A comparison of continuum and discrete modeling techniques of the effects of manufacturing defects common to composite structures
Nelson, Jared William
MetadataShow full item record
Application of different damage modeling approaches for use with composite materials and composite material structures has grown with increasing computational ability. However, assumptions are often made for "worst case" scenarios with these modeling techniques resulting in approximating defects as a hole or notch in a plate instead of modeling actual flaw geometry. These analytical tools have helped bound composite material and structure capabilities, but do not allow for comprehensive understanding of the effects of defects as the characteristic parameters of the defects vary. In order to develop a tool that will allow for accurate analysis of a complete structure, including defects of different parameters, modeling approaches must be optimized. It was the optimization of these approaches that was investigated herein with specific application toward establishing a protocol to understand and quantify the effects of defects in composite wind turbine blades. A systematic, three-round study of increasing complexity was performed to understand the effects of three typical blade manufacturing defects while investigating continuum, discrete, and combined damage modeling. Through the three rounds of the benchmark material testing, significant coupon level testing was performed to generalize the effects of these defects. In addition, material properties and responses were analyzed and then utilized as material inputs and correlation criteria for each analytical technique. Parallel to the material testing, each of the three rounds increased in analytical complexity to ensure that models were only as complex as necessary to achieve acceptable correlation. Correlation was compared both qualitatively and quantitatively for an initial case and other cases were investigated only if initial correlation was acceptable. While each modeling type offered certain attributes, a combined approach yielded the most accurate analytical/experimental correlation. Thus, a unique comparison of several different analytical approaches to composites with respect to manufacturing for consistency, accuracy, and predictive capability allowing for improved blade reliability and composite structural assessment.
This dissertation is accompanied by an errata sheet. Please consult both.