Dynamometer Drag

Chow, Garrick; Rose, Nicholas; Thorne, William

Group 24 Section 12605 October 31, 2019

Abstract—In this project, the drag forces of five different shapes object used in this investigation could be compared to other
at five different Reynolds numbers were measured using a two- existing drag data corresponding to objects of similar geometry.
component dynamometer. In order to achieve this, a calibration
was performed relating the voltage readout on a Linear Voltage As all the objects placed within the wind tunnel have a
Displacement Transducer (LVDT) to a force being exerted on the
circular cross-sectional area, Equation (3) was then used to
dynamometer. This drag force was then used to calculate the drag
coefficient for each of the objects being measured. Specifically,
determine that area A.
Hemisphere A, Hemisphere B, the disk, and the smooth sphere had .
coefficients of drag that were 1.57%, 2.11%, 5.12%, and 1.29% 𝐴 = .25𝜋𝑑 2 (3)
away from the from their published scientific data respectively.
However, the rough sphere had a bigger separation from the In order to measure the desired drag force for each of the
established data because the roughness of the object in this objects, a two-component dynamometer was used in
experiment is not quantifiable and will not be exactly the same as conjunction with a Linear Voltage Displacement Transducer
the roughness of a sphere in published data. (LVDT). In response to the air flow within the test section, the
cantilever beam of the dynamometer will deflect in varying
Index Terms— Differential Transformer, Drag Force, Reynolds
Number, Sphere, Two-Component Dynamometer amounts due to the drag force on the object attached to it. While
there would theoretically be a drag force on the beam itself, a
I. INTRODUCTION protective layer shaped like an airfoil (Fig. 1) surrounding it
nullifies these forces on the lever arm and allows the drag on
T HE primary objective of this experiment was to become
proficient in the use of a two-component dynamometer to
find the drag coefficients of different shapes at differing
each object to be almost entirely isolated.

Reynold’s numbers and compare them with preexisting

published data. In order to satisfy this objective, five objects of
varying shapes at five different Reynolds numbers were tested
within a wind tunnel.
A previous study was used to determine the dynamic pressure
q at each turbine frequency [1]. This was then converted to
velocity through a rearrangement of the definition of dynamic
pressure shown in Equation (1).

2𝑞
𝑈=√ (1)
𝜌

This velocity was then used with the diameter of the object d
in the wind tunnel to calculate the Reynolds number, shown in
Equation (2).
𝜌𝑈𝑑
𝑅𝑒 = (2) Fig. 1. Two-component dynamometer used to measure the drag forces for the
𝜇
objects in the lab [2].

The numerator and denominator in Equation (2) represents

Additionally, since the readout on the dynamometer is not a
the inertial forces and viscous forces acting on the object of
force, a simple calibration curve must be created that relates
focus, respectively. Reynolds number was used to describe the
each weight placed on the dynamometer to the voltages on the
conditions within the wind tunnel test section as it is a
LVDT. This calibration will be performed using known
nondimensional value that describes the flow pattern of the
gravitational forces on the dynamometer and will result in
environment. Via scaling, drag data corresponding to each
Equation (4), where C is the linear calibration constant, V is the
Section12605_Lab4 1

voltage given by the LVDT and D is the resultant drag on the maximum desired voltage was set to 10V and was calibrated to
object. occur at a force of 22.24N or a mass of 2.27kg according to
𝐷 =𝐶×𝑉 (4) design specifications [4]. The mass of 2.27kg was placed on the
lever arm and the reading of the voltage for the dynamometer
Any bias error resulting from the calibration of the LVDT on the transducer was set to 10V. The weights were then
was then removed. This corrected drag force was then used to removed and the core thumbwheel as adjusted to allow for a
determine the nondimensional coefficient of drag for the object zero reading on the transducer. This was repeated until the
within the flow, through Equation (5). voltages were accurate to the millivolt in both scenarios without
adjustment. Weights were then hung from the lever arm in
𝐷 0.5kg increments from 0kg to 2kg and the voltages measured
𝐶𝐷 = 𝑞𝐴 (5)
by the dynamometer were recorded. Each weight then had its
mass measured using a conventional scale and its mass
In order to superimpose data collected during the lab onto converted to a gravitational force. These gravitational forces
preexisting data graphs using a logarithmic scale, logarithmic were then plotted against their corresponding voltages and used
interpolation had to be performed using Equation (6) [3]. to determine a linear calibration curve for the data.
𝑓 1−𝑓 (6)
𝑥 = 𝑥2 𝑥1

