Design by testing of industrial racks

Industrial racks are one of the most common structures for the storage of palletised goods. The behaviour of these structures, which are built-up from thin-walled cold-formed steel profiles, is quite complex. The sensitivity of the uprights to buckling, the presence of the perforations on the uprights, the non linearity of the connections, the frame sensitivity to the second-order effects and the influence of the imperfections are the main sources of complexity. The large variability in terms of geometry of the profiles, of the joints and of the perforations, and the complexity of the phenomena which affects the member behaviour do not yet allow performing a pure numerical design, but call for tests aimed at the characterisation of the structural components. Traditionally, the design of the racks is carried out by a procedure combining experiments and numerical analysis. This approach follows the so-called 'design by testing'. This paper intends to provide an overview of the experimental part of the approach. The specifications in the European and the North American standards are reviewed for the main structural components, i.e., the uprights, the upright frames and the joints. Problems related to the clarity, accuracy and completeness of the specifications are pointed out. The main results of some studies carried out by the authors provide a key to better understanding of the importance of testing.