&EPA
United States Environmental Monitoring and Support EPA-600 4-79-042
Environmental Protection Laboratory June 1979
Agency Research Triangle Park NC 27711
Research and Development
Angular Flow
Insensitive Pitot
Tube Suitable for
Use with Standard
Stack Testing
Equipment
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL MONITORING series.
This series describes research conducted to develop new or improved methods
and instrumentation for the identification and quantification of environmental
pollutants at the lowest conceivably significant concentrations. It also includes
studies to determine the ambient concentrations of pollutants in the environment
and/or the variance of pollutants as a function of time or meteorological factors.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
ANGULAR FLOW INSENSITIVE PITOT TUBE SUITABLE FOR USE
WITH STANDARD STACK TESTING EQUIPMENT
by
W. J. Mitchell, B. E. Blagun, D. E. Johnson, and M. R. Midgett
Quality Assurance Branch
Environmental Monitoring and Support Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring and Support
Laboratory, U.S. Environmental Protection Agency, and approved for publication.
Mention of trade names or commercial products does not constitute endorsement
or recommendation for use.
-------
ABSTRACT
Existing pitot tube designs were tested under various gas flow conditions
for accuracy in measuring static and total pressure. Then the static- and
impact-pressure measuring tubes least affected by angular flow were combined
and evaluated in the presence of particulate sampling nozzles. Tests were per-
formed on "S" pitot tubes, "L" pitot tubes, Kiel pitot tubes, cylinder pitot
tubes and a shielded static-pressure pitot tube. The percent error of each
pitot tube was determined against an ASME "L" pitot tube as a function of
yaw, pitch, orifice size, orifice location, pitot tube size, and velocity.
Secondary factors considered in determining the acceptability of a pitot tube
were: construction cost, ability to align with gas flow, ease of insertion
into small ports, tolerance to flow disturbance by a sampling nozzle, suita-
bility for use with a variety of in-stack and out-of-stack filter assemblies
popular in stack testing, and ability to be used in particulate and moisture
laden gas streams without plugging. The latter point was tested during field
tests at a sewage sludge incinerator, a clay crushing plant, and a power plant.
The Kiel pitot tube was found to be essentially error free (<5%) when yawed
and pitched ±30° even while attached to a standard EPA Method 5 sampling
assembly. Also included are: a summary of the present state-of-the-art in
sampling a stack with cyclonic flow, the errors involved in such sampling, a
recommendation for straightening cyclonic flow by insertion of a venturi throat,
and the effect of Reynold's Number on pitot tube accuracy.
iii
-------
CONTENTS
Abstract
List of Figures v-vi
List of Tables vii
1. Literature Review 1
2. Introduction 6
3. Experimental 7
4. Results and Discussion 11
Shielded Static Pressure Tube 11
"S" Pitot Tubes 12
"L" Shaped Pitot Tubes 13
Kiel Total Pressure Pitot Tubes 15
Cylinder Pitot Tubes 15
Kiel/Cylinder Pitot-Static Tube 16
Using Pitot Tubes to Determine Flow Direction and
Velocity in Angular Flow Situations 21
5. Summary and Conclusions 23
6. Recommendations 27
References 28
Appendices
A. Summary of Existing Data on "L" Pitot Tubes 30
B. Review of Present Methods for Sampling Stacks
with Cyclonic Flow 34
C. Pitot Tube Accuracy at Low Reynold's Number 46
IV
-------
FIGURES
Number Page
1 Merriam and Spaulding's standard "L" pitot tube 53
2 Kiel total pressure tube 53
3 Representation of yaw and pitch 54
4 Cylinder pitot-static tube , 55
5 Pressure distribution about a 25mm diameter cylinder 55
6 Standard "S" pitot tube 55
7 Effect of yaw on different "S" pitot tubes 56
8 Effect of pitch on different "S" pitot tubes 56
9 Design specification for "S" pitot tube after Vollaro 57
10 Relative position of pitot and sampling nozzle as recommended
by Williams and De Jarnette 57
11 Sample of pitch angle profile in stack with cyclonic flow 58
12 Sample of yaw angle profile in stack with cyclonic flow 58
13 Shielded static pitot tube 60
14 Shielded static pitot tube error curve for pitch at 8 m/sec .... 61
15 Effect of yaw on tube 5S 61
16 Effect of yaw on tube 10S 61
17 Variation of pitot coefficient with velocity for tube 1L 63
18 Yaw and pitch error curves for static pressure for tube 8L
at m/sec 63
19 Pitch error curve for static pressure for tubes 3L, 7L, and 8L . . 64
(continued)
-------
FIGURES
(continued)
Number
20
21
Error in
pitch
Relative
at 8
pressure
at 8 m/se
error in
m/sec . .
head
ac . .
for
pressure
tube
for
3L as
tube 1
function
of
K as function
yaw
of
and
pitch
Page
. . 64
. . 64
22 Calibration curve for tube IK as function of velocity 67
23 Calibration curve for "S" pitot 3-04 after Williams and
De Jarnette 67
24 Nozzle/pitot orientations 69
25 Combined effect of yaw and pitch on Kiel/cylinder pitot tube
attached to sampling probe .- 72
26 Orientation of Kiel/cylinder pitot tube with Alundum thimble ... 74
27 Variation of pitot coefficients with velocity for Alundum
thimble sampling assembly 74
28 Standard EPA sampling assembly for particulate 75
lA-a Absolute error in static pressure as a function of
orifice distance from tip of "L" pitot 76
lA-b Absolute error in static pressure as a function of
orifice distance from stem of "L" pitot 76
C-l Flow patterns downstream of a cylinder at high and
at low Reynolds' number 77
C-2 Effect of Reynolds' number of pitot coefficient after MacMillan . . 77
C-3 Correlation of pitot coefficient and Reynolds' number at 35°C
for "S" and Kiel/cylinder pitot tubes 78
C-4 Reynolds' number as a function of velocity for a 0.95 cm O.D.
pitot tube 78
vi
-------
TABLES
Number Page
1 Dimensions of "S" Pitot Tube Legs Studied 59
2 Dimensions of "L" Pitot Tubes Studied 59
3 Dimensions of Kiel Pitot Tubes Studied 59
4 Dimensions of Cylinder Pitot Tubes Studied 60
5 Effect of Yaw and Pitch on Pressure Measurement by
"S" Pitot Tube Legs 62
6 Effect of Yaw on Cylinder Pitot Tube Pressure Measurement 65
7 Effect of Pitch on Cylinder Pitot Tube Wake Pressure Measurement . . 66
8 Relative Effect of Pitch When Direction of Gas Flow is
Determined with Kiel/Cylinder Pitot Tube 68
9 Effect of Yaw on Pitot Coefficient for Wake Pressure Measurement
by Kiel/Cylinder Pitot Tube 70
10 Combined Effect of Yaw and Pitch on Coefficient of Kiel/Cylinder
Pitot Tube Attached to Particulate Sampling Assembly 71
11 Sensitivity of "S" and Kiel/Cylinder Pitot Tube Coefficients
to Nozzle/Thimble Interaction 73
VI1
-------
SECTION 1
LITERATURE REVIEW
The pitot tube, a device used to measure the velocity of fluids, operates
on the principle that a moving fluid exerts pressure on any object placed in
its path. This pressure, which is termed impact pressure or total pressure
(Pj) is the sum of the dynamic pressure (P.) and the static pressure (P_).
Dynamic pressure is a measure of the momentum of the molecules in the fluid
resulting from their movement in the overall direction of fluid flow, while
static pressure is a measure of the random motion of molecules that exists in
all fluids whether moving or at rest. By determining Pg and PT of a flowing
fluid using orifices located on the pitot tube, it is possible to determine
the velocity (V) of a fluid:
PT = Ps + Pd
Pd = PT . PS = * v2
V = [f (PT - Ps)f*
The first reported use of the pitot tube is credited to Henri Pitot, who
used it in 1732 to determine river flow as a function of water depth. Little
additional development work was done on the pitot tube until 1925 when Ower
and Johansen studied the effect of pitot tube shape and orifice location on
2
the measurement of velocity. Parts of their work were checked by Merriam
o
and Spaulding during a more extensive, but similar study reported in 1935.
-------
Merriam and Spaulding used their results and the results of Ower and Johansen
to design and evaluate a pitot/static tube that had a predictable degree of
error under most flow conditions. This pitot tube, shown in Figure 1, became
the basis for the ASME, ASTM and AVA pitot tubes. Unfortunately, they limited
their study to pitot tubes that were very similar in design to that shown in
Figure 1 and because of its long tip and small static orifices, this tube is
not generally applicable for stack testing.
Several other types of pitot tubes have been developed and studied
extensively. Perhaps the most unusual one is the Kiel total impact pitot tube
(Figure 2) which was developed and studied extensively by G. Kiel in 1935. An
outstanding feature of Kiel's tube is that it is accurate to within one percent
for yaw and pitch angles up to 40° over a wide velocity range. This tube has
since been modified by United Sensors, a commercial supplier of pitot tubes
(Figure 3), and is essentially error free for yaw and pitch angles up to 64°.
(For our purposes, pitot tube yaw and pitch are defined as shown in Figure 3.)
Another pitot tube that has been evaluated extensively is the cylinder
c
pitot tube in which the impact and static pressure orifices are located on the
same side of a cylindrical tube (Figure 4). One advantage of this pitot tube
is its easy insertion through small port holes. However, a major disadvantage
is that the exact location of the static pressure orifice with respect to the
impact pressure orifice must be determined with previous knowledge of the
viscosity, density, and approximate velocity of the fluid to be measured.
Also, this pitot tube is extremely inaccurate under conditions of angular flow,
that is, yaw, pitch and swirl (a combination of yaw and pitch). Like the Kiel
tube, pitot tubes of this type are commercially available.
-------
Hemke studied the pressure distribution on the circumference of cylinder
type pitot tubes and determined that the pressure on the back half of the tube
is essentially constant (Figure 5). This observation suggests that if an
orifice is located on the back of a cylinder, the pressure measured will be
insensitive to yaw angles of ±80° when the pitch error is zero. Hemke did
not evaluate the effect of pitch on this type of pitot tube.
The Staubscheibe or "S" pitot tube (Figure 6) has been studied exten-
sively. This tube has received wide application in stack testing for the
following reasons: 1) the distance between the two orifices can be as small
as 21 mm allowing easy insertion through small port holes, 2) it is easy to
construct; and 3) the large orifice size allows it to be used in a stack with
high particulate loading. Unfortunately, the design of the tube is not
standardized in relation to orifice spacing, orifice shape or the angle of
78 9
the bend (0 in Figure 6). Vollaro » and Williams and De Jarnette have shown
that these parameters do not affect the accuracy of an isolated "S" pitot tube
when the pitot tube is perfectly aligned with the flow, i.e., when pitch and
Q
yaw are both zero. However, Williams and De Jarnette have shown that the
orifice spacing can significantly affect the accuracy of this pitot tube when
it is not perfectly aligned with the flow (Figures 7 and 8). The downstream-
facing orifice measures a flowing stream experiencing disturbances from the
upstream-facing orifice. Thus, increasing the orifice spacing reduces this
Q
disturbance, making this pitot tube less sensitive to angular flow and to
Q
misalignment of the orifices. Vollaro has shown that misalignment of the
two pitot tube legs can be a significant source of error for pitot tubes
having an orifice spacing of 21 mm when the tube is perfectly aligned with
stack gas flow. The turbulent effect of the sampling assembly was not con-
sidered in these studies.
-------
Vcllaro , Williams and De Jarnette , and Gnyp10 have shown that the pitot
calibration coefficient of the "S" pitot tube is affected if the spacing between
either the adjacent edges of the nozzle and "S" pitot tube or the thermocouple
and "S" pitot tube is less than 18 mm. Vollaro's work was used to establish
construction guidelines for the EPA Reference "S" pitot tube (Figure 9).
De Jarnette and Williams have shown that this pitot tube is very sensitive to
g
error from pitot misalignment with respect to stack flow. They recommend:
1) that the spacing between the orifices be expanded to the widest practical
limit, 2) that the "S" pitot tube be inserted approximately 20 to 25 mm further
into the stack than the sampling nozzle (Figure 10), and 3) that the thermo-
couple be inserted approximately 20 mm behind the pitot tube orifices. Besides
making the velocity measurement less sensitive to interference from the nozzle
and the thermocouple, their sampling configuration allows the sampling assembly
to be inserted through the 76 mm (3 inch) diameter sampling ports frequently
encountered in stack testing.
