ABSTRACT
A method is proposed for determining the in situ shear modulus of a structural adhesive from a sandwich beam loaded in 3-point bending in which the adhesive is contained as a thin layer. Expressions for calculating the elastic shear modulus of the adhesive layer from compliance data on the beam are derived, and experimental tests to validate the theory are conducted. To verify the test results, tensile tests are also conducted, and the shear modulus for bulk adhesive is determined using the constitutive equation for an isotropic material relating tensile modulus and Poisson’s ratio to shear modulus.
However, the bulk shear modulus as traditionally determined from a tensile test was up to an order of magnitude greater than the in situ shear modulus obtained from the 3-point bend test. A finite element simulation and sensitivity study replicated the experimental results of the 3-point bend tests, and showed that using the shear modulus obtained from the tensile tests would result in significant errors in predicting material and joint behavior. In addition, torsion tests were conducted on bonded cylinders to measure directly the shear modulus. The shear modulus from the torsion test was in agreement with the in situ modulus obtained from the 3-point bend test. This combined experimental-computational approach validated the 3-point bend test as a means to determine the in situ adhesive shear modulus. Finally, micrographs of the interface of the 3-point bend specimen indicated that adhesion occurred by the extension of adhesive pillars to the surface of the adherends. This pillaring phenomenon may have resulted in a lack of bonding along significant portions of the interface, and may explain the compliance of the 3-point bend specimens and, subsequently, the lower shear modulus. The repeatability of the experiments and the substantiation of the results of the experiments by finite element analysis suggest that this pillaring phenomenon may be a mechanism of adhesion.
