Contact:
Christopher Wilson
Tennessee Technological Univ.
Mechanical Engineering Dept.
Box 5077
Cookeville, TN 38505
Tel: (931) 372-3216
chriswilson@tntech.edu

There are distinct differences between a composite structure as typically modeled and the composite structure as it was actually built. Structural analysts often idealize the composite structure making simplifying assumptions that lead to reasonable solutions. A common design assumption (for continuous fiber laminated composites) is that classical lamination theory can be applied to determine the strain and stress states. Further, commonly used failure theories are based on the concept of no initial defects or imperfections in the composite. Finite element analysis, used for more geometrically complex composites components, usually entails similar assumptions. Differences between the as-manufactured and the as-analyzed composite structure are not generally considered until unexpected and unwelcome mechanical behavior is observed. In the worst case, there is no real correlation between actual service life and predicted service life failure. Failure occurs well before expected.

Tennessee Technological University has initiated an effort structured towards "Understanding the Service Life of Composites" which will result in the development of analytical and numerical models of representative as-manufactured composite components or sub-components using continuum mechanics and the finite element method. These models will be used in service life predictions to determine the service life with the consideration of initial material quality.

Two major outcomes of a multi-phased approach will be first, an enhanced method of incorporating initial material quality and stress state in the service life of a composite, namely the enhanced Critical Element Model. The second outcome will be a set of experimental results ranging from quantitative density measurements to fatigue characterization suitable for a benchmark of current and future service life prediction schemes.