Bond-slip parameter estimation in fiber reinforced concrete at failure using inverse stochastic model

Fiber reinforced concrete is an emerging material in civil engineering with many desirable properties. However, material model for this material has not been yet firmly established. Failure model has a particular property: fibers do not break in failure, they got pulled-out of the concrete matrix. Accordingly, the bond between the fiber and the matrix (bond-slip law) strongly influences material behavior in failure. In attempts to experimentally determine the bond-slip law rather large spread in results becomes evident. This variance in results can be simulated by replacing the deterministic model with the stochastic description based on the idea from the fiber bundle model. Stochastic parameters are difficult to guess so mathematical procedure based on inverse analysis has been devised. The inverse model relies on the non-linear least squares approach combined with order statistics. This novel procedure successfully extracts stochastic bond-slip parameters from experimental results.