United States Acid Rain Division EPA/430/R-97-013
Environmental Protection (6204J) February 1997
Agency Washington, D.C. 20460
&EPA Flow Reference Method
Testing and Analysis:
Wind Tunnel
Experimental Results
Volume 1
-------
Flow Reference Method
Testing and Analysis
Wind Tunnel Experimental Results
Preparedfor
U.S. Environmental Protection Agency
Acid Rain Division, 6204J
401 M Street, SW
Washington, DC 20460
Contract No. 68-D2-0168
Work Assignment 5C-01
Prepared by
The Cadmus Group, Inc.
1920 Highway 54, Suite 100
Durham, NC 27713
February, 1997
-------
Wind Tunnel Experimental Results
ACKNOWLEDGMENTS
The principal authors of this report were Dr. Arthur Werner, Ms. Penelope Kellar, and Mr. Scott
Shanklin of The Cadmus Group, Inc. Dr. Elliott Lieberman and Mr. John Schakenbach of the U.S.
Environmental Protection Agency provided continuing guidance throughout this project and the
preparation of this report. Dr. William Warren-Hicks and Ms. Jie Tao of The Cadmus Group, Inc.,
provided the statistical design and data analysis assistance. Mr. Gene Fax of Cadmus provided
constructive review on this report. Experiments were conducted in North Carolina State University's
(NCSU) Merrill Subsonic Wind Tunnel under the critical supervision of Dr. John Perkins, in the
NCSU Department of Mechanical and Aerospace Engineering. Dr. Perkins was assisted by Mr. Kraig
Marquis and Mr. Eric Thirolle (Cadmus), Mr. Ed. Baker (NCSU), and Mr. Keith Hazel (Climax
Engineering). Mrs. Blanche Dean of Cadmus provided production assistance.
-------
Wind Tunnel Experimental Results
TABLE OF CONTENTS
Page
VOLUME 1
1.0 INTRODUCTION 1-1
l.l PURPOSE OF PROGRAM 1-1
1.2 GOALS OF THE WIND TUNNEL TESTS 1-1
1.3 DESCRIPTION OF WIND TUNNEL 1-2
1.4 DESCRIPTION OF TEST PROBES 1-4
1.5 DOCUMENT PURPOSE AND ORGANIZATION 1-9
2.0 OVERVIEW OF DATA COLLECTION,
HANDLING, AND QUALITY ASSURANCE 2-1
2.1 OVERVIEW OF FLOW MEASUREMENT PROCEDURES 2-1
2.2 YAW ALIGNING AND VELOCITY MEASUREMENT PROCEDURES 2-3
2.3 DATA ACQUISITION, HANDLING, AND QUALITY ASSURANCE 2-7
2.4 STUDY TO DETERMINE APPROPRIATE SAMPLE SIZE 2-8
2.5 CALIBRATION OF PROBES 2-9
3.0 MODIFIED KIEL PROBE EXPERIMENTS 3-1
3.1 COMPARISON OF WAKE PORTS AND FECHHEIMER PORTS IN DETERMINING
YAW NULL 3-1
3.2 SENSITIVITY AND ACCURACY OF STATIC PRESSURE DETERMINATIONS
USING THE WAKE PORTS 3-2
3.3 CONCLUSIONS 3-5
4.0 STANDARD CALIBRATION COEFFICIENTS AND
VARIABILITY WITH VELOCITY FOR 2D PROBES 4-1
4.1 TEST PROCEDURES 4-1
4.2 CALIBRATION COEFFICIENTS : 4-1
4.3 VARIATION WITH VELOCITY 4-1
4.4 CONCLUSIONS 4-3
5.0 PITCH ANGLE EFFECTS ON VELOCITY
MEASUREMENTS PERFORMED WITH 2D PROBES 5-1
5.1 TEST PROCEDURES 5-1
5.2 SENsmvrrY OF CALIBRATION COEFFICIENTS TO PITCH ANGLE 5-1
5.3 PITCH ANGLE EFFECTS ON VELOCITY DETERMINATIONS 5-2
5.4 CONCLUSIONS 5-7
6.0 ACCURACY OF YAW ANGLE MEASUREMENTS
AND EFFECT ON VELOCITY MEASUREMENT 6-1
6.1 TEST PROCEDURES 6-1
6.2 ACCURACY OF YAW ANGLE DETERMINATIONS AND
VELOCITY DETERMINATIONS 6-1
111
-------
Wind Tunnel Experimental Results
TABLE OF CONTENTS
(continued)
Page
6.3 FLOW MEASUREMENT ERROR DUE TO YAW NULL MISALIGNMENT
FOR THE AUTOPROBE TYPE S 6-3
6.4 CONCLUSIONS 6-6
7.0 CALIBRATION CURVES AND ACCURACY
OF VELOCITY DETERMINATIONS FOR3D PROBES 7-1
7. l TEST PROCEDURES 7-1
7.2 CALIBRATION CURVES 7-2
7.3 ACCURACY OF VELOCITY DETERMINATIONS 7-3
7.4 ACCURACY OF PITCH ANGLE DETERMINATIONS 7-9
7.5 CONCLUSIONS 7-9
8.0 ENTRY WALL EFFECTS ON YAWNULLING 8-1
8.1 TEST PROCEDURES 8-1
8.2 RESULTS OF ENTRY WALL EFFECTS TESTS 8-1
8.3 CONCLUSIONS 8-4
9.0 CONCLUSIONS AND RECOMMENDED FOLLOW-UP 9-1
VOLUME 2
Appendix A—Study to Determine Sample Size A-l
Appendix B—Study of Entry Wall Effects on Yaw Nulling B-l
Appendix C—Raw Data C-l
Appendix D—-Data Disks •. D-l
IV
-------
Wind Tunnel Experimental Results
LIST OF FIGURES
Page
Figure 1-1 Schematic (top view) of the Merrill Subsonic Wind Tunnel Facility
located at North Carolina State University in Raleigh 1-3
Figure 1-2 Schematic of modified Kiel probe (MK1) 1-5
Figure 1-3 Autoprobe positioned for insertion into the wind tunnel 1-6
Figure 1-4 Front and side views of spherical probe 1-7
Figure 1-5 Top, side, and cross-sectional views of DAT probe 1-7
Figure 2-1 Magnehelic meters used to measure differential pressures (PrP2, P2-Ps> P-rPs)
across various ports of test probes 2-4
Figure 2-2 Schematic (overhead perspective) showing positive and negative pitch
angle and yaw angle orientation of flow probe with respect to
test port for wind tunnel and flow directions 2-5
Figure 3-1 APW and APf vs. Yaw Angle: Modified Kiel -
Wake Port 1 - Velocity = 30 fps 3-6
Figure 3-2 APW and APf vs. Yaw Angle: Modified Kiel -
Wake Port 1 - Velocity = 60 fps 3-7
Figure 3-3 APW and APf vs. Yaw Angle: Modified Kiel -
Wake Port 1 - Velocity = 90 fps 3-8
Figure 3-4 APW and APf vs. Yaw Angle: Modified Kiel -
Wake Port 4 - Velocity = 30 fps 3-9
Figure 3-5 APW and APf vs. Yaw Angle: Modified Kiel -
Wake Port 4 - Velocity = 60 fps 3-10
Figure 3-6 APW and APf vs. Yaw Angle: Modified Kiel -
Wake Port 4 - Velocity = 90 fps 3-11
Figure 3-7 Probe: Modified Kiel MK1 - Wake Port Calibration Coefficients -
Wake Port 1 3-12
Figure 3-8 Probe: Modified Kiel MK1 - Wake Port Calibration Coefficients -
Wake Port 2 3-13
-------
Wind Tunnel Experimental Results
LIST OF FIGURES
(continued)
Page
Figure 3-9 Probe: Modified Kiel MK1 - Wake Port Calibration Coefficients -
Wake Port 3 3-14
Figure 3-10 Probe: Modified Kiel MK1 - Wake Port Calibration Coefficients -
Wake Port 4 3-15
Figure 3-11 Probe: Modified Kiel With Thermocouple - MK2 - Wake Port
Calibration Coefficients - Wake Port 4 3-16
Figure 3-12 Probe: 3/8" Type S-l- Calibration Coefficients 3-17
Figure 3-13 Probe: 3/8" Type S-2 - Calibration Coefficients 3-18
Figure 3-14 Probe: 3/8: Type S-3 - Calibration Coefficients 3-19
Figure 3-15 Probe: Modified Kiel - MK1 - Wake Port
Calibration Coefficients - Wake Port 1 3-20
Figure 3-16 Probe: Modified Kiel - MK1 - Wake Port
Calibration Coefficients - Wake Port 2 3-21
Figure 3-17 Probe: Modified Kiel - MK1 - Wake Port
Calibration Coefficients - Wake Port 3 3-22
Figure 3-18 Probe: Modified Kiel - MK1 - Wake Port
Calibration Coefficients - Wake Port 4 3-23
Figure 3-19 Probe Modified Kiel with Thermocouple - MK2 -
Wake Port Calibration Coefficients - Wake Port 4 3-24
Figure 3-20 Probe: 3/8" Type S-l- Calibration Coefficients 3-25
Figure 3-21 Probe: 3/8" Type S-2 - Calibration Coefficients 3-26
Figure 3-22 Probe: 3/8" Type S-3 - Calibration Coefficients 3-27
Figure 3-23 Probe: Modified Kiel -MK1 - Calibration Coefficients Wake Port 1 3-28
Figure 3-24 Probe: Modified Kiel -MK1 - Calibration Coefficients Wake Port 3 3-29
VI
-------
Wind Tunnel Experimental Results
LIST OF FIGURES
(continued)
Page
Figure 3-25 Probe: Modified Kiel -MK1 - Calibration Coefficients Wake Port 4 3-30
Figure 3-26 Probe Modified Kiel with Thermocouple - MK2 -
Calibration Coefficients - Wake Port 4 3-31
Figure 3-27 Probe: 3/8" Type S-l - Calibration Coefficients 3-32
Figure 4-1 Probe: 3/8" Type S-l - Calibration Coefficients 4-5
Figure 4-2 Probe: 3/8" Type S-2 - Calibration Coefficients 4-6
Figure 4-3 Probe: 3/8" Type S-3 - Calibration Coefficients 4-7
Figure 4-4 Probe: Modified Kiel With Thermocouple -
MK2 - Wake Port Calibration Coefficients (Wake Port = 4) 4-8
Figure 4-5 Probe: Modified S - USS-1 - Calibration Coefficients 4-9
Figure 4-6 Probe: Modified S - USS-2 - Calibration Coefficients 4-10
Figure 4-7 Probe: Modified S - US S-3 - Calibration Coefficients 4-11
Figure 4-8 Probe: Autoprobe - Type S - Calibration Coefficients 4-12
Figure 4-9 Probe: Autoprobe - Modified S - Calibration Coefficients 4-13
Figure 4-10 Probe: French Probe - F2 - Calibration Coefficients 4-14
Figure 5-1 Probe: 3/8" Type S - S-l - Calibration Coefficient vs. Pitch Angle 5-9
Figure 5-2 Probe: Modified S - USS-1 - Calibration Coefficient vs. Pitch Angle 5-10
Figure 5-3 Probe: Modified S - USS-3 - Calibration Coefficient vs. Pitch Angle 5-11
Figure 5-4 Probe: Autoprobe - Type S - Calibration Coefficient vs. Pitch Angle 5-12
Figure 5-5 Probe: Autoprobe - Modified S - Calibration Coefficient vs. Pitch Angle ... 5-13
Figure 5-6 Probe: French Probe - F-2 - Calibration Coefficient vs. Pitch Angle 5-14
vn
-------
Wind Tunnel Experimental Results
LIST OF FIGURES
(continued)
Page
Figure 5-7 Percent Deviation of Measured Velocity from Axial Velocity
vs. Pitch Angle at 30 fps 5-15
Figure 5-8 Percent Deviation of Measured Velocity from Axial Velocity
vs. Pitch Angle at 60 fps 5-16
Figure 5-9 Percent Deviation of Measured Velocity from Axial Velocity
vs. Pitch Angle at 90 fps 5-17
Figure 6-1 Probe: Modified Kiel Probe with Thermocouple - MK2
Theta Null Deviation from the True Yaw Null 6-7
Figure 6-2 Probe: 3/8" Type S-SI
Theta Null Deviation from the True Yaw Null 6-8
Figure 6-3 Probe: Modified S - USS-l
Theta Null Deviation from the True Yaw Null 6-9
Figure 6-4 Probe: Modified S-USS-3
Theta Null Deviation from the True Yaw Null 6-10
Figure 6-5 Probe: Autoprobe Type S
Theta Null Deviation from the True Yaw Null 6-11
Figure 6-6 Probe: Autoprobe Mod S
Theta Null Deviation from the True Yaw Null 6-12
Figure 6-7 Probe: Spherical MS5 - 2
Theta Null Deviation from the True Yaw Null 6-13
Figure6-8 Probe: DAT-3D-1
Theta Null Deviation from the True Yaw Null 6-14
Figure 6-9 Probe: DAT - 3D-2
Theta Null Deviation from the True Yaw Null 6-15
Figure 6-10 Probe: DAT - 3D-3
Theta Null Deviation from the True Yaw Null 6-16
Figure 6-11 C vs. Yaw Angle 6-17
Vlll
-------
Wind Tunnel Experimental Results
LIST OF FIGURES
(continued)
Page
Figure 6-12 Flow Measurement Error from Yaw Error 6-18
Figure 7-1 Probe: DAT 3D-1 - Fl vs. pitch angle 7-13
Figure 7-2 Probe: DAT 3D-1 - F2 vs. pitch angle 7-14
Figure 7-3 Probe: DAT 3D-2 - Fl vs. pitch angle 7-15
Figure 7-4 Probe: DAT 3D-2 - F2 vs. pitch angle 7-16
Figure 7-5 Probe: DAT 3D-3 - Fl vs. pitch angle 7-17
Figure 7-6 Probe: DAT 3D-3 - F2 vs. pitch angle 7-18
Figure 7-7 Probe: Spherical MS5-2 - Fl vs. pitch angle 7-19
Figure 7-8 Probe: Spherical MS5-2 - F2 vs. pitch angle 7-20
Figure 7-9 Probe: DAT 3D-1 - Fl vs. pitch angle 7-21
Figure 7-10 Probe: DAT 3D-1 - F2 vs. pitch angle 7-22
Figure 7-11 Probe: DAT 3D-2 - Fl vs. pitch angle 7-23
Figure 7-12 Probe: DAT 3D-2 - F2 vs. pitch angle 7-24
Figure 7-13 Probe: DAT 3D-3 -Fl vs. pitch angle 7-25
Figure 7-14 Probe: DAT 3D-3 - F2 vs. pitch angle 7-26
Figure 7-15 Probe: Spherical MS5-2 - Fl vs. pitch angle 7-27
Figure 7-16 Probe: Spherical MS5-2 - F2 vs. pitch angle 7-28
Figure 8-1 Flow pattern at yaw angle of-45° 8-3
Figure 8-2 Flow pattern at yaw angle of+20° 8-3
Figure 8-3 Manual S-type: yaw/wall effect - 60 fps, 0 pitch, 1-42" from wall 8-7
Figure 8-4 Manual S-type: yaw/wall effect - 60 fps, 0 pitch, 1" from wall 8-8
ix
-------
Wind Tunnel Experimental Results
LIST OF FIGURES
(continued)
Page
Figure 8-5 Manual S-type: yaw/wall effect - 60 fps, 0 pitch, 3" from wall 8-9
Figure 8-6 Manual S-type: yaw/wall effect - 60 fps, 0 pitch, 6" from wall 8-10
Figure 8-7 Manual S-type: yaw/wall effect - 60 fps, 0 pitch, 9" from wall 8-11
Figure 8-8 Manual S-type: yaw/wall effect - 60 fps, 0 pitch, 12" from wall 8-12
Figure 8-9 Manual S-type: yaw/wall effect - 60 fps, 0 pitch, 42" from wall 8-13
Figure 8-10 Autoprobe S-type: yaw/wall effect - 60 fps, 0 pitch, -10° to + 10° 8-14
Figure 8-11 Autoprobe S-type: yaw/wall effect - 60 fps, 0 pitch, 165° to 185° 8-15
Figure 8-12 Autoprobe S-type: yaw/wall effect - 60 fps, 0 pitch, -50° to +200° 8-16
Figure 8-13 Autoprobe S-type: yaw/wall effect - 60 fps, +10 pitch, -10° to + 10° 8-17
Figure 8-14 Autoprobe S-type: yaw/wall effect - 60 fps, +10 pitch, 165° to 185° 8-18
Figure 8-15 Autoprobe S-type: yaw/wall effect - 60 fps, +10 pitch, -50° to +200° 8-19
Figure 8-16 Autoprobe S-type: yaw/wall effect - 60 fps, -10 pitch, -10° to + 10° 8-20
Figure 8-17 Autoprobe S-type: yaw/wall effect - 60 fps, -10 pitch, 165° to 185° 8-21
Figure 8-18 Autoprobe S-type: yaw/wall effect - 60 fps, -10 pitch, -50° to +200° 8-22
Figure 8-19 Mod Kiel w/therm.: Yaw/wall effect - 60 fps, 1" from wall 8-23
Figure 8-20 Mod Kiel w/therm.: Yaw/wall effect - 60 fps, 5" from wall 8-24
Figure 8-21 Spherical probe: yaw/wall effect - 60 fps, 4" from wall 8-25
Figure 8-22 DAT probe: yaw/wall effect - 60 fps, 3" from wall 8-26
x
-------
Wind Tunnel Experimental Results
LIST OF TABLES
Table 1 -1 Probes Calibrated and Tested in NCSU Wind Tunnel
Page
. 1-8
Table 3-1 AP as a Function of Yaw Angle for Wake Ports 1 and 4 of the
Modified Kiel Probe Prototype 3-2
Table 3-2 Calibration Coefficients Across Pitch Angles for Various
Wake Ports of the Modified Kiel Probe Prototypes
(MK1 and MK2) and for the Type S (S-l) Probe at 0° Yaw 3-5
Table 4-1 Calibration Coefficients Obtained for 2D Probes at Specific Velocities 4-2
Table 4-2 Variability of Calibration Coefficients Across Velocities for
2D Probes Determined at Pitch Angle 0° 4-3
Table 5-1 Sensitivity of Cp to Pitch Angle for 2D Probes 5-2
Table 5-2 Axial Velocity Evaluation for 2D Probes 5-4
Table 5-3 Maximum Percent Deviation of Measured Velocity from Axial Velocity
Across All Velocities at Zero Yaw 5-6
Table 5-4 Maximum Percent Deviation of Measured Velocity from Axial Velocity
Across 60 and 90 fps at Zero Yaw 5-6
Table 6-1 Summary of Yaw Angle Accuracy and Effect on Velocity Determination .... 6-2
Table 6-2 Yaw Angle Error Detailed Data 6-3
Table 7-1 DAT Probe (3D-1) Velocity Accuracy 7-5
Table 7-2 DAT Probe (3D-2) Velocity Accuracy 7-6
Table 7-3 DAT Probe (3D-3) Velocity Accuracy 7-7
Table 7-4 Spherical Probe (MS5-2) Velocity Accuracy 7-8
Table 7-5 3D Probe Pitch Angle Accuracy 7-10
Table 7-6 Summary of 3D Probes Velocity Accuracy Results . . . 7-11
Table 8-1 Summary Table of the Effect of Distance from Wall Entry on the Number
of Yaw Nulls Observed 8-4
XI
-------
1.0 INTRODUCTION
1.1 PURPOSE OF PROGRAM
Electric utilities have recently expressed concerns that continuous emission monitoring systems
(CEMS) installed to comply with regulations implemented under the U.S. Environmental Protection
Agency's (EPA) Acid Rain Program report higher heat input and S02 emissions values than are
estimated using conventional (input/output and output loss) methods. The basis of the concern is that
EPA's reference method for volumetric flow determination (Test Method 2)1 may overestimate flow
rate, which could result in overestimates of sulfur dioxide emissions and heat input.
To address this concern, EPA has begun an experimental field project with The Cadmus Group, Inc.,
to assess Test Method 2 and compare its results with those obtained using alternative methods of
measuring volumetric flow. During the next year, EPA will work with Cadmus to design, coordinate,
and report on a multi-laboratory field study of Test Method 2. The study will address several major
questions. First, what is the accuracy and variability of the reference method? Next, how do
reference method measurements compare with other measurements of flow, particularly in situations
experiencing significant off-axial flow? Finally, do the field results indicate that the procedures for
Test Method 2 require improvements, and if so, in what regard?
As a preliminary step toward evaluating the test method, EPA asked Cadmus to perform wind tunnel
tests of flow measurement probes that are either approved or under consideration for use with Test
Method 2. Under EPA direction, Cadmus designed these tests and conducted them in the Merrill
Subsonic Wind Tunnel at North Carolina State University (NCSU) in Raleigh. The wind tunnel tests
were designed to evaluate how accurately various probes can measure angles and velocity of flow
under prescribed conditions and, additionally, to calibrate the probes for use in planned multi-
laboratory field experiments. To provide a basis for selecting probes for the field test, the wind tunnel
testing was performed over a range of velocity, pitch, and yaw settings approximating the conditions
encountered at actual utility sites.
1.2 GOALS OF THE WIND TUNNEL TESTS
The wind tunnel tests were originally designed to achieve five principal goals: (1) evaluate a
prototype modified Kiel probe; (2) determine the minimum sample size required for subsequent
experiments; (3) evaluate the accuracy of yaw angle nulling procedures and velocity determinations
by various two-dimensional (2D) and three-dimensional (3D) probes; (4) determine pitch angle and
velocity measurement accuracy for selected 3D probes; and (5) calibrate probes that will be used in
the field study. During the course of the wind tunnel study, a sixth objective was added as a result
of preliminary findings revealed in a related flow study conducted by the Electric Power Research
Institute (EPRI). This objective was to examine what effect, if any, proximity of the probe tip to the
wind tunnel entry wall had on determining yaw null. A more detailed summary of the wind tunnel
study goals follows.
1 40 CFR Ch. 1 (7-1 -92 Edition), Pt. 60, App. A, Method 2—Determination of Stack Gas Velocity and
Volumetric Flow Rate (Type S Pitot Tube), pgs. 691-709 (hereinafter referred to as Method 2).
1-1
-------
Wind Tunnel Experimental Results
• Modified Kiel Probe Evaluation. Using the Kiel/cylinder pilot static probe constructed
and tested by Mitchell, et al.2 in the late 1970s as the basis, a standard Kiel probe was
modified by adding four candidate wake ports and two Fechheimer ports to produce a
prototype for testing in the NCSU wind tunnel. The tests were designed to determine
whether the alternative probe was sufficiently promising to warrant further consideration
for use under Test Method 2. If so, additional wind tunnel testing would be conducted to
collect data for use in finalizing the probe's design prior to field testing.
• Required Sample Size Determination. Early in the wind tunnel testing, an experiment was
conducted to determine the minimum number of replicates required for each subsequent
test. A 3/8-inch Type S probe was used in this experiment.
• Yaw Angle and Axial Flow Accuracy Determination. Evaluating the accuracy of flow
angle determinations using various probe types was the third goal of the wind tunnel
experimentation. Flow angles were determined using procedures prescribed for the specific
type of probe under evaluation. All probes were tested to examine accuracy of yaw angle
determination, as well as accuracy of yaw-corrected velocity measurements.
• Determination of Pitch Angle and Velocity Pressure Accuracy for 3D Probes.
Determining pitch angle accuracy and velocity pressure accuracy using DAT and spherical
probes was the fourth goal of the wind tunnel testing.
• Probe Calibration. Calibrating each probe to be used in the field study was an important
goal of the wind tunnel testing. For the 2D probes, single calibration factors at 0° yaw and
0° pitch were derived. Calibration curves as specified in draft Method 2F3 were derived for
the DAT and spherical probes.
• Examination of Wall Effects on Yaw Null Determination. A recent EPRI study indicated
that multiple yaw nulls were located in yaw-angle scans when a Type S probe was placed
near the wall entry point. Therefore, a study to evaluate how proximity of the probe tip to
the wind tunnel entry wall affects the ability to identify yaw null was added as an objective
of this study.
