A research team at Drexel University's College of Engineering, led by Professors Ahmad R. Najafi, Jonathan Awerbuch, and Tein-Min Tan of mechanical engineering and mechanics, has developed a new way to predict how tiny cracks in aircraft fuselages grow and join together over time. The work, sponsored by a grant through the FAA-Drexel Fellowship Research Program, could help engineers design safer aircraft and develop better maintenance schedules.
The research, which appears in the International Journal of Fatigue, addresses a problem that became tragically clear in 1988, when part of an Aloha Airlines Boeing 737 fuselage tore away mid-flight at 24,000 feet. That incident revealed how small cracks at rivet holes can grow toward each other during repeated takeoffs and landings, eventually joining to form dangerous long cracks.
"When multiple fatigue cracks form at adjacent rivet holes, they grow toward each other over thousands of flights, interact, and eventually link up to form longer fatigue cracks," Najafi explained. "Such crack link-up behavior significantly reduces the residual strength of the structure and, if unchecked, can lead to catastrophic failure. Our framework provides a robust tool for damage tolerance assessment and life prediction in aircraft structures."
Xiaomo Zhang, a recent PhD graduate and lead author of the paper, conducted fatigue tests at the FAA Technical Center on scaled-down versions of aircraft fuselage panels, applying force thousands of times to mimic the stress of takeoffs and landings. Periodically, they would apply a different force level for a few cycles, which left visible striations on the cracked metal surface. These marker bands showed the researchers exactly where the crack edges were at specific points in time, creating a detailed history of how the cracks grew. To better replicate the conditions in actual curved fuselage panels, the researchers used a fixture on some specimens to reduce unwanted bending, which improved their accuracy in modeling real-world behavior.
Zhang, and Li Meng, a post-doctoral research associate, then built computer models to simulate the crack growth. The framework uses what's called a phase field approach, which treats cracks not as sharp lines but as zones where damage gradually increases. This is a departure from older finite element methods that require engineers to predict exactly where a crack will form and then track its precise path as it grows – a nearly impossible task when multiple cracks are interacting.
"This approach is particularly powerful because it can simulate multiple cracks, branching, and link-up behavior," Zhang said. "The computer model doesn't need to know in advance where cracks will appear or which direction they'll grow. It calculates damage throughout the material and lets the cracks emerge naturally from the simulation."
The Drexel team also made an important refinement to existing phase field methods. They adjusted how the model accounts for stress during the unloading part of each pressure cycle, preventing the simulation from overestimating damage accumulation. This modification led to predictions that closely matched the experimental results, coming within about 15 percent of the actual number of pressure cycles needed for cracks to link up.
"This work demonstrates the effectiveness of our approach in capturing fatigue crack growth and link-up in aircraft fuselage lap joints," Najafi said. "As computational power continues to grow, this framework has the potential to enable accurate life prediction for large-scale structures, helping engineers design safer aircraft and develop more effective maintenance strategies."