| dc.description |
Product design is often driven by customer needs and desires. In this case the customer is the US military, and the desire is to have a cooling unit small enough and light enough to incorporate into hazard suits for use in desert combat. The Micro technology-based Energy and Chemical Systems (MECS) being developed at Oregon State University (OSU) and other institutions can generate an extraordinary rate of heat and mass transfer capabilities. As this MECS technology is progressing into the application stages, simulation-based design optimization models will provide invaluable information; saving time and money by guiding the direction of prototype creation and validation. This project studies a 2 kW cooling load based on an absorption cycle ammonia-water cooling system. The absorption cycle was chosen because it requires much less work input in the compression phase than the standard compression cycle systems, making it more suitable for portable applications.
Using basic principles of robust design methodology to reduce the sensitivity of the system performance to changes in input conditions, a more robust system is implemented throughout the research. Specifically for the cooling system that means that varying ambient conditions and thermal configurations will have less of an impact on overall system performance and system weight. The thermodynamic system modeling is done using Engineering Equation Solver. A tradeoff study is conducted to determine an appropriate design space, and then a D-Optimal fractional factorial design is selected using Matlab to define a manageable sized data sample within the design space with the eight design variables with five levels each. Based on the EES-based thermodynamics model and sizing results, S-PLUS is the statistical analysis program used to develop the surrogate models of the system performance and weight. Microsoft Excel’s “solver” function was used to optimize and do a sensitivity analysis of the objective function that was created. The results of this process are a range of potential optimal configurations for the system that can be evaluated and selected by a user depending on conditions and the importance of certain factors. The optimization process generated optimal values for the thermal properties of each component based on a range of starting points. Each of these sets of optimal points had a variance of less than 20% when the input parameters were varied in a range of 10%. The resulting data supplies potential users with a good range of reasonable configurations for a 2 kW system that operate within acceptable parameters. |
|