Some of the knock-down factors applied in design of rocket launcher structures are based on design recommendations which rely on lower-bound curves from experimental data. The best known example is the NASA guideline SP 8007, published in 1965 and revised in 1968, which is applied for cylindrical structures in the space industry. This guideline is based on test data, computational methods and resources from the 1930-1960's. At that time the application of less empirical methods for the design of actual cylindrical shells could not count with the current computational power, and the available methods led to quite large discrepancies between experiments and test observations. Significant improvement on the available analyses approaches and manufacturing techniques since 1960's have not been taken into account in design processes using the NASA SP-8007, and many authors have recognized that for the current standards this guideline is leading to conservative structures. Another aspect for attention regarding application of the NASA SP-8007 for composite shells is that it does not consider the laminate stacking sequence. Moreover, physical observations regarding how does the imperfection sensitivity of unstiffened cylindrical shells change with the presence of an induced geometric imperfection have also suggested that the current applied design rules are too conservative. This conservativeness is confirmed by many tests carried out recently. This study presents an overview of the problem and a detailed description of the physical observations regarding the buckling mechanism of the thin shells under consideration. It is discussed how these observations can be used for less conservative, laminate dependent, knock-down factors accounting for geometric imperfections. The single perturbation load approach is studied in detail and a physically based definition for the minimum perturbation load (P1) is given, paving the way for the development of semi-analytical methods to calculate this minimum perturbation load.
- Knock-down factor
- Single perturbation load