An experimental study of the mechanical behavior of cracked aluminum specimens repaired with composite-material patches has yielded findings that may eventually contribute to the development of more- effective composite patches and of techniques for predicting and detecting failures in composite-patched aluminum aircraft panels. Some prior studies have addressed various aspects of composite-patch repair of aluminum specimens, but until now, little attention has been given to such important aspects of mechanical behavior and properties as the relationships among stress, strain, and growth of cracks through-out the lifetimes of specimens. In this study, effects of initiation and growth of cracks on the residual strengths of the patched specimens were characterized. This study established a correlation among damage modes, residual strengths, and evolution of strain inside and outside the patched areas.

Aluminum Dog-Bone Specimens containing deliberately introduced defects were repaired with composite patches and then subjected to tensile tests.
In the experiments, tensile tests were performed on six 7075-T6 aluminum dog-bone specimens. A hole and widthwise notches were machined at the middle of each specimen, then widthwise starter cracks were grown from the notches (see figure). A rectangular patch made of a 16-ply composite of unidirectional (lengthwise) boron fibers in an epoxy matrix was centered at the hole and bonded to the aluminum by use of an adhesive film comprising a toughened adhesive in a polyester knit cloth carrier. Each specimen was instrumented with strain gauges at several locations inside and outside the patch area.

A monotonic tensile test was performed on one specimen to determine baseline strength. Fatigue tests were performed on two other specimens to establish baseline fatigue life. The three other specimens were fatigue-tested to different fractions of the expected fatigue life, introducing damage that was then analyzed by non-destructive evaluation techniques. Then each of these three specimens was subjected to a monotonic tensile test to failure in order to characterize (1) its residual strength, (2) the evolution of strain inside and outside its patch area, and (3) the correlation between debonding of the patch and strain measurements at various locations. Debonding was detected by use of thermal imaging following transient heating. Crack lengths were measured by use of a travelling microscope with a digital readout.

Some of the results of the tests were as anticipated; others were not as anticipated. Two main conclusions were drawn from the results:

  • Strain values were found to be approximately proportional to crack lengths, with no noticeable dependence on the crack-growth rates. For strain gauges close to cracks, crack lengths could be determined to within close approximations from the strain readings alone.
  • It was found that when a cyclically loaded specimen is overloaded in the sense that the load applied during one or a few cycles significantly exceeds the standard cyclic load, then there occurs a retardation effect in the sense that crack growth slows significantly for a while. Once the standard cyclic loading is resumed, the crack-growth rate rises, gradually approaching the rate for standard cyclic loading as though the overload had not occurred.

This work was done by Michael A. Hansen of the Air Force Institute of Technology.


This Brief includes a Technical Support Package (TSP).
Cracking of Aluminum Panels Repaired With Composite Patches

(reference AFRL-0098) is currently available for download from the TSP library.

Don't have an account? Sign up here.