As indicated, comprehensive studies have been done to characterize the be-
havior of several pitot tubes under conditions of angular flow. However, only
q
in the study by De Jarnette and Williams was an attempt made to characterize
the behavior of a specific pitot tube under conditions of swirling flow, that
is, a flow that has non-zero components of both yaw and pitch. (In the other
studies involving angular flow, the error curve for pitch was determined under
conditions of zero yaw and vice-versa.) Yet, recent work done by Phoenix ,
12 13
Peeler, and Lundgren shows that (1) angular flow in stacks has non-zero
components of both yaw and pitch; and (2) the percent contribution that each
component makes to the overall direction of flow changes with distance from the
-------
wall (Figures 11 and 12). In circular stacks, for example, both the yaw and
pitch components of flow increase as the distance from the wall increases.
Consequently, unless its error curves for yaw and pitch are identical or
completely independent of each other, a pitot tube cannot be used to measure
the velocity in field situations where cyclonic flow is encountered. For
q
example, Williams and De Jarnette have shown that the error curve for "S"
pitot tubes under conditions of cyclonic flow (a combination of yaw and pitch
error) are different from those obtained when only yaw or pitch is present,
even in the absence of a sampling nozzle. Thus, unless the true direction of
flow is known and the "S" pitot tube is perfectly aligned with this flow, it
cannot accurately measure the velocity. For this reason, the USEPA limits the
use of the "S" pitot tube to conditions of angular flow less than 10° from the
stack axis.
-------
SECTION 2
INTRODUCTION
Reported here are the results of a study that was done to evaluate the
behavior of different types of pitot tubes as a function of orifice size,
orifice location, velocity, angle of flow (yaw only, pitch only, yaw and pitch
combined), and presence of sampling nozzle. The following types of pitot tubes
were studied: Kiel total pressure pitot tube, "S" pitot tube, "L" pitot tube,
cylindrical pitot tube similar to that of Hemke and a shielded static-pressure
pitot tube. The primary objective of this study was to yield a pitot tube
design that would have exceptional tolerance to flow misalignment and nozzle
and thermocouple interference effects when used with a stack sampling assembly.
Each tube's behavior was studied as a function of yaw, pitch, and combined
conditions of yaw and pitch.
This study resulted in the development of an improved pitot design that
will measure the stack velocity within 5% accuracy under conditions as extreme
as a combination of 30° yaw and 30° pitch, even when attached to a stack
sampling probe. After development, this pitot tube was tested at three
stationary sources to determine its ability to function in dust laden stack
gases.
-------
SECTION 3
EXPERIMENTAL
TRANSIT ASSEMBLY
A standard theodolite (surveyor's transit) was used to generate the yaw,
pitch, and combined yaw and pitch error curves. The yaw and pitch functions
on the theodolite were graduated in 0.1 degree increments.
WIND TUNNELS
Two wind tunnels having similar characteristics, that is, test sections
approximately one meter square, each capable of generating a laminar flow
profile in the velocity range of 1 to 21 m/sec (3 to 70 fps) were used in this
study. Both wind tunnels used an ASME standard "L" pitot tube installed in
the test section to determine the true velocity in the test section. One wind
tunnel was equipped with a door that permitted the entire transit assembly to
be placed in the tunnel. Access to the test section at the other wind tunnel
was through 15 cm (6 inch) portholes. Therefore, it was necessary to locate
the body of the transit assembly outside this wind tunnel and insert the pitot
tube through the porthole.
-------
FIELD TEST SITES
Power Plant (Coal-fired, equipped with an undersized electrostatic precipitator)
The pitot tube was inserted downstream of the ESP where stack conditions
were as follows: 5% moisture, 150°C, particulate concentration between 400
and 1100 mg/m3.
Municipal Incinerator (sewage sludge, equipped with a wet scrubber)
The pitot tube was inserted downstream of the scrubber where stack
conditions were: 8% moisture, 42°C, particulate concentration approximately
200 mg/m3.
Clay Crushing Plant (equipped with baghouse)
The pitot tube was inserted downstream of the baghouse where stack
conditions were: 2% moisture, 35°C, particulate concentration approximately
100 mg/m3.
Source Simulator with Provisions for Cyclonic Flow
A 46 cm diameter experimental test stack equipped with turning vanes
to produce a cyclonic flow velocity profile in the test section was used to
evaluate the applicability of various pitot tubes for measuring flow direction
and velocity in stacks that have a cyclonic flow pattern.
PRESSURE MEASUREMENT SYSTEM
Inclined Manometer
A Dwyer Series 424 inclined manometer was used to measure pressure head
in all field tests.
Electric Manometer
A Datametrics Barocel Model 1023A Electronic manometer was used to measure
the pressure head in the wind tunnel studies.
8
-------
Voltmeter
The output of the Model 1023A manometer was measured using a Thermo-
Systems Model 1076 RMS voltmeter equipped with 0.1, 1, 10 and 100 second time
constant functions. The 10 second time constant function was used here.
PITOT TUBE DIMENSIONS
S-Pitot Tubes
As noted in the Literature Review Section the design of the "S" pitot
tube has not been standardized with respect to 9, 3, and x (Figure 6). Thus,
a study was done to evaluate the effect of 0, e, and x on the measurement of
impact and wake pressure. To achieve this, sets of "S" pitot tube legs were
constructed having the dimensions shown in Table 1. These pitot tube legs
were individually evaluated as a function of yaw and pitch error in an attempt
to explain the varying calibration coefficient* reported for "S" pitot tubes
in the literature.
"L" (Standard) Pitot Tubes
Eight "L"-shaped hemispherical tip pitot tubes having the dimensions
shown in Table 2 were studied. These tubes were obtained from United Sensor
and Control Corp., Watertown, MA.
Kiel Pitot Tubes
Three Kiel pitot tubes having the dimensions shown in Table 3 were
obtained from United Sensor and Control Corp., Watertown, MA, and their
applicability for measuring static and impact pressure was evaluated.
Shielded Static Pitot Tube
A shielded pitot tube having the dimensions shown in Figure 13 was
obtained from United Sensor and Control Corp., Watertown, MA, and its accuracy
in measuring static-pressure was evaluated.
9
-------
Cylinder Pi tot Tubes
Cylinder pitot tubes having the dimensions shown in Table 4 were evaluated
for their applicability in measuring both impact and static pressure.
10
-------
SECTION 4
RESULTS AND DISCUSSION
In the first part of this program, the impact-pressure and static-pressure
error curves for the five types of pressure sensing tubes described in the
Experimental Section of this report were constructed. The static- and impact-
pressure error curves were studied separately to determine the factors that
affected each type of pressure measurement. A major deficiency with most
previous studies to develop and evaluate pitot tubes for stack testing is the
fact that they studied only the pressure difference between the static- and
impact- pressure and thus combined the two error curves.
The results from the laboratory evaluation were used to select the best
pitot tube for further evaluation. In this latter phase, the ability of the
pitot tube to measure velocity when attached to a sampling probe was evaluated
as a function cf nozzle/pitot separation, yaw and pitch. Also, the ability of
this pitot tube to measure the velocity in particulate- and moisture-laden
stacks was evaluated.
The results from both phases of this study are described below as a
function of the individual types of pressure sensing tubes.
SHIELDED STATIC PRESSURE TUBE (FIGURE 13)
This particular tube has two static-pressure sensing orifices, one located
further upstream than the other. Under conditions of zero yaw and zero pitch,
the absolute static-pressures measured by the two orifices were significantly
11
-------
different from each other and from the ASME "L" pitot tube used as a reference.
For example, at 8 m/secs the upstream orifice measured a static pressure 17%
more negative than the "L" pitot tube and the downstream orifice measured a
pressure reading 7% more negative than the "L!! pitot tube. Further, the %
difference between the static pressure measured by these two orifices and the
"L" pitot tube varied widely when testing with velocities between 21 and 2 m/sec.
Figure 14 presents the relative error curve for pitch for each orifice in
relation to the pressure measured by that orifice at zero pitch and yaw at a
velocity of 8 m/sec. From Figure 14 it is obvious that the downstream orifice
is less sensitive to pitch, however, an acceptable reason for this has not been
determined.
"S" PITOT TUBES (TABLE 1)
Two "S" tubes, 5 S (9 = 45°, X = 25 mm, 3 = 45°) and 10 S (9 = 90°, X = 25
mm, 3 = 0) were evaluated extensively as a function of yaw angle (Figures 15 and
16). In addition, the effect of yaw and pitch on the impact-pressure measure-
ment was determined for all tubes under the following conditions: yaw = 0°,
pitch = 20, 0°, -20°; yaw = 10°, pitch =20°, 0°, -20°; and yaw = 20°, pitch =
20°, 0°, -20°. The accuracy of the static-pressure measurement was determined
for all tubes under the following conditions: pitch = 0°, yaw = ±90°, 180°.
The results are summarized in Table 5. For comparative purposes, the correspond-
ing impact and static pressures measured by the ASME "L" pitot tube installed in
the wind tunnel are also presented in Table 5. (The actual pressures measured
are presented in Table 5 to portray the differences in impact- and static-
pressure measured by different "S" tubes under identical flow conditions. For
example, compare the impact-pressure measured at zero yaw and pitch for each
"S" tube to the impact-pressure measured by the Reference "L" Pitot Tube.)
12
-------
Based on the results in Table 5 and in Figures 15 and 16, the following
conclusions can be made concerning the use of the "S" pitot tube for pressure
measurement: (1) decreasing X increases the magnitude of the wake pressure
measured by an "S" pitot leg; (2) changing 9 from 45° to 90° and 3 from 0°
to 45° does not significantly affect the wake pressure measurement; (3) no
definite correlation exists between either X, 3, or 9 and the accuracy of the
impact measurement (compared to the "L" pitot); and (4) the "S" pitot tube
cannot be used to give an accurate measurement of static pressure (compare the
static pressure measured by the "S" pitot tube legs at 90° and -90° with the
corresponding pressure measured by the "L" pitot tube).
These results suggest that subtle differences in the construction of "S"
pitot tubes can significantly affect the calibration coefficient, which may
explain why calibration coefficients of 0.9 to 0.7 have been reported for
isolated "S" pitot tubes. In addition, these results may explain why some "S"
Id.
pitot tubes meeting the specifications in EPA Reference Method 2 do not
have the predicted coefficient of 0.84 when "properly" attached to a sampling
probe.15'16 This is why all "S" pitot tubes should be calibrated with the
sampling nozzle in place.
"L" PITOT TUBES (TABLE 2)
The eight "L" pitot tubes studied differed in relation to stem diameter,
static orifice size and spacing, but had similar static orifice locations, all
between 2.6 and 3.3 tube diameters from the tip and the stem. Based on the
results of Merriam and Spaulding, these pitot tubes should read the static
pressure within 2% of true static, and as expected, all tubes were found to
have an error of less than 2% in the static pressure under conditions of zero
yaw and zero pitch. Also as expected from Merriam and Spaulding's work
13
-------
the three pitot tubes with both impact- and static-pressure orifices OU 2L,
3L) were found to have coefficients (Cp) between 0.99 and 1.00 when properly
aligned with the flow. (These coefficients were calculated on the assumption
that the coefficient for the ASME "L" pitot in the windtunnel test section was
0.99.) The effect of velocity on the coefficient for tube 1L is presented in
Figure 17. As expected, the coefficient varied only slightly with velocity.
Merriam and Spaulding designed their pitot-static tube to minimize the
effect of stem and tip on the static pressure measurement. Decreasing the
spacing between the static orifice and the stem and the tip to 3 tube diameters
should destroy this symmetry a fact confirmed through yaw and pitch studies
on tubes 3L, 7L and 8L. Since the results from the three tubes were similar,
only those for tube 8L are presented in Figure 18.
Within the experimental error of 1%, varying either the number of static
orifices in a row or the number of rows of static orifices did not significantly
affect the yaw and pitch error curves in the static pressure measurement in the
range ± 20°. However, as expected, tube 3L, which had only four static orifices
separated by 90°, was more sensitive to angular flow in excess of 20° than were
tubes 7L and 8L (Figure 19).
Merriam and Spaulding also designed their pitot tube to have offsetting
error curves for static- and impact-pressure measurement under conditions of
moderate angular flow. Thus, as their tube is yawed or pitched, the positive
error in the static pressure is somewhat offset by the negative error in the
impact pressure, so the pitot coefficient varies no more than ± 2% for yaw
3
and pitch errors as large as ± 30°. Figure 20 demonstrates that the pitot-
static tubes in Table 2 also have offsetting error curves for impact and static
pressure; but as expected, the deviation from the true value is larger for
these shortened "L" pitot-static tubes.
