constructed Facilities

Steel tanks contained wheat in Zaïre, West Africa. They were provided with a conical hopper. After one hopper collapsed abruptly during emptying, inspection disclosed that failure had occurred at the bolted seams. Grain-induced pressures and their distributions were determined. Simple membrane theory and a detailed finite-element analysis calculated hoop and meridional stresses and deformations in the hoppers. They showed that underdesigned, overstressed bolted seams caused hopper failure. The remaining hoppers were strengthened by welding them together. Sixteen mm square steel bars were placed along side the seams to prevent the existing sandwiched neoprene-type gasket from burning and contaminating the fillet welds. Laboratory tests indicated that transfer of forces at the welded seams is through the welded fillets; the bolts are bypassed and become irrelevant. This reduces the maximum stress in the steel to less than half and makes nil the probability for a repeat, abrupt failure of a hopper. All strengthened hoppers are presently used; recent observations show them in an excellent state of repair and performing well.

This paper presents the results of a laboratory evaluation of the performance of 104 used adjustable-length wooden formwork shores. The shore samples were randomly collected from eight different construction sites in Chicago, Milwaukee, and Minneapolis-St. Paul. The shores tested are known as Ellis shores. Ellis shores are very popular for shoring and reshoring elevated slabs in multistory concrete construction. An Ellis shore is made of two pieces of 100 mm × 100 mm (4 in. × 4 in.) nominal-size solid lumber attached together with two metal clamps. The collected shore samples were loaded to failure at four different lengths ranging from 2.4 to 4 m (8 to 13 ft). The observed modes of failure included bending failure, crushing of the wood under clamps, clamp failure, and crushing or splitting of the shore ends. The observed load-deformation behaviors and ultimate capacities are presented. Shore ultimate capacities are modeled with appropriate probability distributions.

Vibration and noise are two major factors in the performance assessment of rail transit systems, particularly in an urban environment where elevated systems are used. A program of field measurements was conducted in Singapore on an elevated section of straight tracks to monitor the rail vibrations and noise generated during the passage of trains. Field measurements included impact hammer tests to determine the driving point impedances of the train wheel and track support system. These tests are relatively easy to perform, requiring only an inverted wheel-set-and-bogie assembly at the workshop and the tracks at the test yard. An analytical train model was developed to predict the level of rolling noise at the wayside of the tracks for comparison with measured levels. The analytical predictions from the model were found to be in good agreement with field measurements. A hybrid approach involving a finite-element model of the track support system and impact hammer test results was also presented to determine the effect of using softer rails pads.

Fifty-four existing buildings (five to eight years old) and 54 construction sites were surveyed in Riyadh, Saudi Arabia. Different types of masonry walls were used, mainly as non–load-bearing reinforced or unreinforced and cavity walls. It was found that the causes of cracks and failures were attributed to structural changes, temperature changes, and foundation settlement deficiencies. It was also found that these failures and cracks can be minimized by adopting proper masonry design and construction. This can be accomplished by introducing movement joints, bed-joint reinforcement, adequate support and anchoring, and good artisanship and supervision.

To quantify the impact of tiling workmanship on the performance of tiled facades, the Singapore building authority launched a research project to review and investigate the entire process of external wall tiling. As part of this effort, field investigation and laboratory experiment were conducted. The field investigation was specially featured by conducting in-situ pull-off tests on tiled walls of an 11-year-old building. Evidence was obtained by statistically correlating the bond-strength reduction with various signs of inadequate workmanship. However, the bona fide workmanship impact became confused with the weathering effect. To distinguish the two, tiled specimens were carefully fabricated for laboratory experiments. Before being destructively tested, most of the specimens were subjected to a closed-loop-controlled artificial weathering. These tests revealed that bonding strength plunged as the quality of workmanship declined. Although the weathering effect was quite moderate in the specimens prepared with very good workmanship, the synergistic effect of the weathering together with inadequate workmanship was disastrous.

Numerous roof collapses throughout the Middle Atlantic and New England States followed the occurrence of two winter storms during the period of Jan. 6–15, 1996. Cumulative horizontal ground-measured snow loads for these two storms are derived by analyzing daily liquid-equivalent precipitation measurements and preexisting snow depth observations at approximately 300 stations. The resulting values are compared to published 50-year recurrence interval snow loads that define the minimum design criteria for roof structures. In general, snow loads were found to exceed the 50-year recurrence interval value along a band encompassing the Boston to Washington corridor. Snow loads in excess of the 100-year recurrence interval value were estimated near Baltimore, Md; in central New Jersey and southeastern Pennsylvania; over the Lower Hudson Valley and extreme eastern Long Island in New York; and south of Boston, Mass. Snow loads in southern New Jersey, central Massachusetts, and most of Rhode Island did not to exceed the 50-year return period amount.