Motivation

Thin-walled, unstiffened cylindrical shells made of carbon-fiber-reinforced polymer are used, among other things, as structural elements in the aerospace sector. Under compressive loads, these structures are prone to buckling, so their design is governed by buckling considerations. At the same time, the application imposes stringent requirements on both weight and safety. The currently applicable design guideline, NASA SP-8007, was developed in the 1960s and specifies certain knock-down factors for metallic cylinders under various boundary conditions. As the radius-to-thickness ratio (R/t ratio) increases, the deviation from the theoretical buckling load grows.

The deviation from the theoretical buckling load is primarily caused by geometric imperfections, load imperfections, or disturbances introduced, for example, through the boundary conditions. For fiber-reinforced composites, these knock-down factors lead to highly conservative designs. Moreover, it is not possible to determine the failure probability, since this would require all influencing parameters to be known.

Objective

The objective of this project is to further develop the existing probabilistic design approaches. To expand the statistical database, an experimental campaign is carried out using two different cylinders with distinct R/t ratios. Imperfections measured prior to the buckling tests are published in the form of Fourier coefficients.

Procedure

Building on the results of the cylinder tests conducted within the CompoSeat project, characteristics of the imperfection patterns are extracted from the existing measurement data. Among other aspects, it is investigated how many modes are required to describe the imperfection pattern with sufficient accuracy. To further expand the statistical database, six additional cylinders with the same properties as in the CompoSeat project are manufactured. In addition, six cylinders with the same layup but a larger R/t ratio are produced. Further specifications can be found in the table below.

The cylinders are mounted in a specially developed fixture for testing. Before and after mounting, their geometric deviations are measured using an optical scanning system. At the hexapod test bench, compression and tension tests are subsequently performed to determine the material parameters. Finally, the cylinders are loaded in compression until buckling occurs. The measured data are compared with the simulated buckling loads, and the model uncertainty of various detailed models is quantified. Once the material, structural, load, and model uncertainties have been determined, the reliability can be taken into account in the design process.

The project was funded by the German Research Foundation (DFG) as part of its infrastructure grant. GEPRIS

The project ran from January 2018 to March 2020.