14
-------
KIEL TOTAL PRESSURE PITOT TUBES (TABLE 3)
Three Kiel tubes were evaluated in velocities ranging from 1 to 21 m/sec
to determine their suitability for measuring static as well as impact pressure.
The three tubes were found to measure the impact pressure to within ± 5% under
a condition as extreme as 40° yaw and 30° pitch. However, their ability to
measure the static pressure (in reality a wake pressure), which was evaluated
by facing the tube downstream, was not as good. For example, Kiel tubes IK and
2K, which had a venturi exit and entrance, measured the wake pressure within 5%
of the pressure measured by the tube when properly aligned with the flow but
only up to yaw and pitch angles of ± 15°. At yaw angles of ± 40° and pitch
angles of ± 30°, the relative error was as large as 10%. As expected, tube 3K
(which did not have a venturi entrance when faced downstream) was even more
sensitive to yaw and pitch effects.
Figure 21 presents the results of pitch on the static- and impact-pressure
measurement for tube IK. The curves for yaw error were essentially the same
and thus are not presented. Neither error curve was significantly affected by
velocity in the range 8 to 21 m/sec.
CYLINDER PITOT TUBES (TABLE 4)
The cylinder pitot tubes were evaluated for yaw and pitch error at
velocities of 1, 8 and 21 m/sec and found independent of velocity and orifice
size. The effect of yaw was evaluated by aligning the pitot tube in the wind
tunnel to obtain the maximum impact pressure (0°) and then measuring the
absolute pressure at 5° intervals as the tube was rotated through a 180° arc.
Analogous to Hempke's results6, the absolute pressure was found to vary less
than W, from 100° to 180°, but changed rapidly from 0° to 70°. Some represent-
ative results are presented in Table 6.
15
-------
The effect of pitch on the impact- and static-pressure measurement was
determined as follows: The pitot tube was aligned to obtain the maximum impact
pressure, the absolute pressure was then measured at 10° intervals from ± 40°,
the pitot tube was then returned to 0° pitch, rotated 180° and again pitched
± 40°. As expected, increasing the distance between the orifice and the tube
edge dramatically reduced the effect of pitch on the pressure measurement until
the orifice was located at least three tube diameters from the edge (Tables 4
and 7).
KIEL/CYLINDER PITOT-STATIC TUBE
Based on the above study, the Kiel pitot tube was selected as the best for
measuring impact pressure and the cylinder pitot tube was selected as the best
for measuring static pressure and for determining direction of gas flow. Since
these two tubes were essentially error free (<5%) when yawed and pitched ± 30°,
an attempt was made to combine them into one pitot-static tube. The first
approach involved simply piggybacking a cylinder pitot tube onto the sheath
(stem) of the Kiel tube. However, the presence of the Kiel tube stem made the
cylinder pitot tube much more sensitive to both yaw and pitch error much
like the impact leg of the "S" pitot tube significantly affects the error
curve for the wake-pressure leg.
The next attempt to combine these two pressure sensing tubes involved
placing a series of 3.2 mm orifices on the downstream side of Kiel tube IK at
points 19, 29, 38, 48, 57, and 66 mm from the center of the Kiel impact orifice.
The optimum location for the static orifice was then determined by plugging
all but one orifice and establishing the yaw and pitch error curves for that
orifice location. As expected, the results were similar to those received for
the isolated cylinder pitot tubes, i.e., increasing the distance between this
16
-------
and the Kiel tube shield dramatically improved the tolerance of the tube to
pitch effects until a point 38 mm from the impact orifice was reached. After
38 mm the improvement was modest. Based on this study, it was decided to
locate the wake pressure orifice on the stem of the Kiel tube approximately
38 mm from the center of the impact orifice.
Figure 22 presents the calibration curve (calculated relative to the ASME
"L" pitot tube) for tube IK when the wake pressure orifice is located 38 mm
from the center of the impact orifice and the tube is properly aligned with
the direction of gas flow. The double hump in the curve between 2 and 5 m/sec
is not an artifact of the wind tunnel because it was observed at both wind
tunnels. Similar fluctuations in the pitot tube curve have been reported for
Q in
"S" pi tot tubes (Figure 23) by Williams and De Jarnette , Gnyp , and Brooks
17
and Williams .
This Kiel/cylinder pitot-static tube was then studied to evaluate the
effect of pitch on the tube when it is used to detect gas flow direction when
the wake pressure orifice is faced downstream and used as an impact-pressure
orifice. Three orifice positions were evaluated, i.e., 19, 38, and 57 mm from
the center of the impact orifice. (All orifices were 3.2 mm in diameter.)
The results, which are presented in Table 8, show that the effect of pitch was
small up to ± 20° for all three positions.
The suitability of the Kiel/cylinder pitot tube for use with a stack
sampling probe was then evaluated using EPA Reference Method 5 sampling
probes18 equipped with 4.6 mm ID x 12.5 mm OD and 12.5 mm ID x 15 mm OD
gooseneck-shaped nozzles, and an EPA Reference Method 5 sampling probe equipped
with a 12.5 ID x 15 mm OD, 90°-bend sampling nozzle. The effect of yaw on the
velocity measurement in the velocity range 1-20 m/sec was evaluated as a
17
-------
function of probe sheath to pi tot sheath distance. During this study, the
entrances to the Kiel head and to the nozzle were in the same horizontal plane
and the centerline of the impact orifice was approximately 15 mm in front of
the vertical plane containing the centerline of the nozzle tip (Figure 24).
Also, the static orifice was located 38 mm behind the impact orifice. Probe
sheath to pitot sheath spacings (X in Figure 24) of 5, 8, 18, and 22 mm were
evaluated. The results for the two gooseneck-shaped nozzles are summarized
in Table 9. Results for the 90°-bend nozzle were similar to these and are
therefore not presented. (Pitch studies were not done because pitch is not a
significant source of error until combined with yaw).
Based on the results in Table 9, a probe/pitot sheath spacing of 18 mm
was selected as the best spacing and the combined effects of yaw and pitch on
sampling assemblies having this dimension were evaluated for the two gooseneck-
shaped nozzles and the 90°-bend nozzles described in Figure 24. The results
for the 12.5 mm ID x 15 mm gooseneck-shaped nozzle at 8 m/sec are summarized
in Table 10 and in Figure 25. The results for the gooseneck nozzle at 1.3 and
21 m/sec were similar to those in Table 10 and are not presented. Similarly,
the results for the 90°-bend nozzle differed only by a few percent from the
results for the gooseneck nozzle so only selected results are presented
(Figure 25). All results dramatically demonstrate how tolerant this sampling
assembly is to angular flow.
After completing the studies with the Method 5 probe, the suitability of
the Kiel/cylinder pi tot tube for use with an Alundum in-stack filter holder was
compared to that of the "S" pitot. First, the effect of thimble blockage on the
velocity measurement was evaluated by locating the impact orifice of the pitot
tube 25 mn in front of the nozzle centerline, a distance 225 mm in front of the
18
-------
leading edge of the thimble, and monitoring the impact and wake pressure reading
as the pitot was moved toward the thimble. As shown in Table 11, the Kiel/
cylinder pitot tube could be positioned much closer to the leading edge of the
thimble than could the "S" pitot tube before the velocity reading would be
affected. The pitot tube was then located at a point 15 mm in front of the
nozzle center!ine (Figure 26) and the effect of velocity on the "S" pitot and
Kiel/cylinder pitot tube coefficients determined (Figure 27). In this latter
study no attempt was made to determine the effect of yaw or pitch on the Kiel/
cylinder pitot tube/Alundum thimble assembly because: (1) the thimble and the
pitot were located far enough away from each other that blockage effects were
not present, and (2) the spacing between the nozzle and the pitot tube was
similar to that encountered when the Method 5 probe with 90° bend nozzle was
studied.)
JO 1Q
In the present EPA recommended sampling methods for particulate '
the orifices of the "S" pitot tube are located in the plane containing the
centerline of the nozzle (Figure 28). In contrast, the impact orifice for all
sampling configurations studied here was located approximately 15 mm in front
of this plane to accomplish the following: (1) reduce the chance for the Kiel
shield to perturb the particulate flow to the nozzle, (2) allow the pitot
sheath to be located closer to the probe sheath to decrease the overall width
of the sampling assembly, (3) reduce the chance for the nozzle to strike the
stack wall while testing a stack but still permit sampling in the area near
the wall, and (4) reduce the chance for the nozzle to affect the flow of gas
to the pitot tube.
Field Evaluation of the Kiel/Cylinder Pitot Tube
The performance of the Kiel/cylinder pitot tube under actual sampling
conditions was evaluated by inserting it along with an "S" pitot-static tube in
19
-------
stacks at a coal-fired power plant, a clay crushing plant calciner, and a
sewage sluge incinerator. Both pitot-static tubes operated without plugging
for the length of the test 16 hours at the incinerator, 15 hours at the
calciner, and three hours at the power plant. The actual stack conditions
during the field tests are described in the Experimental Section of this report.
Inclusion of a Thermocouple in the Sampling Assembly
Since the Kiel/cylinder pitot tube evaluated was not equipped with a
thermocouple, the best thermocouple location had to be selected on the basis of
the flow profile about the nozzle and pitot tube. Due to the insensitivity of
the pitot tube to flow disturbance, almost any position that does not interfere
with the flow in the vicinity of the wake orifice is acceptable. Four such
positions are:
1) Run the thermocouple inside the pitot sheath into the Kiel shield
directly behind the impact tube. This is the best location from the view of
thermocouple protection and measuring temperature of the stack gas at the point
of velocity measurement. Kiel pitot tubes with the thermocouple inside the
shield are commercially available, but cost approximately $20 more than those
without the thermocouple.
2) Run the thermocouple down the side of the probe sheath opposite the
pitot tube to a point about 25 mm behind the nozzle. (This location may make
it difficult to attach the nozzle to the sampling probe.)
3) Run the thermocouple through the 12 mm space between the probe and
pitot sheath to a point 25 mm behind the wake orifice and angle it toward the
nozzle.
4) Run the thermocouple along the pitot1s sheath to a point 25 mm behind
the wake orifice and angle it toward the nozzle.
20
-------
USING PITOT TUBES TO DETERMINE FLOW DIERECTION AND VELOCITY IN ANGULAR FLOW
SITUATIONS
The "S" and cylinder pitots were mounted in the surveyor's transit and
used to determine the direction of flow in a circular stack with cyclonic flow.
Fifteen points located at distances of 1.2, 3.8, 5.3, 7.8, 11.4, 16.5, 20.3,
22.9, 25.4, 29.2, 34.3, 37.6, 40.1, 42.6, and 44.7 cm along the stack's 46-cm
in diameter were surveyed. Flow direction was determined with the "S" pitot
using the standard null technique and with the cylinder pitot tube 2C (orifice
two tube diameters from end) by rotating it until the maximum impact pressure
with respect to barometric pressure was obtained.
In general, the two pitot tubes agreed in flow direction within 10° at each
of the six points nearest the walls of the stack, but differed by as much as 40°
for the three points in the center the region where the radial (pitch)
component of flow is largest. For example, the average angle of flow with
respect to the longitudinal axis was 36° and 50° for the cylinder and "S" pitot
tubes, respectively, based on a survey of all fifteen points. In contrast, the
respective values for the twelve points nearest the wall (the points that would
be surveyed if the stack were divided into six concentric zones of equal area)
were 56° and 51°.
A cojaparative test was also done to determine how the impact-pressure
measured by the Kiel pitot tube would compare to that measured by the "S" and
cylinder pitot tubes. Impact-pressure was determined at the 15 survey points
used in the flow direction study using the flow angles previously determined by
that pitot tube. (The Kiel pitot tube was rotated to the flow angle determined
by the cylinder pitot.) The results were as expected, i.e., all three tubes
compared favorably at the four points nearest the wall, but disagreed markedly
21
-------
the closer the survey point was to the center of the stack the points where
radial component of flow was significant.
22
-------
SECTION 5
SUMMARY AND CONCLUSIONS
The primary objective of this study was to develop a pi tot-static tube
possessing certain characteristics:
1. It should have a high tolerance to angular flow and to flow
disturbance;
2. It should be suitable for routine use with the variety of in-stack
and out-of-stack filter assemblies now used in stack testing;
3. The error curve should be predictable as a function of angular flow,
i.e., the error curve should not be affected by minor differences in pitot
construction parameters;
4. It must be able to establish the direction of gas flow;
5. It should be inexpensive to construct; and
6. Like the conventional "S" tube now being routinely used in stack
testing, it should not rapidly plug when being used in particulate- and
moisture-laden gas streams.
