|M.Sc Thesis||Department of Mechanical Engineering|
|Supervisors:||Prof. Altus Eli|
|Prof. Emeritus Ishai Ori (Deceased)|
|Full Thesis text|
Graphite Phenolic (GPh) composites are used for light weight structures, which serve under mechanical and HygroThermal (HT) loadings. Thick laminated GPh panels and beam specimens cut through their thickness were exposed to severe drying conditions for a short duration. Interlaminar cracks were detected during the first two days of the panels drying exposure. The residual interlaminar flexure strength of the beams was found to be significantly reduced to less than half of its ultimate strength during the first hours of the drying exposure but recovering its initial strength when the beam was totally dry. These findings were attributed to the high shrinkage strains which were directly related to the water loss during drying. When shrinkage of the external surface is restrained by the internal wet core of the panel, tensile stresses are induced at the external surface layer. The combination of the time-dependent process with internal stresses near free edges may lead to a "critical time" during which the composite is most vulnerable to failure.
Modeling, analyzing, and predicting of these phenomena were conducted by the following stages:
1) Measuring the Shrinkage and Weight Loss of beam specimens during a controlled drying process, from which the Coefficients of Moisture Expansion, diffusion rate and initial moisture concentration were derived.
2) Modeling of the mechanical and HT stresses of beam specimens by using the HT properties found experimentally.
3) Measuring the residual interlaminar strength at different drying times by three-point bent flexure testing.
4) Combining the hygrothermal stress fields obtained by the beam model in (2) with the mechanical strength measured in (3) to propose a failure criterion which includes mechanical and hygrothermal loads. A local failure criterion is not appropriate and a non local failure criterion based on averaging the stress field over a "characteristic length" in the vicinity of the free edge is proposed.
5) Developing a simplified model for the HT stresses in the vicinity of the free edge region of a thick panel.
6) Predicting the critical time for which the panel is most vulnerable to microcracking by applying the above failure criterion (4) on the simplified panel model (5). It was found that the predicted critical time is in the same order (two days) as seen in the experiments.
The above model may serve as a material selection tool for laminated composite structures under long-term drying conditions based on short-term experiments.