A Study of the Behavior of Reinforced Concrete T-Beams Subjected to Repeated Loads

Abstract

A total of 14 reinforced concrete T-beams and one beam of rectangular cross-section were tested to failure under repeated loading. The 13 inch deep beams were tested as restrained beams and were designed to be weaker in shear than flexure, with the critical section for shear, or shear span, being the length from the section of zero moment to the section of maximum negative moment. The only variables were the shear span-to-depth ratio and the magnitude of the repeated load. The shear span-to-depth ratio ranged from 2.88 to 5.03 and the rate of loading was 250 cycles per minute. The object of this research was to compare the node of failure of repeatedly loaded T-beams to the failure pattern of identical T-beams subjected to static loading. It was also intended to make a comparison with the results of fatigue tests performed on reinforced concrete beams of rectangular cross-section. These companion beams had been tested in two previous investigations at Purdue University. Two of the test specimens, one T-beam and one rectangular, were cast with iron-constantan thermocouples attached to the reinforcing steel at, or near, the point of maximum negative moment. It was hoped to detect a significant increase of temperature due to repeated stressing of the steel bars. It was found that the repeatedly loaded T-beams failed in all cases, regardless of shear span, by fatigue of the longitudinal steel reinforcement if subjected to a sufficient number of loading cycles. This was in direct contrast to the shear type failures of the companion beams of both T-shaped and rectangular cross-section. The stress range was found to have the most influence on the fatigue life of the steel reinforcement while the shear span-to-depth ratio had almost no influence at all. The number of cycles to failure decreased with increasing stress range, and the fatigue limit was found to be approximately 40 ksi. Failure occurred at maximium repeated loads ranging from 52% to 76% of the ultimate static load. The amount of temperature increase detected during testing was negligible.