Different types of pitot tubes were evaluated according to the above
criteria. (The result* are summarized below.) On the basis of these findings,
a tube was developed and field tested that met the performance specifications:
the Kiel/cylinder pitot tube.
23
-------
SHIELDED STATIC PRESSURE TUBE
The static pressure measured was inaccurate and varied with gas velocity
even when properly aligned with the direction of gas flow.
SHORTENED "L" PITOT TUBE
Decreasing the distance between the static holes and the tip and stem
allowed this type pitot to be inserted through standard size portholes. It
measured velocity accurately under ideal flow conditions of zero yaw and pitch,
but the symmetry in the error curves for yaw and pitch, present in the standard
size "L" pitot, was lost when the pitot was shortened. Compared to the
standard "L" pitot tube, the magnitude of the error curve for the shortened
"L" pitot tube was five times larger in the range ± 30° yaw and pitch. However,
these results do show that the shortened "L" pitot-static tube can be used in
source testing when parallel flow situations are encountered. It will also
measure velocity accurately in the range of 1 to 3 m/sec and can be used to
yield an accurate static pressure.
"S" PITOT TUBE
Found to be highly sensitive to angular flow. Subtle differences in
construction between two "S" pitot tubes can radically alter the behavior of
this pitot tube. It does not give an accurate static pressure when the orifice
face is placed parallel to the direction of stack gas flow.
CYLINDER PITOT TUBE
Found to be useful for determining stack gas flow direction when used as
an impact-pressure tube. When used to measure static pressure, in terms of a
24
-------
wake pressure, the tube is essentially error free for yaw angles up to ± 90°
and pitch angles up to ± 20°. When used to measure static pressure it has a
coefficient of 0.74.
KIEL IMPACT-PRESSURE TUBE
Found to accurately measure impact-pressure within ± 2% under a condition
as extreme as a combination of 40° yaw and 40° pitch. The Kiel tubes studied
had coefficients of 0.99 ± 0.01. These tubes can also be used to measure
static pressure, in terms of a wake pressure, when properly constructed.
KIEL/CYLINDER PITOT-STATIC TUBE
This tube combines the best features of the Kiel and cylinder pitot tubes
and meets all the performance characteristics specified above.
(1) Its coefficient varies less than 5% when attached to a standard
particulate sampling probe and the entire assembly is rotated and tilted to
yield conditions as extreme as ± 30 yaw and pitch.
(2) Its coefficient is unchanged when switched from an in-stack filter
probe equipped with a right angle bend nozzle to a standard Method 5 probe
equipped with a gooseneck-shaped nozzle. The in-stack filter assembly can be
located closer to this pitot tube than to the "S" pitot tube before a velocity
disturbance is observed.
(3) It will pass through 2-inch diameter sampling ports even when attached
to a standard Method 5 sampling probe.
(4) It can be used to establish direction of gas flow, has a larger
coefficient than t^e "S" pitot tube (0.74 compared to 0.85 for the "S" pitot)
and its performance in stacks with a high moisture and particulate concentra-
tion is equivalent to that of the "S" pitot.
25
-------
(5) It is relatively inexpensive to construct the 0.7 meter long
Kiel/cylinder pitot tubes used in this study were procured commercially for
$50.00 each.
(6) It can be used to accurately measure static pressure, in terms of a
wake pressure, in a stack, i.e., multiply the wake pressure reading by 0.74.
This tube was evaluated only in the velocity range 1 to 22 m/sec, but it
should perform well up to at least 40 m/sec, based on the reported performance
of the individual Kiel and cylinder pitot tubes. Like all pitot tubes, this
pitot tube should be checked before each use to be sure that the welds are
intact as improper welds may cause erroneous pressure measurements.
26
-------
SECTION 6
RECOMMENDATIONS
Based on the experiments conducted during this study, the Kiel/cylinder
pitot tube should be suitable for use in stack testing. It is more accurate
than the "S" pitot under conditions of moderate and severe angular flow and
is less sensitive to interference by the sampling nozzle. The suitability of
the Kiel/cylinder pitot-static tube should be evaluated further by routine
users of stack testing equipment to confirm that the conclusions of this study
are correct.
Recommended sampling assembly dimensions are shown in Figures 24a, 24b,
and 26 for gooseneck nozzle, 90° bend nozzle and in-stack thimble, respectively.
For all assemblies the edge to edge spacing between the probe sheath and the
pitot sheath (X in Figures 24a and 24b) should be at least 18 mm and the wake
orifice should be located at least 25 mm behind the impact orifice. Suggested
positions for the thermocouple are presented in the text of this report.
If desired, a small Kiel head can be used, e.g., the Kiel head on Tube 2K
in Table 3 performed the same as Tube IK in the laboratory study. Use of this
smaller Kiel tube would decrease the overall length and width of the sampling
assembly, but at the same time, it would increase the chance for the pitot tube
impact orifice to be plugged by particulate. Of course, pitot tube to nozzle
spacings different from those in Figures 24 and 26 may also be acceptable
provided care is taken to ensure that the pitot shield does not interfere with
the pollutant flow to the nozzle.
27
-------
REFERENCES
1. Folson, R. G. Review of the Pitot Tube. Trans. Am. Soc. Mech. Eng.,
78:1447-1459, 1956.
2. Ower, E., and F. C. Johansen. Design of Pitot-Static Tubes. R and M,
No. 981, British Aeronaut. Res. Council, 1925.
3. Merriam, K. F., and E. R. Spaulding. Comparative Tests of Pitot Static
Tubes. Nat. Advis. Comm. Aeronaut., Tech. Note 546, 1935.
4. Kiel, G. Total Head Meter with Small Sensitivity to Yaw. Nat. Advis.
Comm. Aeronaut., Tech. Note 755, 1935.
5. Winternitz, F. A. L. Cantilevered Pitot Cylinder. The Engineer, 729-732,
May 27, 1955.
6. Hemke, P. E. Influence of the Orifice on Measured Pressures. Nat. Advis.
Comm. Aeronaut., Tech. Note 250, 1926.
7. Vollaro, R. F. Use of Type-S Pitot Tubes for the Measurement of Low
Velocities. U.S. Environmental Protection Agency, Research Triangle Park,
NC. In-house Report, January 19, 1977.
8. Vollaro, R. F. A Survey of Geometric and Aerodynamic Factors Which Can
Affect Type-S Pitot Tube Accuracy. U.S. Environmental Protection Agency,
Research Triangle Park, NC. In-house Report, February 17, 1978.
9. Williams, J. C., and F. R. De Jarnette. A Study on the Accuracy of Type S
Pitot Tubes. EPA 600/4-77-030. U.S. Environmental Protection Agency,
Research Triangle Park, NC, 1977. 70 pp.
10. Gnyp, A. W., D. S. Smith, D. Mozzon, J. Steiner. S-Type Pitot Tube
Coefficients - An Experimental Investigation of the Effect of Pitot-Tube
Sampling Probe Configurations on the Magnitude of the S-Type Pitot Tube
Coefficient for Commercially Available Source Sampling Probes. University
of Windsor, Windsor, Canada. February 1975.
11. Phoenix, F. J., and D. J. Grove. Cyclonic Flow-Characterization and Recom-
mended Sampling Approaches. Interim report prepared for U.S. Environmental
Protection Agency, Division of Stationary Source Emissions, Washington, DC,
under Contract 68-01-4148. February 1978
12. Peeler, J. W. Isokinetic Particulate Sampling in Non-Parallel Flow
Systems - Cyclonic Flow. Unpublished report prepared by Entropy
Environmentalists, Inc., Raleigh, NC, 1978.
28
-------
13. Lundgren, D. A., M. D. Durham, and K. W. Mason. Sampling of Tangential
Flow Streams. Am. Ind. Hyg. Assoc. J., 39:640-644, 1978.
14. U.S. Environmental Protection Agency. Standards of Performance for New
Stationary Sources. Federal Register, 42:41758-41768. August 18, 1977
15. Walz, L. R., and J. L. Hatheway. Effects of Probe Blockage on Type-S
Pi tot Tubes in Small Diameter Ducts. Paper 78-35.7. Presented at
National Conference of Air Pollution Control Association, Houston, TX,
June 25-30, 1978.
16. Leland, B. J., J. L. Hall, A. W. Joensen, and J. M. Carroll. Correction
of S-Type Pitot-Static Tube Coefficients When Used for Isokinetic Sampling
from Stationary Sources. Environ. Sci. Techno!., 11:694-700, 1977.
17. Brooks, E. F., and R. L. Williams. Technical Manual for Process Stream
Volumetric Flow Measurement and Gas Sample Extraction Methodology. Report
prepared for U.S. Environmental Protection Agency, Research Triangle Park,
NC, under Contract 68-02-1412, November 1975. 90 pp.
18. U.S. Environmental Protection Agency. Standards of Performance for New
Stationary Sources. Federal Register, 42:41777. August 18, 1977
19. U.S. Environmental Protection Agency. Standards of Performance for New
Stationary Sources. Federal Register, 43:7575. February 23, 1978
29
-------
APPENDIX A
SUMMARY OF THE EXISTING DATA ON "L" PITOT-STATIC TUBES
The Merriam and Spaulding "L" pitot-static tube is not generally suitable
for stack testing because (1) its small static orifice holes plug easily, and
(2) its long head makes its insertion difficult in many stacks. However, this
study as well as other studies show that there may be applications where a
scaled1down or otherwise modified Merriam and Spaulding type "L" pitot-static
tube may be the best choice. Fortunately, there is an extensive body of data
available that allows one to predict the approximate effect of many design
modifications on an ISOLATED "L" pitot-static tube, i.e., one not attached to
a sampling probe. The existing body of data on the "L" pitot-static tube is
summarized below as an aid to those interested in possibly using this type of
pitot-static tube.
STATIC PRESSURE ORIFICE
1) At the velocities encountered in source testing the shape (hemispheri-
cal, square, tapered, ellipsoidal) of the tip does not significantly affect the
static pressure measurement as long as the static orifice is greater than two
tube diameters from the tip.
2) Under conditions of zero angular flow, increasing the orifice size
from the standard of 1.0 mm (0.04 inches) up to 2.0 mm (0.08 inches) does not
decrease the accuracy of the static error measurement, at least up to tube
diameter to orifice diameter ratios of 10. However, as the orifice size
30
-------
increases the magnitude of the error for yaw and pitch increases, probably
due to increased gas circulation through the static holes.
3) The distance of the orifice from the stem and from the tip can be
used in conjunction with the work of Merriam and Spaulding to estimate the
absolute error in the static pressure measurement at zero angular flow.
(Figures A-l, A-2)
4) If static orifices are symmetrically located at a maximum spacing of
90° on the circumference of the tube, and the stem is at least five tube diame-
ters from the static orifices, the yaw and pitch error curves should be similar.
5) If the static orifice holes are asymmetrically distributed on the
circumference of the tube, the yaw and pitch error curves will not be similar.
6) The ratio of the orifice depth to orifice diameter can be varied
from 0.5 to 6 without a significant error in the static pressure measurement
at conditions of zero angular flow. However, some data available indicates
that the sensitivity of the static orifice to angular flow error is decreased
as the depth to diameter ratio increases, possibly because flow through the
orifice is reduced. (This same effect could possibly be obtained by making
the angular spacing between the inner and outer tube as small as possible for
the combined pitot/static tube, but this would also increase the time lag in
the static pressure measurement.)
IMPACT-PRESSURE ORIFICE
1) The sensitivity of the impact orifice to yaw and pitch is lessened
when the orifice size is made as large as the tube will allow.
2) Cylindrical probes with square ends seem to have the same relative
error as tubes with tapered and hemispherical ends under conditions of zero
angular flow.
31
-------
3) The impact orifice should remain unchanged for at least three tube
diameters if an accurate pressure is to be measured, i.e., the orifice opening
cannot change in size for a depth of at least three tube diameters from the
opening.
4) The impact orifice diameter should be at least 3.17 mm (0.125 in.)
to avoid excessively long time constants in the pressure measurement.
5) Reynolds Number effects may be evaluated more accurately, if the
impact orifice size and not the tubing size is used in the equation for Rg.
