Approach for detection of fatigue phases using the example of high-performance concrete

High-performance concretes (HPC) are an emerging material in modern structural design, enabling the development of ever more robust and slender structures. This mode of construction, however, opens up new problems. While gaining compressive strength, HPC loses deformation capacity, ductility and gets more and more prone to fatigue loading. Hence, special attention needs to be paid to ductility and fatigue scenarios. In this case, HPC are just an example. The outlined approach is applicable to any material that shows a three-stage development of damage during fatigue. It is agreed in literature that the gradient of the secondary creep phase is a good indicator to predict failure. However, at present it is only possible to assess the secondary creep phase retrospectively, i.e. after failure. To the author's knowledge, there is no comprehensive indicator or conclusive approach to determine the onset of the second and third phases of a fatigue process during testing without prior knowledge of the point of failure. Such an approach would offer several advantages, e.g.: Firstly, from a material science standpoint, recognition of reaching the third phase of fatigue during testing is beneficial to stop laboratory tests, e.g. for CT measurements. Secondly, for cycle-jump methods to reduce numerical effort in modelling high-cycle fatigue, knowledge of the evolution of deterioration rate allows a reasonable estimation for the length of the cycle jump without losing precision. It could further be very helpful in structural health monitoring when adjusted to a specific structure. This paper outlines an approach to distinguishing between the different stages of fatigue using uniaxial compression fatigue tests on HPC specimens with a compressive strength of about 90 MPa as an example. Both specimens with and without steel fibres are considered. The aim is to have a clear and reproducible separation of the three phases of the classic S-shape degradation curves. It is taken advantage that no major individual events are to be expected during the secondary creep phase, which ensures a sufficiently constant deterioration rate. Accordingly, it is sufficient to detect the beginning and end of this behaviour. It is evident that the moving standard deviation of the stiffness evolution as a damage indicator has a very high significance in detecting both the beginning and the end of the second fatigue phase. Furthermore, it is shown that the approach can be transferred to completely different framework conditions, in this case to flexural fatigue of the mentioned HPC and compressive fatigue tests on normal strength concrete with only slight changes adjustment in parameters. 1 G. Gebuhr, S. Anders, M. Pise, D. Brands and J.