Within Equation (6), 𝑥 is the data of interest, 𝑥2 and 𝑥1 refers

to the upper and lower bound, 𝑥 is the data of interest, and 𝑓 is
the fractional division, which when applied to the range
bounded by 𝑥2 and 𝑥1 , describes the location of 𝑥 on a
logarithmic scale. As such, 𝑓 would be calculated to determine
where the data point would be placed on graphs that employ
logarithmic scale.
To determine the error in the calibration coefficient, a Monty
Carlo simulation was performed. A Monty Carlo simulation
uses the errors in the x and y values of a group of points to create
a large amount of lines based on the theoretical frequency of the
errors within the original data set. The slopes of these lines are
then averaged, and the standard deviation of these slopes was
also determined. This average and standard deviation were then
Fig. 2. The dynamometer mounted horizontally with a calibration weight
used as the value and error of the slope of the data set. attached.

Dynamometer in the Wind Tunnel

II. PROCEDURE The dynamometer was installed in the wind tunnel so that
Determining Reynolds Numbers the drag forces could be calculated and is shown in Figure 3.
To determine the Reynolds numbers within the wind tunnel, the
local density of the air must first be determined. To determine
this value, a barometer was used to measure atmospheric
pressure, and a thermometer and hydrometer were used to
determine temperature and humidity respectively. The
diameters of each of the circular objects placed within the
tunnel were determined and were then used to find the cross-
sectional area of the blocked flow. The dynamic pressures at
this wind were then determined for each of the used frequencies
using a previous calibration [4]. These dynamic pressures were
then converted into velocities and Reynolds numbers using the
respective cross-sectional areas of the objects placed within the
flow.

Dynamometer Calibration
The first step to accurately reading the drag force measured
by the dynamometer was finding the calibration curve that
related the voltage measured by the LVDT to the force on its
lever arm. In order to do this, the dynamometer was mounted
with the lever arm parallel to the ground (Fig 2), and different Fig. 3. Configuration of the dynamometer and attached object within the
wind tunnel. The arrows show the direction of where the flow originated with
weights were hung from the lever arm with the voltage respect to the dynamometer and attached object.
measured by the LVDT for each weight was recorded. The

1
Section12605_Lab4 2

As the weight of the lever arm of the dynamometer itself had These values for cross sectional area were then used with
not been accounted for during the initial calibration and would previously acquired data for dynamic pressure and plugged into
otherwise be present as a bias error, the initial voltage measured Equation (2) to determine the Reynolds number at each turbine
by the vertical dynamometer without any outside interference frequency.
was then recorded and tared out of all future measurements. The masses of the weights used in the calibration of the
As shown in Figure 4, the shapes used in this experiment dynamometer are given in Table II. As each individual weight
were a rough sphere, smooth sphere, hollow hemisphere with had a measurement error of 0.0001, the use of multiple weights
the curved end facing into the flow, hollow hemisphere with the resulted in increased error.
curved end facing away from the flow, and a disk shape.
TABLE II
CALIBRATION MASS VALUES

Given Mass Real Mass Uncertainty

0Kg 0Kg 0 Kg

0.5Kg 0.4994Kg 0.00005 Kg

1Kg 0.9988 Kg 0.00005 Kg

1.5Kg 1.4982 Kg 0.0001 Kg

2Kg 1.99869Kg 0.0001 Kg

2.27Kg 2.26819Kg 0.0003Kg

Fig. 4. This shows the 5 objects used in the wind tunnel. From left to right,
they will be referred to as hemisphere A, hemisphere B, disk, rough sphere, and
smooth sphere for the remainder of this lab.
These values were then graphed and used to calibrate the
The wind tunnel was then run at five different Reynolds LVDT to their respective gravitational forces (Fig. 5).
number for each of the shapes shown in Figure 4 and the voltage
of the dynamometer were recorded. The voltages were then
converted to force values through Equation (4). To give this 25
experiment a less arbitrary result, these drag forces were then 20
converted to drag coefficients using Equation (5). These drag
Force(N)

coefficients were finally compared to external data to check 15

whether the results of this experiment were consistent with the
rest of the scientific community. 10

