1996
SAE International Congress & Exposition
Feb. 26-29, 1996,
Detroit, MI
DESIGN
OF AN AUTOMOTIVE HVAC BLOWER WHEEL FOR FLOW, NOISE AND STRUCTURAL INTEGRITY
C. Toksoy, M. Zhivov, F. Cutler, & A. Ecer, Technalysis
M. Rayhill, S. Guzy, & R. Vasko, Harrison
Division General Motors Corporation
Introduction
Design of a centrifugal
blower wheel for automotive air-conditioning is a complicated process for
several reasons. Although it is a low cost product, the design of a
squirrel-cage type blower wheel, shown in Figure 1, requires a certain level
of precision. Wheels with slightly modified geometries may perform quite
differently in terms of flow efficiency and noise. Usually, the HVAC system
is not designed around a blower wheel; rather the wheel is fitted to a given
system. Centrifugal blowers with forward-curved blades are not
aerodynamically designed to be the most efficient. Instead, they are chosen
for their ability to deliver relatively high flow rates for rather
restrictive packaging requirements [1]. For example, in comparison to a
mixed flow fan, the resulting flow field is more three-dimensional and
complex; resulting in many separated flow regions and noise generation
mechanisms.
The objective of the blower
wheel design is to provide the best wheel to suit these rather non-ideal
conditions. Depending on the specific situation, the design requirements may
vary depending on requirements for air flow performance, noise, structural
integrity and cost. Another important factor is the expectation in terms of
time required to design a new wheel. One has to perform the engineering
tasks described here in a relatively short time, even though the problem to
be solved is rather complex.
In the present application,
we consider a specific design problem. We start with a reference wheel which
provides sufficient flow but it is relatively noisy. The problem is defined
as reducing the noise level while producing at least the same flow
performance. It is also desirable to reduce the cost of the design by using
a less expensive material. The packaging is fixed. While the blower wheel
and the housing have to be ideally considered as a single unit, the housing
design is not to be changed and the new wheel should match the existing
housing. The problem described above is a typical design problem for HVAC
blower wheels. As mentioned above, packaging requirements usually results in
such restricted design conditions.
In the present paper, we
describe a design strategy, which involves the implementation of a
computational procedure complemented with an experimental testing program.
It is demonstrated that by using three-dimensional CFD tools , one can
design blower wheels which will provide the desired objective. The use of
such tools reduce the cost and time required for testing to reach the
desired objective. At the same time, it helps to improve the basic
understanding of many complex situations which may occur inside the rotating
blower wheel. Many times CFD tools are used to analyze a design which is not
the desired one. The strategy here is to use CFD to eliminate rather than
analyze the bad designs and come up with a design to meet the
specifications.
Objectives
The objectives of the
present development were to improve an existing a/c blower wheel along the
following directions:
- Reduce the flow
induced noise generated by the blower,
- Retain and
possibly improve the flow performance,
- Develop a blade
shape which lend itself to a less costly pull-up mold design rather than
the existing split louver design,
- Replace a higher
strength but expensive material ($2.50/lb.) with a less expensive
($0.70/lb.) material.
More specifically, the design
process involved making changes to the blade profile while controlling the
flow performance. It also involved increasing the blade thickness for
improving strength and allowing the less expensive material to be utilized,
while at the same time providing noise reduction which is also sensitive to
such changes.
Design Criteria
The design criteria of a
blower wheel for improved flow performance and noise characteristics can be
summarized as follows:
It is desired to improve the
combination of the static pressure rise across the wheel and the pressure
rise along the housing. In the case of a blower wheel with forward curved
blades, a considerable portion of the pressure rise is developed inside the
housing which is tied to the tangential velocity (swirl) introduced by the
trailing edge of the blade. Thus, proper evaluation and design of the
trailing edges of the blades becomes a critical problem to match an existing
housing. We are interested in evaluating the deviation angles between the
blade trailing edge and the flow exiting the blade. This region is a source
of flow losses and noise as well as determining whether the housing matches
the wheel properly. Of course we are also interested in understanding the
matching between the inlet and the leading edges of the blade, as well as
the velocity distribution within the blade passage. We are concerned about
the pressure distribution along the blade and flow attachment along the
entire blade, flow circulation regions and the variations within the
three-dimensional flow field from the pressure side to the suction side.
Large variations in the circumferential direction will cause unsteadiness in
the wake of the wheel exit resulting in loss of performance and generation
of noise.
In this particular project,
we are trying to increase the blade thickness at areas of high stresses. As
we change the geometry, we are also concerned about improving flow
performance. In most cases, noise and efficiency are directly related.
However, when making design improvements, the priority was given to reduce
the noise rather than increasing the efficiency. It will be shown that the
final design is both more efficient, quiet and less expensive to manufacture
at the end of the process.