Increasing the fatigue resistance of highly ductile concrete by experimental-virtual multiscale material design

SHCC (Strain hardening cement-based composite) shows a high strain capacity and a pronounced strain hardening behavior due to formation of multiple cracks. There are already many findings about the mechanical behavior of the material under cyclic loading regimes on the mesoscopic scale made in the first project phase. Hence, in the second phase it is examined if it is possible to transfer the results and degradation mechanisms of the single fiber level to the fatigue behavior of the composite material under high numbers of loading cycles. To advance the numerical multiscale model of the single fiber level and thus for further basics for the material design, cyclic fatigue tests are carried out in tension-swell and tension-compression-alternating load regime under variation of the strain rates and reversal points. Furthermore, the application of hybrid fiber reinforcement made of highly ductile polymer and micro-steel fibers to reduce the damage caused by crack closure is investigated and the influence of fibers with an end anchorage on the behavior of SHCC is examined. Since the load-bearing capacity of the material decreases with increasing crack width, different maintenance approaches for crack closure are to be investigated.

From the individual fiber tests, it has been shown that the pull-out behavior of the fibers is very sensitive to changes in individual parameters. Therefore, the influences of the fiber type, temperature and fiber orientation on the behavior of the composite under cyclic loading are investigated additionally. In order to understand the damage mechanisms during the fatigue tests, structural-morphological analyses are carried out.

The listed experiments are the basis for the development of a material model on the macro scale. They serve both as a data basis for the model development and for its validation. To describe the material behaviour, the already formulated micro-layer model for the description of spatially distributed damage is further developed. In addition, time homogenization is being further developed for both the meso and the macro model in order to be able to map the properties over large numbers of load cycles. By coupling the composite model on the mesoscale with the macroscopic material model, the damage and deformation mechanisms on the mesoscale are linked to the non-local elasto-plastic fatigue formulation of the macroscale. Using the developed macroscopic model, it is possible to make predictions about the fatigue behaviour of strain hardening cement-based composites under changed material parameters.