Thermoacoustic refrigerators: Time-domain modelling and experimental setup

Licentiatavhandling, 2008

Thermoacoustic refrigerators use acoustic waves to transfer heat. Although their benefits compared with their vapour-compression counterparts are environmental friendliness, robustness and simplicity, they are currently still less efficient than traditional vapour-compression refrigerators. A thermoacoustic refrigerator consists of four parts: a resonance tube, a loudspeaker, two heat exchangers and a stack of plates. Attempts to increase the efficiency have mainly focused on optimising the different parts separately. In order to take the whole thermoacoustic refrigerator into account, there is a need for numerical models and practical experience. The primary aims of this thesis have been: (1) to develop a time-domain model for thermoacoustic refrigerators; and (2) to design and construct an experimental setup of a thermoacoustic refrigerator. In contrast to the established frequency domain models, time-domain modelling was chosen to allow a time varying acoustic field. Detailed time-domain methods are in general computationally expensive. In order to reduce computational cost, the detailed Finite-Difference Time-Domain (FDTD) method was combined with the quicker Equivalent Source Method (ESM). The complex acoustic field in the stack of plates was modelled using the FDTD method. The acoustic field outside the stack was modelled using the ESM. The experimental setup was designed according to existing design strategies, where each part is optimised separately. Limited time and budget constrained the choice of materials and components. The combination of the methods (ESM+FDTD) is shown to be numerically stable and accurate for 1-D energy flow, including viscous attenuation. Numerical errors in the coupling between the two methods caused by dispersion in the FDTD were insignificant in the frequency range of interest. Essential thermoacoustic effects as thermal conductivity, temperature and heat transfer have not yet been included in the model. The experimental setup was capable of creating a significant temperature difference across the stack. However, the heat pumping was insufficient to obtain a measurable cooling power. The time-domain model and the experimental setup provide useful tools to further develop and improve thermoacoustic refrigerators with time-varying boundary conditions.