1.3 DESCRIPTION OF WIND TUNNEL
The Merrill Subsonic Wind Tunnel (Figure 1-1), where these experiments were conducted during the
summer of 1996, is a continuous flow, single-return tunnel with a test section vented to the
atmosphere. The test section is 45 inches wide, 32 inches high, and 46 inches long in the flow
direction. Wind tunnel speeds ranging from 10 to approximately 120 feet per second (fps) are
controlled by a variable-pitch fan driven by a 50-hp constant-speed motor. Vanes located in each
2 Mitchell, W. J., B.E. Blagun, D.E. Johnson, and M.R. Midgett. 1979. Angular Flow Insensitive Pilot Tube for
Use with Standard Stack Testing Equipment. U.S. Environmental Protection Agency, Office of Research and
Development, EPA-600/4-79-00.
3 U.S. Environmental Protection Agency Method 2F—Determination of Stack Gas Velocity and Volumetric
Flow Rate (Three-dimensional Pitot Tube). Draft, June 30,1993 (hereinafter referred to as Method 2F).
1-2
-------
Wind Tunnel Experimental Results
comer of the tunnel reduce losses in total pressure and prevent flow separation from the tunnel walls.
Three screens and one honeycomb section are located just upstream of the entrance cone to the test
section to reduce turbulence and eliminate tunnel surge.
_/ '
c
~-
t
Screens
Test Section
Variable
Pitch Fan
Figure 1-1. Schematic (top view) of the Merrill Subsonic Wind Tunnel Facility
located at North Carolina State University in Raleigh.
Well suited for probe calibration, the wind tunnel maintains steady, low-turbulence, uniform flow in
the test section, which is large enough to avoid flow disturbance by the presence of the probe. Wind
tunnel calibration has shown that the velocity profile is constant to within less than 1% in a 6-inch
cube located in the center of the test section, where all measurements for these experiments, except
for the entry wall effects, were made.
The actual wind tunnel velocity during each test is measured using an elliptical-nose, standard pitot-
static tube manufactured by Airflow Developments Limited, High Wycombe, England. Widely
regarded as the best available, this standard probe is the only one recommended for use without
individual calibration and has a pitot coefficient of unity. All test probes were calibrated against this
standard tube.
A digital inclinometer is used to measure angles of rotation with respect to the wind tunnel
longitudinal axis (yaw angle) to a visually estimated accuracy of ±0.1°. The tunnel also has an
installed device, termed a "pitch board" (see Section 2.1), that allows the pitot tube to be rotated in
the flow direction (pitch angle) to a visually estimated accuracy of ±0.5°.
Gas temperature, used in calculating gas velocity, is measured using a standard chrome-alumel
thermocouple, mounted inside the wind tunnel test section. A digital readout of the temperature is
displayed on the outside wall of the tunnel. Because the test section is vented, the barometric
1-3
-------
Wind Tunnel Experimental Results
pressure4 in the room is equal to the pressure in the wind tunnel test section. This pressure is
measured using a standard mercury-filled barometer (manufactured by Central Scientific Company,
Chicago, Illinois) that can be read with a visually estimated accuracy of 0.005 inches of mercury.
1.4 DESCRIPTION OF TEST PROBES
Probes selected for the wind tunnel experiments included those considered most likely to be tested
in the field either because (1) they have been used historically for field testing or (2) earlier wind
tunnel or field tests have demonstrated the probes' ability to measure both velocity and angle of flow.
These highest-priority probes included three types of 2D probes (three manual probes and two
"Autoprobes") and two types of 3D probes:
2D Probes
Modified Kiel (prototype)
TypeS
Modified S
Autoprobe with Type S probe
Autoprobe with Modified S probe
3D Probes
DAT
Spherical
The modified Kiel probe is shown schematically in Figure 1-2. Figure 1-3 is a photograph of the
Autoprobe being inserted into the wind tunnel. The spherical and DAT probes are shown
schematically in Figures 1-4 and 1-5, respectively.
Initial plans called for testing of two additional probes—the French and Loop S—if sufficient time
were available. Of these, only the French probe—a cylindrical probe with a single impact port—was
tested, however, and testing was limited to calibration and determination of pitch angle accuracy.
The Loop S probe was not tested because the wind tunnel entry port would have required
modification to accommodate the probe head.
Two-dimensional probes are used to measure stack velocity pressure and can be used to determine
yaw angle. In Mitchell's study, the ability of the modified Kiel probe to measure total flow was
unaffected by yaw and pitch angles of up to ±30°. The wake port on the probe was found to have
some ability to measure yaw angle and to accurately measure static pressure. The prototype Kiel
probe tested in this study, shown in Figure 1-2, is a standard Kiel probe to which four candidate wake
ports and two Fechheimer ports were added. The four wake port holes are offset 180° from the
Kiel's impact port and are located 1.375, 2.125, 2.875, and 3.625 inches down the probe shaft from
the center of the impact port. The impact port in the probe's venturi head is 1/8 inch in diameter, each
wake port is 3/32 inches in diameter, and each Fechheimer port is 1/8 inch in diameter. Each
Fechheimer port is offset 30° from, and is located 2.375 inches down the probe shaft from, the center
of the impact port. The modified Kiel probe was also fitted with a mock "thermocouple"—a stainless
4 Barometric pressure is synonymous with ambient pressure.
1-4
-------
Wind Tunnel Experimental Results
3.625
Wake Port-
Section AA
0.125
00.25
< 1.50
A
0 0.094
• 0.50
-------
Wind Tunnel Experimental Results
Figure 1-3. Autoprobe positioned for insertion into the wind tunnel.
1-6
-------
Wind Tunnel Experimental Results
1 1/2" O.D.
Figure 1-4. Front and side views of spherical probe.
Probe
Shaft
1" •>
1
r
o2 J
4 o 1 5 p
03 ]
t
1"
1
Top View
>P
Probe
Shaft
4 1 5 -
Section P-P
Side View
Figure 1-5. Top, side, and cross-sectional views of DAT probe.
1-7
-------
Wind Tunnel Experimental Results
steel wire approximately 3 inches in length attached to the top of the probe near the Kiel head to
simulate the presence of a thermocouple. Also tested were three 3/8-inch diameter Method 2 Type S
pitot tubes, which are the probes most commonly used in flow measurements. Three 3/8-inch
diameter "modified S" probes, which have wider orifice separations than standard Type S probes,
were also included in the testing.
Three-dimensional probes are used to provide measurements of pressure, pitch angle, and yaw angle
of the velocity vector, which are used to calculate axial velocity. For these experiments, three DAT
probes and one spherical probe were tested.
Table 1-1 shows the candidate probes for testing in EPA's planned multi-laboratory field study. Also
shown are their key physical dimensions and the experiments performed on the probes in the NCSU
wind tunnel.
Table 1-1. Probes Calibrated and Tested in NCSU Wind Tunnel
Probe
Type
Modified Kiel Prototype
TypeS
Modified Type S
Autoprobe Type S
Autoprobe Modified S
French
Spherical
DAT
ID
MK-la
MK-2 a
3/8S-1
3/8S-2
3/8S-3
USS-1
USS-2
USS-3
AS
AMS
F-2
MS5-2
3D-1
3D-2
3D-3
Probe Characteristics
Outside
Diameter
(in.)
3/4
3/4
3/8
3/8
3/8
3/8
3/8
3/8
3/8
3/8
3/4
1-1/2
1
1
1
Measurement
Port
Diameter (in.)
3/32
3/32
b
b
b
1/4
1/4
1/4
b
1/4
c
1/8
d
d
d
Length
(ft)
4
4
10
10
10
10
10
10
4
4
6
7
10
10
10
Test
Yaw Nulling
X
X
X
X
X
X
X
X
X
X
X
Calibration
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
MK-1 is the modified Kiel probe; MK-2 is the modified Kiel probe with a mock "thermocouple" added. See section 2.
b Elliptical ports 15/32" x 9/32"
0 Conical impact port—5/16" leading edge tapered to 3/16" orifice; wake port—3/16"
d Ports PI, P2, and P3—1/8"; Ports P4 and P5—3/32"
1-8
-------
Wind Tunnel Experimental Results
1.5 DOCUMENT PURPOSE AND ORGANIZATION
This document presents the results of several experiments conducted at the NCSU Merrill Subsonic
Wind Tunnel during the summer of 1996. The results are being used to help design the field testing
phase of EPA's collaborative flow study, which, as currently planned, will be conducted at utility
boilers having distinct stack flow characteristics. The results of the modified Kiel probe experiments
presented in this report were used to refine the design of that probe. Results of the tests performed
on the other probes, along with the calibration coefficients, are also being used to develop the
experimental design for the field study.
Section 2.1 of this document describes the general test procedures used in the wind tunnel
experiments, including the procedures for yaw nulling and for pitch angle determination. The
procedures used in acquiring and handling the data and the steps taken to ensure data quality during
both the testing and data analysis phases of the project are presented in Section 2.2. Section 2.3
discusses the steps implemented to determine the minimum number of replicates required for each
experiment on each probe (based on the 3/8-inch Type S probe). Procedures used for calibrating
probes are described in Section 2.4.
Sections 3 through 8 present the procedures, results, and conclusions specific to each experiment
conducted in the wind tunnel. Section 3 presents and discusses the experiments performed on the
modified Kiel probe, including a comparison of the wake ports and Fechheimer ports in determining
yaw null and an analysis of the sensitivity and accuracy of static pressure determinations using the
wake ports. Section 4 describes the derivation of the calibration coefficients for the 2D probes and
their variability with velocity. Pitch angle effects on velocity measurements made with the 2D probes
are discussed in Section 5 (including the sensitivity of the calibration coefficients to pitch angle and
how pitch angle influences velocity determination). Section 6 examines the accuracy of yaw angle
measurements and the effects of yaw angle on velocity measurements. Section 7 presents and
discusses calibration curves for the 3D probes and the accuracy of the velocity and pitch angle
measurements obtained with each of these probes. The results of the experiments performed to
examine how the determination of yaw null is influenced by a probe's proximity to the tunnel wall are
presented and discussed in Section 8. Overall conclusions and recommendations for follow-up study
are presented in the final section, Section 9. Three appendices accompany this report: Appendix A
presents and analyzes the experiment conducted to determine the minimum sample size required;
Appendix B contains additional plots of the wall-effect experiments; Appendix C contains the raw
experimental data for the results discussed in this report and Appendix D contains the raw data on
3 Vz inch floppy diskettes.
1-9
-------
2.0 OVERVIEW OF DATA COLLECTION,
HANDLING, AND QUALITY ASSURANCE
2.1 OVERVIEW OF FLOW MEASUREMENT PROCEDURES
Volumetric flow is calculated as the product of the average axial flow velocity and the cross-sectional
area of a stack or duct at the measurement location. The accuracy of this calculation depends on the
accuracy of the individual velocity measurements and on the accuracy of the measurement of the
cross-sectional area at the test location. This study focuses on the accuracy of measuring axial
velocity.
EPA Test Method 2 (Determination of Stack Gas Velocity and Volumetric Flow Rate) is based on
the differential pressure measurement technique first developed by Henri Pitot in 1732. The total
pressure of a fluid in motion is the sum of static pressure and dynamic pressure. Static pressure is
due to the random motion of molecules vibrating with thermal energy, transferring momentum from
molecule to molecule and to surrounding surfaces. Static pressure is the pressure one would feel if
one were moving along with the stream. The static pressure of a gas is determined by its temperature
and density. Dynamic pressure is proportional to the square of the velocity and the gas density;
density is, in turn, proportional to molecular weight. The velocity of a moving stream can be
calculated from the following equation:
M
where:
V = flow velocity
R = the gas constant
P, = total pressure
P8 = the static pressure of the gas
P,-PS = dynamic or differential pressure
T = absolute temperature
M = molecular weight of the gas
Thus, the temperature, static pressure, and molecular weight of the effluent stream and the difference
between the total and static pressure must be measured to calculate velocity.
When a pitot probe is placed into a moving gas stream, the impact opening of the probe will
experience the total pressure of the system. Static pressure must be measured at a point where the
flow is undisturbed. Many velocity probes do not sense the true static pressure directly; instead, for
various design reasons, they sense a pressure related to it. A probe-specific calibration factor Cp is
introduced to account for this difference. Incorporating Cp into Equation 2.1 provides the final
equation as applied in Test Method 2:
2-1
-------
Wind Tunnel Experimental Results
V = KC(JSP)
\
P.M
(Eq. 2.2)
s
where
Kp = 85.49 = pilot tube constant to calculate velocity in fps
AP = P.-P,
A well constructed standard, or "L-shaped", pilot tube can have a calibration coefficienl of unily,
indicating lhal Ihe probe measures Irue lolal and slalic pressures. Unfortunalely, Ihe slandard pilol-
slalic probe is rarely suilable for repealed use in hostile environments such as power-planl slacks.
Because of ils robust design, the probe most commonly used for Test Method 2 is the Stausscheibe,
or "Type S" probe. The static port of Ihe Type S probe is localed on Ihe back side of Ihe probe,
where the pressure is significantly affected by flow over the probe head. Cp values for Type S pitol
lubes therefore generally range from about 0.80 to 0.86. Other types of probes have different Cp
values, depending on probe shape and static port placement.
The NCSU wind tunnel is vented to the atmosphere; thus, slatic pressure is equal to barometric or
ambienl pressure. The gas is air, and Ihe molecular weighl is 28.97 Ib/mole. Thus, Equation 2.2 can
be simplified lo yield Equation 2.3, as follows:
V = 15.9C
p
\
P..
(Eq. 2.3)
bar
where
15.9 = scale faclor lo calculale velocity in fps
AP = differential pressure in inches of waler
TF = lemperalure in °F
Pbar = baromelric pressure in inches of mercury
Equation 2.3 can be rearranged lo calculale AP:
P V2
bar v
,
252.8 C*(TF + 460)
- 2-4)
'
For Ihe experimenls conducled in Ihis sludy, wind lunnel velocities were sel by adjusting Ihe blade
pilch of a variable-pilch, conslanl-RPM (revolutions per minule) fan (see Figure 1-1) until the desired
AP (calculated by selling Ihe nominal velocity in Equation 2.4) is reached. The AP in Ihe wind lunnel
was measured wilh Ihe slandard pilot-static tube (which has a Cp of 1) described in Section 1 . This
2-2
-------
Wind Tunnel Experimental Results
pitot tube was inserted through an opening in the top of the wind tunnel and then lowered into
position in the center of the test section. Temperature was taken with a standard chrome-alumel
thermocouple mounted inside the wind tunnel test section, and barometric pressure was read using
a standard mercury-filled barometer located outside the tunnel in the laboratory.
The pressure drop between two points in the nozzle section of the tunnel was continuously monitored
by an electronic pressure transducer. The signal from the transducer (in millivolts) was displayed
continuously by a digital readout near the tunnel. The signal from the transducer was read and
recorded when the velocity was set using the standard pitot tube. To maintain a constant pressure
drop in the tunnel, the fan blade pitch was adjusted until a constant reading from the electronic
transducer was obtained.
Test probes were secured to portable floor stands and then leveled with an inclinometer. Desired yaw
angles were also established with the use of an inclinometer. Each test probe was connected to a
series of magnehelic differential pressure gauges with brass fittings and 1/8-inch tubing (Figure 2-1).
Pressure differences (in inches of H2O) were read directly from these gauges. Desired pitch angles
were determined using a board located just under the opening to the test port (illustrated in
Figure 2-2). Pitch angles were changed by rotating the probe stands until the probe was in line with
the desired angle delineated on the pitch board.
Leak checks were performed on the probes before they were inserted into the wind tunnel. Each
probe pressure line was connected through a tee to a common pressure vessel and the appropriate
magnehelic gauge port. After sealing each probe pressure orifice with tape, the vessel was
pressurized and the magnehelic gauges were read. After several minutes, the gauges were re-read;
the probes were assumed leak-free if none of the readings changed. The potential for leaks was also
monitored during the test by observing the magnehelic gauges. Following completion of successful
leak checks, the probes were inserted through the test port and the probe tips positioned at the center
of the wind tunnel (approximately 22.5 inches from the inner wall).
2.2 YAW ALIGNING AND VELOCITY MEASUREMENT PROCEDURES
Procedures for yaw nulling and measuring velocity are summarized, by probe, in the following four
subsections.
2.2.1 Modified Kiel Probe
Two alternative procedures were tested to evaluate the accuracy of yaw angle determinations with
the modified Kiel probe. In the first procedure, the probe was rotated so that the wake port was at
a yaw angle of-15° relative to the flow as read by the digital inclinometer. The wake port pressure
minus the Kiel impact port pressure (APW) was read and recorded. This procedure was repeated at
yaw angles of-10°, -5°, 0°, 5°, 10°, and 15°. APW vs. yaw angle was plotted to determine the
position of the probe at which the yaw angle would be zero (i.e., the yaw null position).
The second procedure evaluated the yaw-angle accuracy obtained using the Fechheimer ports. To
do this, the Kiel impact port was oriented into the direction of flow. The probe was rotated so that
the impact port was at a yaw angle of-15° to the flow as read by the digital inclinometer. The
pressure difference across the Fechheimer ports (APF) was read and recorded. This procedure was
2-3
-------
Wind Tunnel Experimental Results
9 0
Figure 2-1. Magnehelic meters used to measure differential
pressures (PrP2, PI-PS* P^Ps) across various ports of test probes.
2-4
-------
Wind Tunnel Experimental Results
Air Flow
Direction
-30
-20
Positive Yaw
Angle Rotation
Test Port
Negative Yaw
Angle Rotation
Pitch Board
Figure 2-2. Schematic (overhead perspective) showing positive and
negative pitch angle and yaw angle orientation of flow probe
with respect to test port for wind tunnel and flow directions.
2-5
-------
Wind Tunnel Experimental Results
repeated at yaw angles of-10°, -5°, 0°, 5°, 10°, and 15°. APF vs. yaw angle was plotted to
determine the angle 0y at which APF = 0.
When the Fechheimer ports were used for yaw nulling (see, e.g., Section 3), the impact port was
aligned with the flow and rotated until APF = 0. The yaw null angle, 6y, relative to the axis of the
duct was then determined with the use of the digital inclinometer. The AP between the impact and
wake ports was used to determine velocity in the direction of flow. Velocity was multiplied by cos6y
to determine axial flow.
2.2.2 Type S, Modified S, and French Probes
Yaw angles for the Type S and modified S probes were determined by rotating the probes until a null
reading was obtained. The probe was then rotated 90° from the null to position the impact port of
the pitot head directly into the flow. Using the digital inclinometer, yaw angle was then determined
from the angle between the pitot tube and the axis of the duct or stack. The differential pressure and
temperature readings were then recorded. The velocity component in the direction of stack gas flow
(i.e., total velocity) was calculated using Equation 2.3. The yaw-corrected axial component of
velocity was determined by multiplying the total velocity by cos0y. Insufficient information was
available to the experimenters to develop a yaw nulling procedure for the French probe. For this
probe, the pitot tube was inserted into the tunnel with the impact port pointed into the flow (i.e., a
nominal yaw angle of zero between the impact port and the axis of flow).
2.2.3 Autoprobe Type S and Modified S
With the Autoprobe, probe rotation is accomplished by a computer-controlled stepper motor and
notched-belt drive. The Autoprobe locates the null angle by (1) rotating the probe by coarse
increments to locate an approximate null, (2) then rotating by finer increments through the
approximate null to verify and improve the approximate null value, and (3) performing a linear
regression to find the most probable null angle. Yaw null values are displayed on the computer
monitor and can be saved in a file. Once the yaw angle has been determined, the probe is rotated 90°
to determine the AP for the velocity measurement. The axial velocity was calculated by multiplying
the measured impact velocity by cos0y.
2.2.4 DAT and Spherical Probes
The five ports on the sensing heads of the DAT and spherical probes are numbered 1 through 5 (see
Figures 1-4 and 1-5), and the pressures measured at each port are referred to as P,, P2, P3, P4, and
P5. Ports P2 and P3 are located symmetrically on either side of impact pressure port Pt. These 3D
probes were yaw nulled by rotating the probe until a zero differential pressure between ports P2 and
P3 was obtained. The yaw angle was then determined with the use of the digital inclinometer that had
been set to zero on the axis of the wind tunnel. Clockwise rotation of the probe was considered
positive, and counter-clockwise rotation, negative.
The pressure differential between P4 and P5 is a function of pitch angle and gas stream velocity. The
pitch angle, 0,,, is obtained from a calibration curve that plots the ratio F, = (P4-P5)/(P,-P2) vs. pitch
angle. Flow toward the tester (P4>P5) is considered positive, and flow away from the tester (P5>P4)
is considered negative. The probe's calibration curve also provides 3D pitot calibration coefficients,
2-6
-------
Wind Tunnel Experimental Results
F2 = Cp^-PJAJVPz)]172, as a function of pitch angle, 0p. (In this case, the standard pilot-static tube
is used, so Cps= 1.) Axial velocity is obtained from 0y, 0p, F 2, Pr P2, temperature, and barometric
pressure (see Section 7, Equations 7.3 and 7.4).
2.3 DATA ACQUISITION, HANDLING, AND QUALITY ASSURANCE
Cadmus developed extensive data-handling and chain-of-custody procedures to ensure that
high-quality information was used in the data analyses. In addition, quality checks of the data
recording procedures, data base handling procedures, and computer code were conducted throughout
this study. General data handling procedures are outlined below.
In the wind tunnel laboratory, data associated with all manual probes (i.e., all probes except the
Autoprobe) were recorded on standardized data collection forms. The data collection forms were
developed to ensure that a consistent and complete set of information was recorded for each run of
a specific experiment. Information that uniquely identified each run was recorded on the data
collection form. These data included date, probe type, probe ID, nominal velocity, pitch angle, yaw
angle, repetition number, and (for entry wall effects experiments) distance from the wall. For the
Autoprobe, this information was recorded in a laboratory notebook and in the electronic file
containing the Autoprobe readings (including run number and count time). This procedure enabled
the Autoprobe electronic data to be associated readily with the information contained in the
laboratory notebook to ensure that the electronic data could be identified.
Completed data sheets and floppy disks were transferred from the wind tunnel to the Cadmus office
in Durham, North Carolina, for quality assurance and analysis, and a hard copy of the information was
retained by the NCSU principal investigator. The information was duplicated, and copies of the data
sheets, laboratory notebook, and floppy disks were stored outside of the Cadmus office.
For the manual probes, data from the data recording sheets were entered into an ASCII file. The
ASCII computer files were checked by two independent observers to ensure the data were transferred
without error. The data were plotted and outliers identified. The outlier data were again checked
to verify they were not the result of a data transfer error. Suspect data were also examined by NCSU
researchers to determine whether either an experimental error or a transcription error could have
occurred. If explainable errors were identified, the data were marked and corrections made, if
appropriate.
Autoprobe data useful for the analyses were extracted from the electronic file using a computer
program written by Cadmus. The computer code was independently checked to ensure that the
proper fields were extracted from the original file. The Autoprobe data and manual probe data were
stored in SAS® data sets for analysis. Autoprobe data were transferred to spreadsheet software files
for analyses and plotting.
All SAS® code used to plot or analyze the data were checked by an independent programmer to
ensure no programming errors had occurred. In addition, the resulting plots and analysis outputs
were examined for validity. When suspect information was found, the computer code and associated
data were again examined to ensure accuracy.
2-7
-------
Wind Tunnel Experimental Results
2.4 STUDY TO DETERMINE MINIMUM SAMPLE SIZE
Early in the wind tunnel experimental program, a study was performed to establish the number of
replicates required for evaluating accuracy of the yaw angle, calibration coefficient, and velocity. The
experimental design and detailed results are presented in Appendix A. A 3/8-inch Type S pitot tube
was tested at
• velocities of 30, 60, and 90 fps;
• pitch angles from -20° to 20° in 10° increments; and
• yaw angles from -15° to 15° in 5° increments.
Eight replicate measurements were taken.
The Type S probe was used in the pilot test because its measurement variance was expected to be
representative of the probes involved in this study. AP for both the Type S and standard pitot tubes
was measured to calculate the calibration coefficient, Cp, using the following equation:
c' = ^
where,
APP = differential pressure for the standard pitot tube, and
AP8 = differential pressure for the Type S pitot tube.
The equation for estimating sample size is
(Eq. 2.5)
where,
n = estimated sample size;
Z^ = Z statistic that indicates the confidence desired (e.g., 90% or 95%) in the
measured calibration coefficient;
ad2 = variance in the calibration coefficient estimates (calculated as sample variance
using the eight replicates); and
E = level of error (e.g., 20% or 10%) in the average calibration coefficient estimation
that is acceptable.
