| About this Abstract |
| Meeting |
2024 ASC Technical Conference, US-Japan Joint Symposium, D30 Meeting
|
| Symposium
|
2024 ASC Technical Conference, US-Japan Joint Symposium, D30 Meeting
|
| Presentation Title |
Bayesian Calibration and Uncertainty Quantification of Cohesive Zone Parameters at an Adhesively Bonded Textile and Fabric Interface |
| Author(s) |
Marcus Stanfield, Matthew Kirby, Erin DeCarlo, Carl Popelar, David Riha |
| On-Site Speaker (Planned) |
Marcus Stanfield |
| Abstract Scope |
The potential of reliable bonded composite structures in advanced aircraft systems lies in enhancing performance, cost-efficiency, and design flexibility. One effective approach involves bonding 3D woven textiles to stiffeners and wing skins, yet conservative certification standards and the need for fail-safe design limit their widespread use. Progressive Damage Failure Analysis methods offer a solution by accurately predicting performance, reducing reliance on costly certification tests. The U.S. Air Force Research Laboratory's OPPERA program focuses on developing predictive tools to assess the performance of bonded composite joints for aircraft systems, particularly the pi-preform joint, using a Verification, Validation, and Uncertainty Quantification framework. Previous sensitivity analyses have identified the cohesive zone disbond properties as significant parameters in the predicted pi-joint performance. This study aims to advance a predictive engineering tool to assess disbond between the pi-preform textile and laminate. Using MATLAB and Abaqus, a parametric finite element model of fracture specimens was created to systematically identify significant phenomena and quantify uncertainties. The model supports three loading configurations: Double Cantilever Beam, End-Notched Flexure, and Single Leg Bend. Material responses are simulated using orthotropic linear elastic properties, with cohesive surfaces modeling disbond failure. Fracture specimens, excised from pi-joints, underwent testing with data collected using standardized methods and Digital Image Correlation to monitor crack tip behavior. Model calibration involved creating Gaussian Process response surface models of the fracture coupon simulations and then using Bayesian methods to quantify uncertainty in the cohesive zone disbond properties by estimating their posterior distributions. |
| Proceedings Inclusion? |
Definite: Post-meeting proceedings |