5 y = 2.2271x - 0.0036
III. RESULTS
It was determined that the temperature was 24.1±0.05oC with a 0
humidity of 39±0.5%. This resulted in a local atmospheric 0 2 4 6 8 10 12
density of 1.190±0.005kg/m3. LVDT Voltage (V)
The diameters and cross-sectional areas of each of the
circular objects placed within the tunnel were recorded as Fig (5). Force vs. Voltage
shown in Table I.

TABLE I This resulted in the calibration Equation (4) with a calibration

PHYSICAL PROPERTIES OF OBJECTS slope C of 2.2271±0.0014.
Hemisphere A Value Uncertainty
The LVDT voltage resulting from the cantilever beam
Diameter 5.31E-02 m 1 E-05 m
Cross sectional Area 2.2E-03 m2 4 E-05 m2 portion of the dynamometer was recorded to be 1.715 V, which
Hemisphere B was used to tare the voltage readings of the shapes to account
Diameter 5.33E-02 m 4 E-05 m for the offset.
Cross sectional Area 2.22E-03 m2 4 E-05 m2
Flat Top After calibration, each of the five objects were placed in the
Diameter 5.042E-02 m 3 E-05 m wind tunnel test section and the resulting voltage was obtained.
Cross sectional Area 2.00E-03 m2 4E-05 m2 Once the cantilever beam portion of the dynamometer was
Rough Sphere
accounted for Drag force and Coefficient of Drag was then
Diameter 5.29E-02 m 2E-04 m
Cross sectional Area 2.194E-03 m2 4.151E-05 m2 calculated via the use of Equations (4) and (5), respectively.
Smooth Sphere
Diameter 5.1E-02 m 3E-02 m
Cross sectional Area 2.0E-03 m2 4E-05 m2

2
Section12605_Lab4 3

The measured values for voltage at each Reynold’s number The measured values for voltage at each Reynold’s number
as well as the corresponding drag force and coefficient of drag as well as the corresponding drag force and coefficient of drag
for hemisphere A are shown in Table III. for the smooth sphere are shown in table VII.

TABLE III TABLE VII

HEMISPHERE A VALUES SMOOTH SPHERE VALUES
Reynolds Number Voltage Drag Force Coefficient of Drag Reynolds Number LVDT Drag Force Coefficient of Drag
2.79×104 0.011V 0.02N 0.283 Voltage
5.41×104 0.054V 0.12N 0.372 2.79×104 0.014V 0.03N 0.395
8.37×104 0.134V 0.30N 0.385 5.41×104 0.06V 0.13N 0.452
1.14×105 0.254V 0.57N 0.391 8.37×104 0.142V 0.32N 0.446
1.45×105 0.412V 0.92N 0.396 1.14×105 0.269V 0.60N 0.454
1.49×105 0.597V 1.33N 0.392 1.45×105 0.438V 0.98N 0.460
1.49×105 0.636V 1.42N 0.457

The measured values for voltage at each Reynold’s number

as well as the corresponding drag force and coefficient of drag
for hemisphere B are shown in Table IV. IV. DISCUSSION
In order to verify proficiency in two-component
TABLE IV dynamometer operation, drag coefficient data was obtained for
HEMISPHERE B VALUES
various shapes. Figure 6 shows the general drag coefficient
Reynolds Number LVDT Drag Force Coefficient of Drag
values for a Reynolds Number with an order of magnitude of
Voltage
2.79×104 0.04V 0.09N 1.02 105.
5.41×104 0.195V 0.43N 1.33
8.37×104 0.486V 1.08N 1.38
1.14×105 0.922V 2.05N 1.41
1.45×105 1.482V 3.30N 1.41
1.49×105 2.172V 4.84N 1.42

The measured values for voltage at each Reynold’s number

as well as the corresponding drag force and coefficient of drag
for the disk are shown in Table V.