For example, a sample size can be estimated that allows us to be 95% confident that the measured
Cp is within 5% of the true mean Cp. Sample sizes were calculated for selected values of Z (80%,
90%, 95%) and E (20%, 10%, 5%). Calculated sample sizes were then examined at different
velocities, pitch angles, and yaw angles. The results of the study (see Appendix A) indicate that the
2-8
-------
Wind Tunnel Experimental Results
variability in the calibration coefficients is very small over all velocities, pitch angles, and yaw angles.
When averaged across the tested yaw angles, fewer than three replicates for each experiment were
sufficient to achieve less than 5% error in the results with 99% confidence. Therefore, to be
conservative, the number of replicates for all experiments was fixed at three (with the exception of
one modified S probe calibration, for which measurements were conducted in duplicate).
2.5 CALIBRATIONS OF PROBES
Two general calibration procedures were used, depending on the type of probe. Each probe to be
tested in the field was calibrated according to the following procedures at three velocities: 30, 60,
and 90 fbs. The modified Kiel and Type S probes were also calibrated at velocities of 40, 50, 70, and
80 fps.
2.5.1 Single Calibration Factors
Single calibration factors at yaw and pitch angles of 0° were derived for the modified Kiel, Type S,
modified S, French, Autoprobe Type S, and Autoprobe modified S probes. The standard pitot-static
tube was inserted into the wind tunnel so that the port used for determining total pressure was located
in the center of the wind tunnel test section; the wind tunnel speed was initially set to an approximate,
desired velocity (usually 30, 60, or 90 fps). The true wind tunnel speed was then calculated using the
APP for the standard pitot-static tube and the gas temperature and barometric pressure (see Equations
2.3 and 2.4). The standard pitot tube was then withdrawn from the wind tunnel test section and the
test probe was inserted at an orientation of 0° pitch and 0° yaw. The AP between the impact and
static ports of the test probe was then measured.
This procedure was repeated for each 2D probe at each test velocity. This procedure was also used
to evaluate the effect of pitch angle on the calibration coefficient, by repeating this process at pitch
angles between ±30° pitch in 10° increments (including 5° and -5°).
2.5.2 DAT and Spherical Probe Calibration Curves
For the DAT and spherical probes, calibration curves were derived as specified in draft Method 2F
for velocities of 30, 60, and 90 fps. Calibration of 3D probes involves generating two types of
calibration curves. The first curve is obtained by plotting the ratio F, of the AP across the fourth and
fifth ports, P4-P5, to the AP across the first and second ports, Pj-P2 [or, F! = (P4-P5)/(P,-P2)] against
known pitch angles. The second curve is constructed by plotting the 3D pitot coefficient, F2, against
known pitch angles; F2 is equivalent to Cpg^-PJ/^i-P^]14 where C^ is the standard pitot tube coef-
ficient (=1 for the standard pitot tube used inthese experiments) and P,-P8 is the differential pressure
across the ports of the standard pitot tube.
The calibration procedure began by inserting the standard pitot-static tube into the wind tunnel so that
the total pressure port was located in the center of the wind tunnel test section. The wind tunnel
speed was set using the APP for the standard pitot-static tube and the gas temperature and barometric
pressure. The standard pitot tube was then withdrawn from the wind tunnel test section, and the 3D
test probe was inserted in its place.
Consistent with draft Method 2F, the 3D probes were first checked to ensure that they were actually
pointing directly into the flow. (The Method 2F calibration procedure requires flow direction to be
2-9
-------
Wind Tunnel Experimental Results
axial.) Probes were yaw nulled by inserting the probe head into the center of the wind tunnel and
rotating the probe until the digital inclinometer read 0°. At that point, the 3D probe was properly
aligned and ready for calibration.
Using the digital inclinometer, pitch angles were established to within ±0.5°. With the 3D probe
oriented at 0° pitch and yaw nulled, readings of Pj-P2 and P4-P5 were taken. These steps were
repeated for each pitch angle (-30, -20, -15, -10, -5, 0, 5, 10, 15, 20, 30°), and the calibrations were
performed at 30, 60, and 90 fbs. The F, and F2 ratios obtained for each pitch angle in the three flow
ranges were averaged, and the average values were plotted against pitch angle. A smooth line was
then drawn through the data points. For each calculated F, value at each velocity, the angles and the
corresponding F2 values were determined from the smoothed calibration curve; differences from the
actual measured angles and F2 values were then determined. For F,, Method 2F requires that the
difference at each comparison be within ±2° for angles between 0° and 40°. For F2 values, the
difference must not exceed ±3.0% of the F2 ratio at the same pitch angle. If these criteria cannot be
met, the applicability of the calibration curve should be limited to a finite velocity range over which
measurements within the specifications can be obtained.
2-10
-------
3.0 MODIFIED KIEL PROBE EXPERIMENTS
The wind tunnel tests of the prototype Kiel probe were designed to:
• evaluate the sensitivity and accuracy of yaw angle determinations using the wake ports,
• evaluate the sensitivity and accuracy of yaw angle determinations using the Fechheimer
ports,
• evaluate the sensitivity and accuracy of static pressure determinations using the wake ports,
• evaluate the sensitivity and accuracy of static pressure determinations at different pitch
angles,
• determine whether the addition of a thermocouple would affect probe performance, and
• obtain a calibration coefficient.
3.1 COMPARISON OF WAKE PORTS AND FECHHEIMER PORTS IN DETERMINING YAW NULL
3.1.1 Test Procedures
Experiments were conducted at velocities of 30, 60, and 90 fps to evaluate the sensitivity and
accuracy of yaw angle determinations using the wake ports and Fechheimer ports. In this
experiment, the yaw angle sensitivities of wake port 1 (closest to the Kiel head) and wake port 4
(farthest from the Kiel head) were tested. The angular sensitivity of wake port 1 was determined
using the following procedure. All wake ports except wake port 1 were covered with tape, and wake
port 1 was oriented into the direction of flow. With the aid of a digital inclinometer, the probe was
rotated so that the wake port was positioned at a yaw angle of -15° relative to the flow. The
difference between the wake port pressure and the Kiel head pressure (APW) was read and recorded.
This procedure was repeated at yaw angles of-10°, -5°, 0°, 5°, 10°, and 15°. APW vs. yaw angle
was plotted for all three velocities. The entire procedure was repeated for wake port 4. No
difference was observed between wake ports 1 and 4; therefore, wake ports 2 and 3 were not tested.
To evaluate the Fechheimer ports, the Kiel head port was oriented into the direction of flow. The
probe was rotated so that the Kiel port was at a yaw angle of-15° to the flow as measured by the
digital inclinometer. The pressure difference across the Fechheimer ports (APF) was read and
recorded. This procedure was repeated at yaw angles of-10°, -5°, 0°, 5°, 10°, and 15°. APF vs.
yaw angle was plotted for all three velocities. The entire procedure was repeated for wake port 4.
3.1.2 Yaw Nulling Results
Figures 3-1 through 3-6 present the results of the yaw angle sensitivity and accuracy tests. Each
figure is a plot of APW and APF vs. yaw angle for a specific wake port and velocity (30, 60 or 90 fps).
In all six plots, APW is approximately constant with yaw angle from -15 ° to 15°. This observation
demonstrates that the wake port cannot be used to find the zero yaw angle (yaw null). Alternatively,
at five of the six tested velocities, APF is shown to be a strong monotonic function of yaw angle and
equals zero when yaw angle equals zero. Even in Figure 3-4 (wake port 4, velocity 30 fps) where
APF vs. yaw angle is not monotonic, APF is zero only at the point where yaw angle equals zero. (The
anomalous, non-monotonic behavior shown in this figure was repeatable but cannot be explained
readily.)
3-1
-------
Wind Tunnel Experimental Results
The data shown in these plots are summarized in Table 3-1. Over the range in yaw angles of-15°
to 15 °, the largest variation in APW is 0.15 inches of water for wake port 1 and 0.05 for wake port 4.
By contrast, the average slope of the APF vs. yaw angle plot is 0.13 inches/degree at 90 fps and
0.05 inches/degree at 60 fps (see also Figures 3-2, 3-3, 3-5, and 3-6). At 30 fps, APF changes by
about 1.5 inches between yaw angles of-5° and 5°.
Table 3-1. AP as a Function of Yaw Angle for Wake Ports 1 and 4
of the Modified Kiel Probe Prototype
Differential
Pressure"
APrwi
APW1
AP™
APW4
Velocity
(fps)
30
60
90
30
60
90
30
60
90
30
60
90
Yaw Angle (°)
-15
-0.20
-0.90
-1.90
0.34
1.50
3.20
-0.20
-0.90
-2.00
0.31
1.43
3.20
-10
-0.15
-0.60
-1.40
0.34
1.50
3.20
-0.17
-0.65
-1.40
0.33
1.44
3.23
-5
-0.07
-0.30
-0.70
0.34
1.47
3.20
-0.85
-0.35
-0.70
0.33
1.44
3.25
0.00
-0.002
-0.01
-0.10
0.34
1.45
3.15
-0.01
-0.02
0.00
0.33
1.40
3.20
5
0.07
0.23
0.50
0.34
1.50
3.30
0.70
0.28
0.60
0.33
1.45
3.20
10
0.12
0.50
1.30
0.35
1.50
3.30
0.12
0.53
1.30
0.33
1.44
3.21
15
0.19
0.74
1.95
0.35
1.50
3.35
0.12
0.71
1.95
0.33
1 .42
3.20
a Nomenclature for this table:
APpwi = APF with wake port 1 open
APW1 = APW between Kiel probe head and wake port 1
APpw4= APF with wake port 4 open
APW4 = APW between Kiel probe head and wake port 4
3.2 SENSITIVITY AND ACCURACY OF STATIC PRESSURE DETERMINATIONS
USING THE WAKE PORTS
3.2.1 Test Procedures
After determining that the Fechheimer ports can be used for yaw nulling (see Section 3.1), a series
of tests was conducted to evaluate the feasibility of using a wake port as the static pressure port for
the Kiel probe across a range of pitch angles. To help assess the performance of the modified Kiel
probe, the sensitivity and accuracy of pressure drop measurements made with the Kiel probe were
compared to measurements of AP made with the standard Type S probe at similar velocities and pitch
angles.
In these experiments, the modified Kiel probe (labeled MK1) was yaw nulled using the Fech-heimer
ports, and APW was measured at pitch angles between -30° and 30° in 10° increments (but including
-5° and 5°) at three velocities (30, 60 and 90 fps). Complete data sets were col-lected using wake
ports 1, 3, and 4. Data for wake port 2 were collected only at 0°. At each measured point, APW, APP
3-2
-------
Wind Tunnel Experimental Results
(the pressure differential across the standard pitot tube), yaw angle as measured by the digital
inclinometer, pitch angle, and temperature were recorded. The Type S probe was yaw nulled using
the procedure described in Section 2.1.1, and AP, APp, yaw angle, pitch angle, and temperature were
recorded. Three replicates of the S probe measurements were taken.
After the initial tests demonstrated that the modified Kiel probe could be used to measure static
pressure and find a yaw null, a stainless steel wire—intended to mimic a thermocouple—was added
to the probe (see Section 1.4) to determine whether a temperature-reading device added to the Kiel
probe would affect probe performance. This modified prototype Kiel probe was identified as MK2.
Only wake port 4 on the MK2 probe was tested. For MK2 only, additional sampling was conducted
at 25, 28, 32, and 3 5 fps.
For each combination of pitch angle, velocity, and wake port, the calibration coefficient was
calculated as:
APp
p (Eq. 3.1)
\j "" w
where
APW = difference between wake port pressure and Kiel head pressure
Cpw, calculated at zero pitch and yaw, was plotted against velocity in the range of 30 to 90 fps in
10 fps increments. These plots were created for all four wake ports of the original probe (MK1), for
wake port 4 of MK2, and for the three replicate Type S probes.
3.2.2 Results of Wake Port Evaluation
Figures 3-7 through 3-10 are plots of C^ vs. velocity over the velocity range 30 to 90 fps for wake
ports 1 -4 for probe MK1. Three measurements were taken at each velocity. A line connects the
means of each set of three measurements taken at each velocity and the bands show maxima and
minima at each velocity. For each wake port, the means of Cpw between 40 and 90 fps are constant
to within 0.02; however, Cpw increases by at least 0.03 as velocity decreases from 40 to 30 fps.
Figure 3-11 is a plot of Cpw vs. velocity for MK2 with additional Cpw determinations at velocities of
25, 28, 32, and 35 fps. The behavior of this probe at velocities less than 40 fps markedly differs from
the behavior at velocities greater than 40 fps, in that Cpw oscillates over a 0.03 range between 40 and
25 fps.
Similar plots for the three Type S probes are shown in Figures 3-12 through 3-14. The mean
calibration coefficient, Cp, for each probe varies by 0.025 and 0.030 from 40 to 90 fps, with a
minimum value at 40 fps and maximum value at 90 fps (but the curves are not monotonic in this
range). For all three Type S probes, Cp increases by about 0.015 between 40 and 30 fps, but the
value of Cp in this range is always less than the value of Cp at 90 fps. This increase at the low end of
the velocity range is about half of the increase observed for the Kiel probes.
3-3
-------
Wind Tunnel Experimental Results
Figures 3-15 through 3-19 show curves fitted to the Cpw vs. velocity data over the range of 40 to
90 fbs for the Kiel probe. (Velocity data at 30 fps were not used because of the apparent instability
of the C,^, measurements below 40 fps noted above.) Data are fit with both a regression line having
slope = 0 (assuming Cpw is independent of velocity) and an unconstrained regression line. Clearly,
little difference between the horizontal and unconstrained fits is evident. Over the velocity range
between 40 and 90 fps, the two regression lines for Cpw differ, at most, by less than 1%. Therefore,
using a constant calibration factor for any of the wake ports in this velocity range is reasonable.
These velocity-independent calibration coefficients, 0.737 and 0.742 for wake port 4 on probes MK1
and MK2, respectively, differ by less than 1%, and in both cases the mean square error (MSE) is less
than IxlO"4 (see Figures 3-18 and 3-19). This comparison demonstrates that the addition of a
thermocouple to the modified Kiel probe would not measurably affect the calibration coefficient at
zero pitch and zero yaw.
Figures 3-20 through 3-22 show similar curve fits to the Cp vs. velocity data for the three Type S
probes over the velocity range from 30 to 90 fps. Regressions constructed with slope = 0 (Cp
constrained to be velocity-independent) for these probes yield calibration coefficients of 0.829, 0.832,
and 0.828, all within less than 1% of each other. Similar to the results for the Kiel probe, mean
square errors for these fits were around 1x10"4.
To determine the robustness of the modified Kiel probe with respect to pitch angle, APW was
measured at pitch angles between -30° and 30° at constant velocity (30, 60, and 90 fps) for wake
ports 1, 3, and 4. Figure 3-23 is a plot of Cpw vs. pitch angle at velocities of 30, 60, and 90 fps for
wake port 1 of probe MK1. Figures 3-24 and 3-25 are similar plots for wake ports 3 and 4 of probe
MK1. Figure 3-26 shows Cpw vs. pitch angle for wake port 4 of probe MK2. For comparison,
Figure 3-27 plots Cp vs. pitch angle for one of the Type S probes (Probe S-l) at the same three
velocities.
Figure 3-23 demonstrates that Cpw for wake port 1 depends strongly on pitch angle at all three
velocities and that the dependence on pitch is similar at all three velocities. Figure 3-24 shows that
the dependence of C^ on pitch angle is much reduced for wake port 3 when compared to wake port
1. Figures 3-25 and 3-26 show that for wake port 4 the Cpw vs. pitch angle curve for wake port 4
on both MK1 and MK2 is relatively flat and that the "thermocouple" does not affect this pitch angle
independence. For wake ports 1 and 3, C^ varies linearly between pitch angles of-20° and 20° and
remains unchanged between -30° and -20° and between 20° and 30°. The dependence of Cp on pitch
angle is similar at each tested velocity.
Table 3-2 quantitatively summarizes the stability of the calibration coefficients across pitch angles
for the modified Kiel probes and the Type S probe. The table lists the range of calibration coefficients
for each wake port and the Type S probe at each velocity across the range of pitch angles. The
stability of the modified Kiel probe with respect to pitch angle improves markedly as the location of
the wake port is moved farther from the Kiel head. The Type S probe is more sensitive to changes
in pitch angle than any Kiel probe configuration, with changes in Cp ranging from 26.52% to 29.51%
over the measured pitch-angle range. The largest change in Cp for the Type S probe occurs at pitch
angles beyond ±20°. In the range of-20° to 20°, the pitch sensitivity of the Type S probe ranges
3-4
-------
Wind Tunnel Experimental Results
from 6.10% to 9.00% (see Figure 3-27), similar to the sensitivity of the modified Kiel probe with
wake port 3, which ranges from 7.49% to 9.91% (see Figure 3-25).
Table 3-2. Calibration Coefficients Across Pitch Angles" for Various Wake Ports of the
Modified Kiel Probe Prototypes (MK1 and MK2)b and for the Type S (S-l) Probe at 0° Yaw
Minimum
Mean0
Maximum
Range
% Change*
Velocity (fps)
30
MK1
Wld
0.71
0.78
0.86
0.14
20.03
W3
0.77
0.80
0.84
0.08
9.91
W4
0.76
0.78
0.81
0.06
7.55
MK2
W4
0.75
0.78
0.80
0.05
6.01
TypeS
0.84
0.90
1.06
0.22
26.52f
60
MK1
Wl
0.67
0.75
0.82
0.15
21.64
W3
0.72
0.75
0.78
0.06
8.09
W4
0.74
0.75
0.77
0.03
3.94
MK2
W4
0.72
0.75
0.76
0.05
6.03
TypeS
0.80
0.86
1.03
0.23
29.43g
90
MK1
Wl
0.67
0.76
0.82
0.15
22.26
W3
0.72
0.75
0.78
0.05
7.49
W4
0.73
0.75
0.76
0.03
3.96
MK2
W4
0.72
0.75
0.76
0.05
6.28
TypeS
0.79
0.87
1.03
0.23
29.51"
Pitch angles from -30° to 30°
MK1 is the modified Kiel probe prototype; MK2
is the modified Kiel probe prototype with mock thermocouple
Average of three replicates
Wl= Wake Port 1
W3= Wake Port 3
W4 = Wake Port 4
f % change = 6.10% between -20° and +20°
8 % change = 7.80% between -20° and +20°
h % change <= 9.00% between -20° and +20°
% change in Cr* >
minC..
100
3.3 CONCLUSIONS
The test results described in this section lead to the following conclusions regarding the prototype
modified Kiel probe:
• Wake ports are too insensitive to be used for yaw nulling.
• Fechheimer ports can be used to yaw null with an accuracy of better than 1 ° at velocities
between 40 and 90 fps and at all pitch angles in the range ±30°.
• At pitch and yaw angles equal to zero, a wake port placed between 1-3/8 inches and
3-5/8 inches from the Kiel head can be used to measure static pressure. For velocities
between 40 and 90 fps, each port can be used to derive a constant calibration coefficient.
For a wake port located 3-5/8 inches from the Kiel port, pressure drop measured by the
probe is independent of pitch angle in the range ±30°; for wake ports placed closer to the
Kiel head, measured pressure drop is dependent on pitch angle.
• Thus, the modified Kiel probe using wake port 4 can be used to determine total stack flow
and yaw angle in the velocity range 40 to 90 fps for pitch angles in the range ±30°. The
probe cannot be used to determine pitch angle.
3-5
-------
Figure 3-1. Delta Pw and Delta Pf vs. Yaw Angle:
Modified Kiel - Wake Port 1 - 30 fps
N^ •
0 3 '
VJ.O
O 0 2
c^j u'^
X 01
. U. i
c
^ 0
g °
0- _0 1
vy . I
CD
"fli -0 9 F
VW \J .£- ti
Q
-0 3
\j . \j
-O4 _
•
T^-^
j
1
r
—--^
I
^-^
_I
1
^^
-• — "
I
— -—•
b^
•
^-^
•
^^
j
^c
— "
^^
^^£
:T^
^^^
•
0 ^
v/.o
0 2 O
r] \J.£. \_J
CM
01 ^
\j . i
0 ^3-
vy ^™^
M—
-01 °-
5
-0 2 "CD
u.z Q
-0 3
\j ,\j
_.n 4
-15
-10
-505
Yaw Angle (Degrees)
10
15
-m- Delta Pw
Delta Pf
-------
Figure
1.5 •
1 -
O
X 0.5 -
^ 0 -
><^ w
a.
m -D S
\\J \J . +J
Q 1 c
-1 f> _
3-2. Delta Pw and Delta Pf vs. Yaw Angle:
Modified Kiel - Wake Port 1 - 60 fps
• ™ ...... • , n , •
^
^
^
• — •
^
•
^
~?T
1
^-^
1
^^
H^"^
_^
^^
3
T
r
^—^
U
'
p
^^
• 1.5
x^~«
: O
05 ^
c
0 ^
H—
CL
-0 S ft
\j . <-* ^_i
"0
Q
-15 -10
-505
Yaw Angle (Degrees)
10 15
Delta Pw -B- Delta Pf
-------
Figure 3-3. Delta Pw and Delta Pf vs. Yaw Angle:
Modified Kiel - Wake Port 1 - 90 fps
O 2
CM
1
0
£
0 -2
Q
-3
-15
-10
-505
Yaw Angle (Degrees)
10
CM
0
-2
-3
Q_
03
15
Delta Pw
Delta Pf
-------
Figure 3-4. Delta Pw and Delta Pf vs. Yaw Angle:
Modified Kiel - Wake Port 4 - 30 fps
8 °-5
0
a.
CO
-O.i
-1
C3-
-15
-10
/
-505
Yaw Angle (Degrees)
\
10
-ED
CM
X
0 ^
M—
Q_
3
-0.5 ®
15
Delta Pw -B- Delta Pf
-------
Figure 3-5. Delta Pw and Delta Pf vs. Yaw Angle:
Modified Kiel - Wake Port 4 - 60 fps
15 i *
i
1
o
I 05-
^L ^ • *J
m
c
^ 0
^^ \j
Q_
m -0 *5
\\j \j.\j
.1 R _
•
^^
•
^
•
^
•
^
•_
3^
1
^
•—
^^
_______•
-E
I
^^
•
5,
-^^
•
3 -~
1
_ -t
O"^^
05 ^
w. sy T^
c:
0 ^
v/ -
QL
_n 5 CO
\j . \j ^_>
"o5
-15 -10
-505
Yaw Angle (Degrees)
10 15
Delta Pw -a- Delta Pf
-------
Figure 3-6. Delta Pw and Delta Pf vs. Yaw Angle:
Modified Kiel - Wake Port 4 - 90 fps
o
CM
^ °
0--1
£
1i5 -2
Q
-3
-15
-10
-505
Yaw Angle (Degrees)
10
i2
1
0
CM
C
3
"(D
Q
15
Delta Pw
Delta Pf
-------
Figure 3-7. Probe: Modified Kiel - MK1
Wake Port Calibration Coefficients
Wake Port=1
Pitch = 0, Yaw=0
to
0.82-
0.81
0.80-
0.79-
0.78 -
0.77
0.76 ^
d °75 H
0.74-
0.73
0.72-
0.71 -
0.70-
0.69
0.68
T
20
25
' I '
30
35
1 I '
40
1 I '
45
1 I '
50
' i
55
60 65
Velocity (fps)
1 i '
70
1 I '
75
80
J\ '
85
' i '
90
1 i '
95
100
Bands Indicate the Minimum and Maximum Values
c:flow\statpres\mk1stata.dat; c:\sas\mk1plot4.sas
-------
Figure 3-8. Probe: Modified Kiel - MK1
Wake Port Calibration Coefficients
Wake Port=2
Pitch=0, Yaw=0
0.82-
0.81 -
0.80-
0.79-
0.78-
0.77-
0.76-
0.75-
0.74-
0.73-
0.72-
0.71 -
0.70
0.69
0.68 J
I
20
25
30
r I '
35
40
45
1 I '
50
1 I '
55
1 l '
60
T \ '
65
' l '
70
75
1 r T
80
1 l '
85
1 l '
90
1 l '
95
100
Velocity (fps)
Bands Indicate the Minimum and Maximum Values
c:flow\statpres\mk1stata.dat; c:\sas\mk1plot4.sas
-------
Figure 3-9. Probe: Modified Kiel - MK1
Wake Port Calibration Coefficients
Wake Port = 3
Pitch = 0, Yaw=0
£
O
0.82
0.81
0.80
0.79
0.78
0.77
0.76
0.75
0.74
0.73
0.72
0.71
0.70
0.69
0.68
20
25
1 I '
30
1 I '
35
40
45
50
' I
55
60
Velocity (fps)
l '
65
1 I '
70
75
80
85
1 I '
90
95
100
Bands Indicate the Minimum and Maximum Values
c:flow\statpres\mk1stata.dat; c:\sas\mk1plot4.sas
-------
Figure 3-10. Probe: Modified Kiel - MK1
Wake Port Calibration Coefficients
Wake Port=4
Pitch = 0, Yaw=0
rt o
0.82
0.81 -
0.80-
0.79
0.78-
0.77-
0.76
0.75-
0.74-
0.73
0.72-
0.71-
0.70
0.69
0.68 H
20
25
30
1 \ T
35
40
45
1 l '
50
1 l '
55
1 l '
60
1 i T
65
1 i '
70
'I '
75
80
85
90
95
100
Velocity (fps)
Bands Indicate the Minimum and Maximum Values
c:flow\statpres\mk1stata.dal; c:\sas\mk1plot4.sas
-------
U)
ON
O
0.82-
0.81 :
0.80
0.79
0.78-
0.77-
0.76-
0.75-
0.74
0.73-I
0.72
0.71 -
0.70-
0.69
0.68 H
20
Figure 3-11.
