ENGINEERING PAPER
5254-08
A Comparison of Flow Testing
Results where Measurements
were Taken by both a Standard
Pitot Tube and a Directional Probe
Harry McArthur, P. Eng.
Senior Staff Engneer
Flowcare Engineering Inc.
Cambridge Ontario
Vern Martin P. Eng.
Flowcare Engineering Inc.
Cambridge Ontario
AMCA International Engineering Conference
Las Vegas, NV, USA
2 – 4 March 2008
A Comparison of Flow Testing Results where Measurements were Taken by both
a Standard Pitot Tube and a Directional Probe
Harry McArthur, P.Eng, Senior Staff Engineer
FLOWCARE Engineering Inc., Cambridge, Ontario
1.0
Background
Fan testing requires the determination of the
volume flow rate. One of the most common
methods to obtain this measurement is the
velocity traverse method.
The velocity traverse method consists of
subdividing a duct cross-sectional area into a
number of elemental areas. The velocity
pressure normal to each elemental area is then
measured, using a suitable probe at the centroid
of each elemental area. The average velocity
normal to the duct cross sectional area is then
calculated using a formula that recognizes the
contributions of each elemental area in
conjunction with the gas density.
This method of testing generally requires that
the following conditions hold true:
• All the flow in the test plane is in one
direction (there is no flow reversal.)
• The flow has a relatively uniform velocity
profile across the test plane.
• The flow is predominately normal to the test
plane, not skewed. Skewed flow is defined
as flow that is in a direction other than
normal to the test plane. Only the
component of velocity normal to the test
plane is pertinent to the calculation of flow. If
the flow is skewed, the test method must be
able to quantify that portion of the flow that is
normal to the test plane.
Recognizing the requirements of field testing
where varied velocity distributions and/or
skewed flow may occur, performance standards
have been developed in an effort to reduce or
eliminate the uncertainty and errors associated
with these factors. These standards either
discourage
testing
at
locations
where
questionable conditions may occur or specify the
use of probes capable of indicating both flow
direction and velocity pressure.
The AMCA Publication 203 ‘Field Performance
Measurements’ and the AMCA Standard 803
‘Industrial Process/Power Generation Fans: Site
Performance Test Standard’ are a test guide
and standard based on the use of nondirectional probes. These documents stipulate
specific requirements for both the traverse test
plane and the test results that are designed to
reduce or eliminate problems associated with
flow uncertainties.
The American Society of Mechanical Engineers
(ASME) Performance Test Code No.11 (PTC11) is a test code that addresses the uncertainty
of field testing by specifying the use of a
directional probe at a relatively large number of
points for the velocity traverse. In this way, the
actual flow vectors can be measured and the
uncertainty due to a lack of uniformity reduced
by testing at a large number of points.
Testing done by FLOWCARE has shown that
velocity traverses with Pitot-static tubes can give
relatively accurate test results for test conditions
that would be considered unacceptable based
on AMCA 803 or even AMCA 203 test criteria.
These are test conditions with significantly
skewed flows.
• This paper will discusses the AMCA and
PTC-11 test standards and procedures used
to conduct a velocity traverse and the
specific requirements of these standards as
they pertain to testing in conditions where
there may be skewed flow in the velocity
traverse test plane.
• A case study detailing the results of fan
performance testing done on two 6,000 hp,
ID fans on Unit 2 of PacifiCorp’s Jim Bridger
Generating Station will be presented. The
test plane chosen for these fans does not
meet AMCA requirements due its geometry
and the flow angles. The velocity traverses
were done following both PTC-11 code
requirements and AMCA 203 guidelines with
equal results.
• The performance and accuracy of a Pitotstatic tube in a skewed flow stream will be
examined.
• The paper will discuss the implications of the
test results on the accuracy and reliability of
the two test probes.
1
2.0
Field Performance Test
Standards for Acceptable
Velocity Traverse
Probably the single greatest factor behind the
specification of test standards and codes as they
pertain to the velocity traverse involves the
direction of the flow stream through the test
plane. A concern is fuelled by the reality that
errors in the assumption of the flow direction can
lead to significant errors in the calculation of the
volume or mass flow rates. AMCA addresses
this issue by the stipulation of specific
requirements for both the traverse test plane
and the test results. These are designed to
eliminate problems associated with flow
uncertainties. ASME addresses this issue in
PTC-11 by the use of a directional probe and a
large number of traverse points.
