PSI - Issue 70

Umeshkumar T. Mourya et al. / Procedia Structural Integrity 70 (2025) 35–42

39

where, F = applied load, l is the lever arm, 1 d and 2 d are the horizontal displacements at the top and bottom of the end-connectors, and h is the distance between the LVDTs used for measuring 1 d and 2 d , M and  are moment and the relative rotation of the connection. Moment versus rotation curve is plotted for all the tested configurations to obtain the looseness in the connection. The looseness is calculated for each loading cycle and their average considered as the actual looseness. The moment rotation curve follows the same path in both the loading cycles as the connection was loaded in the elastic limit and no damage was ensured in the all the elements of the connection. The moment rotation curves obtained from the looseness tests are shown in Fig. 4. The values of the looseness obtained per EN15512 (2020) procedure are presented in Table 2. Equivalent value of imperfection, which is a representation reciprocal of looseness is also shown. From the moment-rotation curves of looseness tests (Fig. 4), it has been observed that the connections with beam B1 having 3 tabs in end-connector exhibit lesser looseness with the exception of configuration B1T2. The looseness on an average increase with the connections having higher beam depths as observed by comparison of the looseness of configurations of beams B1 and B2, both have 3 tabs in end-connectors. This is attributed to the increased lever arm of flange forces being transferred to the end-connectors. Although this is not consistently observed across various configurations with other parameters being same. The looseness further increases significantly by number of tabs in the end-connectors in the connection, as evident from the comparison of configurations of B2 having 3 tabs and B3 having 5 tabs in the end-connector. More number of tabs provide more interaction surfaces and hence more relative slip and looseness. The thickness and depth of the upright also affect the looseness of the connection but the relationship is not apparent. For example, Configurations B1T1 and B1T4 have lesser looseness as compared to B1T2 and B1T3. Overall, all the configurations with T2 and T3 uprights exhibit higher looseness values than the configurations with T1 and T4 uprights. A very important observation is that the uprights having same dimension but different thickness can have significant different value of the looseness in the connection. This can be observed by comparing the looseness values of T3 and T4 uprights, both have same dimensions, except thickness of 2.0 mm and 2.5 mm. T4 configurations have much lesser value of looseness than that of T3 configurations. The authors are of the opinion that the looseness value of the connection significantly depends on geometry of the tab and placement of the perforations in the web of upright. Uprights having same depth and dimensions but different thickness will have slightly different clearance between the end connector tab and the upright perforation edges. Even though this difference in the clearance may of very small magnitude in terms of distances, this will significantly affect slip between the loading, unloading and reversal loading on the connections. This affects the moment-rotation relationship and thus causes looseness in the connection. The minimum and maximum looseness values observed in this work are observed to be 0.005 rad and 0.028 rad, which are equivalent to imperfection amplitudes of are 1/200 and 1/35 as per EN 15512 (2020). This variation is significant and cannot be ignored in design of storage racks.