Minimizing scatter in load-controlled fatigue test data of textile polymer matrix composites using statistical techniques

Abstract

Polymer matrix composites are composed of fiber reinforcement and polymer matrix. Fiber reinforcement is major load carrying component whereas matrix transfers load between the fibers. Fiber reinforcement in the form of textile fabric is popular in the composite industry. Fibers in textile form exhibit good out of plane properties, and good fatigue and impact resistance. The variety of fabric architectures includes weaves, knits, braids, and stitched fabric. Mechanical properties of the composites are dependent on fiber volume percentage and fiber orientation angle. Composites manufactured using low-cost vacuum infusion processes typically have varying fiber volume percentage from location to location and there exists some degree of fiber misalignment. In load controlled fatigue test, fatigue stress is applied as percentage of average ultimate tensile strength (UTS). Due to variation in fiber volume percentage and misalignment in fibers the UTS varies from specimen to specimen. The variation in the UTS between the specimens causes relatively large scatter in the fatigue data. The scatter can be confirmed on stress-cycle (S / Su − N) diagram. Stress-cycle diagram is major tool in predicting fatigue life of the composites. The large scatter in this diagram causes erroneous predictions. The large scatter is because of the considerable difference between average UTS and specimen’s actual UTS. It is necessary to apply fatigue stress based on specimen’s actual UTS. Since, evaluation of UTS is a destructive test there is need to predict UTS of specimen using some analytical tool. In this research braided fabric which is one of the forms of textile fabrics, was used in vacuum assisted resin transfer molding (VARTM) to produce flat composite panels. It was observed that the UTS is a function of the braid angle and fiber volume percentage. Using statistical approach an equation is derived for UTS as function of braid angle and fiber volume percentage. It is suggested that braid angle and fiber volume percentage be measured on each specimen. Then compute UTS of that specimen using derived equation. Then apply fatigue stress based on this computed UTS. This approach would minimize scatter in fatigue data. This approach can be used for any type of composites where UTS is dependent on multiple factors such as fabric orientation angle and fiber volume percentage.