2.1
AMCA 203 - Criteria for the Velocity
Traverse
AMCA Publication 203 ‘Field Performance
Measurements’ is the simplest and least costly
test procedure. AMCA 203 recommends the
following criteria for the traverse test plane and
the resulting velocity measurements:
• The cross-sectional shape of the airway
(duct) should not be irregular. Specifically the
test plane should be rectangular or circular in
cross-section.
• The cross-sectional shape and area of the
duct in the vicinity of the traverse plane
should be uniform.
Divergences and
convergence of the duct should be moderate
so that uniform flow conditions exist.
• The velocity distribution should be uniform
with more than 75% of the velocity pressure
measurements greater than 1/10 of the
maximum measurement.
• The angle between the flow stream and the
traverse plane must be within 10 degrees of
a right angle.
The first two criteria are relatively straightforward
and easy to assess. The third criteria can not be
established until after the test is completed while
the fourth criteria would require the use of a
directional probe to be accurately assessed.
The last two criteria are those most likely to be
affected by the presence of skewed flow. There
is no requirement that individual flow vectors fall
into any range but only that the angle between
the flow stream and the traverse plane must fall
with 10 degrees. There are no guidelines given
as to how this flow stream angle is established.
It appears that these aspects are left up to the
subjective experience and judgment of the
tester. It should be noted that only velocity
pressure measurements are required at each
elemental point. Static pressures can be
measured at each point or obtained by static
pressure taps on the duct side panels.
2.2
AMCA 803 Criteria for the Velocity
Traverse
AMCA 803 test standard has significantly more
stringent requirements then the guidelines
provided by AMCA 203. This standard requires
velocity
pressure
and
static
pressure
measurements be taken at each point in the
traverse plane. It also has a number of
requirements that specifically address the
geometry of the duct at the traverse test plane,
the directionality of the flow through the traverse
test plane and the resulting velocity
measurements. Failure to meet any of these
criteria invalidates the results from qualifying as
an 803 test. These criteria are as follows:
• The cross-section of the duct at the
measurement plane shall be circular or
rectangular with no irregularities.
• The measurement plane shall be free from
any accumulation of dust or debris.
• The measurement plane shall not intersect
any internal stiffeners, supports, splitters,
vanes. It shall clear such internal
obstructions by at least 0.5 times the duct
diameter or equivalent diameter for a
rectangular duct sections.
• Any measurement plane shall be at least
0.5 D upstream and 1.0 D downstream of
any bend or change in cross-sectional
area.
• Standard deviation of the velocity variation
must be less than 10% of the mean
velocity.
• AMCA 803 considers the Pitot-static tube
as the primary probe type. The Pitot tube
shall be parallel to the axis of the duct
within ±7.5°.
• The angle of flow at each measurement
point may not exceed 15° from a right
angle to the measurement plane. This
criterion may require the use of a
directional probe to measure the flow
2
angle. The use of a directional probe is not
a mandatory requirement. The directional
probes recommended by AMCA 803 are
the Fecheimer probe, wedge probe, three
hole cylindrical probe or other anglesensitive measuring device. All of these
probes with the exception of ‘other’ are
probes designed to sense the flow angle in
only a single plane.
• The number of points which may exceed
10° from a right angle to the measurement
plane shall not exceed 10% of the total
number of traverse points.
• Ultimately, an agreement must be reached
between the tester and the customer as to
the acceptability of the test plane and test
results.
The first few criteria dealing with the duct
geometry are relatively straightforward and easy
to assess. That is not to say that meeting these
criteria is always possible. The requirements
pertaining to the distortion of the velocity profile
and/or skewed flow are often difficult or
impossible to meet and can only be assessed
after a test is completed. Therefore, costly and
time consuming tests possibly involving multiple
traverses with multiple sensors may end up
being invalidated due to a single test point
vector that exceeds 15° from the axis of the
duct. There is also the possibility of error due to
the fact that the recommended directional
probes only address flow in one plane.
2.3
Figure 1 – Sketch of Prismatic 5-Hole Probe
-
Port P1 is used to measure total pressure.
Ports P2 and P3 are used to establish the
yaw angle.
Ports P4 and P5 are used to measure the
pitch angle.
Figure 2 shows the relationship of pitch and yaw
to the probe tip.
