Centrifugal Fan

Author: Joshua Phonpadith, Austin Dowell

WSU ME406, Pullman, WA
5/17/2019
Summary
The system we were given to analyze was a squirrel cage centrifugal fan and the overall
goal was to characterize the performance of the fan by varying the rotations per minute (RPM)
of the motor giving power to the fan and the damper position, which varies the maximum
volumetric air flow. We tested a total of 5 different damper positions, 5 different RPM values for
each damper position, which were based off the frequency of the motor set using the system
control panel, and measured the pressures, temperatures, and power of the fan and motor for
each trial. The pressure and temperature data allowed us to calculate the volumetric air flow
and the power data helped us calculate the fan efficiency. Using all of these variables, we were
able to characterize the performance of the fan by plotting the volumetric flow rate on the x-axis,
and pressure rise, efficiency, and power on the y-axis. We were also able to plot several rpm
curves relating the fan efficiency to the volumetric flow rate.

Introduction
A centrifugal fan is a mechanical device that intakes air and then redirects it 90 degrees
as outgoing air where the volume and speed is increased by the kinetic energy of the rotating
impellers. An easy visualization of the air flow can be seen in Fig. 1 below, where the arrows are
used to represent the airflow direction. A motor is connected to the fan to supply power at
various RPMs, which in turn changes how fast the outgoing air velocity is. A picture of the
centrifugal fan we worked with is seen in Fig. 2 below. They are typically used in HVAC systems
where they can easily transport gaseous materials through building ventilation systems or in air
pollution control systems where they perform dust control.

Fig. 1 A model of the airflow in a centrifugal fan.

Fig. 2 A picture of the centrifugal fan and motor used to run experiments in order to characterize
the performance of the fan.
System
In our centrifugal fan system, the motor has a load cell attached on one end, a power
cell on the other, and a magnetic pickup attached along the motor shaft to measure force,
power, and RPMs respectively. A tachometer is located on a nearby wall to show the RPMs of
the motor as a visual check, which is proportional to the frequency set in the control system.
There is a pressure transducer located right after the fan to measure the static pressure of the
air and two pressure transducers for the pitot tube to measure the differential air pressure used
to calculate the velocity of the air. Thermocouples are placed both in the fan inlet and within the
wind tunnel to measure the air temperature, which can be used to calculate the air density.
Finally, at the far end of the system opposite the motor is a damper that can be opened and
closed to vary the flow rate of the air and is controlled by a turning valve. A picture of the
damper system is included below in Fig. 3.

Fig. 3 A picture of the damper completely opened to allow maximum airflow.

The motor can be set to various speeds, controlled by a nearby control system that can
be set from 0 Hz to 60 Hz and corresponds to the motor’s RPM. All of the sensors in the system
were connected to a Agilent 34972A Data Acquisition System and transferred to Excel for
further calculations after all of the necessary data was taken to characterize the system. A
schematic diagram showing various sensors of the centrifugal fan system is shown in Fig. 4 and
the real life system is pictured in Fig 5.
 Fig. 4 A schematic diagram of the centrifugal fan that includes the motor, damper position,
thermocouples, and pressure transducers.

Fig. 5 A panoramic picture of the actual centrifugal system we characterized.

Objectives
Our main objective is to characterize the centrifugal fan by studying the performance of
the fan and creating efficiency curves at various fan speeds. We can accomplish this by
identifying performance curves relating to system pressure, power consumption, and fan
efficiency to volumetric flow rate of air. Analysing the these performance curves will give us a
better understanding of how to improve the system overall through the given relationships from
our data/plots. That is, we can increase efficiency of the system by changing other variables. It
is important to study and characterize the fan’s airflow because if airflow conditions are
incorrect, costs are elevated and equipment life expectancy is shortened.

Approach
The input variable to our system is the input frequency of the motor and the damper
position, while the outputs are the average air speed and the pressure and temperature within
the circular pipe. We began our experiments by setting the damper to a specific opening length,
which has to be greater than 0 to allow airflow, and then varying the input frequency of the
motor from 20 Hz to 60 Hz in 10 Hz intervals. The resulting air speed was measured using the
log-tchebycheff method. Since the air velocity in the pipe is not uniform, the log-tchebycheff
method works by dividing the duct into concentric circles where each has the exact same area
and an equal number of readings is taken from each circular area. The diameter of our duct is
16 inches so we took 6 measurements at depth ratios described in Fig. 3 below, rotated the pitot
tube by 90 degrees, and then took those same 6 measurements again to get a total of 12
measurements for each RPM setting.

