Mechanical Engineering Seminars - Effects of Temperature and Primary Creep in Hybrid Metal to Composite Bolted Joints
Thesis Advisor: Dr. Vincent Caccese
Hybrid metal to composite bolted joints are the focus of much research due the inherent advantages that they present. In particular, they are very attractive to designers and engineers alike due to their simplicity and ease of disassembly. However, hybrid connections are particularly susceptible to metal fatigue, stress relaxation primarily due to viscoelastic creep of the composite, thermal effects due to coefficient of thermal expansion mismatch, galvanic corrosion between the dissimilar constituents of the joint and moisture absorption causing differential strain between the metal and composite.
The study presented in this thesis focuses in an investigation of the effects of temperature and primary creep in hybrid metal to composite bolted connections. The study’s relevance stems from the desire to apply this technology to naval applications, where watertight integrity must be maintained. It was then decided to examine this type of connection at the subcomponent level. Therefore, EGlass/vinyl ester plates ½” thick were bolted to aluminum and steel plates of the same thickness with instrumented steel bolts to determine the primary stress relaxation response. Special attention was placed on the effects of temperature change on the stress relaxation that hybrid connections are particularly susceptible from. A model was developed with the sole purpose of integrating the existing coefficient of thermal expansion mismatch between all the joint parts in the scheme of analysis. Experiments were carried out to obtain the CTE of the composite material used in the hybrid connection tests, and a computer program, GASmooth, was specifically written to correct the thermal effects on the stress relaxation data.
The hybrid connection test ran for a period of at least 3 months. Two test phases are distinguishable. The first phase was performed over a period of 25 days (600 hours). Test articles were left to stress relax at room temperature, undisturbed. In the second phase, the test specimens, were temperature cycled 5 times, every 7 days, from room temperature to 62.5°C ± 1.5°C (144.5°F ± 3°F) in a computer controlled autoclave to observe the effects of larger temperature cycles on the test samples.
The test results presented in this work clearly confirm the dependency of single bolt, hybrid metal to composite connections to temperature changes. They also confirm a moderate dependency on environmental relative humidity fluctuations. Temperature and relative humidity variability have a greater effect in those connections bolting the composite to a metal with a relatively low CTE and Modulus of Elasticity. The analysis tools developed in the effort presented herein corrected the stress relaxation data for temperature shifts. Three different power law expressions were used to analyze the bolt load with respect to time in an attempt to seek alternate and simpler means to fit the data. Second phase test results confirm the importance of post curing the composite material above its service temperature. The first temperature cycle carried with it a dramatic loss of bolt load. Subsequent temperature cycles did not further alter the response of the test articles. Relative humidity changes were seen to affect.