SBIR-STTR Award

A 9 month Phase I research program is proposed for studying the three-dimensional state of stresses at bolted/bonded joints in composite laminates, by using a new layer-by-layer hybrid BEM (Boundary Element method)-Sinc interpolation technique. The primary goal of this research is to (i) mimic the singular behavior of the three-dimensional stresses in the vicinity of a vertex (i.e., junction corner point of a laminated plate or surface corner point of a bi-material interface crack), and (ii) perform elastic stress analyses of bolted/bonded joints of practical composites having anisotropic layers, interacting cracks and arbitrary geometries. A Sinc interpolation coupled with boundary integral formulation is proposed as a replacement of the conventional isoparametric polynomial based BEM. This is intended to capture the accurate distribution of interlaminar stresses in the vicinity of singularities, such as those due to free edge, corner or a combination of both. Finally, since any three-dimensional numerical procedure is highly computation intensive, a singular perturbation method will be employed in conjunction with the aforementioned hybrid BEM-Sinc approach. Neither this type of computational procedure nor any numerical result is currently unavailable in the literature

A 2-year Phase II research program is proposed for studying the three-dimensional state of stresses at bolted/bonded joints in composite laminates, by using layered Green's function based hybrid BEM (Boundary Element method)- Sinc convolution technique. The primary goal of this research is to (i) mimic the singular behavior of the three-dimensional stresses in the vicinity of a vertex (i.e., junction corner point of a laminated plate or surface corner point of a bi-material interface crack), and (ii) perform elastic stress analyses of bolted/bonded joints of practical composites having anisotropic layers, interacting cracks and arbitrary geometries, e.g., holes, cutouts, etc. A Sinc convolution coupled with boundary integral formulation is proposed as a replacement of the conventional isoparametric polynomial based BEM. This is intended to capture the accurate distribution of interlaminar stresses in the vicinity of singularities, such as those due to free edge, open hole edge, corner or a combination of both. Our results for bench-mark problems, such as a composite laminate weakened by a hole, will be compared with those computed by FEM, spline-based and BEM codes in regards to accuracy and speed. Finally, since any three-dimensional numerical procedure is highly computation intensive, a singular perturbation method will be employed in conjunction with the aforementioned hybrid BEM-Sinc approach. Neither this type of computational procedure nor any numerical result is currently unavailable in the literature. The potential exists for a user-friendly, interactive hybrid BEM-Sinc computer code that can accurately predict progressive damage and failure of composite bolted joints of arbitrary geometries, and can aid in load-carrying assessments of bonded joints. As conceived, the proposed code will be a powerful analysis tool with wide applicability and high demand in the commercial and military aerospace industries, as well as in other industries where composites are being utilized, such as automotive, marine and sporting goods.

Keywords:
BOLTED AND BONDED JOINTS BOUNDARY INTEGRAL EQUATIONS FREE EDGE STRESS SINGULARITY HYBRID BEM-SINC