Probabilistic multi-scale characterization and prediction of bimaterial bondline structural reliability
Abstract
Design of new vehicles and repairs for aging vehicles using multiple material types provides a novel solution for Navy structures spanning air, land, and sea operations and supports the transition to affordable, lightweight structure in the construction and repair of Navy ships, patrol crafts, and attritables. In addition to significantly improving affordability, implementation of multi-materialstructure has been proven to produce faster vehicles, reduce fuel consumption, and increase payload capacity. One impediment to widespread use of multi-material construction is that joining dissimilar materials using traditional methods is either impossible (for example composites cannot be welded to metals), results in heavy structure when bolted, increases cost due to manual labor for joining, or causes stress concentrations that can become damage initiateon sites. An enabling technology for joining multi-materials without increasing weight and which overcomes many of the limitations of traditional joining methods is adhesive joining. Adhesive joining uses a polymeric material (adhesive) to bond the two dissimilar materials (adherends) such that theadhesive provides strength and stiffness to the structure. In addition to enabling the bonding of dissimilar materials, other benefits of adhesive joining include minimizing the effects of heat distortion due to differences in thermal expansion, minimal alteration or damage to either material, and increased manufacturing efficiency resulting in considerable cost savings by reducing part count, manual labor, and production time. While the advantages of adhesive joining are evident,predictive methods for structural reliability do not capture the effects of installation quality and microstructure and are based on material properties that must be characterized through experimental testing.A multi-scale, multi-physics computational solution is proposed to probabilistically characterize and predict the process-microstructure-performance relationship of an adhesive bondline between metal and composite materials. Bondline structural reliability will measured by strength andfracture toughness degradation under static loading given uncertain installation quality and subsequent microstructure defects including adhesive thickness, voids (size, shape, and distribution), surface roughness, water intrusion, and contaminants. Analysis on high performance computing (HPC) systems enables the rapid investigation of varying configurations and installation quality, as well as the exploration of this large parameter space in a feasible amount oftime. Multi-physics models will be formulated at three scales spanning methodologies (coarsegrained molecular dynamics, peridynamics, and finite element analysis) using the results and knowledge gained from concurrent experimental testing and parametric analysis. Experimental testing results will be used to validate the models and computational results, and model andmethodology development will inform testing focused on the most influential behaviors and parameters required for physics-based formulation. Parametric analysis will produce surrogate models that will be used for sensitivity analysis to identify the most influential parameters and uncertainty quantification to evaluate the impact of these parameters.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Dec 04, 2020
- Source ID
- N000142112041
Entities
People
- Stephanie Termaath
Organizations
- Office of Naval Research
- United States Navy
- University of Tennessee