TABLE V
DISK VALUES
Reynolds Number LVDT Drag Force Coefficient of Drag
Voltage
2.79×104 0.033V 0.07N 0.944
5.41×104 0.154V 0.34N 1.18
8.37×104 0.384V 0.85N 1.22
1.14×105 0.723V 1.61N 1.24
1.45×105 1.166V 2.60N 1.24
1.49×105 1.706V 3.80N 1.24

The measured values for voltage at each Reynold’s number

as well as the corresponding drag force and coefficient of drag
for the rough sphere are shown in Table VI.

TABLE VI
ROUGH SPHERE VALUES Fig. (6). Drag coefficient values of various shapes under 105 Reynolds
Reynolds Number LVDT Drag Force Coefficient of Drag number conditions [5]
Voltage
2.79×104 0.016V 0.04N 0.416
5.41×104 0.065V 0.14N 0.452
8.37×104 0.114V 0.25N 0.331
Referencing Fig. (6), shapes number 2, 9, and 7 correspond
1.14×105 0.23V 0.51N 0.358 to hemisphere A, hemisphere B, and the circular disk used in
1.45×105 0.384V 0.85N 0.372 the lab, respectively.
1.49×105 0.567V 1.26N 0.376

3
Section12605_Lab4 4

For hemisphere A, hemisphere B, and circular disk, the data Due to the specialized shapes of hemisphere A, hemisphere
was plotted, in Fig. 7-9. Within these figures, the x and y axis B, and the circular disk, preexisting drag coefficient data at a
represent Reynolds number and drag coefficient, respectively. multitude of Reynolds number conditions was not readily
The range of the Reynolds numbers used were of similar available. In contrast, the sphere is a common shape, meaning
magnitude, meaning that the use of a linear scale for the x axis that a larger amount of preexisting data at varying Reynolds
was sufficient. As the resulting drag coefficient values were numbers was available. As such, the experimental data obtained
also of similar magnitude, a linear scale was also used for the in this investigation was superimposed onto previously made
y axis. plots in regard to the rough and smooth sphere. Via Equation
(6), the experimental data was interpolated into a logarithmic
0.6 scale and then superimposed onto the preexisting graphs.
Regarding the rough sphere, the superimposed experimental
0.4 data does not appear to be consistent with the data provided by
CD

NASA. The general shape, however, does match NASA’s data

0.2
(Fig 10).
0
0.0E+0 5.0E+4 1.0E+5 1.5E+5 2.0E+5
Re

Fig.(7) Reynolds number vs drag coefficient in regard to hemisphere A.

1.5

1
CD

0.5

0
0.0E+0 5.0E+4 1.0E+5 1.5E+5 2.0E+5
Re

Fig (8) Reynolds number vs drag coefficient in regard to hemisphere B.

Fig. (10). Reynolds number vs drag coefficient regarding the rough sphere data
superimposed on preexisting data [6].

1.5 This discrepancy is best described by that fact that the

roughness of any given rough sphere could vary. In the case of
1 spheres, at a critical Reynolds number, surface roughness
CD

causes the flow surrounding the sphere to become turbulent,

0.5 effectively delaying flow separation. This causes a less
significant vortex sheet trailing the sphere when compared to a
0
smooth sphere at the same Reynolds number conditions,
0.0E+0 5.0E+4 1.0E+5 1.5E+5 2.0E+5 resulting in less drag [7]. Under this assumption, experimental
Re data and preexisting NASA data would match if the rough
spheres used in both instances have the same roughness.
Fig. (9) Reynolds number vs drag coefficient in regard to the circular disk. Therefore, Figure 10 implies that the rough sphere used by
NASA had a rougher surface than that of the rough sphere used
Via linear interpolation, the drag coefficients of hemisphere in this lab. As a result, producing a percent error would not
A, hemisphere B, and the circular disk were calculated to be produce any meaningful insight as the rough spheres in each
0.386, 1.39, and 1.23, respectively. Comparing the instance are significantly different in terms of roughness.
experimental values to the respective values in Fig. (6), the In terms of the smooth sphere, the preexisting data was
percent error of hemisphere A, hemisphere B, and the circular represented by Equation (7) [8].
disk was 1.57%, 2.11%, and 5.12%, respectively. It also must
be noted that the value for the drag coefficient using the lowest −7.94
𝑅𝑒 𝑅𝑒 𝑅𝑒
Reynolds number is less accurate compared to the other drag 24 2.6( ) 2.6(
2.63×105
) 0.25(
106
)
coefficients. This is because the uncertainty has a high relative 𝐶𝐷 = 𝑅𝑒 + 5.0
1.52 + −8.00 + 𝑅𝑒 (7)
𝑅𝑒 𝑅𝑒
1+(2.63×10 1+(106 )
1+(5.0 ) 5)
value compared to the drag force since the flow affects the
object the least at this wind speed.