SPECIAL NOTES ABOUT USE OF "L" PITOT STATIC TUBES
The above data applies to isolated "L" pitot-static tubes, that is, those
not attached to any sampling probe. If an "L" pitot-static tube will be used
with a sampling nozzle attached, extreme care should be taken to ensure that
turbulence between the nozzle edge and the pitot-static tube edge does not
cause an erroneous static pressure measurement. This situation would be
particularly serious under conditions of angular flow, subisokinetic sampling,
and superisokinetic sampling. To avoid this potential error simply attach the
pitot-static tube tip to the probe so that it is at least 15 mm further into
the stack than the sampling nozzle.
Also, the "L" pitot-static tube is sensitive to instrument vibration.
If a small diameter "L" pitot tube is to be used in a stack containing high
velocities or turbulence, extreme care should be taken to ensure that the "L"
pitot tube does not vibrate. This can easily be achieved by using large
diameter tubing for the pitot stem or by attaching the pitot tube directly
to the sampling probe.
32
-------
BIBLIOGRAPHY
1. Poison, R. 6. Review of the Pitot Tube. Trans. Am. Soc. Mech. Eng.,
78:1447-1459,1956.
2. Gracey, W. B. Measurement of Static Pressure on Aircraft. Nat. Advis.
Comm. Aeronaut., Rep. 1364, 1958.
3. Hemke, P. E. Influence of the Orifice on Measured Pressures. Nat. Advis.
Comm. Aeronaut., Tech. Note 250, 1926.
4. Kettle, D. J. Design of Static and Pitot Static Tubes for Subsonic Speeds.
J. Roy. Aeronaut. Soc., 58:835-837, 1954.
5. Merriam, K. F., and E. R. Spaulding. Comparative Tests of Pitot Static
Tubes. Nat. Advis. Comm. Aeronaut., Tech. Note 546, 1935.
6. Ower, E., and F. C. Johansen. Design of Pitot-Static Tubes. R and M,
No. 981, British Aeronaut. Res. Council, 1925.
7. Ower, E., and R. C. Pankhurst. The Measurement of Air Flow. Pergamon
Press, New York, NY 1966.
8. Shaw, R. The Influence of Hole Dimensions on Static Pressure Measurements.
J. Fluid Mech., 7:550-564, 1960.
9. Schulze, W. M., G. C. Ashby,and J. R. Erwin. Several Combination Probes
for Surveying Static and Total Pressure and Flow Direction. Nat. Advis.
Comm. Aeronaut., Tech. Note 2830, 1952.
33
-------
APPENDIX B
REVIEW OF PRESENT METHODS FOR SAMPLING STACKS WITH CYCLONIC FLOW
INTRODUCTION
Ideally, stack sampling should only be done where the gas flow is parallel
to the stack's longitudinal axis. (Under these conditions, it is assumed
that the pollutant profile is similarly aligned.) Most recently constructed
stationary sources have parallel flow downstream of their control device, but
it is not uncommon to find non-parallel flow upstream of the control device.
Also, at many old and/or small sources non-parallel flow is encountered both
upstream and downstream of the control device. There are several common
reasons why non-parallel flow is encountered at these latter sources. First,
they have been retrofitted with a pollution control device that was installed
using short lengths of duct and sudden changes in direction to accommodate
space restrictions. Second, for space and economic reasons, these plants
use inertia! demisters, cyclones, and common stacks as part of their pollution
control system.
As more and more of these older and smaller sources come under emission
regulations with revisions of State Implementation Plans, non-parallel flow is
being encountered more frequently in compliance testing. This fact may explain
why articles suggesting how to sample in stacks with cyclonic (non-parallel)
flow are starting to appear in the literature.
34
-------
During the development of the angular flow insensitive pitot tube described
in the body of this report, numerous articles on angular flow were reviewed.
This appendix describes the present state-of-the-art in sampling cyclonic flow
and THE AUTHORS' VIEWS ABOUT THE ERRORS PRESENT when measuring velocity, volu-
metric flow, and pollutant emission rate in cyclonic flow situations. The three
major areas that will be discussed are: selecting sampling points, determining
flow direction and volumetric flow, and measuring pollutant concentration. The
possibility of using a venturi to straighten cyclonic flow is also proposed.
SELECTING SAMPLING POINTS IN STACK TESTING
The customary approach to determine stack gas velocity and volumetric flow
810
in stacks is to divide the stack into a number of equal area zones. ~ For
circular stacks, this involves dividing the stack into equal area zones and
measuring the velocity at the centroid of each zone. Consequently, the sampling
points are unequally spaced on the traverse diameter, i.e., skewed toward the
wall. Rectangular-shaped stacks are divided into several equal area zones that
are geometrically similar in shape to the full stack. The velocity is then
measured at the center of each rectangular-shaped zone. In both cases, the
volumetric flow is determined by multiplying the average velocity for all zones
across the stack by the stack cross-sectional area.
Underlying this stack sectioning technique is the assumption that the
velocity at the centroid of each zone is a measure of the average velocity in
that zone. But as Ower and Pankhurst point out, there is no a priori reason
why this assumption should be valid, particularly when the flow is not
parallel to the stack axis.
35
-------
When angular flow is encountered, particularly cyclonic flow, studies show
that it will contain radial (pitch) and tangential (yaw) flow components that
change with distance from the wall3'5'6'11"14 For example, Phoenix has found
that in circular stacks with tangential gas flow inlets, the yaw angle (yaw
component of flow) increased from 30° near the wall to 90° at the center, and
the pitch angle (pitch component of flow) varied from 0° near the wall to 30°
near the center. Since the true direction of gas flow changes rapidly across
the stack, it is less likely that the velocity in the centroid of the zone
represents the average velocity across that zone. This source of error can
be reduced by narrowing the size of each zone, but this increases the number
of points that must be sampled, thus increasing the time and effort required
to sample the stack.
One common compromise made for the sake of time is to sample 48 velocity
points, but only 16 or 24 pollutant points. While this improves the relia-
bility in the volumetric flow measurement, it does nothing to improve the
reliability of the pollutant concentration measurement. Further, in cyclonic
flow there are reasons why non-gaseous pollutants may not follow the velocity
distribution in the stack (see discussion below).
IN SUMMARY, SELECTION OF SAMPLING POINTS IN CYCLONIC FLOW IS A TENUOUS
SITUATION AND THE MAGNITUDE AND SIGN (POSITIVE OR NEGATIVE BIAS) OF THE ERROR
IS NOT PREDICTABLE.
DETERMINING VELOCITY AND VOLUMETRIC FLOW IN CYCLONIC FLOW SITUATIONS
As mentioned, cyclonic flow is characterized by three flow vectors
axial, radial, and tangential and its direction can only be accurately
11 1? i ^ i ^
measured using direction-sensing pitot tubes. ' ' ' However, these
36
-------
pitot tubes are fragile, employ small, easily-plugged orifices and use a
tedious null technique to establish flow direction. Consequently, they
generally are not suitable for measuring flow in stack testing. Because of
this, it is now customary to use the non-directional "S" pitot tube to estab-
lish stack gas flow direction and volumetric flow in angular flow situations.
Flow direction is determined by rotating (yawing) the "S" pitot until a null
reading is obtained and then determining the angle, 9, between the pitot tube
and the stack longitudinal axis. The volumetric flow is then determined after
first multiplying the velocity component in the direction of stack gas flow by
the cos 9 (to yield the axial flow component). It is incorrect to determine
flow direction by rotating the "S" pitot until the maximum pressure drop is
obtained as was done by Lundgren.
Velocity and volumetric flow measurements may be seriously in error when
determined by the null technique because the radial (pitch) flow component
is ignored. Obviously, if the angle and magnitude of flow are measured
incorrectly, the true volumetric flow will also be in error. Since "S" pitot
tubes with different orifice spacing are affected to a different extent by
pitch error, it is difficult to generalize whether the measurements will
be biased high or low.
In contrast, the Kiel/cylinder pitot tube measures the true velocity head
when aligned within ± 40° of the true direction of flow. It also gives a
better estimate of the true flow direction because it is less sensitive to yaw
and pitch error than the "S" pitot. However, it does not measure the actual
pitch component and because of this the volumetric flow measurement will be in
error.
37
-------
The Kiel/cylinder tube should be used to determine flow direction as
follows:
1) Invert the pitot tube to use the wake pressure orifice as an impact
orifice.
2) Connect this orifice to the appropriate leg of a manometer and vent
the other side of the manometer to the atmosphere. (Although the velocity
head is always positive, impact pressure may be negative or positive with
respect to barometric pressure depending on the location of the fan relative
to the measuring point.)
3) Rotate the pitot tube at each sampling point until the Maximum
pressure (versus atmospheric pressure) is obtained and the angle, 9, recorded.
4) After completing the above flow-direction traverse, use the Kiel/
cylinder pitot tube to obtain the true velocity head at each point.
5) Calculate the volumetric flow using the appropriate value of cos 9
and the average velocity across the stack. (Obviously this volumetric flow
would be in error to the extent that the true 9 was not measured and the
selection of velocity sampling points was erroneous.)
Another common approach in dealing with cyclonic flow is to straighten
the flow using straightening vanes and stack extensions before measuring the
fi Q
velocity. ' Although this may not straighten the flow completely, it can
make the flow significantly more unidirectional. However, straightening vanes
can accidently bias the sampling results by removing particulate through
impaction or agglomeration and by changing the efficiency (performance) of
the control device.
ANALOGOUS TO THE SITUATION OF SELECTING THE SAMPLING POINTS, THE MAGNITUDE
AND SIGN OF THE ERROR ARE NOT PREDICTABLE WHEN MEASURING VELOCITY AND VOLUMETRIC
38
-------
FLOW IN A CYCLONIC FLOW SITUATION. Through care in instrument selection and
use of a tedious procedure, the relative error can be reduced, but its magnitude
and sign may still not be certain.
MEASURING POLLUTANT EMISSION RATES IN CYCLONIC FLOW SITUATIONS
When measuring pollutant concentration and emission rate in cyclonic
flow situations two general approaches are used: 1) straighten the flow
mechanically and hope that this also straightens the pollutant flow; and
2) attempt to adapt existing sampling equipment to the character of cyclonic
2
flow. Standard sampling equipment changes include: 1) using special probes;
2) rotating (yawing) the probe to align the nozzle with the tangential vector
of flow; and 3) aligning the nozzle with the longitudinal axis of the stack
and correcting the nozzle area for the reduction in area caused by the angular
l 4
flow. ' As described below, both approaches bias the test results in an
unpredictable way. But in general, straightening the stack flow whenever
possible may be superior to attempting to make existing equipment function
satisfactorily in a cyclonic flow situation.
The three techniques generally used to straighten flow (straightening
vanes, stack extensions equipped with straightening vanes, installing a duct
on the side of the existing stack at the angle the gas is exiting the stack)
have the following disadvantages associated with their use. First, they can
be expensive to install, particularly for a single test. Second, they may
change the performance of an inertial device such as a cyclone by causing
re-entrainment of particles or droplets near the wall or by removing them
through impaction or agglomeratio. Third, there is no a priori reason
why the pollutant flow should follow the same path as the straightened flow,
39
-------
so alignment of the nozzle with the straightened flow may still yield erroneous
results. Fourth, space restrictions may make it difficult to sample the
modified stack, particularly when a stack extension is used.
The three techniques now generally used to sample in cyclonic flow
situations without employing flow straightening devices have been described
in detail elsewhere '' and will only be summarized here. The simplest
approach, termed the Blindman's Approach, involves ignoring all angular flow
components and sampling with the nozzle and pitot aligned with the longitudinal
axis of the stack. The second approach, the Compensation Approach, also
involves aligning the pitot and nozzle with the stack longitudinal axis, but
the sampling rate is corrected for the reduction in nozzle area that is caused
by the angular flow, i.e., the nozzle is multiplied by the cos 9. In the third
approach, termed the Alignment Approach, the nozzle and pitot are aligned with
the direction of stack gas flow, which is determined either from a previous
traverse of the stack or by rotating the sampling assembly to obtain the
minimum pressure head. Although more complex than the other approaches, this
latter approach will still be in error to the degree that the direction of
pollutant flow deviates from the direction of stack gas flow. All three
approaches will be in error to the extent that the radial (pitch) component
of flow affects the velocity measurement and thus the isokinetic sampling rate.
Both the Blindman's Approach and Compensation Approach will also be biased to
the extent the misaligned nozzle perturbs the velocity and pollutant flow
upstream of the sampling assembly. For example, this disturbed flow may
cause particulate to move toward or away from the nozzle entrance.