Fig. 6 A schematic of how we plan to use the log-tchebycheff method for our system. We will
make 8 measurements along one diameter, rotate 90 degrees, then make another 8
measurements at the same depths as before.

Since we tested 6 different RPMs, we took a total of 72 measurements for each damper
position for a total of 5 different damper positions. The velocity of the internal air was calculated
for each frequency by using Eq. (1) for each depth measurement and then averaging all 12
velocities to get a single average velocity for the set frequency.

(1)
Where velocity can be simplified to
V = √2𝑑𝑒𝑙𝑡𝑎𝑃/𝑑𝑒𝑛𝑠𝑖𝑡𝑦 (2)
The density of the air was calculated using Table A-15 and interpolation using Eq. (3) for
each point. Delta P is directly measured from the pressure transducer after the fan and
converted to pressure using a calibration curve and unit conversion to get Pascals.After solving
for velocity, the air flow rate (Q) can then be easily determined by multiplying the air velocity by
the cross sectional area of the pipe, as seen in Eq. (4).
Interpolation equation (3)
Q = VA (4)
The power and efficiency of the fan can then be determined using Eq. (5), where the
differential pressure is measured using a pressure transducer and converted to in H20 using a
calibration curve. Finally, we will graphically characterize the centrifugal fan performance by
graphing flow rate on the x-axis and 3 separate y-axis for pressure, efficiency, and power.
Efficiency equation (5)

Table 1. A table describing how we organized our data collection process. A damper position is
set and for each motor input frequency, 12 pitot tube depth measurements were taken to find
the average velocity at that specific damper position and frequency.
 Damper Position Motor Input Frequency (Hz) Depth Measurements (in)
(2x)

Damper position 1 10, 20, 30, 40, 50, 60 .512, 2.16, 5.136, 10.864,
13.84, 15.488

Damper position 2 10, 20, 30, 40, 50, 60 .512, 2.16, 5.136, 10.864,
13.84, 15.488

Damper position 3 10, 20, 30, 40, 50, 60 .512, 2.16, 5.136, 10.864,
13.84, 15.488

Damper position 4 10, 20, 30, 40, 50, 60 .512, 2.16, 5.136, 10.864,
13.84, 15.488

Damper position 5 10, 20, 30, 40, 50, 60 .512, 2.16, 5.136, 10.864,
13.84, 15.488

Results

Fig. 7 Centrifugal Fan motor performance at 60 Hz.

Figure 7 represents our power, pressure, and fan efficiency curves for a frequency of 60
Hz. Considering this, we can see the relationship each has with volumetric flow rate. Power
steadily increases as flow rate increases, pressure decreases as flow rate increases, and
efficiency increases then decreases after it hits a peak around 0.95 m^3/s flow rate. All of this
data is consistent with our predictions and match up with the references given in Cengel’s 4th
ed fluids textbook. So, we know our data is accurate and illustrates direct relationships between
volumetric flow rate and power, pressure, and efficiency.
 Fig. 8 Varying Frequencies with Fan Efficiency and Flow Rates.

Each curve above represents a negative parabola for each frequency from 20 Hz to 60
Hz in a 10 Hz interval. As the frequency increases, the higher the volumetric flow rate reaches
and the higher the efficiency can become. Also, each legend has its own peak where the
efficiency increases then follows a decreasing trend after the peak. We can conclude that for
each frequency there is an ideal point of volumetric flow rate that we can reach which will give
the highest fan efficiency for the system. However, if we has too high of a volumetric flow rate
then we will pass this point and end up with a lower frequency than the ideal.
 Fig. 9 Varying Frequencies with Power and Flow Rates.

Figure 9 shows us the connection between power input of the system, volumetric flow
rate, and frequency change. For each trendline, as flow rate increases then so does the power
on a slow increasing trend. Also, as we increases frequency then we can observe the power
increasing, which makes logical sense in itself. The purpose of this graph is to illustrate the
analogous affects volumetric flow rate and frequency has on power. So, in this case we can
clearly see the increasing characteristics our trendlines cary when evaluating left to right.
 Fig. 10 Varying Frequencies with Static Pressure and Flow Rates.

Figure 10 above is a good depiction of the affects mass flow rate has on static pressure
after the inlet fan for different frequencies. One trend we see right away is the slow decline each
curve has, so as the volumetric flow rate increases then the static pressure drops. This is
reasonable considering bernoulli’s equation dictates that the pressure after the inlet fan would
decrease as velocity increases. FInally, the higher the frequency then the higher the static
pressure.

