The purpose of this research was to quantify the strength of novel composite repair methods for thin-walled composite structures. Carbon/epoxy plates were manufactured and repairs were made at Gloyer-Taylor Laboratories. At BYU, specimens were damaged in a controlled and repeatable process. Three damage modes were implemented: impact, groove, and abrasive damage. Tensile strength tests were performed on control, damaged, and repaired specimens. Four 24 x 24 in (60 x 60 cm) carbon/epoxy plates were received. Each plate was made up of seven plies cured together with epoxy resin for a nominal total thickness of 0.04 in (1.02 mm). The thickness, however, was not uniform: each plate had a smooth side and a wavy side. This resulted in inconsistent damage depth. The plates were cut at BYU using a water-jet cutter into 1 in. (25.4 mm) wide by 8 in. (203.2 mm) long test specimens. Test specimens were grouped into four categories: control specimens, specimens inflicted with damage by machining a shallow groove ranging from 0.012 — 0.018 in. (0.30 — 0.46 mm) deep, specimens inflicted with an abrasive-type damage ranging from 0.006 — 0.012 in. (0.15 — 0.30 mm) deep, and specimens subjected to impact damage ranging from 1.47 — 3.23 J. Five specimens were placed in the control group. Ten specimens were placed in each of the remaining damage groups. All ten specimens were damaged, but only five of each were sent to be repaired. The randomization of the thickness variable was prevented by the desire to repair damaged specimens as a group with a single repair rather than performing repairs on every individual specimen. The stress-strain behavior confirm the control specimens generally exhibited the best overall behavior, as expected. Most damaged specimens, including the repaired specimens, exhibited lower ultimate stress than the undamaged control specimens. The repaired specimens exhibited a higher initial stiffness than either the control or damaged specimens, due to the stiffness of the composite patch. Although repaired specimens should exhibit higher strength than damaged specimens, but this however, was not always the case. In particular, repairs did not improve the ultimate strength of the specimens damaged by abrasion. Correlations between the different damage types were developed, relating damage intensity and strength was approximately. This suggests further investigation is needed.



College and Department

Ira A. Fulton College of Engineering and Technology; Civil and Environmental Engineering



Date Submitted


Document Type





composite, carbon fiber, damage, repair



Included in

Engineering Commons