One further problem in attempting to sample cyclonic flow without first
straightening the flow is caused by the fact that nongaseous pollutants will
40
-------
tend to move toward and concentrate at the wall.4'7'14 This requires
sampling in a turbulent region, which will further aggravate the problem of
obtaining a representative pollutant sample.
IN SUMMARY, IN CYCLONIC FLOW SITUATIONS IT IS IMPOSSIBLE TO PREDICT THE
MAGNITUDE AND SIGN OF THE ERROR ASSOCIATED WITH EACH POLLUTANT SAMPLING APPROACH
because the sources of error are confounded, i.e. not independent of each other.
Thus, attempting to say that one approach will yield a positive bias and another
a negative bias is tenuous.
USING A VENTURI TO STRAIGHTEN CYCLONIC FLOW
As discussed above, the size and sign of the error associated with
standard approaches to sampling cyclonic flow are unpredictable. However, it
is still necessary to sample stacks with cyclonic flow and although other more
exotic sampling approaches can be used, ' the improvement in the relia-
bility of the result may be insignificant with respect to the additional cost
involved.
However, there is one simple, practical, well-characterized device that
may straighten cyclonic flow at low cost without also causing a severe pressure
drop. In addition, this device a venturi installed in the stack or in a
stack extension may allow stacks with very low velocity to be sampled using
standard sampling equipment. The practicality of using an in-stack installed
18
venturi has been demonstrated by Thompson on alfalfa dehydrators.
Although it does not seem to have been evaluated outside the alfalfa in-
iQ ?n ?i
dustry, the available literature '' indicates that the in-stack venturi
would have the following advantages:
19
1) Easy and flexible design criteria.
41
-------
2) Ability to straighten the flow to meet EPA Method 1 specifications
JP
within one duct diameter after the diffusion section.
3) Would permit a decrease in the number of sampling points required
compared to sampling the unstraightened flow. This would decrease the time
required for a test run and at the same time improve the reliability of the
velocity and volumetric flow measurement. It would also allow standard test
equipment to be used without modification.
4) Would permit flow in a rectangular duct to be converted to laminar
flow in a circular duct in a very short distance.
5) Can increase the velocity in a stack if size of the diffuser cone is
reduced compared to the entrance cone (most pi tot tubes are inaccurate below
3 m/sec).
6) Can cause particulate concentrated near the wall to redistribute it-
self more evenly across the stack without loss of particulate through impaction.
Thus, it would seem prudent for researchers to start evaluating the use
of the venturi for straightening cyclonic flow. The following areas should be
investigated:
1) The effect on particulate distribution and concentration.
2) The maximum angle that can be tolerated in the diffusion section
21
before the onset of turbulance. (Gibson has found that in liquids this
angle can be as large as 40° if the inlet to the diffuser is curved. This
alone is a significant improvement over the original venturi design limitation
of 7°.)
3) The variation in the convergent section angle that can be tolerated
before the onset of turbulent flow in the venturi throat.
42
-------
4) The development of adjustable (in length and angle) convergent and
diffuser sections for use with several standard sized venturi throats.
5) The effect of venturi design on the volumetric flow exiting the stack.
43
-------
REFERENCES
1. Phoenix, F. J., and W. S. Smith. Isokinetic Particulate Sampling in Non-
Parallel Flow Systems A Theoretical Solution. Stack Sampling News,
6(9):5-8, 1978.
2. Test Procedure for Cyclonic Flow. Connecticut Department of Environmental
Protection. Method 8, FE 409.
3. Phoenix, F. J., and D. J. Grove. Cyclonic Flow-Characterization and
Recommended Sampling Approaches. Interim report prepared for U.S.
Environmental Protection Agency, Division of Stationary Source Emissions,
Washington, D.C., under Contract 68-01-4148. February 1978.
4. Goerner, C. L., F. H. Hartman, and J. B. Draper. A Method for Stack
Sampling Cyclonic Flow. Paper 78-35.2. Presented at the National
Conference of Air Pollution Control Association, Houston, Texas,
June 25-30, 1978.
5. Hanson, H. A., and D. P. Saari. Effective Sampling Techniques for
Particulate Emissions from Atypical Stationary Sources Interim Report,
EPA-600/2-77-036. U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina 27711. February 1977. 120 pp
6. Peeler, J. W. Isokinetic Particulate Sampling in Non-Parallel Flow
Systems Cyclonic Flow. Unpublished report prepared by Entropy
Environmentalists, Inc., Raleigh, North Carolina, 1978.
7. Lundgren, D. A., M. D. Durham, and K. W. Mason. Sampling of Tangential
Flow Streams. Am. Ind. Hyg. Assoc. J., 39:640-644, 1978.
8. Jorgensen, R., ed. Fan Engineering. Chapter 3. Buffalo Forge Co.,
Buffalo, New York, 7th edition, 1970.
9. Ower, E., and R. C. Pankhurst. The Measurement of Air Flow. Pergamon
Press, New York, New York, 1966. pp. 117-131.
10. U.S. Environmental Protection Agency, Standards of Performance for New
Stationary Sources. Federal Register, 42:41754-41768. August 18, 1977.
11. Fabris, G. Probe and Method for Simultaneous Measurements of Instantaneous
Temperature and Three Velocity Components in Turbulent Flow. Rev. Sci.
Instrum., 49:654-663, 1978.
44
-------
12. Winternitz, F. A. L. Probe Measurements in Three-Dimensional Flow. A
Comparative Survey of Different Types of Instrument. Aircr. Enq.
28:273-278, 1956.
13. Meyer, C. A., and R. P. Benedict. Instrumentation for Axial Flow Com-
pressor Research. Trans. Am. Soc. Mech. Erg., pp. 1327-1336. November
1972.
14. Rajagopalan, S., and S. K. Basu. Theory and Design of Cyclones. Chem.
Age India, 27:42-54, 1976.
15. Chigier, N. A. Velocity Measurement in Vortex Flow. In: Flow: Its
Measurement and Control in Science and Industry, Vol. I, Part 1.
Instrument Society of America, pp. 399-400, 1974.
16. Williams, J. C., and F. R. De Jarnette. A Study on the Accuracy of Type-S
Pitot Tubes. EPA-600/4-77-030. U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina 27711, 1977. 70 pp
17. Rouillard, E. E. A., and R. E. Hicks. Flow Patterns Upstream of
Isokinetic Dust Sampling Probes. J. Air Pollut. Control Assoc., 28:599-601,
1978.
18. Thompson, S. Thompson Dehydrating Co., Topeka, St. Mary's, Kansas.
Private communication. September 1978.
19. Fluid Meters. Their Theory and Application. American Society of
Mechanical Engineering, 5th edition, 1959. pp. 43-49.
20. Ower, E., and R. C. Pankhurst. The Measurement of Air Flow. Pergamon
Press, New York, New York, 1966. pp. 110-116 and pp. 187-203.
21. Gibson, A. H. On the Flow of Water Through Pipes and Passages Having
Converging or Diverging Boundaries. Proc. Royal Soc. A-83. p. 366, 1910.
45
-------
APPENDIX C
PITOT TUBE ACCURACY AT LOW REYNOLD'S NUMBER
INTRODUCTION
Reynold's Number, Re, is an important dimensionless number that describes
the relationship between the inertia! and frictional forces present in a fluid
and indicates both the nature of fluid motion, i.e., laminar, transitional or
turbulent and the energy distribution associated with these stages. It is
represented by the relation
P inertia! forces
Ke a frictional forces
More specifically:
Re = - Eq. C-l
v
where:
p: density, a function of temperature
V: velocity of the fluid
D: diameter
y: dynamic or absolute viscosity, a function
of temperature
Temperature affects the Re because both density and viscosity change with
temperature. For gases, under isobaric conditions, as temperature increases
density decreases but viscosity increases. Therefore, an increase in tempera-
46
-------
ture under isobaric conditions results in a decrease in the inertia! forces
(density) and an increase in the frictional forces (viscosity), for an overall
decrease in Re.
HOW RE DETERMINES FLUID CHARACTERISTICS
The frictional force (viscosity) of a fluid inhibits the onset of turbu-
lence while the inertial force promotes the propagation of turbulence. These
two opposing forces influence the character of the fluid which is indicated by
Reynold's number, Re, a dimensionless ratio of inertial to frictional forces.
The magnitude of this value reflects which force dominates and therefore the
stage of flow development. At low Re the frictional forces dominate and
inhibit turbulent propagation thus forming a laminar regime. At high values
of Re, the inertial forces dominate over the frictional forces and promote
turbulent propagation and the formation of a turbulent regime.
A critical Reynold's number (Re ) is associated with the onset of turbu-
c
lence, i.e., the point where the inertial forces start to dominate. This
critical ratio of inertial to frictional forces, Rec, is also the critical
minimum for amplification and propogation of oscillatory motion with energy
transferring wave characteristics. At values greater than Rec, viscosity
effects which are associated with the frictional forces (the resistance
of movement of one layer of fluid over another) become negligible. The
transference of energy by inertial effects causes interaction of eddies of
different wave numbers and the associated convective spread of energy. The
smaller eddies form concentrated but weak vortex sheets, superimposed as
perturbations on the main vorticity field of the larger eddies. The system
is in energy equilibrium because of the balance between eddy stretching and
47
-------
viscosity effects. Therefore, the total energy of the fluid is constant and
evenly distributed throughout the fluid when inertial forces dominate over
frictional forces, i.e., at a Re greater than Re . At values of Re less than
c
Re , the fluid has a non-uniform energy distribution concentrated in pockets
\f
of eddy turbulence.
A fluid can be divided into an infinite number of layers, each representing
a streamline (See Figure. C-l). If layers B, C, and D encounter an obstruction
in their path, the molecules in these layers are reflected by the obstruction
and bombarded by the oncoming molecules in the fluid creating excess kinetic
energy in the system. Eventually, the molecules in layers B, C, and D are
refracted into the unobstructed layers, A and E, to which they transfer some
of their additional kinetic energy creating a system that is no longer in
equilibrium. At high Re, where viscosity is small there is little resistance of
movement of one layer over another so layers B, C, and D are easily joined and
separated from layers A and E with a limited energy loss. High energy swirls
form downstream at a distance, d, from the obstacle. This distance is a factor
of the separation time, T . At low Re, where viscosity is greater, there is
more resistance of movement of one layer over another and more energy is
required for one layer to join another layer. Since it takes more energy for
the layers to join, it also takes more to separate them.
If we consider a constant energy gain and a constant energy release value
in two systems, i.e. System I with a high Re and System II with a low Re, we
conclude
Energy of the System: E > E
dE11 dE1
Release Value:
48
-------
Separation Time: T Ir > T I
s s
Distance from Obstruction to Swirls: d11 > d1
where: d = T V
V = fluid velocity
Energy of Swirls: E n < E I
sw sw
dE
where: E$w = E$ -
Thus, at low Re, the swirls or eddies, will form further downstream with less
intensity and lower energy than for high Re. These results are portrayed
graphically in Figure C-l.
HOW REYNOLD'S NUMBER AFFECTS PITOT ACCURACY
The Reynolds number has been found to characterize the type of boundary
layer and correspondingly, the frictional losses in the flow around a pitot
tube in a flow stream. The Reynold's number in question is not the one that
describes the stack flow across the stack but rather, the Reynold's number that
characterizes the flow over the pitot tube. Therefore, the value for the
parameter D in equation C-l should be the diameter of the pitot not the diameter
of the stack when considering Re effect of the pitot tube.
At low Re, streamline flow will curve around the pitot tube (Fig. C-2a).
Small eddies or pockets of energy form downstream from the obstruction out of
the range of detection of the impact orifice. The viscosity effects dominate
49
-------
and resist redistribution of this energy causing a lower than true energy to
exist in the vicinity of the impact orifice. Since pressure is by definition
a force per unit area and in fluid mechanics is used to denote energy per unit
volume of fluid, the lower energy in the orifice region results in a lower
pressure and a high pitot coefficient. This has been confirmed experimentally
by MacMillan, who reported on the accuracy of flat-nosed and hemispherical-
nosed impact-pressure pitot tubes at Re from 15 to 1,000. His results
(Figure C-3) showed that the pitot coefficients, Cp, were greater than 1.0 at
Re less than 1000. For example, the coefficient for a hemispherical-nosed
pitot tube (Curve B in Figure C-3) was 1.35 at a Re of 15, but 1.01 at a Re of
1000. The differences between the curves were attributed to the shape of the
pitot tube nose by MacMillan.