4
Section12605_Lab4 5

As shown by Fig. (11), the superimposed experimental data TABLE VIII

GENERAL UNCERTAINTY VALUES
was consistent with the drag coefficient data represented by
Parameter Units Value Uncertainty
Equation (7).
Temperature C 24.1 0.05
Percent Humidity % 39 0.5
3
Density kg/m 1.19 0.005
Viscosity Pa×s 1.83×10-5 5×10-8
Calibration Weight Mass Kg N/a 0.000005
0Kg Calibration Measurement Kg 0 0
0.5Kg Calibration
Measurement Kg 0.4994 0.00005
1Kg Calibration Measurement Kg 0.9988 0.00005
1.5Kg Calibration
Measurement Kg 1.4982 0.0001
2Kg Calibration Measurement Kg 1.9986 0.0001
2.27Kg Calibration
Measurement Kg 2.2681 0.0003
LVDT Measurements V N/a 0.005
Calibration Curve Slope N/V 2.227 0.002
Drag N N/a 0.02
Fig. (11). Reynolds number vs drag coefficient regarding the smooth Dynamic Pressure at 10Hz Pa 39 20
sphere superimposed on preexisting data [8]. Dynamic Pressure at 20Hz Pa 146 20
Dynamic Pressure at 30Hz Pa 350 20
Dynamic Pressure at 40Hz Pa 652 20
Dynamic Pressure at 50Hz Pa 1046 20
By averaging the experimental data and comparing the Dynamic Pressure at 60Hz Pa 1529 20
result to the average of the preexisting data on the same range, Velocity at 10 Hz m/s2 8.09 2
the percent error within the context of the smooth sphere was Velocity at 20 Hz m/s2 15.7 1
Velocity at 30 Hz m/s2 24.2 0.7
1.29%.
Velocity at 40 Hz m/s2 33.1 0.5
Velocity at 50 Hz m/s2 41.9 0.4
Velocity at 60 Hz m/s2 50.7 0.3

V. CONCLUSION
The data determined within this study was consistent with
predetermined values, if not with the exact values, then in how Table IX shows the values and uncertainties of the physical
the data behaved over the range of Reynold’s Numbers. The properties of hemisphere A as well as the Reynolds number and
data for hemisphere A, hemisphere B, the circular disk, and coefficient of drag at each frequency value used by the turbine.
smooth sphere had a percent error of 1.57%, 2.11%, 5.12%, and
1.29%, respectively when compared to preexisting data. The
data for the rough sphere followed the same trend as preexisting TABLE IX
data with a sharp drop in drag coefficient values in the range HEMISPHERE A UNCERTAINTY VALUES
between 104 and 105 but had different values for the coefficient
Parameter Units Value Uncertainty
in that range. The empirical data for the smooth sphere both
follows the same trend as preexisting data and had consistent Diameter m 5.31E-02 1 E-05
values to said preexisting data. Cross sectional Area m 2
0.00222 4 E-05
Reynolds Number at 10 Hz N/a 2.8E+04 7 E+03
Reynolds Number at 20 Hz N/a 5.4E+04 4E+03

APPENDIX Reynolds Number at 30 Hz N/a 8.4E+04 2E+03

Table VIII shows the uncertainty values used in the calibration Reynolds Number at 40 Hz N/a 1.14E+05 2E+03
of the LVDT as well as the uncertainty values for all values Reynolds Number at 50 Hz N/a 1.45E+05 2E+03
common to all objects placed within the flow. Reynolds Number at 60 Hz N/a 1.75E+05 2E+03
Coefficient of drag at 10Hz N/a 0.283 0.2
Coefficient of drag at 20Hz N/a 0.372 0.06
Coefficient of drag at 30Hz N/a 0.385 0.03
Coefficient of drag at 40Hz N/a 0.391 0.02
Coefficient of drag at 50Hz N/a 0.396 0.01
Coefficient of drag at 60Hz N/a 0.392 0.01

