Cylindrical shell structures buckle at service loads which are much lower than their associated theoretical buckling loads. The main source of this discrepancy is the presence of various imperfections which are created on the cylinder body during different processes as manufacturing, handling, assembling and machining. Many cylindrical shell structures are still designed against buckling based on the experimental data introduced by NASA SP-8007 as conservative lower bound curves. This study employed the numerical based Linear Buckling mode shape Imperfection (LBMI) method and modified it using a stochastic method to assess the effect of geometrical imperfections in more details on the buckling of cylindrical shells with and without the cutout. The comparison of results with those obtained from the numerical Simcple Perturbation Load Imperfection (SPLI) method for cylinders with and without cutout revealed a good correlation. The effect of two parameters of size and number of cutouts on the buckling load was investigated using the linear buckling and Modified LBMI methods. Results confirmed that in cylinders with a small cutout inserting geometrical imperfection using either SPLI or modified LBMI methods significantly reduced the value of the predicted buckling load. However, in cylinders with larger cutouts, the effect of the cutout is dominant, thus considering geometrical imperfection had a minor effect on the buckling loads predicted by both SPLI and modified LBMI methods. Furthermore, the modified LBMI method was employed to evaluate the combination effect of cutout numbers and size on the buckling load. It is shown that in small cutouts, an increasing in the cutout size up to a certain value resulted in a remarkable reduction of the buckling load, and beyond that limit, the buckling loads were constant against D/R ratios. In addition, the cutout number shows a more significant effect on decreasing the buckling load at small D/R ratios than large D/R ratios.
Read full abstract