Although analogous data for Reynold's number less than 1000 is not avail-
able for the "S" and Kiel/cylinder pitot tubes, they also seem to experience
a shift in Cp in the lower ranges of Reynold's number (Figure C-4). The
sinusoidal pattern evident in both curves between Reynold's numbers of 1000
345
and 3500 has been observed by Williams and De Jarnette , Lei and , Smith , and
in this study; and therefore, should be considered a real effect and should
not be dismissed as a result of data scatter. The two local maxima in both
curves may be associated with separate Reynold's number effects on the impact-
and the static-pressure sensing orifices but at present this is pure conjecture
and the need for more precise studies is indicated.
Figure C-5 shows that low Reynold's numbers can be encountered when
sampling stacks with low flows and high temperatures and demonstrates that
care should be used in this sampling situation. Quite possibly, the higher
flow sometimes measured downstream from a control device compared to that
50
-------
measured upstream may be due to a Reynold's number effect and not to air in-
leakage as is usually assumed.
51
-------
REFERENCES
1. MacMillan, F. A. Viscous Effects on Pitot Tubes at Low Speeds. J. Roy.
Aeronautical Soc., 58, 570-572, 1954.
2. This work.
3. Williams, J. C., and F. R. De Jarnette. A Study on the Accuracy of Type S
Pitot Tubes. EPA-600/4-77-030. U.S. Environmental Protection Agency,
Research Triangle Park, NC 27711.
4. Leland, B., J. Hall, A. Joensen, and J. Carroll. Correction of S-Type
Pitot-Static Tube Coefficients When Used for Isokinetic Sampling from
Stationary Sources. Env. Sci. & Tech., 11_, 694-700, July 1977.
5. Smith, M. L. Velocity Calibration of EPA Source Sampling Probe. Presented
at the 68th Annual Meeting of the Air Pollution Control Association, Boston,
MA, June 1975.
52
-------
H 1/8" DIA.
ID = 5/16"
5/3Z" RAD
8 HOLES 0.04" DIA.
EQUALLY SPACED
FREE FROM BURRS
SECTION A-A
STATIC PRESSURE
INNER TUBING
1/8" O.D. x 21 B.&S. GA. COPPER
*~ OUTER TUBING
5/16" O.D. x 18 B.&S. GA. COPPER
^- TOTAL PRESSURE
Figure 1. Merriam and Spaulding standard "!_'
pitot tube.
Figure 2. Kiel total pressure tube.
53
-------
90°
0°
(YAW)
, 4 _
'GAS FLOW
FRONT VIEW OF INSERTED PROBE
-90°
I GAS FLOW
SIDE VIEW OF INSERTED PROBE
Figure 3. Representation of yaw and pitch.
STACK WALL
/ '
\
1 ,
"*"
1
1
1
f
M
/
1
1
MH^IHHH
mm
\
1
g
V
1 (PITCH)
1
1
1
1
1
1
1
54
-------
0.6
Figure 4. Cylinder pitot tube.
180°
20° 40° 60° 80° 100° 120° 140° 160° 180°
Figure 5. Pressure distribution about a 25 mm
diameter cylinder.
DOWNSTREAM PITOT LEG
UPSTREAM PITOT LEG
I GAS FLOW
Figure 6. Standard "S" pitot tube.
55
-------
0.900
0.880
0.8EO
0.840
o.
u
2 0.820
LU
u
It 0.800
UJ
O
o
o 0.780
K
a.
0.760
0.740
0.720
0.700
O -ORIFICE SPACING 22mm
O -ORIFICE SPACING 32mm
A -ORIFICE SPACING 107mm
30° -20° -10° 0° 10° 20°
YAW ANGLE
Figure 7. Effect of yaw on different "S" pitot tubes'"'.
30°
1.000
0.980
0.960
0.940
0.920
0.900
0.880
0.860
0.840
0.820
ORIFICE
SPACING
O -ORIFICE SPACING 22mm
O - ORIFICE SPACING 32 mm
A -ORIFICE SPACING 107mm
I
I
20°
-10°
0°
PITCH ANGLE
Figure 8. Effect of pitch on different "S"
10°
20°
56
-------
TRANSVERSE
TUBE AXIS
FACE
-OPENING-
PLANES
(a)
1
Dt A
A-SIDE PLANE
. t
J 1
S ,
v T -\ '
-\ \
1NOTE:
1.05Dt
-------
STACK WALL
-10° -
CENTER OF STACK
STACK WALL
0°
10°
20°
30° l-
Figure 11. Sample of pitch angle profile in stack
with cyclonic flow. (Data near walls and at center
is unreliable^!)).
STACK WALL
40°
60°
60°
70°
80°
90° L_
CENTER OF STACK
STACK WALL
Figure 12. Sample of yaw angle profile in stack
with cyclonic flow. (Data near walls and at
center is unreliable(H)).
58
-------
Table 1. DIMENSIONS OF "S"
PITOT TUBE LEGS STUDIED
Table 2. DIMENSIONS (IN mm) OF "L" PITOT TUBES STUDIED
TUBE NO.
1S
2S
3S
4S
5S
6S
7S
8S
9S
10S
11S
12S
13S
0
45°
45°
45°
45°
45°
60°
60°
60°
60°
90°
90°
90°
90°
X, mm
12
25
12
25
25
12
25
12
25
25
12
25
25
(3
0°
0°
20°
20°
45°
0°
0°
20°
20°
0°
20°
20°
45°
TUBE
NO.
1L
2L
3L
4L
5L
6L
7L
8L
D
8
8
9.5
4.8
4.8
6.4
6.4
9.5
a
20
20
24
13
16
20
20
30
b
24
26
29
16
16
18
18
25
IMPACT
ORIFICE
ID
4.6
4.6
6.4
-
-
-
-
STATIC ORIFICE DIMENSIONS
NO.
ROWS
1
1
1
1
2
1
2
1
ORIFICES
ROW
6
4
4
4
3
4
3
6
c
_
_
_
5
_
5
-
OFFSET OF
ADJ. ROWS
_
_
_
_
60°
_
60°
-
ORIFICE
ID
1.01
1.60
2.05
1.01
1.60
1.60
2.05
2.05
IMPACT r
TAP e
X
u SN
STATIC \
TAP
-IT
\ \
x^ i
H
i
o
Pi r\
1 1
..-D-.-
. J
D
3
L_
Table 3. DIMENSIONS (IN mm) OF KIEL PITOT TUBES STUDIED
Tl IDC
NO.
1K
2K
3K
IMPACT
MODEL3
KFF
KFF
KRF
ORIFICE
3.3 (ID)
2.3 (ID)
3.3 (ID)
x5.0
(OD)
x 3.3 (OD)
x5.0
(OD)
DIMENSIONS
a
19
13
19
b
38
27
38
c
12.5
10.0
12.5
d
23
15
23
e
5.0
1.5
5.0
f
10
10
10
9
9.5
6.2
9.5
aKFF TYPE HAS VENTURI EXIT AND ENTRANCE. KRF TYPE
HAS VENTURI ONLY AT ENTRANCE.
59
-------
I - = = = = -
1mm ORIFICE
z
SHIELD: 8 mm O.D. x 6 mm I.D. TUBING
5.0 mm O.D. BY 3.5 mm I.D. TUBING CONTAINING
A 1mm I.D. TUBE AT THE UNION WITH SHIELD
Figure 13. Shielded static pitot tube.
Table 4. DIMENSIONS (IN mm) OF CYLINDER
PITOT TUBES STUDIED
NO.
1C
2C
3C
4C
5C
6C
7C
8C
9C
ORIFICE SIZE,
mm
2.60
2.60
2.60
1.58
1.58
1.00
2.75
2.75
3.20
ORIFICE LOCATION,
TDa
3
2
1
3
2
3
5
4
5
aTD=TUBE DIAMETERS FROM END OF TUBE.
ALL TUBES WERE 9.5 mm IN DIAMETER.
ORIFICE
SIDE VIEW
END VIEW
60
-------
e
&
o
50° 40° -30° 20° -10°
10° 20° 30° 40° 50"
PITCH ANGLE
Figure 14. Shielded static pitot tube error curve for
pitch at 8 m/sec.
0*8%
IMPACT PRESSURE
-- STATIC PRESSURE
J_
\
0° 20° 40° 60° 80° 100° 120° 140° 160° 180°
YAW ANGLE
Figure 15. Effect of yaw on tube 5S.
IMPACT PRESSURE
STATIC PRESSURE
I '
40° 60°
100° 120° 140" 160° 180°
Figure 16. Effect of yaw on tube 10S.
61
-------
TABLE 5. EFFECT OF YAW AND PITCH ON PRESSURE MEASUREMENT BY "S" PITOT TUBE LEGS
(mm H20)
Tube
IS
2S
3S
4S
5S
6S
7S
8S
9S
10S
US
12S
13S
Yaw 0°
Pitch -20° 0° +20°
-3.25 -4.11 -7.52
-3.10 -2.82 -3.63
-3.02 -2.82 -4.09
-3.51 -6.53 -13.9
-2.74 -2.99 -6.25
-3.18 -3.05 -4.34
-3.02 -2.89 -3.61
-3.28 -3.84 -7.29
-3.20 -3.35 -6.29
-3.94 -3.78 -5.11
-3.18 -4.14 -7.69
-3.05 -3.35 -6.35
-3.71 -6.45 -12.2
10°
-20° 0°
-3.12 -4.29
-3.23 -2.87
-3.10 -2.82
-3.78 -6.88
-2.67 -3.25
-3.18 -3.18
-3.28 -3.00
-3.48 -4.52
-3.23 -3.71
-3.86 -3.73
-3.56 -4.72
-3.23 -3.81
-4.19 -7.11
+20°
-7.59
-3.23
-4.29
-13.6
-6.60
-4.27
-3.78
-7.65
-6.68
-5.13
-8.33
-6.81
-13.6
20°
-20°
-3.18
-3.33
-3.12
-4.39
-2.74
-3.12
-3.35
-3.58
-3.40
-4.01
-4.04
-3.40
-5.23
0°
-4.52
-3.00
-2.97
-7.95
-3.84
-3.15
-3.02
-4.65
-4.19
-3.05
-5.79
-4.34
-8.48
+20°
-7.92
-4.85
-4.39
-14.9
-6.99
-4.62
-4.17
-8.03
-7.59
-5.38
-9.37
-7.67
-15.2
+90°
0°
-20.9
-24.6
-16.1
-23.9
-23.0
-19.7
-20.9
-20.5
-24.0
-22.3
-24.0
-23.9
-24.5
-90°
0°
-22.7
-20.2
-23.1
-20.9
-21.8
-22.4
-23.1
-22.2
-21.7
-24.2
-18.0
-21.7
-18.5
180°
0°
-21.0
-17.6
-19.6
-17.4
-17.3
-19.5
-17.4
-20.8
-17.5
-19.3
-21.0
-17.1
-17.5
ASME
pi tot
Pn-
-3.48
-3.38
-3.45
-3.43
-2.97
-3.43
-3.45
-3.47
-3.40
-4.01
-3.30
-3.51
-3.51
a Ref.
value
PS
-12.8
-12.6
-12.6
-12.3
-12.1
-12.3
-12.4
-12.3
-12.3
-12.1
-12.0
-12.1
-12.3
en
ro
aASME - American Society of Mechanical Engineering
-------
1.02 -
-0.98
4 6 8 tO 12 14 16 18 20
m/sec
Figure 17. Variation of pitot coefficient with velocity for tube 1L.
I
o
CC
IU
N
-10
S-20
a
UJ
cc
a.
8 -30
EC
O
CC
CC
111
o
CC
UJ
1
1
"xl
PITCH ERROR
- YAW ERROR
I I
10
20
30
40
40° -30°
-20°
-10°
10°
20"
30° 40°
50"
Figure
ANGLE OF MISALIGNMENT
18. Yaw and pitch error curve for static pressure for tube 8L (8 m/sec).
63
-------
cc
1
cc a
O LU
CC CC
-IB
cc
LU
a.
20
J L
50° -40° -30° -20° -10° 0° 10° 20° 30°
ANGLE OF MISALIGNMENT
Figure 19. Error curve for static pressure
for tubes 3L, 7L, and 8L
40° 50°
1.10
1.00 -
t 0.90
0.80
PITCH ERROR
- YAW ERROR
I
-40°
-30°
-20° -10
40°
ANGLE OF MISALIGNMENT
Figure 20. Error in pitot coefficient for tube 3L as function of
yaw and pitch at 8 m/sec.