5
Section12605_Lab4 6

Table X shows the values and uncertainties of the physical Table XII shows the values and uncertainties of the physical
properties of hemisphere B as well as the Reynolds number and properties of the rough sphere as well as the Reynolds number
coefficient of drag at each frequency value used by the turbine. and coefficient of drag at each frequency value used by the
turbine.
TABLE X TABLE XII
HEMISPHERE B UNCERTAINTY VALUES ROUGH SPHERE UNCERTAINTY VALUES
Parameter Units Value Uncertainty Parameter Units Value Uncertainty
Diameter m 0.05334 4 E-05 Diameter m 0.0529 0.0002
Cross sectional Area m2 0.00222 4 E-05 Cross sectional Area m2 0.00219 4E-05
Reynolds Number at Reynolds Number at
10Hz N/a 2.8E+04 7E+03 10Hz N/a 2.8E+04 7E+03
Reynolds Number at Reynolds Number at
20Hz N/a 5.4E+04 4E+03 20Hz N/a 5.4E+04 4E+03
Reynolds Number at Reynolds Number at
30Hz N/a 8.4E+04 3E+03 30Hz N/a 8.3E+04 2E+03
Reynolds Number at Reynolds Number at
40Hz N/a 1.15E+05 2E+03 40Hz N/a 1.14E+05 2E+03
Reynolds Number at Reynolds Number at
50Hz N/a 1.45E+05 2E+03 50Hz N/a 1.44E+05 2E+03
Reynolds Number at Reynolds Number at
60Hz N/a 1.76E+05 2E+03 60Hz N/a 1.74E+05 2E+03
Coefficient of drag at Coefficient of drag at
10Hz N/a 1.02 0.5 10Hz N/a 0.416 0.3
Coefficient of drag at Coefficient of drag at
20Hz N/a 1.331 0.2 20Hz N/a 0.452 0.07
Coefficient of drag at Coefficient of drag at
30Hz N/a 1.384 0.08 30Hz N/a 0.331 0.02
Coefficient of drag at Coefficient of drag at
40Hz N/a 1.410 0.05 40Hz N/a 0.358 0.02
Coefficient of drag at Coefficient of drag at
50Hz N/a 1.412 0.04 50Hz N/a 0.373 0.01
Coefficient of drag at Coefficient of drag at
60Hz N/a 1.416 0.03 60Hz N/a 0.376 0.009

Table XI shows the values and uncertainties of the physical Table XIII shows the values and uncertainties of the physical
properties of the disk as well as the Reynolds number and properties of the smooth sphere as well as the Reynolds number
coefficient of drag at each frequency value used by the turbine. and coefficient of drag at each frequency value used by the
turbine.
TABLE XI
TABLE XIII
DISK UNCERTAINTY VALUES
SMOOTH SPHERE UNCERTAINTY VALUES
Parameter Units Value Uncertainty Parameter Units Value Uncertainty
Diameter m 0.05042 3E-05 Diameter m 0.05 0.03
Cross sectional Area m2 0.00200 4E-05 Cross sectional Area m2 0.00203 4E-05
Reynolds Number at Reynolds Number at
10Hz N/a 2.7E+04 7E+03 10 Hz N/a 2.7E+04 2E+04
Reynolds Number at Reynolds Number at
20Hz N/a 5.1E+04 4E+03 20 Hz N/a 5.2E+04 3E+04
Reynolds Number at Reynolds Number at
30Hz N/a 7.9E+04 2E+03 30 Hz N/a 8.00E+04 4E+04
Reynolds Number at Reynolds Number at
40Hz N/a 1.08E+05 2E+03 40 Hz N/a 1.1E+05 5E+04
Reynolds Number at Reynolds Number at
50Hz N/a 1.37E+05 2E+03 50 Hz N/a 1.4E+05 7E+04
Reynolds Number at Reynolds Number at
60Hz N/a 1.66E+05 1E+03 60 Hz N/a 1.7E+05 8E+04
Coefficient of drag at Coefficient of drag at
10Hz N/a 1.022 0.5 10Hz N/a 0.395 0.2
Coefficient of drag at Coefficient of drag at
20Hz N/a 1.331 0.2 20Hz N/a 0.452 0.07
Coefficient of drag at Coefficient of drag at
30Hz N/a 1.384 0.08 30Hz N/a 0.446 0.03
Coefficient of drag at Coefficient of drag at
40Hz N/a 1.410 0.05 40Hz N/a 0.454 0.02
Coefficient of drag at Coefficient of drag at
50Hz N/a 1.412 0.04 50Hz N/a 0.460 0.01
Coefficient of drag at Coefficient of drag at
60Hz N/a 1.416 0.03 60Hz N/a 0.457 0.01

