Investigation of Automotive Cooling Fan Performance

This work was conducted in collaboration with DaimlerChrysler Canada through an NSERC CRD Grant held with Dr. R. Barron. The project focussed on the development of a new methodology for simulating fan aerodynamic performance. Both experimental and computational investigations were conducted and the experimental data used to validate both a full CFD fan simulation used to generate the fan blade data needed for the new methodology.

The experimental investigations were the responsibility of my research group and hence are mentioned now in more detail. A M. A. Sc. student by the name of Philip Nourse obtained his thesis based on the design, construction and testing of a 3m x 3m x 3m fan test facility which contained a fan and shroud combination located in the ceiling. The fan was attached to a vertical rotating shaft, which was driven by a variable speed DC motor arrangement. An X-array hot-wire probe was used to determine the three-components of velocity. The “ram air” effect was simulated by drawing additional air through the fan using a blower on the exhaust from the test chamber. An orifice plate was used to determine the overall flow rate through the facility and the entire operation automated using a microcomputer.

A second M.A.Sc. student by the name of Michael Brown continued the work by modifying the facility to allow measurement of the three-components of the upstream velocity field using a Laser Doppler Anemometer.

This project resulted in the establishment of an excellent facility for use in further fan studies. The expertise gained also led to the establishment of research collaboration with Dr. S. Kouidri, of Ecole Nationale Supérieure d’Arts et Métiers (ENSAM), Paris, France in the area of axial fan aerodynamics and noise generation.

Investigation of Axial Cooling Fan Noise

Two students from ENSAM have worked in this area during 5 month appointments at the University of Windsor as Visiting Research Students.

The first study involved the development and implementation of an engineering model for predicting the spectrum of noise emanating from an axial flow fan. An existing semi-empirical method was modified for this purpose. It is based on the assumption that the radiated acoustic pressure is primarily due to the fluctuating pressures exerted on the fluid medium by the rotating fan blades. Simplified techniques were used to determine the blade surface force distribution for input to the model. In this initial work, no effort was made to account for the effects of the shroud, radiator, condenser or engine compartment shape. Characteristic dimensions and shape parameters as well as operating conditions such as rotational speed and volumetric flow rate were inputs to the model. The model was compared with narrow band measurements of the radiated noise from an axial flow fan.

The second investigation involved a computational fluid dynamic (CFD) study of the flow field and noise generated by the cooling fan of an electronic computer board. The flow measurements were in a reasonable agreement with the predictions. Due to the extreme limitations of time and the oversimplification of the noise prediction models used, however, comparisons of the noise predictions with measurements were not in good agreement.

The fan noise work is being conducted in collaboration with Dr. Smaine Kouidri of ENSAM in Paris, France.