COMPOSITE METALS

Metal to Composite Bonded Joint Subjected to Heat and Load

Metal to Composite Bonded Joint Subjected to Heat and Load

Problem Statement

A bonded joint finite element (FE) for a symmetric double lap joint is developed that is capable of predicting field quantities in the lap region. The element is a hybrid method and incorporates features of classical analytical and numerical methods. The element stiffness and load vector formulations have unique, load dependent, non-linear shape functions based on an analytical solution. The adaptive shape functions are formulated in terms of the dimensionless mechanical load fraction and total load and are capable of predicting the thermal and mechanical load response.

The bonded joint element has been implemented as a user element in the Abaqus R commercial FE code. A comparison of the stress predictions for the bonded joint element and a conventional 2D FE model is presented and are found to be in good agreement. Therefore, the element provides a computationally efficient and mesh-independent stress prediction. The single element reproduces the analytical solution with minimal analyst input and can be easily incorporated into early design and sizing studies.

Aims of the Dissertation

The objective of this dissertation is to develop an element capable of predicting basic joint performance with a limited number of degrees of freedom (DOF) and with little meshing overhead. As a result, this element could be adopted for initial joint sizing in FE models at all system levels. The element is formulated to predict stress and strain fields of orthotropic constituents in thermal (or any scalar) and mechanical loading environments. The orthotropy of a joint is of particular concern in laminated composite materials since transverse properties are often significantly lower than in-plane properties in a laminate . Temperature dependence is included since anisotropic materials (such a long fiber-reinforced composites) often require high-temperature curing cycles.

Proposed Dissertation approach

There are many factors that affect the stress field and associated bonded joint failure. Theseinclude adhesive spew and the geometric discontinuity and unbounded stresses associated withstepwise geometries . Additionally, material non-linearity has a significant effect on the stressfield and requires a level of material characterization that is often unavailable early inan analysis cycle. All of the specialized joint analysis techniques (cohesive elements, the virtualcrack closure technique and others) require material properties that can be difficult to obtain (suchas critical energy release rates and cohesive strengths). In many circumstances, a designer hasinsufficient information or time to obtain a highly accurate solution and instead would prefer asimple analysis that captures primary effects.

These types of analyses are often useful in tradestudies and to identify likely problem areas needing further study.With that goal in mind, it might be considered adequate to perform a conventional linear elasticFE analysis on an idealized geometry such as a joint with square corners (i.e. the double lap jointin Figures 1 and 2). In that solution type, however, the singular stress field causes a broad rangeof predicted stresses near the edges, particularly at the material interfaces. This is an undesirableand an unavoidable feature that emerges ...