5
4
3
2
1
0
-1
2
3
4
-5
6
7
8
9
10
-11
\
\
IMPACT PRESSURE
STATIC PRESSURE
50 40 30 20 1U 0 10
PITCH ANGLE
20 30
40
so
Figure 21 . Relative error in pressure for tube
1 K as function of pitch at 8 m/sec.
64
-------
TABLE 6. EFFECT OF YAW ON CYLINDER PITOT TUBE PRESSURE MEASUREMENT3
Yaw Angle
IMPACT PRESSURE
0
5
10
15
20
WAKE PRESSURE
180
170
160
150
140
130
120
no
100
90
85
80
75
70
65
60
50
1C
1
7
22
47
0
0
0
0
0
0
-1
-1
-1
-1
1
4
10
12
4
-15
Tube
4C
Reference Point
0
0
2
14
Reference Point
0
0
0
0
0
0
-1
0
4
10
14
12
8
1
-8
-25
5C
0
2
8
28
1
1
3
3
2
0
0
0
4
7
13
14
12
7
0
-20
error in pressure measurement at 8 m/sec.
65
-------
TABLE 7. EFFECT OF PITCH ON CYLINDER PITOT TUBE WAKE PRESSURE MEASUREMENT0
Pitch
Angle
-40
-30
-20
-10
0
10
20
30
40
Tube
1C 2C
3 7
0 0
-1 -1
-2 -2
5 4
5 12
8 13
8 16
3C
15
10
6
2
8
19
21
25
4C
0
-2
-3
-3
Rnfr
l\UTl
3
4
7
10
5C
6
-1
-3
-3
jrence P
3
9
10
14
6C
-1
-3
-2
-2
AT n4~
U 1 II U
4
5
7
9
7C
4
0
-1
-1
2
4
7
9
8C
4
0
-1
-1
2
4
7
13
9C
3
-2
-1
-1
2
3
7
10
% error in pressure measurement at 8 m/sec. The results for impact-pressure
were similar to those for wake-pressure except that all tubes were in error
by more than 5% for pitch angles greater than ± 25°.
66
-------
0.79
0.78
0.77
&
u
H 0.76
z
UJ
if 0.75
a 0.74
CJ
0.73
0.72
0.71
0.70
I I
I
I
I
6
8
18
10 12 14 16
m/sec
Figure 22. Calibration curve for tube 1K as function of velocity.
20
0.90
5 0.85
u
01
o
o
0.80
0.75
10
I
15
m/sec
20
25
Figure 23. Calibration curve for "S" pitot 3-04 after Williams and
DeJarnetteO).
30
67
-------
TABLE 8. RELATIVE EFFECT OF PITCH WHEN DIRECTION OF GAS FLOW IS DETERMINED
WITH KIEL/CYLINDER PITOT TUBE'
a
Pitch
Angle
25
20
15
10
5
-5
-10
-15
-20
-25
Distance Between Centers of Static and Impact Orifice
19 mm
3
2
2
1
0.5
1
1
3
6
8
38 mm
6
3
2
2
1
2
3
5
7
10
57 mm
2
1
1
0.5
0.5
0.5
0.5
2
5
10
error in the velocity measured as function of pitch.
68
-------
26
140
12.5mm I.D. x 15 mm O.D.
! 12
10mm O.D.
' '
i I
38
15
(a) 90° BEND NOZZLE
NOZZLE
YAW ' 0° YAW
FLOW
I I
\/
NOZZLE
(b) GOOSENECK SHAPED NOZZLE
(ALL DIMENSIONS IN mm)
Figure 24. Nozzle/pitot orientations.
69
-------
TABLE 9. EFFECT OF YAW ON PITOT COEFFICIENT FOR WAKE PRESSURE MEASUREMENT BY KIEL/CYLINDER PITOT TUBE1
Nozzle ID
(mm)
6.4
X
(mm)
8
+50
1
Yaw Angle
+40
2
(nozzle
+30
-1
upstream)
+20
-2
Yaw
+10 0 -10
-1 t
£ 7
Angle
-20
6
(nozzle
-30
4
downstream)
-40 -50
2
1
o
12.5
5
2
-4
-4
-3
-2 °- 8
8
8
3
0
eu
6.4
22
4
0
-3
-1
t
J
-1 £ 7
<
9
4
3
1
QJ
12.5
12.5
18
22
-1
-2
-3
-2
-3
-3
-3
-3
-2 «
o
-2
u -2
8
-2
9
-3
6
-3
4
-2
1
The tests were conducted with the sampling probe attached.
-------
TABLE 10. COMBINED EFFECT OF YAW AND PITCH ON COEFFICIENT OF KIEL/CYLINDER
PITOT TUBE ATTACHED TO A PARTICULATE SAMPLING ASSEMBLY9
Pitch
Angle
45
40
30
20
10
0
-10
-20
-30
-40
-45
Yaw Angl
+50
6
4
2
0
-1
-1
-1
_1
0
4
6
+40
6
3
0
-2
-3
-3
-2
0
3
6
8
+30
7
4
1
-2
-3
-3
-2
0
3
6
7
+20
2
1
-1
-2
-2
-3
-4
-5
-3
-1
1
+10
11
8
3
-1
-2
-2
-1
-1
0
3
5
0
15
9
4
1
0
1
2
4
5
6
e
-10
11
8
2
-1
-1
-2
0
3
5
8
9
-20
12
8
2
_1
-1
-2
0
3
6
10
12
-30
12
9
3
0
-2
-3
-1
2
6
9
11
-40
11
7
2
-1
-2
-3
-2
1
3
6
9
-50
10
7
2
-1
0
-2
-1
1
3
6
11
a
'% error in pitot coefficient with respect to zero yaw and pitch.
71
-------
LLI
U
o
o
K
O
DC
ee
u
ce
ui
14
12
10
8
6
4
2
0
-2
-4
-6
GOOSENECK NOZZLE
90° BEND NOZZLE
I I I I
20° 30° 40°
-50° -40° -30° -20° -10° 0° 10°
YAW ANGLE
Figure 25. Combined effect of yaw and pitch on Kiel/cylinder pitot
tube attached to sampling probe (8 m/sec) (Reference point Cp at
zero yaw and pitch).
50°
72
-------
TABLE 11. SENSITIVITY OF "S" AND KIEL/CYLINDER PITOT TUBE COEFFICIENTS
TO NOZZLE/THIMBLE INTERACTION3
Zb (mm)
+25
0
-25
-50
-IOC
C^ for "S" PI tot
P
Impact
.89
.89
.96
1.0
Wake
1.28
1.25
1.39
1.52
C for Kiel/Cylinder Pitot
P ₯
Impact
.99
.98
.98
.98
.98
Wake
1.33
1.34
1.34
1.34
1.28
Measured relative to the Reference ASME Pitot Tube, C = 0.99
Z = Distance between the pitot impact orifice and the leading edge of the
nozzle. The thimble was equipped with a 12.5 mm ID x 15 mm OD, 90°
bend nozzle, the spacing between the nozzle edge and the pitot sheath
was 18 mm, and the nozzle centerline was 200 mm ahead of the leading
edge of the thimble body. A negative "Z" value means the pitot tube
orifice was between the nozzle centerline and the leading edge of the
thimble body.
73
-------
^
1
t
1
in
«H
|
^^^1
1 ALUNDUM
65 THIMBLE
^r
«nn fej
r
=^
L 1
1
N
!
3 '
16 CO
r 18 Ix
PirnTTiinc ( (
ALL DIMENSIONS IN mm
Figure 26. Orientation of Kiel/cylinder pitot tube with alundum thimble.
o
i-
UJ
O
O
WITH "S" PITOT
WITH KIEL/CYLINDER PITOT
0.76
0.70
Figure 27. Variation of pitot coefficient with velocity for alundum
thimble sampling assembly.
74
-------
TYPE SPITOT TUBE
7
x >1.90 cm (% in.) FOR Dn = 1.3 cm 04 in.)
SAMPLING NOZZLE
A. BOTTOM VIEW: SHOWING MINIMUM PITOT-NOZZLE SEPARATION.
SAMPLING
PROBE
SAMPLING
NOZZLE
TYPES
PITOTTUBE
NOZZLE ENTRY
PLANE
LANE V
STATIC PRESSURE
OPENING PLANE
IMPACT PRESSURE
OPENING PLANE
Figure 28. Standard EPA sampling assembly for particulate.
75
-------
V)
u
cc
I
I
I
I
I
4 8 12 16
TUBE DIAMETERS FROM TIP
Figure 1 A-a. Absolute error in static pressure as a function of
orifice distance from tip of "L" pitot.
TUBE DIAMETERS FROM STEM
8 12 16 20
24
I
I
LU
3
CC
+2
+3
Figure 1 A-b. Absolute error in static pressure as a function of
orifice distance from stem of "L" pitot.
76
-------
SYSTEM I. HIGH REYNOLDS' NUMBER
SYSTEM II. LOW REYNOLDS' NUMBER
Figure C-1. Flow patterns downstream of a cylinder
at high and at low Reynolds' number.
1.45
1.40
1.35
o 1.30
H
| 1-25
!Z
It 1.20
o
o
£ 1.15
H
1.10
1.05
1.00
_L
I I I
CURVE A - THEORETICAL CURVE FOR SPHERE
CURVE B - ASME "L" PITOT TUBE
CURVE C - FLAT-NOSED PITOT TUBE
I I
I
_L
10 20 30 40 60 80 100
Re
J_
J L
J_
200 300 400 600 800
Figure C-2. Effect of Reynolds' number of pitot coefficient after MacMHIan.
77
-------
0.95
Re
1000 2000 3000 4000 5000 6000 7000 8000 9000 10,000
0.90
0.85
0.80
0.75
KIEL/CYLINDER
PITOT
7.0 1 I I I I | I 1 I I I I I I
0 4 8 12 16
m/sec
Figure C-3. Correlation of pitot
coefficient and Reynolds' number
at 35QC for "S" and Kiel/cylinder
pitot tubes.
6000
4000
3000
2000
1000
I I
i I I
I I I I I
50° 150° 250" 350° 450° 550° 660° 750°
"C
Figure C-4. Reynolds' number as a function
of velocity for a 0.95 cm O.D. pitot tube.
78
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing
REPORT NO.
EPA 600/4-79-042
3. RECIPIENT'S ACCESSIOfVNO.
4. TITLE AND SUBTITLE
ANGULAR FLOW INSENSITIVE PITOT TUBE SUITABLE FOR USE
WITH STANDARD STACK TESTING EQUIPMENT
5 "'EPPRT HATE
June 1979
6. PERFORMING ORGANIZATION CODE
7. AUTHORIS)
W. J. Mitchell, B. E. Blagun, D. E. Johnson, and
M R
8. PERFORMING ORGANIZATION REPORT NO.
9 PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
Environmental Monitoring & Support Laboratory
Quality Assurance Branch
Research Triangle Park, NC 27711
10. PROGRAM ELEMENT NO.
1AD800
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Monitoring and Support Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA 600/08
115. SUPPLEMENTARY NOTES
To be published in Environmental Monitoring Series
16. ABSTRACT
Five pitot tube designs were tested under various gas flow conditions for
accuracy in measuring static and total pressure. The static- and impact-pressure
measuring tubes least affected by angular flow were combine and then evaluated in
the presence of standard particulate sampling nozzles. Tests were performed on "S"
"L", Kiel and cylinder pitot tubes and a shielded static-pressure pitot tube. The
percent error for each pitot tube was determined as a function of yaw, pitch,
orifice size, orifice location, pitot tube size, and velocity. A pitot tube was
developed that is accurate within 5% when yawed and pitched ± 30° even while
attached to a standard EPA Method 5 sampling assembly. This pitot tube was field
tested at a sewage sludge incinerator, a clay crushing plant and a power plant.
Also included in the report are: a-summary of the existing literature on design
of "L" pitot tubes; a summary of the present state-of-the-art in sampling stacks
with cyclonic flow and the errors involved in such sampling; a recommendation
for straightening cyclonic flow by insertion of a venturi throat; and the effect
of Reynolds Number on pitot tube accuracy.
til
k
DESCRIPTORS
Velocity measurement
"S", "L", ASME, Kiel/cylinder pitot tubes
Cyclonic Flow
Reynolds Number
KEY WORDS AND DOCUMENT ANALYSIS
^-IDENTIFIERS/OPEN ENDED TERMS
COSATl Field/Group
Yaw Error
Pitch-Error
68A
43 F
* DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
79
20. SECURITY CLASS (This pagel
22. PRICE
orm 2220-1 (9-73)
------- |