PTC-11 Test Criteria
PTC-11 is primarily a fan test code. As such,
the test planes are at the fan inlet(s) and fan
outlet. PTC-11 states ‘Due to the highly
disturbed flow at the fan boundaries and the
errors obtained when making measurements
with probes unable to distinguish directionality,
probes capable of indicating gas direction and
speed, hereinafter referred to as directional
probes, are generally required. Only the
component of velocity normal to the elemental
area is pertinent to the calculation of flow.’ While
allowing other flow measurement methods, the
velocity traverse is considered the primary
method by PTC-11.
One of the most common types of directional
probes is the prismatic 5-hole probe. Figure 1
shows a sketch of the tip of a 5-hole prismatic
probe and the location of the pressure taps
(holes) on the probe tip.
Figure 2 – Yaw and Pitch Referenced to
Probe Tip
•
•
The traverse plane suggested by PTC-11
assumes that the traverse plane contains
skewed flow. The skewed flow is accounted
for by using a directional probe.
The directionality of the flow vector at
individual elements is measured in terms of
yaw and pitch angles. The yaw angle is
determined by rotating the probe about its
axis until the pressure (P1 – P2) is zero. The
angular rotation of the probe is the yaw
angle.
The pitch angle is generally
3
•
determined by measuring the differential
pressure across the pitch test ports and then
interpolating the pitch angle from calibration
curves.
All pertinent pressures must be interpolated
as a function of the pitch and/or yaw angles
and velocity from calibration curves. Each
calibration curve is for a specific velocity and
covers a range of pitch angles generally
from 0° to ±40°. A number of calibration
curves covering the anticipated velocity
range of the test plane are required.
Finding the yaw angle at each traverse point
greatly increases the amount of time required to
perform the test and thus keep the system at a
constant operating point. PTC-11 requires that
five different test values be recorded at each test
point. These parameters are the yaw angle or
yaw pressure differential, pitch pressure
differential, the velocity pressure, total and/or
static pressure and the temperature. Testing to
this code is generally costly, time consuming
and onerous to complete. The accuracy of the
measured data is very dependent on proper use
of the probe in the field as well as accurate
calibration over the measured velocity range.
The time required to log and record all required
field test parameters has resulted in heavy
reliance on electronic instrumentation and data
logging technology. Once field data collection is
complete, the flow calculation procedure is also
complex and onerous.
3.0
Description of Actual Field Test
Procedures Used for Case
Study
The selection of a test plane on the ID fans at
PacifiCorp’s Jim Bridger Generating Station
presented a difficult challenge. Figure 3 shows
the configuration of the fans and surrounding
ducts. It also shows the location of the test plane
in one of the fan inlet boxes.
The fans are double width drawing flue gas from
an electrostatic precipitator. The flue gas feeds
into a crossover header that in turn feeds into
the transition above the fan inlet boxes from
three different directions. The pants split the flow
from the three converging ducts into the inlet
boxes. The pants have both converging and
diverging sides. Two sides converged on the
inlet box at 30 degree angles while the opposite
two sides diverged at approximately 15 degrees.
Figure 3 – Fan and Duct Arrangement and Location of Test Plane
4
The velocity traverse test planes chosen for
these fans were in the fan inlet boxes at a plane
approximately 18 inches below the inlet box
flange. Consideration for a test plane in the
discharge evase was rejected as the evase was
diverging in all planes, had an unknown internal
geometry due to a false bottom, terminated in a
45° degree elbow and was difficult to access.
None of the potential test planes met AMCA 803
requirements in terms of the duct configuration.
It was also anticipated that there would be
significantly skewed flows at the inlet test plane
that would also disqualify an AMCA 803 test.
Due to the poor traverse plane location and the
anticipation of significantly skewed flows, a
velocity traverse with a directional probe had
been specified. The inlet boxes have a crosssectional area of 121 ft2. A total of 64 velocity
pressure readings were obtained in each box.
The velocity traverses were carried out using a
prismatic 5-hole probe. For three of the inlet
boxes, data were first collected using the 5-hole
probe first and then using the Pitot tube.
Two fan tests were carried out using both the 5hole probe and the Pitot tube. Table 1 lists the
tests and operating conditions in terms of the
Variable Inlet Vane (VIV) damper setting and the
gross megawatt production rate over the test
period. It should be noted that for Test 21A, both
probes were only used on one inlet box.
