Establishment of an experimental method for a grooved composite subjected to out-of-plane contact loading

Abstract

A specimen and an experimental method to observe the behavior of a grooved composite subjected to out-of-plane contact loading is established and verified, and its response is examined. The specimen is designed so that the variability of stress-strain state is negligible across the width of the specimen. The dominant concept of the design is to isolate the response of the specimen around the groove from any other effects. Geometric parameters, stacking sequence (layup), and boundary conditions are determined for the specimen. With simply-supported boundary conditions, specimens fail in a simple beam shear mode as determined from the overall structural response of the specimen, thereby indicating that this configuration is not appropriate for the primary design goal. Thus, the rigid backface boundary condition is chosen and verified as the appropriate configuration. Contact, load transfer, and alignment issues arose in the first set of rigid backface tests and were solved by introducing finer machining, harder material for the indenter, and overall alignment with better accuracy. This resulted in the final test specimen configuration and associated test method, consisting of a specimen with a length of 56.00 mm, a width of 25.00 mm, an approximate thickness of 12.5 mm, and a maximum groove depth of 3.48 mm. The standard layup used for the tests is [F45/0/90]1os, while an alternate layup of [T30/0]13S was also used. In these tests, a number of key behaviors were observed: mode of failure, load-per-stroke slope, and "knee load".

(cont.) Specimens failed in two different modes: a delamination near the bottom of the groove (Mode A), and a crack under the groove propagating to a delamination near the midplane (Mode B). From observations, it is concluded that damage is generated at the bottom of the groove and then propagates in the longitudinal and the thickness direction, resulting in Mode A or B respectively. A "knee load" is defined as a point where the load-per-stroke slope deviates from linear behavior. Failure Mode B and the presence of the "knee load" are observed in the standard layup, but are not applicable to the alternate layup. The presence of 90' plies is indicated as the main cause of the observed differences. The test results clearly show that a specimen and a test method is established and verified for the objectives of the current work, and furthermore is valid for tests with different test parameters. Recommendations are made with regard to extension of the basic testing established herein.