Time-dependent response of flat plate structures under high sustained load

Abstract

[EMBARGOED UNTIL 8/1/2023] Reinforced concrete (RC) structures may be exposed to high levels of sustained stresses resulting from errors in construction or design, abnormal loading, and material degradation. Although most RC structures are designed to withstand service level load with a large safety factor, errors may cause failure or overloading in a RC member. This in turn may lead to a high sustained load in RC members and nonlinear concrete creep that may eventually lead to partial or complete collapse. In the past decade, previous research has considered disproportionate collapse by looking at the instantaneous removal of one member. However, the overall time-dependent response of the member and how that response can lead to structural collapse is unknown. This research aims to understand the impact of high sustained loads on the evolution of structural collapse in reinforced concrete flat-plate buildings and develop modeling methods capable of analyzing RC structures under high sustained load. In order to isolate and study aged concrete behavior under high sustained load, ASTM C512 (2010) compressive tests on plain concrete cylinders were conducted under sustained load levels ranging from 68 percent to 76 percent of the compressive strength of concrete. Concrete subjected to a sustained load level [greater than or equal to] 71 percent failed when reaching the inelastic strain capacity. The average times to failure for specimens were inversely proportional to the load levels that they were 105 and 35 hours for concrete subjected to sustained load levels of 71 percent le and 76 percent le (Group E), respectively. A model was presented to predict concrete creep's linear and nonlinear behavior under high sustained stresses. A comparison of the model with test results showed good agreement. In order to isolate and study bond behavior under high sustained load, ASTM A944 (2004) beam-end tests were subjected to sustained load levels ranging from 71 percent and 100 percent of their ultimate capacity and observed until failure or at least 20 days. Four-specimen groups consisting of 21 specimens constructed with scaled properties and average concrete age of 407 days were experimentally tested. Three of the specimens failed under sustained load at load levels as low as 80 percent of the control specimens. Subsequent loading of the remaining specimens showed that the application of sustained load did not reduce the ultimate residual capacity. The results also show that all specimens experienced timedependent bond-slip, which increased as the sustained load level increased. The study experimentally investigated the effect of concrete cover and bonded length on the bondslip. The results show that failure under sustained load may occur if sustained bond stress to concrete tensile strength exceeds 2.29. A simple time-dependent model for determining the bond-slip under high sustained loading was developed and validated with the available data. A finite element modeling method for the time-dependent response of RC members under high sustained load was developed and verified with experimental data. The models were built using a specially written user concrete material subroutine to simulate the concrete behavior under short-term and high sustained loading. The increase in concrete tension strain caused by time-dependent bond-slip between reinforcement steel and concrete was taken into account and incorporated into the concrete subroutine. A comparison of the model with experimental results of RC beam and slab-column tests (tested previously by the research team) as well as a RC substructure test (tested by the Defense Threat Reductions Agency (DTRA)) showed good agreement. The comparison of the FE model, with and without the time-dependent bond-slip effect, with experimental results, demonstrated the significance of taking time-dependent bond-slip into account. For example, The FE model's 25-day creep deflection in beam B7-SL was about 2 percent lower (when including time-dependent bond behavior) and 15 percent lower (when excluding timedependent bond behavior) than the experimental results. A parametric analysis was conducted considering the effect of concrete compressive strength, steel reinforcement ratio, concrete cover, and sustained load level on the behavior of RC beams and slab column connections. The FE failure of RC beams was observed at a sustained load level [greater than or equal to] 94 percent of the ultimate load. The FE failure of RC slab-column connections was observed at a sustained load level [greater than or equal to] 94 percent of the ultimate load. Overall, this research resulted in a more thorough understanding of the behavior of RC members under high sustained load and a FE modeling method capable of analyzing RC structures.