• When using the 5-hole probe, difficulty was
encountered in establishing the yaw angle at
some traverse points as turbulence resulted
in fluctuations of up to 15 degrees. A 15°
error in the yaw angle can result in a 5% or
greater error in establishing the actual
velocity pressure.
• Flow measurements with the Pitot tube could
be carried out in approximately 2 hours while
the tests with the directional probe took up to
8 hours.
Fan
Test No
Date
Description
21
22
21A
22A
Nov. 6, 2007
Nov. 7, 2007
Test of fan with motor fully loaded
Test of fans at MCR conditions
VIV
(% Open)
69
80
Gross
MW
540
540
Table 1 – Fan Performance Test Schedule, Nov. 6 to Nov. 8, 2007
For the Pitot tube measurements, data readings
were averaged over a period of 10 to 20
seconds. With the 5-hole probe, due to the time
necessary to zero the yaw and the greater
number of parameters that had to be recorded,
data were obtained using a shorter averaging
period. This period could be as short as 3
seconds.
The flow distribution across the inlet boxes was
relatively uniform. This is illustrated in Figure 4
which shows a comparison of the flow traverses
in the north inlet box of Fan 22 for both the 5-hole
probe and the Pitot tube. The velocity profile
measured by the pitot tube appears smoother
than that obtained by the 5-hole probe; probably
because the pitot tube measurements were
averaged over a longer time period.
ƒ
ƒ
ƒ
ƒ
Table 2 shows a comparison between data
obtained using the 5-hole probe and that
obtained using the Pitot tube for Test 21 of
north inlet box.
Table 3 shows the maximum, minimum and
average pitch, yaw and absolute angles of
the 5-hole probe. All the angles are
referenced to the axis normal to the traverse
plane. The sign on the yaw and pitch
indicate the quadrant. The absolute angle is
the angle between the velocity vector and
the axis normal to the test plane.
Table 4 shows a comparison of the flow and
pressure data between the two types of
probes used for Test 22 of ID Fan 22.
Table 5 shows the maximum, minimum and
average pitch and yaw angles of the 5-hole
probe for test 22 of ID Fan 22.
5
Figure 4 - Fan 22 North Inlet Box Velocity Profiles Measured by a Pitot Tube and 5-Hole Probe
Fan 21A Nov. 6, 2007
Static Pressure (in.wg)
Velocity Pressure (in.wg)
Total Pressure (in.wg)
Volume (acfm)
North Inlet Box
5-hole Probe Pitot Tube
-18.71
-18.17
1.11
1.06
-17.61
-17.11
687,162
678,793
%Difference
2.9
4.4
2.8
1.2
Table 2 – Test 21, Comparison of 5-Hole Probe and Pitot Tube Data
Maximum
Minimum
Average
North Inlet Box
Pitch Angle Yaw Angle
(Deg.)
(Deg.)
15.2
17.8
-13.8
-41.0
0.8
-5.5
Absolute Angle
(Deg.)
41.1
0.2
12.1
Table 3 – Test 21, North Inlet Box, Max., Min. and Average Pitch, Yaw and Absolute Angles
North Inlet Box
5-hole Probe Pitot Tube
-18.69
-18.50
1.07
1.13
-17.62
-17.37
692,574
716,421
South Inlet Box
Fan 22A Nov. 7, 2007
5-hole Probe Pitot Tube
Static Pressure (in.wg)
-18.53
-18.54
Velocity Pressure (in.wg)
1.09
1.11
Total Pressure (in.wg)
-17.44
-17.43
Volume (acfm)
686,657
694,562
Average/Total
Fan 22A Nov. 7, 2007
5-hole Probe Pitot Tube
Static Pressure (in.wg)
-18.61
-18.52
Velocity Pressure (in.wg)
1.08
1.12
Total Pressure (in.wg)
-17.53
-17.40
Total Volume (acfm)
1,379,231
1,410,983
Fan 22A Nov. 7, 2007
Static Pressure (in.wg)
Velocity Pressure (in.wg)
Total Pressure (in.wg)
Volume (acfm)
% Difference
1.0
-5.9
1.4
-3.4
% Difference
0.0
-1.6
0.1
-1.2
% Difference
0.5
-3.7
0.7
-2.3
Table 4 – Test 22, Comparison of 5-Hole Probe and Pitot Tube Data
6
Maximum
Minimum
Average
North Inlet Box
Pitch
Yaw Absolute
Angle Angle
Angle
(Deg) (Deg)
(Deg.)
