Life-cycle robustness of fastening systems

Fastening systems are an important part of the construction industry. The requirements of the future —faster and automatized construction, improved life-cycle performance and sustainability of infrastructure — will further promote their significance. However, the challenges associated with the safe and efficient design and assessment of such systems in the light of life cycle concepts requires new concepts. Life–cycle robustness is achieved when a (fastening) system is designed to maintain its intended function and required safety level for the full duration of its intended life time. The different characteristics of processes each element will face in course of time (damage, ageing, extreme events, changes in usage) in conjunction with the diversity in the intrinsic material properties form a demanding problem. Further complexity emerges when one realises that time is not simply a variable, but a factor permeating model choices. Different effects on the load side and properties on the resistance side develop differently in time, as does the dependence structure between variables. Spatial randomness of materials, such as concrete, requires careful modelling, even at a meso–scale. For a long–term analysis, where the influence of uncertainty on the safety level may dominate over the mean trend, robust design can prove decisive. On the computational side, challenges often appear since the computational costs of simulations and non–linear analyses may quickly prove infeasible. Suitable numerical techniques for small scale sampling, accounting for arbitrary distribution types and dependence structures, are yet to be developed. The realistic prediction of spatial randomness for now fails due to a lack of understanding regarding the physical basis of main input parameters. Within this contribution the authors present the general concept of life–cycle robustness and the expected prospects that arise from its application to fastening systems. A detailed discussion of the aforementioned challenges and review of the state of the art complement the paper