Rate effect on the pull-out load of individual fibers in concrete: A probabilistic modeling approach

Abstract

Understanding of the load-rate dependence of fiber pull-out from concrete has been hindered by the large scatter of test data. Relatively little attention has been given to the statistical nature of single-fiber pull-out under different loading rates. The present study is based on a probabilistic interpretation of the pull-out phenomenon, assessing the statistical fluctuations observed in the common finding that complete pull-out occurs at greater load levels under higher loading rates. Since progressive debonding along the fiber–matrix interface affects the pull-out behavior, interfacial crack growth is modeled using a fracture mechanics approach. Based on a theoretical justification, the transition probability to debonding is devised and implemented within a Markov chain model. The Markov model accounts for debonding propagation with each stress increment until complete pull-out. The results demonstrate that the probability distributions of stress at complete pull-out differ from a normal distribution. The coefficient of variation is also independent of the loading rates. These findings show reasonable agreement with statistical variations observed from test data.