Abstract

Composite materials are increasingly being used in the aerospace industry today due to their excellent specific strength and stiffness characteristics. Unfortunately the full potential of efficient composite structures is yet to be realised due to limitations of analytical and numerical models to accurately predict material failure. Aerospace structures tend to comprise of thin plates or skins relying upon stiffeners for lateral and in-plane stability. Under in-plane compressive loading conditions, such structures tend to fail due to the microbuckling of 0 degree plies and subsequent skin/stiffener separation. The aim of this study is to extend our knowledge in the field of strength prediction in stiffened carbon fibre reinforced plastic (CFRP) panels subjected to in-plane compressive loading. The panels may include stress concentrators in the form of open holes or low velocity impact damage.

The suitability of the Soutis-Fleck fracture model coupled with finite element analysis in predicting the microbuckling failure loads of various CFRP structures is explored. The predictions made by the fracture model are fair, albeit conservative, even in the case of a complex structure such as a fighter aircraft wingbox.

The study progresses onto the phenomenon of skin/stiffener interface failure. The phenomenon is explored using detailed two and three-dimensional (2-D, 3-D) finite element modelling incorporating novel interface elements. A global/local approach is taken in simulating skin/stiffener debonding failure. This is further explored experimentally using the four and recently proposed seven-point bending tests. It is known that skin/stiffener failure usually occurs at a location of maximum bending or twisting moments. Data extracted from a full-size stiffened panel compression test is compared with experimental and computational submodel predictions.

The four and seven-point bending tests display a mildly non-linear loading response highlighting stiffener web/cap effects. The failure loads of the four-point bending specimens were on average 30% lower than the seven-point specimens suggesting edge effects dominated their failure. However, strains measured during the seven-point bending tests exceed similar strains measured during the full-size panel test. The study therefore concludes that a global/local approach to predicting the failure load of stiffened CFRP panels is feasible provided the failure mode consists of fibre microbuckling or skin/stiffener debonding failure.