Investigating the Reinforcement Architecture Dependency of Failure for Composites
Heim, Frederick, Mechanical and Aerospace Engineering - School of Engineering and Applied Science, University of Virginia
Li, Xiaodong, EN-Mech/Aero Engr Dept, University of Virginia
Ozbulut, Osman, EN-Eng Sys and Environment, University of Virginia
Green, David, EN-Mat Sci/Engr Dept, University of Virginia
Xu, Baoxing, EN-Mech/Aero Engr Dept, University of Virginia
Opila, Elizabeth, EN-Mat Sci/Engr Dept, University of Virginia
The properties of composites are tunable via the constituents, but failure is often related to nonuniformity of the constituents, including irregular size, shape, density or orientation. Textile composites featuring continuous fiber reinforcements, bundled into tows, are set to replace many metallic structural components to extend performance and safety in extreme material applications. The intricate manufacturing process of braided and woven composites produces inconsistency in both the microstructure and mechanical properties, particularly, along the length of long components. Some of this variability is attributed to nonuniform tow placement, producing systematic and stochastic distortions in the tow trajectories, causing unit cells of irregular shape and size, and ultimately, influencing stress redistribution behavior. Much information about these nonperiodic defects is unknown, including how they are spatially distributed at the macroscale, how the specific braid/weave architecture or local geometry influences their development, and how they dictate mechanical failure. To address this need, stereoscopic digital image correlation was utilized in an unconventional manner to develop a multitude of scalable systems to quantify nonuniform reinforcement distributions and investigate the reinforcement architecture dependency of failure for textile composites.
In this dissertation, a brief literature review of fundamental background information on the targeted ceramic and polymer matrix composite systems and stereoscopic digital image correlation is given. In the second chapter, an overview is given of preceding work utilizing customized reinforcement phase manufacturing to improve mechanical performance in aluminum matrix composites, which served as the inspiration for the pursuit of the work completed for textile composites. In the third chapter, a scalable system to triangulate the surface of complexly shaped materials or objects is developed and evaluated using a series of fundamental and computational experiments. In the fourth chapter, the scalable system is applied separately to both carbon fiber and silicon carbide fiber braided composite systems to reveal the first measurements of long-range tow spacing variability and through a comparison of two braids, the spatial influence of each tow orientation family. In the fifth chapter, targeted mechanical experiments and numerical simulations are performed on carbon fiber braided composites to quantify the influence of tow structure irregularity on local stress/strain concentration leading to failure. In the sixth chapter, novel stereoscopic digital image correlation techniques and tow location segmentation schemes developed in previous chapters are applied to braided and woven silicon carbide fiber / silicon carbide matrix composites to assess the influence of underlying tow structure on hermetic seal and cracking behavior during mechanical loading. In the final chapter, a summary is given of the completed research as well as several recommendations for future pursuits and applications to additional material systems. The investigation detailed herein is expected to reveal new fundamental information on tow trajectory deviations in textile composites and their critical implications on mechanical performance, which will serve to guide manufacturing improvements of defect sensitive composite materials.
PHD (Doctor of Philosophy)
Digital Image Correlation, Braided Composites, SiC/SiC
U.S. National Science Foundation (CMMI- 1537021)Commonwealth Center for Advanced Manufacturing (CCAM D-073)U.S. Department of Energy’s Nuclear Energy University Program (Project 17-13080)Westinghouse Electric Company LLC (PO 4500695139),