Probe: Modified Kiel With Thermocouple - MK2
Wake Port Calibration Coefficients
Wake Port=4
Pitch = 0, Yaw=0
25
30
35
40
1 i '
45
1 I r
50
55 60 65
Velocity (fps)
70
75
80
85
90
95
100
Bands Indicate the Minimum and Maximum Values
mk2ca)a.dat mk2plot1.sas
-------
Figure 3-12. Probe: 3/8" Type S-1
Calibration Coefficients
Pitch=0, Yaw=0
0.90
0.89-
0.88
0.87-
0.86-
0.85
0.84 H
0.83
0.82 H
0.81
0.80-
0.79-
0.78-
0.77-
0.76
\
20
1 I r
30
40
T r r
50
60
Velocity
1 i '
70
80
90
100
Bands Indicate the Minimum and Maximum Values
3-81cal2.dat 3-81c2pl.sas
-------
oo
0.90-
0.89-
0.88
0.87-
0.86-
0.85
0.84-
0.83-
0.82-
0.81
0.80 H
0.79
0.78 H
0.77
0.76 H
20
Figure 3-13. Probe: 3/8" Type S-2
Calibration Coefficients
Pitch = 0, Yaw=0
30
40
50
60
Velocity
1 i '
70
80
90
100
Bands Indicate the Minimum and Maximum Values
3-82cal2.dat 3-82c2pl.sas
-------
Figure 3-14. Probe: 3/8" Type S-3
Calibration Coefficients
Pitch=0, Yaw=0
0.90-
0.89-
0.88
0.87-
0.86
0.85-1
0.84
3- 0.83
^
0.82
0.81
0.80
0.79
0.78-
0.77-
0.76
20
30
40
1 I '
50
60
Velocity
70
80
1 l '
90
100
Bands Indicate the Minimum and Maximum Values
3-83cal2.dat 3-83c2pl.sas
-------
Figure 3-15. Probe: Modified Kiel - MK1
Wake Port Calibration Coefficient
Wake Port=1
Pitch = 0, Yaw=0
1.0
0.9
o O
0.8
0.7-
0.6
20
30
40
50 60
Velocity (fps)
70
• Cpw vs. Velocity
— Cpw=0.7588-0.000175*Velocity, MSE = 0.00028
80
90
100
Cpw=0.7474, MSE = 0.00027
mklstatascC j:rsgavg.sas
-------
Figure 3-16. Probe: Modified Kiel - MK1
Wake Port Calibration Coefficient
Wake Port=2
Pitch = 0, Yaw=0
1.0-
0.9-
Is)
0.7-
0.6-
20
30
40
50 60
Velocity (fps)
70
80
90
100
• • • Cpw vs. Velocity
Cpw=0.7495 + 0.000302*Ve!ocity, MSE = 0.00003
•-- Cpw=0.7692, MSE = 0.000056
mk1stata.sd2 JTsgavg.sas
-------
Figure 3-17. Probe: Modified Kiel - MK1
Wake Port Calibration Coefficient
Wake Port=3
Pitch = 0, Yaw=0
1.0-
0.9-
d
0.7-
0.6-
20
30
40
50 60
Velocity (fps)
I
70
80
90
100
Cpw vs. Velocity
Cpw=0.7554 + 0.000031 *Velocity, MSE = 0.00004
Cpw=0.7574, MSE = 0.00004
cnklstatn Rfg J:regavg.sas
-------
Figure 3-18. Probe: Modified Kiel - MK1
Wake Port Calibration Coefficient
Wake Port=4
Pitch = 0, Yaw=0
1.0
0.9
I
O
0.8
NJ
u>
0.7
0.6-
20
30
I
40
50 60
Velocity (fps)
70
80
i
90
100
Cpw vs. Velocity
Cpw=0.7196 + 0.000267*Velocity, MSE=0.00008
Cpw=0.7369, MSE = 0.000096
mklstaJascC i:regavg.sas
-------
Figure 3-19.
Probe: Modified Kiel With Thermocouple - MK2
Wake Port Calibration Coefficient
Wake Port=4
Pitch = 0, Yaw=0
1.0-
0.9
r 8™
NJ
0.7
0.6-
i
20
30
40
50 60
Velocity (fps)
70
80
90
100
Cpw vs. Velocity
Cpw=0.7430-0.000011*Velocity, MSE = 0.000014
Cpw=0.7423, MSE = 0.000014
mk2cala.sd2 j:regavg.sas
-------
Figure 3-20. Probe: 3/8" Type S-1
Calibration Coefficient
Pitch=0, Yaw=0
1.0-
0.9-
0.8
0.7-
0.6 -I
20
30
40
50
i
60
70
Velocity (fps)
i
80
Cp vs. Velocity
Cp = 0.8108 + 0.000307*Velocity, MSE=0.00008
Cp = 0.8293, MSE=0.0001145
i
90
100
ca!2_381.sd2 j:ragavg.sas
-------
ON
1.0
0.9-
O
0.8-
0.7-
0.6-
20
30
Fgure 3-21. Probe: 3/8" Type S-2
Calibration Coefficient
Pitch = 0, Yaw=0
40
50 60
Velocity (fps)
70
80
Cp vs. Velocity
Cp = 0.8181 + 0.000238*Velocity, MSB = 0.00005
Cp = 0.8323, MSB = 0.0000745
90
100
cat2_3S2.sd2 j:regavg.sas
-------
Figure 3-22. Probe: 3/8" Type S-3
Calibration Coefficient
Pitch = 0, Yaw=0
1.0-
0.9-
O 0.8
0.7
0.6-
20
i
30
40
50
i
60
70
80
90
100
Velocity (fps)
• Cp vs. Velocity
— Cp=0.8134+0.000235*Velocity, MSE = 0.00008
Cp = 0.8275, MSE = 0.000098
cat2_383.ad2 j:regavg.sas
-------
00
0.90-
0.88-
0.86-
0.84-
0.82-
0.80-
0.78
0.76-
0.74
0.72-
0.70
0.68-
0.66-
Figure 3-23. Probe: Modified Kiel - MK1
Calibration Coefficients
Wake Port=1
e
i ^
-30
-25 -20 -15 -10 -5 0
Pitch (degrees)
r
5
10
l
15
l
20
I
25
l
30
•*-* Velocity = 30 (fps) © -^© Velocity = 60 (fps) Q -H- a Velocity = 90 (fps)
Bands Indicate the Minimum and Maximum Values
\statpres\mk1statc.dat mk1cp!02.sas
-------
Figure 3-24. Probe: Modified Kiel - MK1
Calibration Coefficients
Wake Port=3
0.90-
0.88-
0.86
0.84 H
0.82
0.80 H
0.78
0.76
0.74
0.72-
0.70-
0.68-
0.66-
to
I ^
-30
-l | I I- -TTyiiTr|-TTTij -r —I I I i I I I I i I I I I |—
-25 -20 -15 -10 -505
Pitch (degrees)
10
15
r
20
25
30
Velocity = 30 (fps)
Velocity = 60 (fps) D -H- D Velocity = 90 (fps)
Bands Indicate the Minimum and Maximum Values
\statpres\mk1statc.dat mk1cpl02.sas
-------
Figure 3-25. Probe: Modified Kiel - MK1
Calibration Coefficients
Wake Port=4
0.90
0.88
0.86-
0.84-
0.82-
0.80
0.78-
0.76-
0.74-
0.72-
0.70-
0.68-
0.66-
i ^
-30
-25
-20
-15 -10 -5 0
Pitch (degrees)
i
5
10
15
i
20
Velocity = 30 (fps)
Velocity = 60 (fps) D -H- Q Velocity = 90 (fps)
Bands Indicate the Minimum and Maximum Values
\statpres\mk1statc.dat mk1cp!02.sas
^ i
25
I
30
-------
U)
0.90-
0.88-
0.86-
0.84-
0.82
0.80
O 0.78
0.76-
0.74
0.72-
0.70-
0.68-
0.66-
-30
Figure 3-26. Probe: Modified Kiel With Thermocouple - MK2
Calibration Coefficients
Wake Port=4
-25 -20 -15 -10 -5 0
Pitch (degrees)
I
5
r
10
i
15
I
20
I
25
I
30
Velocity = 30 (fps)
Velocity = 60 (fps) D -0- Q Velocity = 90 (fps)
Bands Indicate the Minimum and Maximum Values
\statpres\mk2stat.dat mk2stp02.sas
-------
Figure 3-27. Probe: 3/8" Type S-1
Calibration Coefficients
1.08-
1.04-
1.00
0.96
<§• 0.92-
0.88-
0.84
0.80-
0.76-
-30
-25 -20 -15 -10
-5
I
0
I
5
r
10
i
15
i
20
I
25
i
30
Pitch (degrees)
Velocity = 30 (fps) © -Q-O Velocity = 60 (fps) D -H- D Velocity = 90 (fps)
Bands Indicate the Minimum and Maximum Values
318s1ang.dat 38s1pl08.sas
-------
4.0 STANDARD CALIBRATION COEFFICIENTS AND
VARIABILITY WITH VELOCITY FOR 2D PROBES
4.1 TEST PROCEDURES
All 2D probes tested were calibrated at zero pitch and yaw angles to derive standard (Method 2)
calibration coefficients. The procedure is similar to that specified in EPA Reference Method 2 for
a Type S pilot tube. Method 2 calls for averaging three replicate calibrations at a velocity of 50 fps
and states that the resulting calibration coefficient will generally be valid to within ±3% above
velocities of 17 fps. Alternatively, probes may be calibrated at a minimum of four velocities between
10 and 83 fps.
In the calibrations conducted for these studies, the probes were not yaw nulled. The three modified
S probes, the French probe, and the Autoprobe Type S and modified S were calibrated in triplicate
at velocities of 30, 60, and 90 fps. The three manual Type S probes were calibrated in triplicate at
30,40, 50, 60, 70, 80, and 90 fps. For the modified Kiel probe, discussed in Section 3, the calibration
coefficient became unstable at velocities less than 40 fps. Thus, for the purpose of standard
calibration, only data collected at 40, 50, 60, 70, 80, and 90 fps are used, with the understanding that
the calibration coefficient so derived is not valid below 40 fps.
4.2 CALIBRATION COEFFICIENTS
Figures 4-1 through 4-10 are plots of Cp vs. velocity for each probe calibrated during these tests.
The vertical bar at each velocity denotes the range of the three replicates, and the solid line connects
the means of Cp at each velocity. Table 4-1 shows the mean and range of the calibration coefficients
for each probe at each velocity setting. Also shown is the range as a percent of the mean value. The
largest range around any point for either Autoprobe is less than 0.37% of the mean Cp. For all other
probes, except one of the manual modified S probes, the largest range around any point in the
calibration curve is less than 2.43% of the mean Cp. The ranges around the mean Cp values for the
modified S probe USS-1 are between 3.30 and 3.93%.
4.3 VARIATION WITH VELOCITY
Table 4-2 summarizes across all velocities the data collected during the calibration coefficient
determinations. All reported results, except the coefficients of variation, are derived from the means
of the three data points shown in the plots. The coefficients of variation are derived from the
individual, unaveraged Cp values. The table lists velocity ranges and specific velocities over which
each probe was calibrated, the mean Cp over the velocity range, the minimum and maximum values
of Cp over the velocity range, the coefficient of variation, and the sample size used to calculate the
coefficient of variation. For all probes except the modified S probes, the coefficient of variation is
1.29% or less. The largest coefficients of variation are 3.06% and 3.59%, both for modified S
probes.
4-1
-------
Wind Tunnel Experimental Results
Table 4-1. Calibration Coefficients Obtained for 2D Probes at Specific Velocities
(Pitch Angle = 0°, Yaw Angle = 0°)
Probe Type
TypeS
Modified Kiel
Modified S
French
Auto Type S
Auto
Modified S
Probe ID
3/8 S-l
3/8 S-2
3/8 S-3
MK2-WP4
USS-1
USS-20
USS-3
F-2
AS
AMS
Cp"
Mean
Range
% of Mean
Mean
Range
% of Mean
Mean
Range
% of Mean
Mean
Range
% of Mean
Mean
Range
% of Mean
Mean
Range
% of Mean
Mean
Range
% of Mean
Mean
Range
% of Mean
Mean
Range
% of Mean
Mean
Range
% of Mean
Velocity (in fps)
30
0.832
0.019
2.27
0.835
0.001
0.18
0.831
0.006
0.70
0.752
0.00
0.00
0.939
0.037
3.93
0.863
0.010
1.15
0.914
0.004
0.42
0.740
0.008
1.12
0.814
0.001
0.10
0.864
0.001
0.12
40
0.815
0.015
1.89
0.820
0.016
1.93
0.816
0.012
1.42
0.741
0.017
2.32
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
50
0.821
0.005
0.61
0.826
0.005
0.62
0.822
0.010
1.22
0.746
0.004
0.51
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
60
0.830
0.00
0.00
0.833
0.002
0.22
0.828
0.011
1.28
0.742
0.005
0.70
0.886
0.029
3.30
0.858
0.008
0.94
0.853
0.00
0.00
0.760
0.014
1.82
0.802
0.00
0.02
0.852
0.003
0.37
70
0.822
0.003
0.31
0.826
0.00
0.00
0.821
0.013
1.54
0.740
0.006
0.76
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
80
0.840
0.00
0.00
0.840
0.00
0.00
0.835
0.020
2.43
0.742
0.004
0.58
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
90
0.844
0.002
0.20
0.845
0.00
0.00
0.841
0.016
1.95
0.742
0.003
0.46
0.878
0.030
3.39
0.838
0.00
0.00
0.875
0.00
0.00
0.755
0.004
0.48
0.804
0.00
0.17
0.850
0.001
0.15
% Change in
Mean Cp Across
All Velocities"
3.56
3.06
3.05
1.62
6.95
2.98
7.15
2.70
1.50
1.65
Values for"% of mean" may not agree with those obtained by dividing the value for "Range" by the value for "Mean" due to rounding.
Range and Mean values were rounded to three significant digits, and values for % of Mean were calculated from values with more than
three significant digits.
(Percent Change^
in Mean C
Across Velocities^
max {Mean Cv} - min {Mean Cv}
v-30...90frj
v-30...90frj
100
min {Mean Cv}
Based on duplicate measurements at each velocity, measurements for all other probes based on three or more measurements at each
velocity.
4-2
-------
Wind Tunnel Experimental Results
Table 4-2. Variability of Calibration Coefficients Across Velocities
for 2D Probes Determined at Pitch Angle 0°
Probe Type
TypeS
Modified Kiel
Modified S
French
Auto Type S
Auto
Modified S
Probe
ID
3/8 S-l
3/8 S-2
3/8 S-3
MK2-WP4
USS-1
USS-2
USS-3
F-2
AS
AMS
Velocities
(fps)
30-90°
30-90°
30-90C
40-90d
30,60,90
30,60,90
30,60,90
30,60,90
30,60,90
30,60,90
Statistics of Cp"
Min.
0.815
0.820
0.816
0.740
0.878
0.838
0.853
0.740
0.802
0.850
Mean
0.829
0.832
0.827
0.742
0.901
0853
0.881
0.752
0.807
0.855
Max.
0.844
0.845
0.841
0.746
0.939
0.863
0.914
0.760
0.814
0.864
Coefficient
of Variation
(%)"
1.29
1.04
1.20
0.50
3.59
1.45
3.06
1.22
0.71
0.79
Sample
Size
21
21
28
36
9
6
9
9
9
9
" These statistics were calculated using the mean values of Cp at each velocity.
b Coefficient of variation = (standard deviation/mean) x 100.
c Velocity = 30,40, 50,60,70, 80, and 90 fps.
" Velocity = 40, 50,60,70, 80, and 90 fps.
4.4 CONCLUSIONS
The data presented in this section support the following findings:
• Except for one manual modified S probe, the range of variation in any single-velocity
calibration is less than or equal to 2.43% of the mean Cp value. The range of variation in
the single-velocity calibration of one modified S probe is higher (from 3.30 to 3.93%). For
the Autoprobes, the range is less than or equal to 0.37% of the mean Cp value.
For the three modified S probes, the maximum percentage changes in mean Cp over the
range of tested velocities are 7.15, 6.95, and 2.98%. For all other probes, the percentage
change in mean Cp over the range of tested velocities is 3.56% or less. (See last column in
Table 4-1.)
• Except for the manual modified S probes, the coefficient of variation with respect to velocity
over the range 30 to 90 fps (40 to 90 fps for the modified Kiel probe) is 1.29% or less.
4-3
-------
Wind Tunnel Experimental Results
These findings point to the following conclusions:
• For all probes except the manual modified S, variability in the calibration coefficients derived
at individual velocity settings was low. The variability was especially low for the
Autoprobes.
• As reflected in the percentage change in mean Cp and in the low coefficients of variation,
the calibration coefficients for all the tested 2D probes, except for the modified S probes,
are stable in the 30 to 90 fps range (40 to 90 fps for the modified Kiel probe).
• Across all velocities, the greatest variation in calibration coefficients was exhibited by the
manual modified S probes, reaching as high as 7.15% across the tested range. It may be
advisable to require this probe to be calibrated at a wind tunnel velocity close to the
prevailing velocity where the probe will be used.
• The modified Kiel probe should not be used at velocities less than 40 fps because its
calibration coefficient appears to be very sensitive to velocity changes below this level. This
specific velocity cutoff applies only to the modified Kiel probe dimensions tested here;
modified Kiel probes of the same design but with a different diameter and port sizes may
have different velocity cutoffs.
4-4
-------
Figure 4-1. Probe: 3/8" Type S-1
Calibration Coefficients
Pitch = 0, Yaw=0
0.90-
0.89-
0.88
0.87-
0.86
0.85-
0.84-
0.83
0.82
0.81
0.80
0.79 H
0.78
0.77 H
0.76
I ' ' ' '
20
30
40
1 I '
50
60
Velocity
1 i '
70
80
90
100
Bands Indicate the Minimum and Maximum Values
-------
t
0.90
0.89
0.88
0.87
0.86
0.85
0.84
0.83
0.82
0.81
0.80
0.79 H
0.78
0.77 H
0.76
20
Figure 4-2. Probe: 3/8" Type S-2
Calibration Coefficients
Pitch = 0, Yaw=0
30
40
50
60
Velocity
70
80
1 i '
90
100
Bands Indicate the Minimum and Maximum Values
-------
Figure 4-3. Probe: 3/8" Type S-3
Calibration Coefficients
Pitch=0, Yaw=0
0.90
0.89-
0.88
0.87
0.86-
0.85
0.84-)
0.83
0.82-
0.81
0.80
0.79
0.78
0.77-
0.76-
T
20
T T ' ' ' '
30
T I . I T J 1
40
50
60
Velocity
70
80
1 i '
90
i
100
Bands Indicate the Minimum and Maximum Values
-------
t I
0.82-
0.81 -
0.80
0.79 H
0.78
0.77 H
0.76
0.75
0.74-
0.73-
0.72 :
0.71
0.70
0.69
0.68
20
Figure 4-4.
Probe: Modified Kiel With Thermocouple - MK2
Wake Port Calibration Coefficients
Wake Port=4
Pitch=0, Yaw=0
25
30
35
40
1 l '
45
50 55 60 65
Velocity (fps)
70
75
80
85
1 l '
90
1 i '
95
100
Bands Indicate the Minimum and Maximum Values
-------
f
1.16
1.12
1.08
1.04
1.00
0.96
0.92
0.88
0.84
0.80
0.76
i
30
Figure 4-5. Probe: Modified S - USS-1
Calibration Coefficients
Pitch = 0, Yaw=0
60
Velocity (fps)
i
90
Bands Indicate the Minimum and Maximum Values
-------
0.92
0.91
0.90
0.89-
0.88-
f
o
t O 0.87-
0.86
0.85
0.84
0.83
0.82
Figure 4-6. Probe: Modified S - USS-2
Calibration Coefficients
Pitch = 0, Yaw=0
i
30
Bands Indicate the Minimum and Maximum Values
uss212ca.dat uss-2plt.sas
60
Velocity (fps)
90
-------
Figure 4-7. Probe: Modified S - USS-3
Calibration Coefficients
Pitch = 0, Yaw=0
0.96-
0.95
0.94
0.93 H
0.92-
0.91
O 0.90
0.89
0.88
0.87
0.86-
0.85-
0.84
I
30
r
60
r
90
Velocity (fps)
Bands Indicate the Minimum and Maximum Values
-------
f
i—*
to
0.84
0.83
0.82-
0.81-
0.80
0.79
0.78
i
30
Figure 4-8. Probe: Autoprobe - Type S
Calibration Coefficient
Pitch = 0, Yaw=0
60
Velocity (fps)
90
Bands Indicate the Minimum and Maximum Values
-------
f
0.88
0.87
0.86
0.85-
0.84
0.83
0.82
i
30
Figure 4-9. Probe: Autoprobe - Modified S
Calibration Coefficient
Pitch=0, Yaw=0
60
Velocity (fps)
i
90
Bands Indicate the Minimum and Maximum Values
-------
f
0.80
0.79
0.78
0.77
0.76
0.75
0.74
0.73
0.72
0.71
0.70
I
30
Figure 4-10. Probe: French Probe
Calibration Coefficient
Pitch = 0, Yaw=0
- F-2
60
Velocity (fps)
i
90
Bands Indicate the Minimum and Maximum Values
-------
5.0 PITCH ANGLE EFFECTS ON VELOCITY
MEASUREMENTS PERFORMED WITH 2D PROBES
5.1 TEST PROCEDURES
A series of tests was conducted to determine the sensitivity of velocity measurements to changes in
pitch angle for each probe tested in this study. Probes included the Type S, modified S, Autoprobe
fitted with a Type S probe, Autoprobe fitted with a modified Type S probe, French, and modified Kiel
(see Section 3.2). Data on pitch-angle sensitivity were collected along with data on the sensitivity
and accuracy of yaw-angle nulling using 2D probes. (Note that 2D probes cannot be used to
determine pitch angle.)
For these experiments, each test probe (except the French probe) was yaw nulled using the procedure
described in Section 2.1.1, and AP was measured at pitch angles between -30° and 30° in 10°
increments (but including -5° and 5°) and at three velocities (30, 60, and 90 fps). The Type S,
modified S, and Kiel probes were yaw nulled manually, and the Autoprobes were yaw nulled
automatically. No procedure has been established for yaw nulling the French probe. Thus, all
measurements with this probe were taken at a nominal yaw angle of zero. For each probe, AP, APP
(the pressure differential across the standard phot-static tube), yaw angle as read by the digital
inclinometer, pitch angle, and temperature were recorded.