15.6
39.7
40.1
-24.3
-25.0
0.86
-0.5
0.3
11.7
South Inlet Box
Pitch
Yaw Absolute
Angle Angle
Angle
(Deg) (Deg)
(Deg.)
25
11.5
25.0
1.2
-12.0
-13.3
8.6
0.1
-1.8
Table 5 – Test 22, Maximum, Minimum and Average Pitch and Yaw Angles
The test results from the 5-hole probe and the
standard Pitot tube showed very close
correlation in spite of the fact that the pitch and
yaw angles covered a wide range of values.
4.0
Effects of Pitch and Yaw on Pitotstatic Tube Pressure Measurements
The Pitot-static tube is the primary sensing
instrument for the AMCA 203 and the AMCA
803 test standards. One of the concerns in using
the Pitot-static tube is the direction of the flow
vector impacting on the tube. This concern is
evidenced in the criteria given for acceptable
traverse plane configurations and the limitations
placed on the allowable angularity of the flow
stream in the AMCA test guide and standards.
Figure 5 shows a graph illustrating the errors in
static, total (stagnation) and dynamic pressures
when a Pitot tube is subjected to increasing yaw.
The graph in Figure 5 was created from data in
NACA Technical Note No. 546 ‘Comparative
Tests of Pitot-static Tubes’ By Kenneth G.
Merriam and Ellis R. Spaulding published in
1937. Part of the analysis carried out by
Merriam and Spaulding was the effect of yaw on
the Pitot tube. This paper presents a
recommended design for Pitot tubes that is the
basis of many Pitot tubes sold today.
The percentage errors shown in Figures 5
pertain to the error between the actual total,
static and dynamic pressures of the flow stream
as compared to what the pressure gauges
attached to the Pitot tube will be indicating. The
errors are shown as a percentage of the
dynamic
pressure.
They
are
not
Pitot-static Tube Percent Error in Dynamic Pressure
5
Static Pressure
Total Pressure
Dynamic Pressure
Percentage Error
0
-5
-10
-15
-20
-25
0
5
10
15
20
25
Yaw Angle Degree
Figure 5 – Relation between Percent Error and Degrees of Yaw
7
absolute errors. It is interesting to note that the
dynamic pressure error will result in slightly
higher than actual dynamic pressure readings
for angles of yaw up to approximately 18° of
yaw. This is a interesting observation due to the
fact that Section 8.3 of AMCA 203 states ‘the
angle of the flow stream in any specific location
is indicated by the orientation of the nose of the
Pitot-static tube that produces the maximum
velocity pressure readings at that location’. It
can be seen from Figure 5 that this exercise
may result in the Pitot tube pointing
approximately 10° to 12° off of the direction of
the flow stream.
Figure 5 shows the errors as a function of yaw
angle. In reality, skewed flow will have both pitch
and yaw components. Figure 6 shows a Pitot
tube in a skewed flow stream where the flow
stream impacts the Pitot tube in a plane that
does not fall on either the pitch or yaw planes.
Therefore, if V and Phi are known, Vz can be
calculated as follows:
Vz = V * cos(Φ)
This is the rational behind the use of the
directional probe. The velocity pressure of vector
V is Pv and the equivalent velocity pressure of
velocity component normal to the test plane (Vz)
is Pvz.
Pv = Density * (V/C)2
V = C * √(Pv/density)
Pvz = Density * (Vz/C)2
Pvz = Density * [(V * cos(Φ) )/C]2
Pvz = Pv * cos2 (Φ)
Where: C = Constant
4.1 Effect of Skewed Flow on Measurement
of Velocity Pressure Normal to test
Plane
The errors shown in Figure 5 are the errors
relative to the actual flow stream velocity vector
V. However, only the velocity component normal
to the test plane is of interest. Therefore, the
velocity pressures that the Pitot tube indicates
should be compared against the equivalent
velocity pressure (Pvz) of the velocity vector (Vz)
normal to the test plane. Figure 7 shows the
velocity pressure indicated by a Pitot tube and
the actual equivalent velocity pressure (Pvz) of
the Vz velocity component. Also shown on
Figure 7 are the errors between the indicated
velocity pressure and the actual (Pvz) and the
error in the velocity Vz calculated from the
indicated velocity pressure.
