Recent Advances in Bolted composite Joint Strength Prediction

AMPTIAC Directory ... 15 The beginning In late 1993, the Mechanics and Surface Interactions Branch of the Air Force Research Laboratory’s Materials Directorate (AFRL/MLBM) was tasked to investigate the accuracy of the analysis methods used to predict the strength of bolted composite joints used on high-performance fighter aircraft such as the F-22 Raptor. The strength of bolted joints is essential to the long-term reliability of a structure and the concern at the time was that the inaccuracy of the method could lead either to increased risk or excessive structural weight. There are a number of different methods being used by industry to model the strength of bolted joints. Lockheed Martin uses a method known as IBOLT. Boeing developed an approach known as BEARBY. Still others use a method known as BJSFM. These approaches are two-dimensional semi-empirical models that incorporate Lekhnitski’s classical solution of holes through laminated plates [1] and they all assume a fixed contact zone between the bolt and the composite. Each method relies upon extensive and costly experimental testing to develop the empirical parameters required to operate the models. The parameters are essentially used to calibrate the models so they can approximate failure in more complex composite joints. The results of MLBM’s investigation indicated that these methods could predict strengths that vary considerably from statistically measured joint test configurations. The only direct comparisons of predictions and measurements that could be made were for very simple uniaxial cases. Due to the difficulty in conducting combined loading tests, it was not possible to compare test results to predictions of simulated structures that were subjected to multiaxial loading and shear. Hence, the amount of conservatism or risk associated with the predicted strengths of joints subjected to complex stress fields could not be ascertained. The semi-empirical methods discussed above represented the best available technology for predicting joint strength. However, after many years the power and availability of high-speed computers was such that new approaches could be considered. Following the initial study into the accuracy of these methods, the author, then a materials research engineer assigned to AFRL/MLBM, proposed developing a new advanced method based upon accurate 3-D stress analysis. Due to the known inaccuracies in conventional 2-D finite element analysis (FEA) solutions and the difficulty and computational power required to develop 3-D composite finite element models, the decision was made early on to seek a fresh approach to the problem. The decision to not use 3-D FEA was later validated