5. Results and discussion
The modeling results of fatigue dependent single-fiber pullout behavior are compared against the experimental measurements to verify the validity of the new model. The micromechanical parameters as the input of the model are given in Table 1. Single-fiber pullout curves after two different fatigue loading schemes are includedinFig. 8. In the experimental study,when the fatigue loading level was relatively low (Pmax b 0.4 N), the full-debonding was never reached during the fatigue preloading and the load drop from Pa to Pb at u = u0 during monotonic reloading still existed. When the fatigue loading level was relatively high (Pmax N 0.5 N), the full-debonding had been reached during the fatigue preloading and the load drop during monotonic reloading was not observed. In can be seen in Fig. 8a and b, both effects can be well captured by the current model. In thefirst stage of Eq. 14, it is noticed that the tensile force P must reach a certain level to overcome the frictional bond, which is quantified by the second term in Eq. 9. The fiber displacement u can be activated after only the frictional bond is overcome, which is reflected in the modeling curve Fig. 8. However, the experimental P-u curve initiates from (0,0) as the stretch of the free fiber was inevitable in the actual testing [20]. Table 2 compares the Pa and Pb measured in the experimental study and those predicted by the current model at different fatigue loading levels and fatigue cycles. It can be seen that the value of Pa and Pb can be well predicted after various fatigue preloading schemes.
