Optimization of the Stack Unit in a Thermoacoustic Refrigerator

Abstract

A thermoacoustic refrigerator is a device which uses acoustic power to pump heat in the absence of harmful refrigerants with no or few moving parts. However, the performance of the thermoacoustic refrigerator, particularly the standing wave types, is currently not competitive compared to its counterpart conventional vapor-compression refrigerator. Presently, thermoacoustic refrigeration prototypes only achieved 0.1-0.2 relative coefficient of performance compared with that of 0.33-0.5 for the conventional vapor-compression refrigerators. Past optimization efforts had been completed based on parametric studies where individual parameters are discretely varied and the final optimized outcome was based on the Downloaded by [University of Lethbridge] at 22:37 25 June 2016 ACCEPTED MANUSCRIPT 2 limited series of numerical/experimental tests. This paper discussed the initial investigation of the optimization of the thermoacoustic refrigerator stack parameters using multi-objective genetic algorithm. The desired output, the maximization of the cooling load and the minimization of the acoustic power at the stack, are obtained with the parameters to be optimized set within some range of values. The stack length and centre position are then optimized simultaneously. The optimized results showed that, the coefficient of performance of the thermoacoustic refrigerator improves from the published value of 1.3 to 1.37.

Optimization of standing-wave thermoacoustic refrigerator stack using genetic algorithm

“…After 6188s and 6802s, we obtain the output of four-and five-variable models, respectively. These values are close to those obtained by Zolpakar et al [23] , where the best normalized cooling power was predicted to be 1.6e-6, when the stack length = 0.24 and the stack position = 0.22 under the condition of a fixed stack spacing = 3 and a porosity Br = 0.75. Between our results and those from Zolpakar et al [23] , the differences in the stack length and the stack position are about 3% and 15%, respectively.…”

“…These values are close to those obtained by Zolpakar et al [23] , where the best normalized cooling power was predicted to be 1.6e-6, when the stack length = 0.24 and the stack position = 0.22 under the condition of a fixed stack spacing = 3 and a porosity Br = 0.75. Between our results and those from Zolpakar et al [23] , the differences in the stack length and the stack position are about 3% and 15%, respectively. The short-stack approximation was used by Zolpakar et al [23] , which is suspected to be the cause of the difference.…”

“…As mentioned earlier in Section 1, several artificial intelligence algorithms were able to realize global optimization, with only the evaluation of objective function necessary. Among these algorithms, the genetic algorithm has been successfully utilized in the area of thermoacoustic optimization [5], [22], [23] , which shows its attractive capability.…”

“…The small displacement amplitude as a result of a high operating frequency also increases the difficulty in installing efficient heat exchangers. An operating frequency ( ) of 400 Hz was chosen in the experimental investigation and optimization for standing-wave refrigerators [14], [18], [23]. Wetzel and Herman adopted the operating frequency of 325 Hz for the TALSR under examination [13] .…”

“…The outcome, therefore, could often be a local minimum/maximum. In order to find global optimum, various intelligent algorithms, including genetic algorithm [23] , particle swarm method [31] , machine learning [32] and teachinglearning-based optimization algorithm [30] have been applied to optimize the performance of the stack. Among all these attempts, the objective functions are all evaluated using the shortstack boundary layer approximation, which can be overly simple for practical design and can introduce errors to some degree to the optimum parameters.…”