Logo image
Back
Failure prediction in large scale composite laminates containing bonded repairs
Dissertation   Open access

Failure prediction in large scale composite laminates containing bonded repairs

Tim Michael Labik
Doctor of Philosophy (Ph.D.), Drexel University
Jun 2024
DOI:
https://doi.org/10.17918/00010484
pdf
Labik_Tim_202420.27 MBDownloadView
PDF Open Access Open Access (License Unspecified)

Abstract

Laminated materials--Deterioration Laminated materials--Testing Composite materials--Testing Failure prediction Scarf repairs
Over the past several decades, aircraft companies have increasingly incorporated carbon fiber reinforced polymers (CFRP) composites in their aircraft structures, replacing conventional aluminum alloys. However, CFRP structures invariably are subjected to damage (e.g., handling and bird impact), rendering them in need of repairs. The primary method of repairing CFRP structures involves sanding away the damaged area in a taper and installing a bonded repair patch flush with the removed area to maintain aerodynamic efficiency. A key disadvantage of such bonded repairs is the lack of reliable methods to detect weak bonded repairs. As such, the Federal Aviation Administration (FAA) has issued in 2014 a policy memo, stating that: "bonded repair must be further limited to a maximum size whereby limit load residual strength can be demonstrated with a complete or partial failure of the bond within the repair of base structure arresting design features." The FAA William J Hughes Technical Center, in collaboration with The Boeing Company, has conducted an extensive testing program using its unique Airframe Beam Structural Test Fixture (ABST), testing composite laminates containing different scarfs (both with and without repairs). In this study, composite laminate panels which contain various scarf configurations (i.e., full-depth scarf, double-sided scarf, open-hole, fiber-oriented scarf), with and without repairs, are subjected to quasi-static failure testing. These panels simulate the geometry and repairs of current transport category aircraft mid-span wing panels and are tested under uniaxial in-plane tensile loading conditions. Nondestructive inspection technologies, such as strain gages, digital image correlation, and flash thermography, are employed to record strain distributions and to monitor damage initiation and progression up to the final failure. The test results provide a multifaceted dataset on damage initiation and progression and fracture loads of such composite laminates under the prescribed loading conditions. Simultaneously, a finite element based predictive framework, incorporating NASA's CompDam composite damage model (which itself is based in continuum damage mechanics (CDM)), has been developed to simulate damage initiation and progression in the test panels. This framework predicts initiation and progression of fiber damage, matrix damage, and disbonding (in cases where repairs are implemented), with different, advanced failure criteria and damage evolution laws for each respective damage type. The framework has been applied to two thermoset CFRP laminated composite containing double-sized scarfs, one with no repairs, and one with a single sided repair. The results show a good agreement between experiments and predictions in terms of strain fields, initial damage loads, and corresponding damage progression paths. The good agreement demonstrates the ability of this methodology to supplement or replace the need for experimental testing on CFRP panels, saving time and money when designing aircraft and repairs. A feasibility study is also presented on replacing the Newton-Raphson convergence routine in CompDam with an explicit Newmark algorithm. Several analytical models run within this thesis were terminated by nonconvergence, and adjusting the convergence routine would alleviate this. Future work includes further validations of the framework by studying similar configured panels with multiple patch repairs, incorporating disbond prediction capabilities, and experimental and analytical testing of various scarfs in different material systems. Further future work would also include implementing a similar framework on the experimentally observed honeycomb panel configurations, as well as the experimentally tested automated scarf thermoset panels. Additional future work includes refining the explicit Newmark algorithm to replace the current implicit Newton-Raphson convergence routine of CompDam, providing a quicker, more robust convergence scheme.

Metrics

98 File views/ downloads
121 Record Views

Details

Logo image