The calibration coefficient was calculated as
(Eq. 5.1)
AP
where
AP = the pressure drop across the 2D probe
5.2 SENSITIVITY OF CALIBRATION COEFFICIENTS TO PITCH ANGLE
Figure 5-1 shows Cp vs. pitch angle for the Type S probe (Probe S-l) at velocities of 30, 60, and
90 fps. The construction of these curves was discussed in Section 3.2.1. Curves for the modified S
probes (USS-l and USS-3) in Figures 5-2 and 5-3, the Autoprobe Type S in Figure 5-4, and the
Autoprobe modified S in Figure 5-5, are qualitatively similar to Figure 5-1 and to each other. In all
cases, Cp decreases rapidly over the pitch angle range from -30° to -10° and then remains relatively
constant from -10° to 20° before decreasing beyond 20°. (The Autoprobe modified S remains
almost constant above 20° at 30 fbs). The sharp demarcation in pitch angle sensitivity at -10° is less
pronounced with the French probe, as is evident from Figure 5-6. However, the three Cp vs. pitch
angle curves for the French probe differ greatly from each other. At 90 fps, the Cp vs. pitch angle
curve for the French probe is similar to those for the Type S and modified S probes, with Cp showing
little change in the -10° to 20° pitch range. However, the situation changes markedly at 30 to 60 fbs.
At 30 fps, the Cp for the French probe is highly pitch angle-dependent over the entire -30° to 30°
pitch range. At 60 fps, Cp is erratically pitch angle-dependent, snowing high dependency below -5°,
between 5° and 10°, and above 20°. The change in Cpwith pitch angle for the modified Kiel probe
(MK2) is shown in Figure 3-26 in Section 3.2.2. For this probe, Cp remains essentially constant in
5-1
-------
Wind Tunnel Experimental Results
the -30° to -10° pitch angle range and then decreases slightly (between 3% and 4%) in the -10
20° pitch angle range for each of the tested velocities.
to
Table 5-1 summarizes the behavior of Cp with pitch angle for all 2D probes tested, including the
modified Kiel probe (MK2) discussed in Section 3.2. 1 . The Type S and modified S probes in both
manual and Autoprobe configurations, show strong sensitivity to pitch at angles between -30° and
-10° and then relatively pitch-independent behavior (< 5% change) between -10° and 20°. The
changes in the Autoprobe coefficients are roughly half those of the corresponding manual probes over
the -10° to 20° regime. The French probe is highly pitch angle-dependent over small changes in pitch
angle. The modified Kiel probe, on the other hand, is relatively robust across the pitch angle range
studied (see Figure 3-26).
Table 5-1. Sensitivity of Cp to Pitch Angle for 2D Probes
Probe
Type S (S-l)
Modified Sb
Autoprobe Type S (AS)
Autoprobe Modified S (AMS)
French Probe (F-2)
Kiel Probe (MK2)
% Change in Cpa
-30° to -10°
15.9
16.8
18.4
21.0
9.1
0.7
-10° to 20°
6.3
3.2
3.0
2.2
8.6
3.4
" % change is calculated as the difference between the largest and smallest mean Cp in the
angular pitch range divided by the average of those means. These values are then averaged for
30,60, and 90 fps.
b Average of USS-1 and USS-3.
5.3 PITCH ANGLE EFFECTS ON VELOCITY DETERMINATIONS
The usefulness of a probe in determining flow depends on the probe's ability to measure total velocity
and axial velocity accurately at different pitch angles. As stated earlier, 2D probes cannot be used
to determine pitch angle. Instead, pitch angle determination must be made with a 3D probe such as
a DAT probe or spherical probe. Pitch angle determinations with 3D probes are discussed later in
this report.
The issue addressed in this section is how much measurement error is potentially associated with
using a 2D probe to determine axial velocity in the absence of any information about the pitch angle
of flow. To estimate this uncertainty, the data presented in the previous subsection were used to
simulate the use of each probe in determining axial velocity where the unknown pitch angle of flow
is constrained to lie between -30° and 30°.
5-2
-------
Wind Tunnel Experimental Results
The data for the simulations were developed as follows. First, actual total velocity (V8 in feet per
second) was calculated from APP measured by the standard pilot-static probe using the following
equation:
Vs = 15.9
(7*+460)
(Eq. 5.2)
bar
The simulated axial component of the actual velocity at a specific pitch angle, 0p (yaw angle = 0°),
was then calculated by multiplying V8 by the cosine of the pitch angle (set on the pitch board), i.e.,
= cose, vt
(Eq. 5.3)
These values were compared to Vnun, the velocity measured by each of the tested 2D probes at each
pitch angle after yaw nulling. Vnun was derived from the measured AP using the following equation:
V = 15
V null *->•
AP (7V460)
(Eq. 5.4)
bar
where Cp is the average of the calibration coefficients obtained at zero pitch and zero yaw over the
applicable velocity range (i.e., 40-90 fbs for the modified Kiel probe; 30-90 fps for all the other tested
probes). Each of the constituent values of Cp is derived using Equation 5.1 and then averaged.
Table 5-2 compares the velocities measured by the test probes with the velocity determined from the
standard probe corrected for pitch angle. The first three columns list the pitch angle, the cosine of
the pitch angle (cos6p), and the percent of the total velocity that exceeds the angle-corrected
("pseudo" axial) velocity. This latter quantity is the percentage difference between Vnun and VK
where Cp is independent of pitch angle. In other words, because a 2D probe cannot measure pitch,
a probe whose calibration coefficient is independent of pitch would provide a velocity that differs
from axial velocity by (l-cos0p)V8. The fourth column lists the nominal test velocity, Vnom.
Four columns appear in Table 5-2 under each tested probe. The first shows the values of the actual
total velocity V8 as measured by the standard pilot (see Equation 5.2). The second shows Ihe aclual
axial velocity VK derived from Ihe slandard pilol measuremenls (see Equation 5.3). The Ihird shows
Ihe yaw-nulled velocity V^ oblained by Ihe lesled 2D probe (see Equation 5.4). The fourth column
shows Ihe percenl deviation of Vnu]1 from Vm, i.e.,
%Dev =
null
100
(Eq. 5.5)
5-3
-------
Table 5-2. Axial Velocity Evaluation for 2D Probes (a)
Pitch
Angle
-30
-20
-10
-5
0
5
10
20
30
cos theta
sub p
0.866
0.94
0.985
0.996
0
0.996
0.985
0.94
0.866
l-cos
theta p
13.4%
6.0%
1.5%
0.4%
0.0%
0.4%
1.5%
6.0%
13.4%
Vnom (b)
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
Modified Kiel Probe (MK2)
Vs
30.1
60.3
90.6
30.1
60.3
90.6
30.1
60.3
90.6
30.1
60.3
90.7
30.1
60.3
90.7
30.1
60.3
90.7
30.1
60.3
90.8
30.1
60.3
90.8
30.1
60.3
90.8
Vax
26.1
52.2
78.5
28.3
56.7
85.1
29.6
59.4
89.2
30.0
60.1
90.4
30.1
60.3
90.7
30.0
60.1
90.4
29.6
59.4
89.4
28.3
56.7
85.3
26.1
52.2
78.6
Vnull
28.0
59.2
89.1
27.9
58.7
88.6
28.1
58.9
88.6
28.3
59.3
88.4
28.5
59.7
89.6
28.6
60.0
90.0
28.9
60.5
91.3
29.3
61.1
91.4
29.6
62.0
93.4
% Dev (c)
7.4%
13.4%
13.6%
-1.4%
3.6%
4.1%
-5.2%
-0.8%
-0.7%
-5.6%
-1.3%
-2.2%
-5.3%
-1.0%
-1.2%
-4.6%
-0.1%
-0.4%
-2.5%
1.9%
2.1%
3.6%
7.8%
7.1%
13.5%
18.7%
18.8%
Type S (3/8 S-l)
Vs
30.0
60.3
90.9
30.0
60.3
91.0
30.0
60.3
91.1
30.0
60.4
91.2
30.0
60.4
91.3
30.0
60.4
91.3
30.1
60.4
91.4
30.1
60.4
91.5
30.0
60.4
91.5
Vax
26.0
52.2
78.7
28.2
56.7
85.5
29.5
59.4
89.7
29.9
60.2
90.9
30.0
60.4
91.3
29.9
60.2
91.0
29.6
59.5
90.0
28.3
56.8
86.0
26.0
52.3
79.2
Vnull
23.5
48.4
72.2
26.3
54.5
80.8
27.7
57.3
86.3
28.2
58.7
88.1
28.4
59.9
89.4
28.9
61.3
91.5
29.0
61.3
92.6
28.2
59.7
89.9
29.5
62.5
95.1
% Dev
-9.6%
-7.3%
-8.3%
-6.7%
-3.8%
-5.5%
-6.2%
-3.5%
-3.8%
-5.6%
-2.4%
-3.0%
-5.3%
-0.8%
-2.1%
-3.3%
1.9%
0.6%
-2.2%
3.1%
2.9%
-0.3%
5.2%
4.6%
13.5%
19.5%
20.0%
Modified S
USS-1
Vs
30.1
60.2
90.6
30.1
60.2
90.7
30.0
60.3
90.8
30.0
60.3
91.0
30.0
60.3
91.0
30.0
60.3
91.0
30.0
60.3
91.1
30.0
60.4
91.2
30.0
60.4
91.2
Vax
26.1
52.1
78.5
28.3
56.6
85.2
29.5
59.4
89.4
29.9
60.1
90.7
30.0
60.3
91.0
29.9
60.1
90.7
29.5
59.4
89.7
28.2
56.8
85.7
26.0
52.3
79.0
Vnull
24.7
51.7
77.6
27.0
56.9
85.5
29.1
61.2
91.8
28.8
61.3
92.7
28.8
61.3
93.3
28.9
61.4
93.0
28.8
62.2
94.1
29.1
63.0
95.9
30.2
67.3
102.9
% Dev
-5.3%
-0.8%
-1.1%
-4.5%
0.6%
0.3%
-1.5%
3.1%
2.7%
-3.6%
2.0%
2.3%
-4.0%
1.7%
2.5%
-3.3%
2.2%
2.6%
-2.5%
4.7%
4.9%
3.2%
11.0%
11.9%
16.2%
28.7%
30.3%
USS-3
Vs
30.2
60.5
90.8
30.2
60.5
90.8
30.2
60.5
90.9
30.2
60.5
91.0
30.2
60.5
91.0
30.2
60.5
91.1
30.2
60.5
91.1
30.2
60.5
91.2
30.2
60.5
91.2
Vax
26.2
52.4
78.6
28.4
56.9
85.3
29.7
59.6
89.5
30.1
60.3
90.7
30.2
60.5
91.0
30.1
60.3
90.8
29.7
59.6
89.7
28.4
56.9
85.7
26.2
52.4
79.0
Vnull
25.2
51.6
75.6
27.2
56.3
82.3
29.2
62.0
90.9
29.2
62.5
91.6
29.2
62.5
91.6
29.1
62.4
90.7
29.2
62.4
90.7
29.6
65.7
94.6
30.9
67.3
100.8
% Dev
-3.7%
-1.5%
-3.9%
-4.2%
-1.0%
-3.5%
-1.8%
4.1%
1.5%
-2.9%
3.7%
1.0%
-3.3%
3.3%
0.7%
-3.3%
3.5%
-0.1%
-1.8%
4.7%
1.1%
4.3%
15.6%
10.4%
18.1%
28.4%
27.6%
Pitch angle expressed in degrees. All velocities in fps.
Vnom = nominal velocity.
% Dev. = percent deviation = [(Vnul,
J 1 00.
-------
Table 5-2 Axial Velocity Evaluation for 2D Probes (a) (cont.)
Pitch
Angle
-30
-20
-10
-5
0
5
10
20
30
cos theta
subp
0.866
0.94
0.985
0.996
0
0.996
0.985
0.94
0.866
1-cos
theta p
13.4%
6.0%
1 .5%
0.4%
0.0%
0.4%
1.5%
6.0%
13.4%
Vnom (b)
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
Autoprobe Type S (AS)
Vs
30.1
59.7
89.6
30.1
59.7
89.6
30.1
59.7
89.6
30.1
59.7
89.6
30.1
59.7
89.6
30.1
59.7
89.6
30.1
59.7
89.6
30.1
59.7
89.6
30.1
59.7
89.6
Vax
26.1
51.7
77.6
28.3
56.1
84.2
29.6
58.8
88.2
30.0
59.5
89.3
30.1
59.7
89.6
30.0
59.5
89.3
29.6
58.8
88.2
28.3
56.1
84.2
26.1
51.7
77.6
Vnull
24.5
49.8
72.2
27.5
53.4
80.8
29.6
58.9
87.5
30.0
59.8
88.6
29.8
60.7
90.8
30.1
60.9
90.8
30.1
60.9
90.9
29.9
60.5
89.9
30.8
61.7
90.9
% Dev (c)
-6.0%
-3.7%
-7.0%
-2.8%
-4.8%
-4.0%
-0.1%
0.2%
-0.8%
0.0%
0.5%
-0.7%
-1.0%
1.7%
1.3%
0.4%
2.4%
1.7%
1.5%
3.6%
3.0%
5.7%
7.8%
6.8%
18.1%
19.3%
17.1%
Autoprobe Modified S (AMS)
Vs
30.1
59.7
89.7
30.1
59.7
89.6
30.1
59.7
89.6
30.1
59.7
89.6
30.1
59.7
89.6
30.1
59.7
89.6
30.1
59.7
89.6
30.1
59.7
89.6
30.1
59.7
89.6
Vax
26.1
51.7
77.7
28.3
56.1
84.2
29.6
58.8
88.2
30.0
59.5
89.3
30.1
59.7
89.6
30.0
59.5
89.3
29.6
58.8
88.2
28.3
56.1
84.2
26.1
51.7
77.6
Vnull
24.0
48.9
73.1
28.3
56.2
83.3
29.5
59.9
91.3
30.2
60.9
90.8
29.8
60.5
91.0
29.6
60.6
89.8
29.7
60.2
89.8
29.9
60.4
91.9
30.1
61.9
92.9
% Dev
-7.9%
-5.4%
-5.9%
0.1%
0.2%
-1.1%
-0.5%
1.9%
3.5%
0.7%
2.4%
1.7%
-1.0%
1.3%
1.6%
-1.3%
1.9%
0.6%
0.2%
2.4%
1.8%
5.7%
7.7%
9.1%
15.5%
19.7%
19.7%
French (F2)
Vs
30.1
60.4
90.7
30.1
60.4
90.7
30.1
60.4
90.7
30.1
60.4
90.7
30.0
60.4
90.7
30.0
60.4
90.7
30.0
60.4
90.7
30.0
60.4
90.7
30.0
60.4
90.7
Vax
26.1
52.3
78.6
28.3
56.8
85.2
29.6
59.5
89.3
30.0
60.2
90.4
30.0
60.4
90.7
29.9
60.2
90.4
29.5
59.5
89.3
28.2
56.8
85.2
26.0
52.3
78.6
Vnull
25.5
52.4
82.1
26.1
54.5
82.9
28.3
58.3
87.0
29.1
59.3
89.8
30.5
59.8
90.3
31.4
59.9
90.7
31.2
63.8
91.3
30.1
64.1
92.0
27.6
61.4
91.9
% Dev
-2.2%
0.2%
4.5%
-7.7%
-4.0%
-2.7%
-4.5%
-2.0%
-2.6%
-3.0%
-1.4%
-0.6%
1.7%
-1.0%
-0.4%
5.1%
-0.4%
0.4%
5.6%
7.3%
2.2%
6.8%
12.9%
7.9%
6.2%
17.4%
17.0%
Pitch angle expressed in degrees. AH velocities in fps.
Vnom = nominal velocity.
c % Dev. = percent deviation = [(Vnun -
-------
Wind Tunnel Experimental Results
To facilitate interpreting the data in Table 5-2, Tables 5-3 and 5-4 summarize the maximum deviation
from axial velocity (in percent) measured for each probe over selected pitch-angle ranges at zero yaw.
At pitch angles between -10° and 10°, the difference between Vnull and V^ is within the range -6.2%
and 7.3%. Over all velocities in this pitch angle range, the smallest variations are seen with the two
Autoprobes. Over the two higher velocities, 60 and 90 fps, the smallest variation at pitch angles from
-10° to 10° is seen with the modified Kiel probe. The French probe exhibited the highest variation
in both Tables 5-3 and 5-4 over this pitch-angle range.
Table 5-3. Maximum Percent Deviation of Measured Velocity
from Axial Velocity Across All Velocities at Zero Yaw
Probe
MK2
3/8 S-l
USS-1
USS-3
AS
AMS
F-2
Pitch Angle Ranges
-10 to +10
-5.6
-6.2
+4.9
+4.7
+3.6
+3.5
+7.3
-10 to +20
+7.8
-6.2
+11.9
+15.6
+7.8
+9.1
+12.9
-20 to +20
+7.8
-6.7
+11.9
+15.6
+7.8
+9.1
+12.9
Table 5-4. Maximum Percent Deviation of Measured Velocity
from Axial Velocity Across 60 and 90 fps at Zero Yaw
Probe
MK2
3/8 S-l
USS-1
USS-3
AS
AMS
F-2
Pitch Angle Ranges
-10 to +10
-2.2%
-3.8%
+4.9%
+4.7%
+3.6%
+3.5%
+7.3%
-10 to +20
+7.8%
+5.2%
+11.9%
+15.6%
+7.8%
+9.1%
+12.9%
-20 to +20
+7.8%
-5.5%
+11.9%
+15.6%
+7.8%
+9.1%
+12.9%
Figures 5-7 through 5-9 are plots of percent deviation from axial velocity vs. pitch angle at each
nominal velocity. Each plot shows the variation of percent deviation with velocity for all six probe
types (data for the two modified S probes in Table 5-2 have been averaged). From the figures it is
apparent that, except for the French probe at 30 fps, the percent deviation from actual axial velocity
5-6
-------
Wind Tunnel Experimental Results
increases significantly between pitch angles of 10 and 30°. In the ±10° pitch-angle range, except for
the French probe at 60 fps, deviations from axial velocities were all between ±5%. At 30 fps in this
same pitch-angle range, except for the French probe, deviations from axial velocities were between
-7% and +2%.
5.4 CONCLUSIONS
5.4.1 Calibration Coefficients
Analyzing the impact of pitch angle on calibration coefficients leads to the following findings:
• Except for the Type S and French probes, the calibration coefficients changed by less than
3.5% over the pitch angle range of-10° to 20°. The Type S and French probes' calibration
coefficients changed by 6.3% and 8.6%, respectively, over this range.
• Over the -10° to 20° range, the percent change in the calibration coefficients of the Type
S and modified S pilots using the Autoprobe (3.0% and 2.2%, respectively) was less than
that found when these probes were used in the manual mode (6.3% and 3.2%).
Over the pitch angle range of-30° to -10° the percent change in calibration coefficients was
above 9.1% for all probes except the modified Kiel. The percent change in Cp ranged from
9.1% for the French probe to 21.0% for the Autoprobe using the modified S probe head.
The extent of change indicates that the calibration coefficients of all test probes except the
modified Kiel are highly pitch angle-dependent in the -30° to -10° pitch range.
• The modified Kiel probe's calibration coefficient changed by 3.4% or less across the -10°
to 20° pitch range, and by less than 0.7% in the -30° to -10° range.
These findings lead to the following conclusions:
• The calibration coefficients of all probes, except for the Type S and French, were stable to
within 3.5% in the -10° to 20° pitch range.
None of the probes, except the modified Kiel, maintained a stable calibration coefficient in
the -30° to -10° pitch range.
5.4.2 Velocity Determination
Analyzing the pitch effects on velocity determination leads to the following findings:
Within the -10° to +10° pitch angle range, consistently high or consistently low velocity
measurements were not evident. However, the manual Type S and modified Kiel
measurements were consistently low between -10° and 0° pitch and -10° and 5° pitch,
respectively.
At positive pitch angles of 10° and greater at the higher tested velocities (60 and 90 fps),
all probes measured high relative to actual axial flow. The degree of the disparity increased
as the pitch angle increased.
5-7
-------
Wind Tunnel Experimental Results
At pitch angles of -10° and less, all probes, except the modified Kiel, measured
predominantly low relative to actual axial flow. The measurements were low by as much
as 9.6%. Five out of nine velocity measurements from the modified Kiel were high by as
much as +13.6%. Four out of nine velocity measurements were low by as much as 5.2%.
Aggregating results across all three tested velocities, measurements by the Autoprobe were
least affected by pitched flow in the -10° to +10° range. The maximum deviation for the
Autoprobe from the actual axial velocity was 3.5% with the modified S probe head and
3.6% with the Type S probe head.
• At the higher tested velocities (60 and 90 fps), the modified Kiel probe's velocity
measurements were least affected by pitched flow in the -10° to +10° range. Its maximum
deviation from the actual axial flow was -2.2 % followed by the Autoprobe modified S
(3.5%), Autoprobe S (3.6%), and manual Type S (-3.8%).
• In the -20° to +20° pitch angle range, the maximum deviation from actual axial velocity for
all tested probes across all velocities was never better than -6.7%. The manual Type S had
the lowest maximum deviation: -6.7% over all three velocities and -5.5% at the higher
velocities. The maximum deviation for the other probes over all velocities ranged from
+7.8% for the modified Kiel and Autoprobe with the Type S probe head to 15.6% for one
of the manual modified S probes.
• Over the ±10° pitch angle range, across all velocities, there was no discernible advantage
of equipping the Autoprobe with the modified S (+3.5% maximum deviation) as compared
to the Type S probe (+3.6% maximum deviation).
These findings support the following conclusions:
• In the ±10° pitch angle range, no single probe excelled under all velocity conditions.
Maximum deviations from axial velocity ranged from +3.5% for the Autoprobe operated
with a modified S pitot head to +7.3% for the French probe.
• In the ±10° pitch angle range at the higher tested velocities (60 and 90 fps), all probes
except the French measured velocities that were less than or equal to 4.9% of the axial
velocity. The modified Kiel had the lowest deviation at -2.2%.
• In the ±20° pitch angle range, across all velocities tested, none of the tested 2D probes kept
deviations from actual axial flow within ±5%.
In the ±20° pitch angle range, over all velocities, the manual Type S probe had the smallest
maximum deviation (-6.7%) and smallest maximum overestimation (+5.2%). At 60 and 90
fps, the maximum deviation for the manual Type S probe (-5.5%) was also lowest of all the
tested probes.
• At negative pitch angles, the manual Type S probe always underestimated true axial
velocity—the more negative the pitch, the larger the underestimate.
5-8
-------
Figure 5-1. Probe: 3/8" Type S - S-1
Calibration Coefficient vs. Pitch Angle
1.12
1.08
1.04
1.00-
0.96
0.92
0.88
0.84
0.80
0.76
Q.
o
-30
-25
-20
-15
-10
i '
-5
i
0
\
10
i
15
I
20
i
25
\
30
Pitch (degrees)
Velocity = 30 (tps) © -G-O Velocity = 60 (fps) D -EJ- D Velocity = 90 (fps)
Bands Indicate the Minimum and Maximum Values
-------
Figure 5-2. Probe: Modified S - USS-1
Calibration Coefficient vs. Pitch Angle
1.12
108
1.04
1.00-
0.96-
0.92
0.88
0.84
0.80
0.76
o.
o
-30 -25 -20 -15
-10
I
0
-505
Pitch (degrees)
I
10
15
r
20
i
30
Velocity = 30 (fps) © -*>-& Velocity = 60 (fps) Q -B- D Velocity = 90 (fps)
Bands Indicate the Minimum and Maximum Values
-------
Figure 5-3. Probe: Modified S - USS-3
Calibration Coefficient vs. Pitch Angle
Q.
o
1.12
1.08
1.04-
1.00
0.96
0.92
0.88
0.84
0.80
0.76-
'- -B-
. - -B
-30
-25
-20
-15
-10
-5
i
0
i
5
\
10
i
15
i
20
i
25
i
30
Pitch (degrees)
^^ Velocity = 30 (fps) © -€^» Velocity = 60 (fps) Q -Q- D Velocity = 90 (fps)
Bands Indicate the Minimum and Maximum Values
-------
K)
Q.
O
1.12
1.08
1.04
1.00
0.96
0.92
0.88
0.84
0.80
0.76
Figure 5-4. Probe: Autoprobe-Type S
Calibration Coefficient vs. Pitch Angle
— — •»=•— -^•^ — — • —
^^^^-•e.-i.-;
I , , , , I , , , , I 1 1 , , 1 . , , , 1 , , , , 1 , , , , 1 , ,
-30 -25 -20 -15 -10 -5 0
. - -^ir— 4- z^rrrruuT^rr^^^-jj
, , i , , , , i , , , , | , , , , i , , , , i , , , , i
5 10 15 20 25 30
Pitch (degrees)
000 Velocity = 30 {fps) © -O-3 Velocity = 60 (fps) D -GJ- D Velocity = 90 (fps)
Bands Indicate the Minimum and Maximum Values
-------
Figure 5-5. Probe: Autoprobe - Modified S
Calibration Coefficient vs. Pitch Angle
1.12-
1.08-
1.04
1.00
0.96
0.92
0.88-
0.84
0.80
0.76-
QL
O
-30
-25 -20 -15
-10 -505
Pitch (degrees)
10
15
i
20
25
^ ' i
30
•e—* Velocity = 30 (fps) © -0-G Velocity = 60 (fps) D -Q- D Velocity = 90 (fps)
Bands Indicate the Minimum and Maximum Values
-------
Figure 5-6. Probe: French Probe - F-2
Calibration Coefficient vs. Pitch Angle
o.