6
Section12605_Lab4 7

The diameter of the circular shapes used within the flow were Drag on objects in the wind tunnel was then determined by
determined empirically using calipers, and therefore the error Equation (4), where C is the calibration coefficient determined
was half of the smallest measurable value. The cross-sectional by the Monty Carlo simulation.
areas of the blockages were therefore determined using
Equation (3). The error in the cross-sectional area was therefore
determined using equation (8). The error in drag was therefore Given by Equation (13).

(8) (13)
𝜕𝐴 2 𝜕𝐷 2 𝜕𝐷 2
𝑈𝐴 = √( ) 𝑈𝐷 2 𝑈𝐷 = √( ) 𝑈𝑉 2 + ( ) 𝑈𝐶 2
𝜕𝐷 𝜕𝑉 𝜕𝐶

The dynamic pressure and the error within the dynamic The coefficient of drag was then determined by Equation (5).
pressure associated with each frequency was determined in a The error in the coefficient of drag was therefore given by
prior calibration of the wind tunnel [1]. The velocity was then equation (14).
determined using Equation (1). The error in the velocity was
therefore determined using Equation (8). 𝜕𝐶𝑑 2 2 𝜕𝐶𝑑 2 2 𝜕𝐶𝑑 2 2
𝑈𝐶𝑑 = √( ) 𝑈𝐷 + ( ) 𝑈𝑞 + ( ) 𝑈𝐴 (14)
𝜕𝐷 𝜕𝑞 𝜕𝐴
2 2
𝜕𝑈 𝜕𝑈
𝑈𝑈 = √( ) 𝑈𝑞 2 + ( ) 𝑈𝜌 2 (9)
𝜕𝑞 𝜕𝜌 REFERENCES
[1] G. Chow, N. Rose, and W. Thorne “Wind Tunnel Calibration” Sep. 2019
[2] J. Abbit and L Ukeiley, Lab 4, October 2019
The Reynold’s number of the flow was determined using [3] “Linear and Logarithmic Interpolation” Internet:
Equation (2). Therefore, the Uncertainty in the Reynold’s https://www.cmu.edu/biolphys/deserno/pdf/log_interpol.pdf
number was determined using Equation (10). [4] G. Chow, N. Rose, and W. Thorne “Pressure Cylinder” Oct. 2019
[5] “Drag of Cylinders and Cones” Internet:
http://www.aerospaceweb.org/question/aerodynamics/q0231.shtml
[6] “Drag of a Sphere” Internet: https://www.grc.nasa.gov/www/k-
(10) 12/airplane/dragsphere.html
[7] “Why Do Golf Balls Have Dimples?” Internet:
https://www.livescience.com/32446-why-do-golf-balls-have-dimples.html
[8] “Data Correlation for Drag Coefficient for Sphere” Internet:
The weight of the calibration masses was determined using https://pages.mtu.edu/~fmorriso/DataCorrelationForSphereDrag2016.pdf
Equation (11).
𝐹 =𝑚×𝑔 (11)

Therefore, the error in force was determined using Equation

(12).
𝜕𝐹 2 𝜕𝐹 2
𝑈𝐹 = √( ) 𝑈𝑚 2 + ( ) 𝑈𝑔 2 (12)
𝜕𝑚 𝜕𝑔

A Monte Carlo simulation using 6 points was used to determine

the slope and error in the slope of the calibration curve (Fig.
12). This resulted in a slope of 2.227±0.0014.
25

20

15
y data

10

0
0 5 10 15
-5
x data

Fig (12). Monty Carlo Simulation for calibration of LVDT