| Summary: | This thesis presents an investigation of the working principles of an acoustic resonance anemometer. To this end, the Navier-Stokes equations were used to generate an expression for the slowly changing, incompressible viscous flow field around and through an anemometer, and an expression for the rapidly changing, compressible inviscid convected acoustic field. The slowly changing field was simulated using a commercial computational fluid dynamics package called FLUENT. The rapidly changing, convected acoustic field was simulated using a finite element solver written for the project, capable of parallel execution using the 'Message Parsing Interface'. Measurements performed in the '18 inch' wind-tunnel at Imperial College, London, were used to validate the simulation of the slowly changing field and provide a means to select a representative turbulence model. Measurements performed in the FloTek wind-tunnel at FT Technologies, Teddington, using a working anemometer were used to validate results generated by the finite-element solver. The finite element solver was further validated against a commercial finite element solver for acoustics in the absence of convective flow, called PAFEC. Steady flow simulation results indicated the acoustic resonance anemometer generates a linear variation of acoustic phase with free stream velocity when operating at resonance: in agreement with measurements. The unsteady flow simulations showed fluctuations of the slowly changing unsteady flow through the duct. These were partly due to a separation region at the entrance of the acoustic duct; vortex shedding from the support pillars used to maintain the separation between the duct faces; and shedding from the anemometer body. The measurement errors caused by these fluctuations were quantified. The findings in this project show that the simulation approach adopted is valid and that the software developed could be used for further study of acoustic resonance anemometry.
|
|---|