O
1.02-
0.98-
0.94-
0.90
0.86
0.82 H
0.78-
0.74-
0.70-
0.66
-30 -25
-20
-15 -10
-5
i
0
i
5
i
10
i
15
i
20
i
25
i
30
Pitch (degrees)
Velocity = 30 (fps) © -G-G Velocity = 60 (fps) Q -H- D Velocity = 90 (fps)
Bands Indicate the Minimum and Maximum Values
-------
Wind Tunnel Experimental Results
Figure 5-7.
% Deviation of Measured Velocity From
Axial Velocity vs. Pitch Angle- 30 fps
30.0%
.1 20.0%
ra
0)
Q 10.0%
-20
Mod Kiel
Autoprobe S
-10 0 10
Pitch Angle
20
Type S -A- Mod S
Autoprobe Mod S -^- French
30
5-15
-------
fF/W Tunnel Experimental Results
Figure 5-8.
Percent Deviation
A _». |VJ CA>
o o o o o
% Deviation of Measured Velocity From
Axial Velocity vs. Pitch Angle- 60 fps
0%
0%
E
0%
0%'
V,
C
«;
0%-
S
X
N^5
\
>
^
~-^s
^=^^
^
^^
^—~~^
1 *
£=^
F=^
F^
^
U^
£=^
^
p^
^^
*^1
. £
s
/
/t
¥
/
V
J
~^r
?
-30 -20 -10 0 10 20
Pitch Angle
-a- Mod Kiel -55- Type S ^^ Mod S
-x- Autoprobe S -e- Autoprobe Mod S -v- French
^
5
30
5-16
-------
Wind Tunnel Experimental Results
Figure 5-9.
30.0%
.1 20.0%
-4—>
.55
o>
Q 10.0%
I
0>
CL
0.0%
% Deviation of Measured Velocity From
Axial Velocity vs. Pitch Angle- 90 fps
-10.0%
-30
-20
Mod Kiel
Autoprobe S
-10 0 10
Pitch Angle
20
Type S -A- Mod S
Autoprobe Mod S -v- French
30
5-17
-------
6.0 ACCURACY OF YAW ANGLE MEASUREMENTS
AND EFFECT ON VELOCITY MEASUREMENT
6.1 TEST PROCEDURES
The objectives of the experiments described in this section were to determine how accurately each
probe could find a yaw null and to determine the effect on velocity measurement of errors in yaw
nulling. Data on yaw nulling accuracy were collected during the tests described in Sections 3, 5,
and 7 using the test procedures described in those sections. During the experiments, the wind tunnel
was set at a velocity of 30, 60, or 90 fps, and the test probe was inserted into the wind tunnel at a
preset pitch angle. The probe was yaw nulled by rotating it until AP across the appropriate ports
(Fechheimer ports for the modified Kiel, impact and static ports for the manual and Autoprobe
Type S and modified S, and P2 and P3 ports for the DAT and spherical probes) read 0. The angle,
read by an inclinometer with a visually estimated accuracy of 0.1 °, was then recorded as 0null.
Because flow in the wind tunnel was at a yaw angle of zero, the measurement error is equivalent
to the measured yaw angle 0null. The Type S and modified S were yaw nulled manually, and the
Autoprobes were yaw nulled automatically. No procedure has been established for yaw nulling
the French probe, and, therefore, it was not included in this analysis.
In each replication of the experiment, the yaw null angle was read with the digital inclinometer
and recorded. The 2D probes (except the modified Kiel probe) were then rotated 90° to align the
measurement ports with the flow and AP across the probe was measured and recorded. The
modified Kiel probe did not require rotation; AP was simply measured and recorded at the yaw
null position. For the 3D probes, AP across ports Pl and P2 and across ports P4 and P5 were
measured and recorded. Velocities at each point were then calculated using Equation 5.4 for the
2D probes and Equation 7.3 for the 3D probes. These steps were conducted three times to
provide replicate runs.
6.2 ACCURACY OF YAW ANGLE DETERMINATIONS AND VELOCITY DETERMINATIONS
Figures 6-1 through 6-10 are plots of the deviation of the measured yaw null (0nuB) vs. pitch angle.
Nine observations were made at each pitch setting, representing measurements taken at three
velocities with three replicates performed at each velocity. Error bars show the maxima and minima
of the nine measurements at each pitch. The means are connected by straight lines. Positive or
negative deviations refer to clockwise or counterclockwise probe rotation, respectively.
Table 6-1 summarizes the data shown in Figures 6-1 through 6-10. For each probe, the table lists
the maximum yaw null error (O^nm^), in degrees, over the tested velocity and pitch ranges, and the
associated percentage error in velocity that would have resulted if the actual yaw angle of the flow
were 0° and 20°. Also shown is the pitch angle at which the maximum yaw angle and maximum
deviation occurs.
6-1
-------
Wind Tunnel Experimental Results
Table 6-1. Summary of Yaw Angle Accuracy and Effect on Velocity Determination
Probe
Modified Kiel (MK2)
Type S (S-l)
Modified S (USS-1)
Modified S (USS-3)
Autoprobe Type S (AS)a
Autoprobe Mod S (AMS)a
Spherical (MS5-2)
DAT (3D-1)
DAT (3D-2)
DAT (3D-3)
"nullmai ( )
1.1
1.3
6.6
2.8
-3.6
5.4
-0.5
1.0
-0.3
0.7
% Error in Velocity
6^=0°
0.02
0.03
0.66
0.12
0.20
0.44
0.00
0.02
0.00
0.01
6^=20"
0.7
0.9
4.8
1.9
2.5
3.9
0.3
0.7
0.2
0.5
Pitch Angle (°)
30
30
-30
-30
5
20
-30
30
-20
30
The Autoprobe tested in this study employed a "sweep" procedure to determine the yaw null. A curve-fitting
algorithm has subsequently been implemented by the manufacturer to find the yaw null.
The errors in velocity were calculated from the following equation:
Error . 100
Vnull cos(0y)
(Eq. 6.1)
«>s
(ey)
where
V,
nun
= velocity calculated from AP
= actual yaw angle (0° or 20°)
= maximum error in measuring 0y
EPA Method 2, the only method currently approved for flue gas flow measurement, is based on
measuring flow with a Type S probe without yaw nulling. EPA Method 1' permits use of the Type S
probe only if the yaw angle of flow determined by averaging yaw angle over traverse points is less
than 20°. Yaw angles are measured at each traverse point by yaw nulling with the Type S probe with
1 40 CFR Ch. 1 (7-1-92 Edition), Pt. 60, App. A, Method 1—Sample and Velocity Traverses for Stationary Sources,
pgs. 682-688.
6-2
-------
Wind Tunnel Experimental Results
the same technique used in this study. Examining errors in velocity attributable to yaw angles of up
to 20° is thus instructive about the range of errors that can potentially be observed.
Table 6-2 shows additional summary statistics for the data displayed in Figures 6-1 through 6-10.
For each probe at each velocity, the table lists the minimum, mean, and maximum deviations from 6null
calculated across all pitch angles tested, and the range and standard deviation across all pitch angles.
The table also shows these parameters calculated across all velocities. Across all pitch angles and all
velocities, the largest absolute mean yaw angle error was 1.5 ° for one of the modified S probes. The
smallest was 0.1° for the spherical probe and one of the DAT probes. The largest standard deviation
across all pitch angles and all velocities was 2.8° for one modified S probe; the smallest was 0.3 ° for
the spherical probe and one of the DAT probes.
6.3 FLOW MEASUREMENT ERROR DUE TO YAW NULL MISALIGNMENT FOR THE AUTOPROBE
TYPES
An issue related to the accuracy of yaw nulling procedures is the magnitude of the error that would
be produced in flow measurements when probes are not yaw nulled or are misaligned relative to the
yaw component of flow. A previous study of Type S probes found that the estimated error was as
high as 18% at a 20° yaw misalignment.2 If the error in velocity measurements were due strictly to
the inability to measure and account for yaw, the overestimation of flow due to a 20° misalignment
of the probe would be (l-cos20°)/cos20°, or only 6.42%. That is, a probe that could measure total
velocity perfectly would overestimate the flow velocity by 6.42% strictly due to the inability to
determine the yaw angle. Thus, a deviation as high as that reported in the previous study would
indicate that the error could not be fully explained by the probe's inability to measure the yaw angle
and derive the axial component of velocity. It would imply that physical characteristics of the probe
were causing a further overestimation of the flow velocity.
Table 6-2. Yaw Angle Error Detailed Data
Probe
Modified Kiel
(MK2)
Type S
(S-l)
Velocity
(fps)
30
60
90
All
30
60
90
All
Across All Pitch Angles (-30° to +30°)
Sample
Size
27
27
27
81
27
27
27
81
Minimum
(°)
-0.4
-0.2
0.1
-0.4
-1.0
-1.0
-1.0
-1.0
Mean
0
0.4
0.4
0.5
0.4
0.1
0.3
0.4
0.3
Maximum
(°)
2.2
1.5
1.3
2.2
1.7
1.4
1.5
1.7
Range
(°)
2.6
1.7
1.4
2.6
2.7
2.4
2.5
2.7
Standard
Deviation
(°)
0.5
0.4
0.3
0.4
0.7
0.6
0.7
0.7
Williams, IE, J.C., and F.R. DeJarnette. 1977. A Study on the Accuracy of Type S Pilot Tubes. EPA Report No.
EPA-600/4-77-030. 70 pgs.
6-3
-------
Wind Tunnel Experimental Results
Table 6-2. Yaw Angle Error Detailed Data (continued)
Probe
Modified S
(USS-1)
Modified S
(USS-3)
Autoprobe
Type S (AS)
Autoprobe
Mod S (AMS)
Spherical
(MS5-2)
DAT
(3D-1)
DAT
(3D-2)
Velocity
(fps)
30
60
90
All
30
60
90
All
30
60
90
All
30
60
90
All
30
60
90-
All
30
60
90
All
30
60
90
All
Across All Pitch Angles (-30° to +30°)
Sample
Size
27
27
27
81
27
27
27
81
26
26
27
79
27
27
29
83
27
27
27
81
27
27
27
81
27
27
27
81
Minimum
(°)
-10.8
-5.1
-4.5
-10.8
-0.7
-3.0
-4.1
-4.1
-8.1
-7.6
-6.1
-8.1
-0.2
0.1
0.5
-0.2
-0.8
-0.6
-0.7
-0.8
-0.3
-0.2
-0.3
-0.3
-1.0
-0.8
-0.8
-1.0
Mean
(°)
1.5
1.6
1.4
1.5
0.6
0.5
0.2
0.5
0.6
-1.1
-2.1
-0.9
1.2
1.2
1.7
1.4
-0.1
-0.2
-0.1
-0.1
0.0
0.4
0.7
0.4
-0.3
0.0
0.0
-0.1
Maximum
(°)
9.0
7.0
5.3
9.0
3.0
4.7
3.1
4.7
3.9
3.0
1.7
3.9
4.7
4.4
8.1
8-1
0.6
0.5
0.6
0.6
0.2
1.6
1.8
1.8
0.4
0.7
0.6
0.7
Range
(°)
19.8
12.1
9.8
19.8
3.7
7.7
7.2
8.8
12.0
10.6
7.8
12.0
4.9
4.5
7.6
8.3
1.4
1.1
1.3
1.4
0.5
1.8
2.1
2.1
1.4
1.5
1.4
1.7
Standard
Deviation
(°)
3.6
2.6
2.2
2.8
0.8
1.6
1.9
1.5
2.5
2.8
2.0
2.7
1.4
1.2
2.2
1.7
0.3
0.3
0.4
0.3
0.1
0.5
0.5
0.5
0.4
0.3
0.3
0.3
6-4
-------
Wind Tunnel Experimental Results
Table 6-2. Yaw Angle Error Detailed Data (continued)
Probe
DAT
(3D-3)
Velocity
(fps)
30
60
90
All
Across All Pitch Angles (-30° to +30°)
Sample
Size
27
27
27
All
Minimum
(°)
-0.9
-0.4
-0.8
-0.9
Mean
(°)
0.0
0.3
0.3
0.2
Maximum
(°)
0.6
1.1
1.1
1.1
Range
(°)
1.5
1.5
1.9
2.0
Standard
Deviation
(°)
0.4
0.5
0.5
0.5
To investigate this issue, the current study used the Autoprobe, because it could take measurements
quickly at predefined yaw angles. Operating in the non-yaw nulled mode, the Autoprobe measured
AP in 5° yaw increments. The measurements were successively carried out at pitch angle settings of
-10°, 0°, and +10° at a nominal velocity of 60 fps. At each yaw angle 0y, the calibration
coefficient C was calculated using Equation 6.2.
\
AP,
AP
(Eq. 6.2)
where
APf
AP.
= differential pressure for the standard pitot tube at yaw angle = 0°
= differential pressure for the Type S pitot tube at yaw angle, 0y
Figure 6-11 is a plot of C vs. yaw angle. The shapes of the curves at all three pitch angles are
similar in the -20° to +20° range. Using the data shown in Figure 6-11, flow measurement errors due
to errors in yaw angle measurement were calculated according to Equation 6.3.
% Error =
-2-1 100
(Eq. 6.3)
where
C
calibration coefficient of Type S pitot tube at yaw angle = 0°
calibration coefficient of the Type S pitot tube at yaw angle, 0y
6-5
-------
Wind Tunnel Experimental Results
These errors are plotted in Figure 6-12. In the -10° to +10° pitch-angle range typically found in
stacks covered under the Acid Rain Program, the maximum flow measurement error in the ±20° yaw
angle range (the range in which Method 2 is applicable) is 1.98%. This error is well below the 6.42%
error resulting at 20° yaw when the yaw angle is not measured at all.
6.4 CONCLUSIONS
The data presented in this section support the following findings:
• For all probes employing Fechheimer-type ports to measure yaw (i.e., the modified Kiel, the
spherical, and DAT probes), the error in the derived yaw angle was 1.1° or less (see
Table 6-1). This translates into a 0.7% or less error in velocity measurements in a flow
profile displaying 20° yaw.
• For the manually operated Type S probe, the yaw angle error was 1.3°, resulting in a 0.9%
velocity error in a flow profile displaying 20° yaw.
• For the manual and automated modified S probes, the absolute value of the yaw angle error
ranged from 2.8° to 6.6° with a corresponding 1.9% to 4.8% error in velocity
measurements in a flow profile with a 20° yaw.
• Maximum error in determining yaw angle with the Autoprobe was -3.6° using the Type S
head and 5.4° using the modified S head. This translates into a 2.5% and 3.9% velocity
error in a flow profile displaying 20° yaw.
These findings lead to the following conclusions:
• Use of Fechheimer ports produces consistently better yaw angle measurements than other
procedures for measuring yaw angles.
• In the wind tunnel, the manually operated Type S probe gave yaw null and velocity accuracy
results comparable to those of the probes employing Fechheimer ports.
• The maximum yaw angle error using the Autoprobe3 was higher than the maximum errors
produced by either the manually operated Type S probe or any of the manually operated
probes that employ Fechheimer ports.
• Among the manually operated probes, the largest error in yaw angle measurements was
produced by the modified S probe: 6.6°. Similarly, the largest error with the Autoprobe
was produced using the modified S probe head: 5.4°.
3 As noted in Table 6-1, the Autoprobe tested in this study employed a "sweep" procedure to determine the yaw null rather
than the curve-fitting algorithm subsequently implemented by the manufacturer.
6-6
-------
Figure 6-1. Probe: Modified Kiel Probe With Thermocouple -MK2
Theta null Deviation from True Yaw Null
2.6
2.2
1.8
1.4
1.0
0.6
0.2
-0.2
IT
s>
§?
o
-1.0
-1.4-
-1.8-
-2.2-
-2.6-
i '
-30
-25
-20
-15
-10
-505
Pitch (degrees)
i
10
15
i
20
25
1 i
30
Bands Indicate the Minimum and Maximum Values
-------
r
oo
2.0-
1.6
1.2-
^ 0.8
to
o>
.§ 0.4 -]
.o
•5 0.0
Q
c. -0.4
(0
o>
-0.8
-1.2-
-1.6-
-2.0
-30
Figure 6-2. Probe: 3/8" Type S-S1
Theta null Deviation from True Yaw Null
-25
-20
-15
-10
l '
-5
l
0
[
5
i
15
I
20
i
25
I
30
Pitch (degrees)
Bands Indicate the Minimum and Maximum Values
-------
I
T -
!<
"55
12-
10-
8:
6
2
0
-2
-4
-6-
-8
-10-
-12 :
-30
Figure 6-3. Probe: Modified S-USS-1
Theta null Deviation from True Yaw Null
-25
-20
-15
-10
-5
i
0
i
5
l
10
i
15
l
20
T
25
I
30
Pitch (degrees)
Bands Indicate the Minimum and Maximum Values
-------
Figure 6-4. Probe: Modified S-USS-3
Theta null Deviation from True Yaw Null
tn
-------
Figure 6-5. Probe: Autoprobe Type S
Theta null Deviation from True Yaw Null
4-
3-
2-
1 -
o-
-1 -
8>
.§ -2
Q -3:|
c
0)
I -4-
-5
7
8
9
-30
-25
-20
-15
-10
-505
Pitch (degrees)
10
i
15
i
20
i
25
i
30
Bands Indicate the Minimum and Maximum Values
-------
Figure 6-6. Probe: Autoprobe Mod S
Theta null Deviation from True Yaw Null
10
9
8
7
co
o
I 6
2. 5
4H
Q
=5 3
1 -
o-
-1 -j
-2-
-30
-25
-20
-15
-10
-5
\
0
].
5
I
10
15
\
20
25
30
Pitch (degrees)
Bands Indicate the Minimum and Maximum Values
-------
Figure 6-7. Probe: Spherical MS5-2
Theta null Deviation from True Yaw Null
2.0
CO
1.6-
1.2
0.8-
0.4
0.0
c -0.4
cO
-0.8
-1.2
-1.6-
-2.0-
I ' ' ' ' i '
-30 -25
-20 -15
-505
Pitch (degrees)
10
15 20 25 30
Bands Indicate the Minimum and Maximum Values
-------
Figure 6-8. Probe: DAT-3D-1
Theta null Deviation from True Yaw Null
2.0-
1.6 :
1.2-
^ 0.8-
} 0.4
'•8
1
Q
G>
o.o
-0.4 -I
-0.8
-1.2-
-1.6
-2.0-
-30 -25 -20 -15 -10
-5
l
0
r
5
10
15
I
20
25
30
Pitch (degrees)
Bands Indicate the Minimum and Maximum Values
-------
Figure 6-9. Probe: DAT-3D-2
Theta null Deviation from True Yaw Null
2.0
0)
2
O
'
1.6-
1.2-
0.8
0.4-
0.0
-0.4-
-0.8-
-1.2-
-1.
-2.0-
i ^
-30
-25 -20 -15 -10
-5
i1 ' ' ' i
0 5
i^
10
15
T
20
l
25
30
Pitch (degrees)
Bands Indicate the Minimum and Maximum Values
-------
w
o
I 1
-------
Wind Tunnel Experimental Results
Figure 6-11.
o
Cp vs. Yaw Angle
Autoprobe Type S- 60 feet per second
0.76
-40
-20 0 20
Yaw Angle (degrees)
+10 Pitch
-10 Pitch
0 Pitch
40
6-17
-------
Wind Tunnel Experimental Results
Figure 6-12.
Flow Measurement Error From Yaw Error
Autoprobe Type S - 60 feet per second
4%
m 2% -
V
^^^ X
C
-------
APP
7.0 CALIBRATION CURVES AND ACCURACY
OF VELOCITY DETERMINATIONS FOR 3D PROBES
7.1 TEST PROCEDURES
7.1.1 Calibrating 3D Probes
The three DAT probes and the spherical probe were calibrated using the procedures specified in draft
Method 2F. The calibration procedure described in the draft method calls for yaw nulling by rotating
the probe until AP across the Fechheimer ports (P2 - P3) equals zero. The pressure differentials, P4-P5
and P,-P2, are then measured to determine the following ratios (see Figures 1-4 and 1-5 for locations
of numbered measurement ports):
(Eq. 7.2)
where C^ is the calibration coefficient of the standard pilot tube (which is equal to unity), and APP
is the pressure differential across the standard pilot tube. Values of Fj and F2 are obtained al a
predefined series of velocities and pilch angles. Then, Iwo calibration curves are plotted: one curve
(F, vs. pilch angle) is used lo determine pilch angle; Ihe olher curve (F2 vs. pilch angle) is used lo de-
termine ihe 3D probe's calibration coefficient F2.
Draft Melhod 2F calls for calibration curves of F, and F2 vs. pilch angle lo be developed by laking
measurement of P4-PS and Pj-P2 al Iwo velocities: one velocity belween 20 and 40 fps and another
velocity between 40 and 60 fps. Each calibration curve is created by calculating and Ihen averaging
Ihe F, or ihe F2 ratio al each pilch angle over ihe Iwo velocities. For Ihe experimenls conducted here,
Iwo sels of calibration curves were created for each probe, one sel by averaging dala oblained al Iwo
velocities, 30 and 60 fps, and anolher sel by averaging dala al Ihree velocities, 30, 60, and 60 fps.
The Ihree-velocity compulation was performed lo see if it yielded more accurate values of velocity
oulside ihe range of calibration velocities diclaled by Melhod 2F.
The draft melhod calls for F, and F2 lo be determined al pilch angles belween -60° and 60° in 5°
incremenls. For Ihe experimenls conducted in ihis sludy, Ihe geomelry of ihe wind lunnel limited
pilch angles lo Ihe range -30° lo +30°. Pilch angles were Iherefore sel al ±30°, ±20°, ±10°, ±5°,
and 0°. All measuremenls were made in Iriplicale; ihe draft melhod calls for only one measuremenl
al each poinl in this angle range.
7.1.2 Velocity Determination with 3D Probes
Once calibration curves have been derived for a 3D probe, velocity measuremenls can be oblained
in a particular duel or slack by yaw nulling ihe probe until P2-P3 equals zero and Ihen measuring
P4-P5, Pj-P2, and the corresponding yaw angle, fy. The measurements of P4 -P5 and P, -P2 are used
7-1
-------
Wind Tunnel Experimental Results
in Equation 7.1 to calculate a value for F,. The pitch angle, 0p, associated with this value of ^ is
then found on the first calibration curve (F, vs. pitch angle). The resulting pitch angle is used to find
the 3D pilot calibration coefficient (F2) from the second calibration curve (F2 vs. pitch angle).
Using the derived values of P4-P5, P,-P2, 6y, 6p, and measurements of temperature and barometric
pressure, the total velocity, V8, and axial velocity, V^, are then determined from the following equa-
tions:
Vs = 15.9 F2
(P, - P2) (T V 460) (Eq. 7.3)
Pbar
Vax = Vs COS0y COS 0p
7.2 CALIBRATION CURVES
Figures 7-1 through 7-8 are plots of the data used to construct the calibration curves for each probe.
Each point is the average of three runs performed at each pitch angle and velocity. The draft method
calls for a "smooth line to be drawn between points." For the curves here, a cubic spline method with
continuous second derivatives was used to fit each curve. This method was chosen because it is the
easiest and most appropriate for the data structure. For each set of two adjacent points, the method
uses a piecewise third-degree polynomial, which passes through the data points and matches the first
and second derivatives of neighboring segments at the points in order to obtain the smoothest line
possible.
The three F, curves in each plot overlap significantly and show no velocity-dependent separation.
The F2 curves for two of the three DAT probes (3D-2 and 3D-3) shown in Figures 7-4 and 7-6 show
clear velocity dependence, because they do not intersect and the spread between the curves is
significant. For probe 3D-2, the spread between the F2 ratios at 30 and 90 fps is about 5%, with the
60 fps curve between the two extremes; for probe 3D-3, the spread at 30 and 90 fps is about 4%.