Figure 6 – Pitot-static Tube in a Skewed Flow
Stream
The absolute value of the flow stream (V)
impacts the Pitot tube at an angle phi that is
relative to the axis (Z) normal to the test plane.
For flow testing purposes, only the velocity
normal to the test plane (Vz) is of interest.
It can be seen from Figure 7 that the error in the
calculated velocity is approximately 5% for
skewed flow angles of 20°. Tests done by
FLOWCARE under non-laboratory conditions
indicate that the error in the calculated velocity is
approximately 10% at angles up to 40°. The
error will result in the calculated velocity being
higher than the true velocity normal to the test
plane.
8
1.10
35%
1.00
30%
0.90
25%
0.80
20%
Measured Vp
Actual Vpz Normal to Test Plane
Error in Measured Velocity
Error in Vp Measurement
0.70
15%
0.60
10%
0.50
5%
0.40
Percentage Error
Unit Velocity Pressure
Comparsion Between Measured and Actual Velocity Pressures Normal to Test
Plane
0%
0
5
10
15
20
25
Angle of Skewed Flow
Figure 7 - Comparison Between Measured and Actual Velocity Pressures Normal to Test Plane
Table 6 shows the results of the ID fan tests at
PacifiCorp’s Jim Bridger Generating Station. The
table shows average angle of skew, the
percentage error in the measured flow rate as
compared to the 5-hole probe and the predicted
Fan
Fan 21 North Inlet Box
Fan 22 North Inlet Box
Fan 22 South Inlet Box
Average
Angle of Skew
12.1
11.7
8.6
error based on the data in Figure 7. This table is
based on the assumption that the velocity
pressure measured by the 5-hole probe is more
accurate than that obtained by the Pitot tube.
Percentage
Error in Flow
1.2% Low
3.4% High
1.2% High
Predicted Error
From Figure 7
3.0% High
2.9% High
1.7% High
Table 6 – Comparison of Actual Results to Predicted Results
It can be seen from Table 6 that the predictions
based on Figure 7 are close for two cases. The
lack of agreement between the results for the
test on Fan 21 and Figure 7 can potentially be
explained by the time lapse between the 5-hole
test and the Pitot test. The 5-Hole probe test
was carried out over an eight hour period. When
the 5-hole probe test was complete, it was
followed by the Pitot tube test. The fan was
ostensibly operating at a fixed point over this
whole period of time but there undoubtedly was
some variation in the actual volume flow rates.
4.2 Effect of Skewed Flow on Measurement
of Total and Static Pressures with a Pitot
Tube
It appears from Figure 5 that the errors in the
total and static pressures increase rapidly and
exponentially with increasing degrees of yaw.
The errors are such that indicated total and
static pressure readings will be lower than the
actual flow stream pressures. However, as was
previously stated, the errors shown in Figure 5
are errors as a percentage of the velocity
pressure. They are not absolute errors. Figure 8
shows a graph of the total pressure error as a
function of velocity pressure.
9
Total Pressure Error as a Function of Velocity Pressure
110%
Maximum Possible Total Pressure Error (%)
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0 Vp
0%
5xVp
500%
10xVp
1000%
15xVp
1500%
20xVp
2000%
25xVp
2500%
Total Pressure as Multiple of Vp
Figure 8 – Total Pressure Error as a Function of Velocity Pressure
For flow vector angles between phi = 0° and phi
= 90°, the error in the total pressure reading of
the Pitot tube will range from 0 at 0° to
approximately (Pt – Pv)/Pt at phi = 90°. At the
90° position, the total pressure tap will be
oriented perpendicular to the flow stream and
will be functioning essentially as a static
pressure tap. If the velocity pressure is a small
fraction of the total pressure, the absolute error
in the total pressure indicated by a Pitot tube in
skewed flow will be minimal. For example, the
maximum total pressure error in a duct with a
velocity pressure of 1 in.wg and a total pressure
of 20 in.wg is 5%. This error will occur with the
Pitot tube oriented 90° to the air stream. If the
flow is at 25° to the Pitot tube, the total pressure
error as indicated by Figure 5 is approximately
25%. This translates into an error in the total
pressure indicated by the Pitot tube of only
1.3%. However, for ducts with high velocities
and low total pressures, the error due to skewed
flows can become very significant. For a duct
with both the velocity and total pressures equal
to 1.0 in.wg, the error in total pressure reading
due to skewed flow can be as high as 100%.
