A thermoacoustic refrigerator was designed using a dimensionless parameter approach. Some basic insight into thermoacoustic design principles was obtained. The resulting device was used as a test bed for three different control schemes. The first was a phase-locked loop, which is the control method most often used in the literature; the second controller utilized a gradient ascent algorithm to track the operating frequency of maximum acoustic pressure; and the third utilized the same gradient ascent architecture to track the operating frequency corresponding to maximum acoustic power transfer to the resonator. The three controllers, tracking different parameters associated with a strong thermoacoustic effect, were compared in simulations and experiments. Difficulties in collecting data for the power controller resulted in unreliable data. Therefore, the power controller was not compared quantitatively with the other two. The PLL performed best in terms of thermoacoustic efficiency, but the acoustic pressure controller was able to produce more cooling power and converted electrical power to cooling power more efficiently due to the amplitude of the input voltage to the driver being held constant. The major short-coming of the gradient ascent approach was the relatively long convergence time. However, convergence time is not always relevant to refrigerator operation. The maximum acoustic pressure control scheme was determined to be the best controller considered because it has fewer sensors than the other two controllers, involves less computational effort than the power controller, and yielded better electrothermal performance than the PLL.
展开▼
