3D Stress Analysis of Adhesively Bonded Composite Joints

James Zhang, Brett Bednarcyk, Craig Collier, Phil Yarrington, Yogesh Bansal, Marek-Jerzy Pindera
2005 46th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference   unpublished
A robust and rapid analytical method for 3D stress analysis of composite bonded joints has been recently developed based on Mortensen's unified approach, with considerable extension to accommodate hygrothermal loads and most importantly, to compute the inplane and out-of-plane, through-the-thickness interlaminar peel and shear stresses in the laminate adherends. Compared to other analytical methods for bonded joint analysis the present method is capable of handling more general situations,
more » ... ding various joint geometries, both linear and nonlinear adhesive, asymmetric and unbalanced laminates, and more general loading and boundary conditions. The formulation has been extended from strict cylindrical bending to consider generalized cylindrical bending that allows an arbitrary constant strain to be applied in the out-of-plane direction. Other analytical methods, such as Hart-Smith's, are 1-D and mainly focus on obtaining adhesive stresses, while generally ignoring stresses in laminate adherends, particularly interlaminar stresses, which are known to be key contributors to failure. Joining composite structures using adhesive bonding remains a challenging problem because performance is severely influenced by the characteristics of the composite laminate adherends, which usually have low interlaminar strengths. This new method, most importantly, computes local 3D stress fields in each ply of each adherend, which vary along the joint. Given the realistic 3D local stress fields at the ply level within each adherend, failure criteria can be employed to predict joint strength, which can facilitate better joint designs. This paper addresses the computation of stresses for composite bonded doubler joints. A companion paper, also presented in this conference, addresses failure prediction.
doi:10.2514/6.2005-2021 fatcat:k5esrnonorc4febzxdutcpokfa