< Back to previous page
Cyclic Behaviour of Steel Fibre Reinforced Concrete with Acoustic Emission-based Multi-scale Analysis
Book - Dissertation
The limited tensile resistance of concrete stimulated the development of steel fibre reinforced concrete (SFRC), consisting of a concrete matrix with randomly distributed steel fibres and characterised by an improved post-cracking behaviour. Despite the various non-structural SFRC applications, its use is still restricted with respect to its potentials, which is mainly caused by the lack of international design codes. Moreover, structural elements are frequently subjected to cyclic loading, e.g. by traffic, leading to cyclic damage development and eventually fatigue failure. Hence, the cyclic behaviour is an important design characteristic of construction materials. Although the presence of fibres has proven to retard the crack growth in concrete, the application and fundamental material knowledge remain scarce. Lastly, most experimental research is limited to only measuring load-displacement curves. However, the nature of cyclic deterioration lies within the progressive micro-cracking. Advanced non-destructive techniques are required for an in-depth investigation of damage initiation and propagation. Within this field, acoustic emission (AE) monitoring has proven its value by continuous real-time damage detection, characterisation and localisation.While SFRC, cyclic loading and AE monitoring have been studied individually before, their combination is a novelty of this research. The aim is to investigate the cyclic behaviour of SFRC with an AE-based multi-scale analysis. Therefore, a combined methodology of experimental work on three scales and analytical modelling is adopted.Firstly (scale 1), individual fibre pull-out tests are combined with AE monitoring and micro-CT scanning. Two fibre types, 3D and 5D, are embedded with varying length and inclination. The combination of AE and micro-CT provides a more profound damage assessment. The monotonic behaviour is divided in pull-out stages, both by material behaviour as by AE activity, and the cyclic displacement rate is related to the AE activity. The AE results such as source location are validated with the micro-CT scans. Deviating results can be attributed to the position of air voids visible on the scans.Secondly (scale 2), direct tension tests (DTT) of notched SFRC cylinders are performed with AE monitoring. It is found that fibre distribution and orientation have a large impact on the uniaxial tensile behaviour, and that the monotonic load-displacement curves are in good agreement with the cyclic envelope curves. Furthermore, the cyclic results validate the linear relation between unloading and plastic crack opening, and confirm that this relation is rather independent from fibre type or dosage and can be used to construct the damage curves. This scale also confirms the complementarity of AE monitoring and traditional test setups. Stages of damage development with micro- and macro-cracking are determined by AE analyses (activity, localisation, failure modes). An experimental power law is established which relates AE activity with mechanical damage evolution based on the stiffness response during cycles. Thirdly (scale 3), three-point bending tests (3PBT) of notched SFRC prisms are performed and the AE monitoring is again upscaled. The main conclusions of scale 2 remain valid for the 3PBT setup. AE signal characteristics are applied to improve the AE source localisation and to separate new crack formation from frictional damage. Both the cyclic crack opening evolution and AE activity follow the Paris' law, with distinction between fatigue of the uncracked and cracked matrix. In general, fatigue failure is obtained when the load cycles reach the envelope post-cracking curve. The damage curves are found to be identical for progressive cyclic or fatigue loading, which eliminates the necessity of long-term fatigue tests. Lastly, it is shown that the neutral axis position during monotonic or cyclic loading is independent of the residual strength class and increases during unloading.Finally, an analytical model is developed to relate the tensile and bending behaviour based on sectional analysis, both for monotonic and cyclic loading. This approach enables to obtain material behaviour information without performing the corresponding test, i.e. tensile or bending curves can be obtained based on 3PBT or DTT respectively. Also the deformation or stress profiles in bending are obtained based on DTT results. As such, the experimental observations of the neutral axis position are confirmed. Through validation with experimental data, it is found that the analytical model results have a high degree of accuracy for the little amount of input and at reasonable calculation times. Furthermore, the fatigue behaviour of SFRC is predicted and validated by estimating SN-curves and the fatigue life (after pre-cracking), without the need to perform long-term fatigue tests.