These differences between the 30 and 60 fps F2 curves and between the 60 and 90 fps F2 curves are
within the range of ±3% allowed by draft Method 2F for F2 values between 30 and 60 fps. The data
for each velocity used to calculate F, and F2 ratios for the spherical probe overlap, except for a
divergence of the F2 ratio measured at 30 fps and pitch angles of-30° and +30°.
The velocity dependence of F2 for DAT probes 3D-2 and 3D-3 (across all pitch angle settings) and
for the spherical probe MS5-2 (at -30° and +30° pitch angles) suggests that it may be advisable to
require calibrations of 3D probes at velocities close to those prevailing in the situations where such
probes will be used. The plot of F2 vs. pitch angle for DAT 3D-1 shows much less velocity
dependence than the curves for the other three probes, for which no readily apparent explanation is
available.
7-2
-------
Wind Tunnel Experimental Results
Figures 7-9 through 7-16 are calibration curves for the three DAT probes and the spherical probe.
Each curve is derived from point-by-point averages of the F, and F2 ratios determined at all three
velocities shown in Figures 7-1 through 7-8.
7.3 ACCURACY OF VELOCITY DETERMINATIONS
An attempt was made to estimate the extent to which the axial velocity derived from the 3D
calibration curves deviated from the axial velocity derived from the standard pitot that was used to
calibrate the wind tunnel. The analysis was performed twice: once using the calibration curves
(shown in Figures 7-9 through 7-14) that were derived by averaging the results across three velocities
(30, 60, and 90 fps), and once using the calibration curves derived using two velocities (30 and 60
fps), i.e., in accordance with draft Method 2F. Repeating the analysis using two separate sets of
calibration curves was intended to provide a sense of the range in error inherent in the 3D calibration
curves.
Note, however, that the data used to perform this error analysis are the same as the data used to
derive the calibration curves. Therefore, the resulting analysis represents only an estimate of the error
inherent in the calibration curves as originally produced, not an independent bounding of the error
using new measurements. Furthermore, the data used to produce the second set of calibration curves
were a subset of the data used to derive the first set of calibration curves, which may limit the
variation in the error detected between the two analyses. Rather than establishing definitive bounds
on the likely error to be found in the 3D calibration curves, this analysis should be viewed as
indicating the extent to which error can be reduced by deriving calibration curves from an average
over three velocities rather than over just two velocities as currently stipulated in draft Method 2F.
The results of the analysis are shown in Tables 7-1 through 7-4. The first four columns on the left
list the pitch angle, nominal velocity, total velocity (V8, measured with the standard pitot probe), and
resulting axial velocity, Vm. The axial velocity is calculated simply as VM = V, cos6p, because the yaw
angle component of flow within the wind tunnel was negligible, and, consequently, cos0y = 1 (see
Equation 7.4). The middle group of five columns lists the values derived from the first set of 3D
probe calibration curves, i.e., those obtained by averaging over 30, 60, and 90 fps. The first column
in this group lists the pitch angle, 6p, calculated from the F, vs. pitch angle calibration curve. The
next column lists the average of the three pitch angles in the preceding column. The third and fourth
columns show the total, V8, and axial, V^, velocities calculated using Equations 7.3 and 7.4. The
fifth column shows the percent deviation of the axial velocity calculated from the 3D probe calibration
curves and the axial velocity calculated for the standard pitot. Percent deviation was computed as
follows:
% Dev =
100
The righthand group of columns is similar to the center group, except that all calculations were made
from the calibration curves determined from the 30 and 60 fps curves. The columns below the table
list the average percent velocity deviations over all pitch angles for each velocity for the entire pitch
7-3
-------
Wind Tunnel Experimental Results
angle range and for pitch angles in the ±20° and ±10° ranges. These values are based on the absolute
values of the percent deviations and show how far, on average, the 3D axial velocity departed from
the standard pitot axial velocity.
DAT Probes (Tables 7-1 to 7-3): Whether using the 30/60 or 30/60/90 fps calibration curves, the
errors in velocity found with probe 3D-1 were less than 1.4%, except at -30° pitch, where the errors
rose to a maximum of 3.0% at 90 fps when using the 30/60 calibration curves and -2.2% when using
the 30/60/90 curves. The largest difference in average errors across the pitch angle ranges
determined from the two calibration curves is 0.2%. Figure 7-2, showing F2 vs. pitch angle for the
3D-1 probe, illustrates that F2 is largely independent of velocity across the pitch angle range of-20°
to +30°. This observation is in marked contrast to the results for the other two DAT probes (see
Figures 7-4 and 7-6.)
With probe 3D-2, the largest errors are found at 30 fps and 20° pitch angle for both sets of
calibration curves, namely, 5.5% using the 30/60/90 curves and 4.4% using the 30/60 curves. Errors
in velocity measurements at 30 fps are consistently high with absolute deviations averaging 2.9%
when the 30/60/90 fps calibration curves were used, and 1.7% when the 30/60 fps curves were used.
Errors in velocity determinations at 90 fps are consistently low. At this velocity, deviations averaged
-2.2% with the 30/60/90 fps curves and -3.4% with the 30/60 fps curves across all velocities. At 10°
pitch across all three velocities, anomalous negative errors occurred between the average calculated
pitch angle (Table 7-2, columns 5 and 10) and the actual pitch angle of 10° (Table 7-4, column 1).
When using the 30/60/90 calibration curves, the difference between the actual and average calculated
pitch angle was -2.3°. With the 30/60 curves, the difference was -2.6°. Nevertheless, the velocity
errors at this pitch angle are consistent with velocity errors at other pitch angles, reflecting the small
difference in cosine values between 10° and 7° (0.8%).
The velocities measured with probe 3D-3 were consistently high at 30 fps and consistently low at
90 fps (Table 7-3). The largest deviation (-3.8%) occurred at 90 fps and 10° pitch using the 30/60
fps calibration curves. This deviation is similar to those found with probe 3D-2.
Spherical Probe (Table 7-4): Over the entire pitch angle velocity range, the largest single velocity
deviation is -5.2% at 30 fps and -30° pitch angle using the 30/60/90 fps calibration curves. Spherical
probe velocities computed using the 30/60/90 curves were consistently higher than the standard pitot
measurements at the 90 fps level (by 1.6% on average) and consistently lower at the 30 fps level (by
1.5% on average). Using the 30/60 curves, the largest velocity deviations at each pitch angle were
found at 90 fps and averaged 2.4% over the entire pitch angle range. Considering the results for both
the 30/60 and 30/60/90 curves together, the largest deviations at each velocity were -5.2% at 30 fps,
3.8% at 60 fps, and 4.7% at 90 fps, occurring at pitch angles of-30°, 30°, and -20°, respectively.
Over each of the pitch ranges shown at the bottom of Table 7-4, the difference between the average
absolute deviations produced using the 30/60/90 calibration curves and those produced using the
30/60 curves never exceeded 0.8%. For example, in the -30° to +30° pitch range, where the
differences produced using the two sets of calibration curves were greatest, the differences were -
0.4% at 30 fps, 0.6% at 60 fps, and 0.8% at 90 fps.
7-4
-------
Wind Tunnel Experimental Results
Table 7-1. DAT Probe (3D-1) Velocity Accuracy *
Reference
Pitch
Angle
(deg)
-30
-20
-10
-5
0
5
10
20
30
Nominal
Velocity
(fps)
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
Vs
Total
Velocity
(fps)
30.0
60.1
90.4
30.0
60.1
90.4
30.0
60.1
90.4
30.0
60.1
90.4
30.0
60.1
90.4
30.0
60.1
90.4
30.0
60.1
90.4
30.0
60.1
90.4
30.0
60.1
90.4
Vax
Axial
Velocity
(fps)
25.98
52.05
78.29
28.19
56.48
84.95
29.54
59.19
89.03
29.89
59.87
90.06
30.00
60.10
90.40
29.89
59.87
90.06
29.54
59.19
89.03
28.19
56.48
84.95
25.98
52.05
78.29
Calculated Values for DAT Probe (3D-1)
Calibrated at 30-60-90 fps
Pitch
Angle
(deg)
-30.0
-28.6
-28.5
-22.4
-18.7
-18.5
-9.6
-10.9
-9.6
-4.9
-5.1
-5.0
0.8
0.6
1.9
5.3
3.4
5.4
10.5
9.5
10.4
18.1
20.0
21.2
28.6
29.6
30.0
Avg
Pitch
Angle
(deg)
-29.0
-19.9
-10.0
-5.0
1.1
4.7
10.1
19.8
29.4
Vs
Total
Velocity
(fps)
29.34
60.22
90.88
30.10
60.15
90.43
29.89
60.26
90.51
29.95
60.16
90.37
30.05
59.75
90.52
30.08
60.01
90.38
29.92
60.33
90.16
29.87
60.39
91.24
29.82
59.94
90.69
Vax
Axial
Velocity
(fps)
25.41
52.87
79.87
27.83
56.97
85.76
29.47
59.17
89.24
29.84
59.92
90.03
30.05
59.75
90.47
29.95
59.90
89.98
29.42
59.50
88.68
28.39
56.75
85.07
26.18
52.12
78.54
%
Dev.
-2.2
1.6
2.0
-1.3
0.9
1.0
-0.2
0.0
0.2
-0.2
0.1
0.0
0.2
-0.6
0.1
0.2
0.1
-0.1
-0.4
0.5
-0.4
0.7
0.5
0.1
0.8
0.1
0.3
Calibrated at 30-60 fps
Pitch
Angle
(deg)
-30.0
-28.1
-27.8
-21.7
-18.0
-17.9
-9.4
-10.6
-9.4
-4.9
-5.0
-5.0
0.4
0.9
2.5
5.5
3.9
5.5
10.7
9.6
10.6
19.0
20.6
21.8
28.9
29.7
30.0
Avg
Pitch
Angle
(deg)
-28.6
-19.2
-9.8
-5.0
1.3
5.0
10.3
20.5
29.5
Vs
Total
Velocity
(fps)
29.53
60.45
91.13
30.09
60.19
90.49
29.91
60.30
90.58
29.95
60.17
90.38
30.09
59.79
90.52
30.06
59.98
90.32
29.88
60.24
90.02
29.86
60.56
91.56
29.87
60.00
90.72
Vax
Axial
Velocity
(fps)
25.57
53.32
80.61
27.96
57.24
86.11
29.51
59.27
89.36
29.84
59.94
90.04
30.09
59.78
90.43
29.92
59.84
89.90
29.36
59.40
88.48
28.23
56.69
85.01
26.15
52.12
78.57
%
Dev.
-1.6
2.5
3.0
-0.8
1.4
1.4
-0.1
0.1
0.4
-0.2
0.1
0.0
0.3
-0.5
0.0
0.1
-0.1
-0.2
-0.6
0.4
-0.6
0.2
0.4
0.1
0.7
0.1
0.4
Pitch
Angles
-30
to
30
-20
to
20
-10
to
10
Total
Velocity
30.0
60.1
90.4
30.0
60.1
90.4
30.0
60.1
90.4
Average Absolute Deviation
0.7
0.5
0.5
0.5
0.4
0.3
0.2
0.3
0.2
%
0.5
0.6
0.7
0.3
0.4
0.4
0.3
0.2
0.2
* Note: Data represent the average of triplicate measurements made at each pitch angle and velocity.
7-5
-------
Wind Tunnel Experimental Results
Table 7-2. DAT Probe (3D-2) Velocity Accuracy
Reference
Pitch
Angle
(deg)
-30
-20
-10
-5
0
5
10
20
30
Nominal
Velocity
(fps)
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
Vs
Total
Velocity
(fps)
30.0
60.1
90.3
30.0
60.1
90.3
30.0
60.1
90.3
30.0
60.1
90.3
30.0
60.1
90.3
30.0
60.1
90.3
30.0
60.1
90.3
30.0
60.1
90.3
30.0
60.1
90.3
Vax
Axial
Velocity
(fps)
25.98
52.05
78.20
28.19
56.48
84.85
29.54
59.19
88.93
29.89
59.87
89.96
30.00
60.10
90.30
29.89
59.87
89.96
29.54
59.19
88.93
28.19
56.48
84.85
25.98
52.05
78.20
Calculated Values for DAT Probe (3D-2)
Calibrated at 30-60-90 fps
Pitch
Angle
(deg)
-29.9
-29.8
-29.5
-21.9
-18.0
-18.8
-13.8
-9.8
-7.8
-4.5
-5.4
-4.8
1.0
0.7
0.6
4.9
5.5
5.0
7.2
7.7
8.2
16.9
19.4
21.2
28.0
30.0
30.0
Avg
Pitch
Angle
(deg)
-29.7
-19.6
-10.5
-4.9
0.8
5.1
7.7
19.2
29.3
Vs
Total
Velocity
(fps)
30.95
59.68
87.90
31.03
60.13
87.99
30.74
60.42
88.02
30.60
60.45
88.06
30.61
60.35
88.32
30.73
60.19
88.22
30.70
60.30
88.20
31.07
60.18
88.55
30.83
59.62
88.20
Vax
Axial
Velocity
(fps)
26.83
51.79
76.50
28.79
57.19
83.30
29.85
59.54
87.21
30.51
60.18
87.75
30.61
60.35
88.32
30.62
59.91
87.88
30.46
59.76
87.30
29.73
56.76
82.56
27.22
51.63
76.38
%
Dev.
3.3
-0.5
-2.2
2.1
1.3
-1.8
1.0
0.6
-1.9
2.1
0.5
-2.5
2.0
0.4
-2.2
2.4
0.1
-2.3
3.1
1.0
-1.8
5.5
0.5
-2.7
4.8
-0.8
-2.3
Calibrated at 30-60 fps
Pitch
Angle
(deg)
-29.7
-29.7
-29.2
-21.5
-17.0
-18.0
-12.4
-8.9
-7.3
-4.5
-5.3
-4.7
1.0
0.7
0.6
5.1
5.4
5.1
7.0
7.4
7.8
16.5
20.5
21.8
28.4
30.0
30.0
Avg
Pitch
Angle
(deg)
-29.5
-18.8
-9.5
-4.8
0.8
5.2
7.4
19.6
29.5
Vs
Total
Velocity
(fps)
30.56
58.95
86.76
30.60
59.34
86.84
30.32
59.64
86.89
30.21
59.69
86.95
30.25
59.65
87.29
30.35
59.45
87.13
30.32
59.55
87.10
30.70
59.36
87.57
30.48
58.85
87.06
Vax
Axial
Velocity
(fps)
26.55
51.21
75.73
28.47
56.75
82.59
29.61
58.92
86.19
30.12
59.43
86.66
30.25
59.65
87.29
30.23
59.19
86.79
30.09
59.05
86.29
29.44
55.60
81.31
26.81
50.97
75.40
%
Dev.
2.2
-1.6
-3.2
1.0
0.5
-2.7
0.2
-0.4
-3.1
0.8
-0.7
-3.7
0.8
-0.8
-3.3
1.2
-1.1
-3.5
1.9
-0.2
-3.0
4.4
-1.5
-4.2
3.2
-2.1
-3.6
Pitch
Angles
-30
to
30
-20
to
20
-10
to
10
Total
Velocity
30.0
60.1
90.3
30.0
60.1
90.3
30.0
60.1
90.3
Average Absolute Deviation
2.9
0.6
2.2
2.6
0.6
2.2
2.1
0.5
2.1
%
1.7
1.0
3.4
1.5
0.8
3.3
1.0
0.7
3.3
Note: Data represent the average of triplicate measurements made at each pitch angle and velocity.
7-6
-------
Wind Tunnel Experimental Results
Table 7-3. DAT Probe (3D-3) Velocity Accuracy
Reference
Pitch
Angle
(deg)
-30
-20
-10
-5
0
5
10
20
30
Nominal
Velocity
(fps)
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
Vs
Total
Velocity
(fps)
30.1
60.2
91.0
30.1
60.2
91.0
30.1
60.2
91.0
30.1
60.2
91.0
30.1
60.2
91.0
30.1
60.2
91.0
30.1
60.2
91.0
30.1
60.2
91.0
30.1
60.2
91.0
Vax
Axial
Velocity
(fps)
26.07
52.13
78.81
28.28
56.57
85.51
29.64
59.29
89.62
29.99
59.97
90.65
30.10
60.20
91.00
29.99
59.97
90.65
29.64
59.29
89.62
28.28
56.57
85.51
26.07
52.13
78.81
Calculated Values for DAT Probe (3D-3)
Calibrated at 30-60-90 fps
Pitch
Angle
(deg)
-30.0
-28.8
-28.4
-21.7
-20.0
-17.0
-12.1
-10.1
-8.3
-6.2
-3.0
-4.9
2.3
1.6
1.0
4.3
4.7
5.6
10.0
7.7
12.0
18.1
20.4
21.3
28.5
30.0
30.0
Avg
Pitch
Angle
(deg)
-29.1
-19.6
-10.2
-4.7
1.6
4.9
9.9
19.9
29.5
Vs
Total
Velocity
(fps)
30.39
59.81
89.22
30.68
60.70
88.91
30.61
59.95
89.65
30.52
59.92
89.75
30.51
60.03
89.62
30.71
59.89
89.56
30.57
60.16
89.32
30.46
60.74
89.72
30.72
59.85
89.17
Vax
Axial
Velocity
(fps)
26.32
52.41
78.48
28.51
57.04
85.03
29.93
59.02
88.71
30.34
59.84
89.42
30.49
60.01
89.61
30.62
59.69
89.13
30.11
59.62
87.37
28.95
56.93
83.59
27.00
51.83
77.22
%
Dev.
1.0
0.5
-0.4
0.8
0.8
-0.6
1.0
-0.4
-1.0
1.2
-0.2
-1.4
1.3
-0.3
-1.5
2.1
-0.5
-1.7
1.6
0.6
-2.5
2.4
0.6
-2.2
3.6
-0.6
-2.0
Calibrated at 30-60 fps
Pitch
Angle
(deg)
-30.0
-28.7
-28.2
-20.7
-18.9
-15.7
-11.1
-9.3
-7.8
-6.0
-2.8
-4.8
2.0
2.1
1.4
4.6
5.1
6.0
11.0
8.6
12.9
18.9
21.0
21.8
29.0
30.0
30.0
Avg
Pitch
Angle
(deg)
-29.0
-18.4
-9.4
-4.5
1.8
5.2
10.8
20.6
29.7
Vs
Total
Velocity
(fps)
30.23
59.45
88.63
30.29
60.11
88.09
30.38
59.54
89.07
30.33
59.53
89.18
30.31
59.57
88.96
30.47
59.41
88.84
30.29
59.64
88.44
30.14
60.27
89.02
30.44
59.19
88.19
Vax
Axial
Velocity
(fps)
26.18
52.15
78.11
28.33
56.87
84.80
29.81
58.76
88.25
30.16
59.46
88.87
30.29
59.53
88.93
30.37
59.17
88.35
29.73
58.97
86.21
28.52
56.27
82.65
26.62
51.26
76.37
%
Dev.
0.4
0.0
-0.9
0.2
0.5
-0.8
0.6
-0.9
-1.5
0.6
-0.9
-2.0
0.6
-1.1
-2.3
1.3
-1.3
-2.5
0.3
-0.5
-3.8
0.8
-0.5
-3.3
2.1
-1.7
-3.1
Pitch
Angles
-30
to
30
-20
to
20
-10
to
10
Total
Velocity
30.1
60.2
91.0
30.1
60.2
91.0
30.1
60.2
91.0
Average Absolute Deviation
1.6
0.5
1.5
1.5
0.5
1.6
1.4
0.4
1.6
%
0.8
0.8
2.3
0.6
0.8
2.3
0.7
0.9
2.4
* Note: Data represent the average of triplicate measurements made at each pitch angle and velocity.
7-7
-------
Wind Tunnel Experimental Results
Table 7-4. Spherical Probe (MS5-2) Velocity Accuracy *
Reference
Pitch
Angle
(deg)
-30
-20
-10
-5
0
5
10
20
30
Nominal
Velocity
(fps)
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
30
60
90
Vs
Total
Velocity
(fps)
30.2
60.4
91.0
30.2
60.4
91.0
30.2
60.4
91.0
30.2
60.4
91.0
30.2
60.4
91.0
30.2
60.4
91.0
30.2
60.4
91.0
30.2
60.4
91.0
30.2
60.4
91.0
Vax
Axial
Velocity
(fps)
26.15
52.31
78.81
28.38
56.76
85.51
29.74
59.48
89.62
30.09
60.17
90.65
30.20
60.40
91.00
30.09
60.17
90.65
29.74
59.48
89.62
28.38
56.76
85.51
26.15
52.31
78.81
Calculated Values for S
Calibrated at 30-60-90 fps
Pitch
Angle
(deg)
-27.5
-26.0
-29.7
-18.4
-20.0
-21.8
-9.2
-9.5
-11.6
-4.7
-4.7
-5.5
0.6
0.5
0.3
5.1
4.8
5.2
10.4
10.1
9.5
20.0
19.7
20.1
28.3
28.0
30.0
Avg
Pitch
Angle
(deg)
-27.7
-20.1
-10.1
-5.0
0.5
5.0
10.0
19.9
28.8
Vs
Total
Velocity
(fps)
27.96
57.99
93.34
30.08
60.72
94.88
29.87
60.56
92.39
29.86
60.63
91.97
29.97
60.47
91.82
29.97
60.47
91.86
30.06
60.35
91.79
30.04
60.52
91.80
28.65
60.91
93.42
Vax
Axial
Velocity
(fps)
24.80
52.12
81.08
28.54
57.06
88.09
29.49
59.73
90.50
29.76
60.43
91.55
29.97
60.47
91.82
29.85
60.26
91.48
29.57
59.41
90.53
28.23
56.98
86.21
25.23
53.78
80.90
%
Dev.
-5.2
-0.4
2.9
0.6
0.5
3.0
-0.9
0.4
1.0
-1.1
0.4
1.0
-0.8
0.1
0.9
-0.8
0.1
0.9
-0.6
-0.1
1.0
-0.5
0.4
0.8
-3.5
2.8
2.7
pherical Probe (MS5-2)
Calibrated at 30-60 fps
Pitch
Angle
(deg)
-27.8
-26.5
-29.7
-19.2
-20.8
-22.5
-9.9
-10.2
-12.4
-5.0
-5.0
-5.9
0.6
0.6
0.4
5.1
4.8
5.2
10.2
9.8
9.4
20.1
19.8
20.2
28.8
28.8
30.0
Avg
Pitch
Angle
(deg)
'
-28.0
-20.8
-10.8
-5.3
0.5
5.0
9.8
20.0
29.2
Vs
Total
Velocity
(fps)
28.48
59.13
94.91
30.40
61.76
96.89
30.06
60.90
93.14
30.01
60.94
92.45
30.10
60.73
92.22
30.09
60.72
92.25
30.16
60.58
92.14
30.16
60.73
92.16
29.09
61.98
94.54
Vax
Axial
Velocity
(fps)
25.19
52.92
82.44
28.71
57.73
89.51
29.61
59.94
90.97
29.90
60.71
91.96
30.10
60.73
92.22
29.97
60.51
91.87
29.68
59.70
90.90
28.32
57.14
86.49
25.49
54.31
81.87
%
Dev.
-3.7
1.2
4.6
1.2
1.7
4.7
-0.4
0.8
1.5
-0.6
0.9
1.4
-0.3
0.5
1.3
-0.4
0.6
1.3
-0.2
0.4
1.4
-0.2
0.7
1.1
-2.5
3.8
3.9
Pitch
Angles
-30
to
30
-20
to
20
-10
to
10
Total
Velocity
30.2
60.4
91.0
30.2
60.4
91.0
30.2
60.4
91.0
%
1.5
0.6
1.6
0.7
0.3
1.2
0.8
0.2
1.0
Average Absolute Deviation
%
1.1
1.2
2.4
0.5
0.8
1.8
0.4
0.6
1.4
* Note: Data represent the average of triplicate measurements made at each pitch angle and velocity.
7-8
-------
Wind Tunnel Experimental Results
7.4 ACCURACY OF PITCH ANGLE DETERMINATIONS
The wind tunnel test data include pitch angle measurements made by each 3D probe. These angles
are determined from the Fj vs. pitch angle calibration curves established for each probe. Table 7-5
presents the pitch angle measurements and calculated differences from the reference pitch angle
(determined by comparing the secured probe orientation with a protractor during the wind tunnel
tests). Performance of each probe in determining pitch was similar. The largest difference in pitch
angle was 4.3°, which occurred for the 3D-3 probe at -20° pitch and 90 fps using the 30/60 fps
calibration curves. An error of this magnitude in the angle measurement produces an error of only
0.3% error in the axial velocity.
