DESIGN OF Si/SiC HYBRID STRUCTURES FOR ELEVATED TEMPERATURE MICRO-TURBOMACHINERY by

Silicon, the material of choice of the first (demonstration) microengine, exhibits strong thermal softening behavior at temperatures above 900 K. This thermal softening behavior limits the turbine inlet temperature, which in turn significantly degrades the overall engine efficiency. Previous studies have shown that hybrid structures of silicon and silicon carbide have good potential for improved engine performance. Detailed design of Si/SiC hybrid structures for high temperature micro-turbomachinery, however, has been hampered by the relatively poor performance of single crystal Si at elevated temperatures and high stresses and by the unavailability of accurate material properties data for both Si and SiC at the temperatures of interest. From previous work, the critical structures and materials issues to be resolved, in order to proceed with the design of high temperature Si/SiC hybrid structures, were identified as follows: 1. the safety margin of the Si/SiC hybrid structures based on the upper yield strength of Si 2. reliable estimation of the service life of the Si/SiC hybrid structures 3. structural instabilities caused by the combination of stress concentrations and strain softening. In the course of this thesis, these issues provided the key motivations of the work, and have been substantially resolved. As a first step, it is critical to obtain a better understanding of the mechanical behavior of this material at elevated temperatures in order to properly exploit its capabilities as a structural material. Creep tests in simple compression with n-type single crystal silicon, with low initial dislocation density, were conducted over a temperature range of 900 K to 1200 K and a stress range of 10 MPa to 120 MPa. The compression specimens were machined such that the multi-slip <100> or <111> orientations were coincident with the compression axis. The creep tests reveal that the response can be delineated into two broad regimes: (a) in the first regime rapid dislocation multiplication is responsible for accelerating creep rates, and (b) in the second regime an increasing resistance to dislocation motion is responsible for the decelerating creep rates, as is typically observed for creep in metals. An isotropic elasto-viscoplastic constitutive model that accounts for these two mechanisms has been developed in support of the design of the high temperature turbine structure of the MIT microengine. From the experimental observations and model validation, basic guidelines for the design of Si/SiC hot structures have been provided. First, the use of the upper yield strength of single crystal Si for design purpose is non-conservative. Also from the perspective of the design of Si hot structures, the lower yield strength is insufficient, particularly for micro-turbomachinery operating at elevated temperatures and high stresses. The recommended approach to the design of Si hot structures is to use the Si model for extracting appropriate operating conditions, and to reinforce the Si structures with SiC in strategic locations. Second, at high temperatures, the effect of stress concentrations is not crucial. Unlike the low temperature Si structures, the plasticity present adjacent to the sharp corners reduces the effect of stress concentrations. Third, the Si/SiC hybrid structures concept was verified. The considerable increase in the load carrying capability of the Si/SiC hybrid specimens encourages the development of Si/SiC hybrid structures for elevated temperature micro-turbomachinery in order to extend the available design space. Finally, the FE results for the creep life estimation of the Si/SiC hybrid turbine rotor identified the limit of the all-Si turbine rotor of the current microengine as well as the superior performance of the Si/SiC hybrid rotor in terms of creep life. Thesis Supervisor: S. Mark Spearing Title: Associate Professor, Department of Aeronautics and Astronautics