|dc.description.abstract||Fiber-reinforced rubberlike materials (FRRM) commonly used in tires undergo large deformations, and exhibit different response in tension and compression along the fiber direction. Assuming that the response of a fiber-reinforced rubberlike material can be modeled as transversely isotropic with the fiber direction as the axis of transverse isotropy, we express the stored energy function, W, in terms of the five invariants of the right Cauchy-Green strain tensor and the fiber direction, and account for different response in tension and compression along the fiber direction. It has been shown in the literature that in shear-dominated deformations, the 5th invariant, I5, significantly contribution to the stress-strain curve. We have implemented the constitutive relation in the commercial software, LS-DYNA. The numerical solutions of several boundary value problems studied here agree with their analytical solutions derived by using Ericksen's inverse approach, in which a part of the solution is assumed and unknowns in the presumed solution are then found by analyzing the pertinent boundary value problem. However, computed results have not been compared with experimental findings.
For W of the FRRMs an expression that is a complete quadratic function of the five invariants is also examined. Homogeneous deformations such as simple extension, simple shear, and biaxial loading problems are studied to delineate the mechanical behaviors of FRRMs. Consistency with the infinitesimal deformation theory requires that linear terms in the 4th and 5th invariants, I4 and I5, be included in the expression for W. Stability analysis of deformations reveals the qualitative changes triggered by the second order terms of the quadratic function. Analytical solutions for inflation, extension and twist deformations caused by internal pressure, end torque, and axial force for a pressurized cylindrical laminate are derived using Ericksen's inverse method. Effects of fiber orientations on the mechanical behaviors of a +/-α angle-ply cylindrical tube are investigated using the derived analytical solutions.
The T-peel test, widely used for characterizing adhesion across a plethora of adhesives, adherends, and geometries, results in a range of responses that may complicate meaningful interpretation of the test data. This research effort, involving several specific specimen types, was undertaken to investigate concerns that commonly used configurations may not always result in plateaus in the force-displacement response. We experimentally and numerically study debonding of T-peel specimens having 75 mm bond length and 0.81 mm thick adherends made of either 6061 aluminum (Al) or one of the three steels (G70 70U hot dip galvanized, E60 elctrogalvanized (EGZ), 1010 cold-rolled steel (CRS) bonded with either LORD® 406 or Maxlok™ acrylic adhesive. For the EGZ and the Al adherends, specimens with a bond length of 250 mm and adherend thickness of 1.60 mm are also examined. Effects of adherend materials and thicknesses, bond lengths, and adhesives on test results are examined using three metrics to interpret the T-peel bond performance. We find a limited correlation between the commonly used "T-peel strength" and the energy dissipated per unit debond area. For those two metrics, the relative performances of the CRS and the Al specimens are quite different. Quasi-static plane strain deformations of the test specimens are analyzed by the finite element method (FEM) and a cohesive zone model using the commercial software, ABAQUS, to help interpret the test data. Numerical results provided energies required to elastically and plastically deform the adherends, and help determine the transition from non-self-similar to self-similar debonding. The FE simulations also facilitate determination of the fraction of the crosshead displacement at which self-similar debonding occurs. Results reported herein should help practitioners select appropriate specimen dimensions for extracting meaningful data for adhesive performance.||en