| Finite element analyses with experimental verification on adhesive joint behavior have been conducted in an effort to enhance the engineering design process for adhesives. This research focused on lap shear tests modeled after ASTM D 1002, D 3528 and associated standards. Lap joints were designed, analyzed and tested for different geometries and with various bond termini constraints applied during these uni-axial pull tests. The basic premise of this dissertation is that modern computational analysis, in conjunction with fracture mechanics, can be used to elucidate the significant role of subtle changes in adhesive joint geometry on joint strength. These changes are often ignored in the adhesive joint design.; Joints were modeled using finite element software, and the adhesive bond strength predicted. The joints were then manufactured and tested to compare the finite element-fracture mechanics predictions with experimental observations.; This research examined the energy release rate (ERR) as it applies to crack propagation in adhesive joints. The ERRs were calculated as a function of adhesive thickness, starter crack length and location, bond termini constraint, and other geometric alterations. Methods to strengthen adhesive joints by such alterations were explored both computationally and experimentally.; Adhesive joint design is still more art than science and typically does not take advantage of nor always follow modern engineering methods.; This research clearly demonstrates that calculating the required bond area by dividing the force at failure by the tabulated adhesive shear strength value is a gross misconception and potentially dangerous. This research further demonstrates that (1) cleavage stresses and shear stresses, near the bond termini, contribute to the adhesive joint failure mechanism(s); (2) constraining the lateral motion of the bond termini with a lateral force decreased the ERR increasing the joint strength; (3) increasing the bond area by increasing the bond overlap only marginally increases the joint strength, thus violating the average stress rule commonly used in adhesive joint design; (4) the unconstrained bond termini always had a higher ERR than the constrained bond termini; (5) the value of the highest ERR was a strong function of the adhesive thickness and (6) that decreasing the adhesive thickness increased the joint strength. This and other work in this laboratory demonstrated that cracks tend to grow from regions of highest ERR's and follow the path of highest ERRs. |