Therefore, when assessing the possible error in
a field test, the relationship between the velocity
and total pressure must be taken into account.
velocity pressure. However, unlike the total
pressure error, the error in the static pressure
does not increase as quickly as the total
pressure error. The static tap in the Pitot tube
consist of 8 ports evenly spaced around the
perimeter of the Pitot tube at a location 8 tube
diameters downstream of the Pitot tube tip. With
flow axial to the Pitot tube, these static ports are
perpendicular to the flow stream and only sense
the static component of the flow. As the flow is
skewed, the ports facing the flow stream will see
an increase of pressure as flow begins to
stagnate on the upstream side of the tube.
Meanwhile, the ports on the sides and
downstream side of the tube will either see just
the static or a pressure below static as flow
starts to separate on the downstream side of the
tube. Figure 9 shows part of the results of a
Computation Fluid Dynamics (CFD) analysis
done by FLOWCARE on a Pitot tube subjected
to various degrees of skewed flow. The Pitot
tube in Figure 9 is positioned 45° to the flow
stream. While the condition shown in Figure 9 is
an extreme case, it illustrates the conditions
around the Pitot static ports. It can be seen from
Figure 9 that the area around the static pressure
ports is subjected to a wide range of pressures.
The net result will be flow into and out of these
ports.
The indicated static pressure from a Pitot tube in
skewed flow is limited by the magnitude of the
10
Figure 9 – Flow Past Static Ports with Pitot
45° to Flow Stream
The net effect will be measured static pressures
below the actual static pressure in the duct. The
CFD analysis, while not conclusive or definitive,
indicates that as the skew angle increases, the
error in the static pressure reading will reach a
limiting value. This limiting value will be
dependent on both the flow stream velocity and
flow angle.
6.0
Conclusions
• Due to concerns regarding the directionality
of the flow, a 5–hole probe was used to
measure both the velocity pressure and flow
direction at each traverse point. The results
of the 5-hole probe confirmed the presence
of skewed flows. A Pitot-static tube was also
used to measure fan performance. A
comparison between the two test methods
indicated that the Pitot-static tube gave
comparable results. Indications are that a
Pitot-tube can be used over a wider range of
skewed flows than originally thought.
• Errors in the total and static pressure
measurements obtained by a Pitot tube in
skewed flows are dependent on the actual
values of these pressures and the velocity
pressures in the duct. The error in the total
pressure will be less for ducts with higher
pressure and lower velocity pressures.
• It is not possible to obtain the true static
pressure at each traverse point in a skewed
flow using just a Pitot tube. Comparing the
Pitot static pressure with static pressure
measurements obtained by some other
means such as duct wall taps should yield
reasonably good results. A total pressure can
then be calculated using Pvz and the static
pressure obtained from side wall taps. The
total pressure calculated in this manner will
not be the true total pressure of the flow
stream as it will not contain the velocity
pressure component of the flow parallel to
the test plane. However, the error will be
relatively small.
• Calibrating Pitot-static tubes over a range of
Yaw/Pitch angles may allow for the
determination of the angle of a flow stream
relative to the Pitot tube axis. This
determination would be accomplished by
recording indicated Pt, Ps and Pv pressures
and comparing the error in the values
obtained by the Pitot tube against Ps and/or
Pt values measured by wall taps or total
pressure from a total pressure probes
unaffected by skew. Using the error between
the calculated and measured values and the
calibrated error curves it may be possible to
estimate the angle of the flow relative the
axis of the Pitot tube.
References:
Merriam, Kenneth, G., Spaulding, Ellis R.
‘Comparative Tests of Pitot-static Tubes’ NACA.
Technical Note No. 546, 1937.
Folsom, R. G., ‘Review of the Pitot tube’, ASME
Fluid Meters Research Committee, ASME, 1955
IP-142
AMCA Publication 203. Field Performance
Measurement of Fan Systems
AMCA Standard 803, Industrial Process/Power
Generation Fans: Site Performance Test
Standard
ANSI/ASME PTC 11-1984, ‘Performance Test
Codes, Fans,’
ASME PTC 19.5-2004, Performance Test Codes
Flow Measurement
11