Discussion
The very first issue we encountered with our system was related to the accuracy of the
calculated pitot tube depth measurements. As an example, the .032 * diameter calculation gave
us a depth measurement at .032 * 16 in = .512 inches. We were unable to get this accurate of a
position because the depth of the pitot tube was read visually using a ruler with ⅛ inch
increments as seen in Fig. 5 below. This only introduced a small amount of variation between
different frequency settings and should not impact the overall data acquired.

Fig. 11 The pitot tube adjuster that allows us set to set how far the pitot tube is in the pipe.

Conclusion
The purpose of this experiment was to characterize the performance of a centrifugal fan
system. In order to do so we needed to understand the relationship between fan efficiency,
volumetric flow rate, power, and static pressure. We were able to accomplish this by taking
several different data points, while changing frequency and the damping position/flow rates, and
plotting them into excel to illustrate the affects flow rate had on each variable. The most
important variable we can use to characterize performance is the efficiency of the fan. Going
back to figure 8, we can see that efficiency increases as frequency increases. However, the
trendlines are negative parabolas that peak at a certain motor frequency, which is equivalent to
a specific RPM. To conclude, if we can identify the right volumetric flow rate for the highest
frequency (6
0 Hz in this case), then we can keep the system at optimal performance with the highest
efficiency.

References
[1] Cengel, Y. A. & Ghajar A. J., 2015, “Heat and Mass Transfer Fundamentals and
Applications 5ed,” pp. 926.
[2] Airflow Instruments,2014, “Traveling a Duct to Determine Average Air Velocity or Volume
Application Note AF-109,” tsi, pp. 1-4
[3] dam-assets.fluke, 2007, “ How to make a duct traversal airflow measurement,dam-assets-
fluke,” Fluke, pp. 1-3

Appendices

Apparatus Dimensions
Diameter of Tube: 0.4064 m = 16 in
Diameter of Pitot Tube: 0.0080518 m = .317 in

Sample Data

Fig. 12 Sample data collected from channels 101-110 for various motor frequencies and pitot
tube depth measurements
 Fig. 13 Calculated values based off of the data collected from Fig. 9.

Sample Calculations
Sample calculations
Frequency: 60 Hz
Damper: Position #2

Density:
Internal Thermocouple: 22.804 °C
Interpolation: (((1.184-1.204)/(25-20))*(22.804-20))+1.204 = 1.192784 kg/m 3

Pitot Tube differential pressure (V->pa):

Pitot Tube differential pressure voltage: 0.89350607 V
Pitot Tube pressure Calibration: ((0.22580142*0.89350607)-0.1293651)/0.0040146 =
18.03164434 pa

Velocity:
V60Hz = (2*(18.03164434 pa)/(1.192784 kg/m 3))^(1/2) = 5.498595367 m/s

Motor Power(V->W):
Motor Power Cell: 2.2455205 V
Power Calibration: (713.75025*2.2455205)+8.5484 = 1611.289218 W

Motor Power(W->Hp):
Motor Power Watts: 1611.289218 W
Motor Power Hp: (1611.289218 W)*(1Hp/745.7W) = 2.160774062 Hp
Static Pressure after fan(V->pa):
Static Pressure voltage: 4.3877361 V
Static Pressure Calibration: (((1.991502*4.3877361)-5.086312)/0.0040146) = 909.6480891 pa

Average Q (volumetric flow rate m^3/s):

Q = V*A
Average Velocity: 5.332432656 m/s
Pipe Area: A = pi*(0.2032)^2 = 0.1297171146 m^2
Average Q = V*A = (5.332432656 m/s)*(0.1297171146 m^2) = 0.6917077782 m^3/s

Fan Efficiency:
E = (((Average Static Pressure after fan)*(Average Q))/Average Power)*100%
Average Static Pressure after fan = 920.3318753 pa
Average Q = 0.6917077782 m^3/s
Average Power = 1609.082969 W
Fan Efficiency (E) = ((920.3318753 pa*0.6917077782 m^3/s)/1609.082969 W)*100% = 39.56%
efficiency

Assumptions
Atmospheric pressure is constant on the fan intake
Pitot tube diameter = 0.317 inch or 0.0080518 m
Channel 101 reads the differential pressure (pressure total - static pressure) of the pitot tube
Don’t use channel #6 Pressure Transducer

Uncertainties
Device Uncertainty

Pressure Transducer ± 2.26 % FS

Thermocouple ±1℃

Agilent 34972A LXI Data Acquisition/Switch ± 1 % FS

Unit

Panel Mount Tachometer Model DT-5RL- ± 0.01 % FS

1,2,3*