While these wind tunnel tests showed the pitch angle accuracy of the spherical probe to be
comparable to that of the DAT probes, a previous study1 indicated that scratches on the probe head
can significantly alter the pitch angle calibration characteristics of the spherical probe design. No
attempt was made in this study to evaluate this property. This characteristic of spherical probes,
however, may merit investigation before such probes are approved for use under field conditions.
7.5 CONCLUSIONS
Cross-probe velocity accuracy results are summarized in Table 7-6. Recognizing the methodological
limitations of the approach employed in this study (as described in Section 7.3 above), the results
support the following findings on velocity accuracy:
• Overall performance: Across all pitch angles and all velocities, whether using the 30/60 or
30/60/90 fps calibration curves, the average absolute deviations from the actual axial
velocities never exceeded 3.4% for the spherical probe and the three DAT probes. For one
of the DAT probes (3D-1), the average absolute deviations were 0.7% or less.
• Comparative results using 30/60 and 30/60/90 fps calibration curves: For all probes over
all tested velocities, the difference between the average absolute deviations obtained with
the 30/60 and that obtained with the 30/60/90 calibration curves never exceeded ±1.2%.
• Patterns in velocity deviations (I): When using the 30/60 fps calibration curves, the greatest
average absolute velocity deviations for all probes over all pitch angles (-30° to + 30°) and
velocities always occurred at the 90 fps velocity setting.
• Patterns in velocity deviations (II): Except for one occurrence with DAT probe 3D-3 at
+10° using the 30/60 fps curves, the largest velocity errors occurred at the pitch angle
settings of-30°, -20°, 20°, and 30°.
Klinge, M. D. 1987. Detailed Near-Wake Flowfield Surveys and Comparison to an Euler Method of an
Aspect Ratio 4 Rectangular Wing. M .S. Thesis, Mechanical and Aerospace Engineering Department, North
Carolina State University, Raleigh.
7-9
-------
Table 7-5. 3D Probe Pitch Angle Accuracy
Reference
Pitch Nominal
Angle Velocity
(deg) (fps)
30
-30 60
90
30
-20 60
90
30
-10 60
90
30
-5 60
90
30
0 60
90
30
5 60
90
30
10 60
90
30
20 60
90
30
30 60
90
3D-1 Probe
30/60/90 Cal
Pitch
Angle Difference
(deg) (deg)
-30.0 0.0
-28.6 1.4
-28.5 1.5
-22.4 -2.4
-18.7 1.3
-18.5 1.5
-9.6 0.4
-10.9 -0.9
-9.6 0.4
-4.9 0.1
-5.1 -0.1
-5.0 0.0
0.8 0.8
0.6 0.6
1.9 1.9
5.3 0.3
3.4 -1.6
5.4 0.4
10.5 0.5
9.5 -0.5
10.4 0.4
18.1 -1.9
20.0 0.0
21.2 1.2
28.6 -1.4
29.6 -0.4
30.0 0.0
30/60 Cal
Pitch
Angle Difference
(deg) (deg)
-30.0 0.0
-28.1 1.9
-27.8 2.2
-21.7 -1.7
-18.0 2.0
-17.9 2.1
-9.4 0.6
-10.6 -0.6
-9.4 0.6
-4.9 O.I
-5.0 0.0
-5.0 0.0
0.4 0.4
0.9 0.9
2.5 2.5
5.5 0.5
3.9 -1.1
5.5 0.5
10.7 0.7
9.6 -0.4
10.6 0.6
190 -1.0
20.6 0.6
21.8 1.8
28.9 -1.1
29.7 -0.3
30.0 0.0
3D-2 Probe
30/60/90 Cal
Pitch
Angle Difference
(deg) (deg)
-29.9 0.1
-29.8 0.2
-29.5 0.5
-21.9 -1.9
-18.0 2.0
-18.8 1.2
-13.8 -3.8
-9.8 0.2
-7.8 2.2
-4.5 0.5
-5.4 -0.4
-4.8 0.2
1.0 1.0
0.7 0.7
0.6 0.6
4.9 -0.1
5.5 0.5
5.0 0.0
7.2 -2.8
7.7 -2.3
8.2 -1.8
16.9 -3.1
19.4 -0.6
21.2 1.2
28.0 -2.0
30.0 0.0
30.0 0.0
30/60 Cal
Pitch
Angle Difference
(deg) (deg)
-29.7 0.3
-29.7 0.3
-29.2 0.8
-21.5 -1.5
-17.0 3.0
-18.0 2.0
-12.4 -2.4
-8.9 1.1
-7.3 2.7
-4.5 0.5
-5.3 -0.3
•4.7 0.3
1.0 1.0
0.7 0.7
0.6 0.6
5.1 0.1
5.4 0.4
5.1 0.1
7.0 -3.0
7.4 -2.6
7.8 -2.2
16.5 -3.5
20.5 0.5
21.8 1.8
28.4 -1.6
30.0 0.0
30.0 0.0
3D-3 Probe
30/60/90 Cal
Pitch
Angle Difference
(deg) (deg)
-30.0 0.0
-28.8 1.2
-28.4 1.6
-21.7 -1.7
-20.0 0.0
-17.0 3.0
-12.1 -2.1
-10.1 -0.1
-8.3 1.7
-6.2 -1.2
-3.0 2.0 .
-4.9 0.1
2.3 2.3
1.6 1.6
1.0 1.0
4.3 -0.7
4.7 -0.3
5.6 0.6
10.0 0.0
7.7 -2.3
12.0 2.0
18.1 -1.9
20.4 0.4
21.3 1.3
28.5 -1.5
30.0 0.0
30.0 0.0
30/60 Cal
Pitch
Angle Difference
(deg) (deg)
-30.0 0.0
-28.7 1.3
-28.2 1.8
-20.7 -0.7
-18.9 1.1
-15.7 4.3
-11.1 -1.1
-9.3 0.7
-7.8 2.2
-6.0 -1.0
-2.8 2.2
-4.8 0.2
2.0 2.0
2.1 2.1
1.4 1.4
4.6 -0.4
5.1 0.1
6.0 1.0
11.0 1.0
8.6 -14
12.9 2.9
18.9 -1.1
21.0 1.0
21.8 1.8
29.0 -1.0
30.0 0.0
30.0 0.0
Spherical Probe
30/60/90 Cal
Pitch
Angle Difference
(deg) (dee)
-27.5 2.5
-26.0 4.0
-29.7 0.3
-18.4 1.6
-20.0 0.0
-21.8 -1.8
-9.2 0.8
-9.5 0.5
-11.6 -1.6
-4.7 0.3
-4.7 0.3
-5.5 -0.5
0.6 0.6
0.5 0.5
0.3 0.3
5.1 0.1
4.8 -0.2
5.2 0.2
10.4 0.4
10.1 0.1
9.5 -0.5
20.0 0.0
19.7 -0.3
20.1 0.1
28.3 -1.7
28.0 -2.0
30.0 0.0
30/60 Cal
Pitch
Angle Difference
(deg) (deg)
-27.8 2.2
-26.5 3.5
-29.7 0.3
-19.2 0.8
-20.8 -0.8
-22.5 -2.5
-9.9 0.1
-10.2 -0.2
-12.4 -2.4
-5.0 0.0
-5.0 0.0
-5.9 -0.9
0.6 0.6
0.6 0.6
0.4 0.4
5.1 0.1
4.8 -0.2
5.2 0.2
10.2 0.2
9.8 -0.2
9.4 -0.6
20.1 0.1
19.8 -0.2
20.2 0.2
28.8 -1.2
28.8 -1.2
30.0 0.0
-J
o
Angle Nominal
Range Velocity
-30 30
to 60
30 90
Average Absolute Difference
(degrees) (degrees)
0.9 0.7
0.8 0.9
0.8 I.I
Average Absolute Difference
(degrees) (degrees)
1.7 1.5
0.8 1.0
0.9 1.2
Average Absolute Difference
(degrees) (degrees)
1.3 0.9
0.9 1.1
1.3 1.7
Average Absolute Difference
(degrees) (degrees)
0.9 0.6
0.9 0.8
0.6 0.8
-------
Wind Tunnel Experimental Results
Table 7-6. Summary of 3D Probes Velocity Accuracy Results
Probe ED
DAT Probe (3D- 1)
DAT Probe (3D-2)
DAT Probe (3D-3)
Spherical (MS5-2)
Calibration
Curve
30/60
30/60/90
Difference
30/60
30/60/90
Difference
30/60
30/60/90
Difference
30/60
30/60/90
Difference
Average Absolute
Deviation
(over the ±30°
pitch range)
30fps
0.5%
0.7%
0.2%
1.7%
2.9%
1.2%
0.8%
1.6%
0.8%
1.1%
1.5%
0.4%
60fps
0.6%
0.5%
0.1%
1.0%
0.6%
0.4%
0.8%
0.5%
0.3%
1.2%
0.6%
0.6%
90fps
0.7%
0.5%
0.2%
3.4%
2.2%
1.2%
2.3%
1.5%
0.8%
2.4%
1.6%
0.8%
Maximum Deviation and
Conditions When it Occurs
Deviation
+3.0%
-2.2%
Pitch
Angle
-30°
-30°
Nominal
Velocity
90fps
30fps
+4.4%
+5.5%
+20°
+20°
30fps
30fps
-3.8%
+3.6%
+10°
+30°
90fps
30fps
+4.7%
-5.2%
-20°
-30°
90fps
30fps
Key findings for the examination of pitch angle accuracy are:
• Overall performance: Across all probes, all velocities, and all pitch angle settings, the
largest error in measuring pitch angle was 4.3 °. The average absolute difference between
derived and actual pitch angles never exceeded 1.7°.
• Other considerations: The potential impact of probe head scratches on the accuracy of pitch
angle determinations using the spherical design probe (which were reported in a previous
study) were not evaluated in the current study.
These findings point to the following conclusions:
Over the pitch angle ranges ±30° and ±20°, the velocity accuracy of all tested 3D probes
was within 5.5% of the true axial velocity. In the ±10° pitch-angle range, the velocity
accuracy of all tested 3D probes was within 3.8% of the true axial velocity.
7-11
-------
Wind Tunnel Experimental Results
No significant gain in accuracy is attained by developing calibration curves at three velocities
rather than at the two specified in draft Method 2F.
In order to minimize error in the flue gas velocity measurement, calibration curves for 3D
probes should be developed at velocity settings close to, or possibly bracketing, the
prevailing velocity where the probe will be used.
The largest error in measuring pitch angle for all four probes was 4.3°, which occurred at
90 fps and -20° pitch and produced a velocity error of 0.3%.
7-12
-------
OJ
CM
Q.
I
0.5
0.4
0.3 H
0.2
0.1 H
0.0
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
-30
Figure 7-1. Probe: DAT 3D-1
F1 vs. Pitch Angle
-20
-10
10
20
30
Pitch (degrees)
• • • Velocity = 30 (fps) e-€>-0 Velocity = 60 (fps) *-*-* Velocity = 90 (fps)
-------
Figure 7-2. Probe: DAT 3D-1
F2 vs. Pitch Angle
CM
Q.
I
-------
Figure 7-3. Probe: DAT 3D-2
F1 vs. Pitch Angle
CM
Q.
I
1
0.6 d
0.5
0.4
0.3 H
0.2
0.1
0.0
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
-30
-20
-10
\
10
~
20
\
30
Pitch (degrees)
• • • Velocity = 30 (fps) e-€>-© Velocity = 60 (fps) *-*-* Velocity = 90 (fps)
-------
Figure 7-4. Probe: DAT 3D-2
F2 vs. Pitch Angle
-& Velocity = 60 (fps)
-»* Velocity = 90 (fps)
-------
Figure 7-5. Probe: DAT 3D-3
F1 vs. Pitch Angle
I
"OL
0.5 ^
0.4
0.3
0.2 H
0.1
0.0
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
-0.7
-30
-20
-10
10
20
i
30
Pitch (degrees)
• • • Velocity = 30 (fps) e-€>-© Velocity = 60 (fps) *-*-* Velocity = 90 (fps)
-------
oo
Figure 7-6. Probe: DAT 3D-3
F2 vs. Pitch Angle
1.09
1.08
c\T
I 1.06
|j 1-05
| 1-04
B- t03
-e Velocity = 60 (fps) *-^K-* Velocity = 90 (fps)
-------
Figure 7-7. Probe: Spherical MS5-2
F1 vs. Pitch Angle
I 1
•8.
ii
LL
-1
-2
-3
-30
-20
-10
I
10
Pitch (degrees)
I
20
I
30
• • • Velocity = 30 (fps) ©
Velocity = 60 (fps)
-* Velocity = 90 (fps)
-------
Figure 7-8. Probe: Spherical MS5-2
F2 vs. Pitch Angle
2
o
CM
Q.
-0 Velocity = 60 (fps) *-*-* Velocity = 90 (fps)
-------
-4
•u
0.6
0.5
0.4
0.3
0.2 H
0.1
0.0
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
-0.7
-30
Figure 7-9. Probe: DAT 3D-1
F1 vs. Pitch Angle
-20
-10
10
20
30
Pitch (degrees)
-------
S)
N>
1.30
1.28^
1.26
1.24
1.22
1.20 H
1.18
1.16
1.14
1.12
i—
w 1.10
1.08
1.06
1.04
1.02
1.00
0.98
I
H
J
CD
-a
II
CM
Figure 7-10. Probe: DAT 3D-1
F2 vs. Pitch Angle
-30
-20
-10
i
10
I
20
30
Pitch (degrees)
-------
Figure 7-11. Probe: DAT 3D-2
F1 vs. Pitch Angle
s
I
I
0.6
0.5
0.4
0.3
0.2
0.1
0.0
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
-0.7
-30
-20
-10
10
20
30
Pitch (degrees)
-------
Figure 7-12. Probe: DAT 3D-2
F2 vs. Pitch Angle
I
11
~s.
s'
o
T3
CM
1.30
1.28
1.26
1.24
1.22
1.20
1.18
1.16
1.14
1.12
1.10
1.08
1.06
1.04
1.02 H
1.00
0.98
-30
-20
-10
10
20
30
Pitch (degrees)
-------
Figure 7-13. Probe: DAT 3D-3
F1 vs. Pitch Angle
I
"EL
0.6
0.5
0.4
0.3 H
0.2
0.1
0.0
-0.1
-0.2
-0.3 H
-0.4
-0.5
-0.6
-0.7
-30
-20
-10
10
r
20
30
Pitch (degrees)
-------
tb
ON
1.30
1.28
1.26
1.24
1.22
1.20
1.18
1.16
1.14
^ 1.12
w" 1.10
" 1.08
^ 1.06
1.04
1.02
1.00
0.98
-------
Figure 7-15. Probe: Spherical MS5-2
F1 vs. Pitch Angle
i
II
C
-1
-2
-30
-20
-10
10
Pitch (degrees)
20
30
-------
to
00
1.30
1.28
1.26
1.24
^ 1.22 -\
csT
Q- 1.20 H
A
"cl
-------
8.0 ENTRY WALL EFFECTS ON YAW NULLING
The experiments described in this section were conducted to confirm swirl tunnel findings obtained
by Fossil Energy Research Corporation (FERCo) in an unpublished study funded by the Electric
Power Research Institute (EPRI) in 1996. FERCo found that a Type S probe yielded multiple yaw
nulls within several inches of the wall through which the probe was inserted into the tunnel. This
effect is called an "entry wall" effect, because it appears to be associated with flow disturbance caused
by a combination of the probe shaft and proximity to the entry wall.
8.1 TEST PROCEDURES
All experiments were conducted at 60 fps and at pitch angles of-10°, 0°, or 10°. The distance from
the entry wall is the shortest (perpendicular) distance between the end of the probe head and the entry
wall. Distances were measured with a ruler. For manual probes, multiple yaw nulls were sought
using the standard yaw nulling techniques for each probe. Manual probes tested were the Type S,
modified Kiel (MK2), DAT, and spherical probes. To look for multiple yaw nulls, AP was measured
and plotted as a function of yaw angle in 5° increments from -20° to 20°. Yaw angles were
measured with the digital inclinometer. Points were connected by straight lines. A yaw null was
defined as the angle at which any line crossed the AP = 0° line. Multiple yaw nulls occurred when
AP = 0 at more than one yaw. When multiple yaw nulls were observed, three AP = 0 values typically
occurred. In some cases, AP remained at 0 for 5° or more. In this case, the yaw angle determination
was said to be flat and the end points were recorded.
The vast majority of measurements were taken with the Autoprobe fitted with the Type S probe.
Advantage was taken of the Autoprobe's ability to scan automatically over a yaw angle range and
record AP at specified angles. All Autoprobe runs represent two replicate scans, each of which was
made from -10° to 190°. Each run consisted of two replicate scans made consecutively. Data were
collected on disk and downloaded to Quattro Pro®, and AP vs. yaw angle was plotted. Plots over
three different yaw angle ranges were generated: a complete scan of AP vs. yaw from -10° to 190°;
a plot of AP vs. yaw centered at 0°; and a plot of AP vs. yaw from 165° to 190°.
8.2 RESULTS OF ENTRY WALL EFFECTS TESTS
Before conducting quantitative tests of entry wall-related multiple yaw nulls, a qualitative test was
performed to examine whether rotation of a Type S probe near an entry wall would cause a flow
disturbance. To make this observation, strings were taped to the entry wall surrounding the probe
shaft. With no flow in the tunnel, the strings were limp. The fan motor was increased slowly until
a flow velocity of about 40 fps was reached, at which time all strings were lifted into a horizontal
position in the direction of flow. At this point, the Autoprobe Type S probe was extended into the
wind tunnel for a distance of about 1 inch. The probe was then rotated to a yaw angle of about -45 °.
As shown in Figure 8-1, the string immediately downstream of the probe at this angle (the flow
direction is right to left) is horizontal, indicating that flow moves around the probe symmetrically at
this yaw angle. Figure 8-2 shows the same configuration after the probe has been rotated to a yaw
angle of about 20°. At this yaw angle, the direction of flow immediately downstream of the probe
is indicated by the string to the left of the probe, which is deflected by about 45 ° from the flow
direction. No quantitative analyses were conducted from these photographs.
8-1
-------
Wind Tunnel Experimental Results
Figure 8-3 shows plots of AP vs. yaw angle determined with the manual Type S probe at 0° pitch
at 1", 3", 6", 9", 12", and 42" from the entry wall. Each curve is the average of three replicate runs.
Figures 8-4 through 8-9 are plots for each distance showing the three replicate runs. The largest
distance, 42", is 3" from the far wall of the wind tunnel. The figures clearly show pronounced
multiple yaw nulls at distances of 1" through 6" from the entry wall. At 12" and 42", unique yaw
nulls are found. The transition from multiple to unique yaw nulls is apparent at 9".
Figures 8-10 through 8-12 contain plots for the Autoprobe Type S at 0° pitch and distances from
1" to 44.5" from the wall. Figure 8-10 includes data for yaw angles from -10° to 15°, Figure 8-11
data from 165° to 190°, and Figure 8-12 for the complete yaw angle scan. For yaw angles around
0°, the thermocouple is on the upstream side of the probe, while at 180° the thermocouple is
downstream. These data show more than one crossing of the AP = 0 line or cases where AP remains
near 0 over a substantial yaw angle range.
Figures 8-13 through 8-15 and Figures 8-16 through 8-18 are similar plots at pitch angles of 10°
and -10°. A similar trend is apparent here. The only exception is the sharp yaw null at 1" and 10°
pitch observed at both 0° and 180°. Individual plots for each distance/pitch angle plot are presented
in Appendix B.
The observation reported in the EPRI-sponsored study is confirmed by this work. With the manual
Type S probe, multiple yaw nulls are clearly observed at distances from the entry wall of up to about
9". Thereafter, unique yaw nulls are found. Multiple yaw nulls are not found near the far wall,
indicating that the effect is related to an interaction between the probe shaft and the entry wall.
Multiple yaw nulls are observed with the Autoprobe Type S at essentially all distances from the entry
wall. The unique design of the Autoprobe may be responsible for the occurrence of multiple yaw
nulls at all distances from the wall. Whereas the manual Type S probe is attached to a 1" shaft and
has only a thermocouple in addition to the pilot head, the Autoprobe has a 2" diameter clamp located
within 2" of the pilot head and has a dilution probe parallel lo the Ihermocouple on Ihe opposite side
of the probe. The clamp may in facl be causing a "pseudo-enlry wall effecl", which is evident at all
dislances from the entry wall.
To determine whelher multiple yaw nulls at dislances near Ihe enlry wall are uniquely associated wilh
Type S probes, AP vs. yaw angle curves were obtained for the modified Kiel probe and the two 3D
probes (i.e., DAT and spherical). Each probe uses Fechheimer-type ports lo locale Ihe yaw null
point. Figures 8-19 and 8-20 are plots of APf (across Ihe Fechheimer ports) vs. yaw angle over the
range -20° to 20° for the modified Kiel probe (MK2). Each plot presents curves, determined in
Iriplicale, al pitch angles of-10°, 0°, and 10°, for a tola! of 9 curves. The dala shown in Figure 8-19
were taken at 1" from the entry wall, and the dala in Figure 8-20 were laken al 5" from Ihe wall.
Bolh plols clearly indicate that the Fechheimer ports on the modified Kiel probe locate a unique yaw
null at each dislance independenl of pilch. In facl, Ihe curves al bolh dislances from Ihe wall are very
similar. Al 1", a yaw null is found al aboul 3.5°; al 5" from Ihe wall, Ihe yaw null is found al
about 2°.
8-2
-------
Wind Tunnel Experimental Results
Figure 8-1. Flow pattern at yaw angle of -45 °.
Figure 8-2. Flow pattern at yaw angle of +20c
8-3
-------
Wind Tunnel Experimental Results
Figures 8-21 and 8-22 are similar plots for the DAT and spherical probes. Because of the length of
these probes, the data for the spherical probe were taken at 4" from the wall and at 0° pitch only.
Data for the DAT probe were taken at 3" from the wall at pitch angles of-10° and 10°. In both
cases, unique yaw nulls are found—at about 0.5° for the spherical probe and 2° for the DAT probe.
Table 8-1 summarizes data collected from all probes that were tested. (The modified S probe was
not tested due to time constraints.) As noted from the figures, multiple yaw nulls are observed only
for the Type S probes. Data in the last column of Table 8-1 indicate that unique yaw nulls and the
middle yaw null, when three yaw nulls are present, are within 4° of the true yaw zero. The yaw angle
range column shows that the range in which multiple yaw nulls is found is symmetric about yaw angle
zero.
8.3 CONCLUSIONS
The following conclusions can be drawn from the data presented in this section:
• In a laminar-flow wind tunnel, three yaw nulls are observed at wall entry distances of up to
about 9" using a Type S probe. The center yaw null is found within 1 ° of the true yaw zero.
• At entry wall distances greater than about 9", unique yaw nulls are found with the manual
Type S probe.
• Multiple yaw nulls are found with the Autoprobe Type S at all distances from the wind
tunnel center line. A "pseudo-entry wall effect" appears to be associated with the
Autoprobe.
• At distances near the entry wall, unique yaw nulls are found with all probes that use
Fechheimer-type ports to determine the yaw angle: the modified Kiel, spherical, and DAT
probes.
It should be noted that probes that detect multiple yaw nulls near the entry wall in a controlled
environment such as a wind tunnel may not exhibit the same behavior under field conditions, where
the increased variability of the gas stream flow may prevent the occurrence of this behavior. Field
test data should be observed to see whether this is the case.
Table 8-1. Summary Table of the Effect of Distance from Wall Entry
on the Number of Yaw Nulls Observed
Probe
Type S (3/8 S-3)
Pitch
Angle (°)
0.00
0.00
0.00
0.00
0.00
0.00
Distance from
entry wall (in.)
1
3
6
9
12
42
Number of
yaw nulls
3
3
3
3
1
1
Yaw angle null
range (°)
-10to+15
-12to+17
- 8 to 12.5
-5to+10
Central
yaw null (°)
+1
0.00
0.00
0.00
+1
+2.5
8-4
-------
Wind Tunnel Experimental Results
Table 8-1. Summary Table of the Effect of Distance from Wall Entry
on the Number of Yaw Nulls Observed (continued)
Probe
Autoprobe Type S
(AS)"
Autoprobe Type S
(AS)"
Autoprobe Type S
(AS)'
Autoprobe Type S
(AS)b
Autoprobe Type S
(AS)'
Pitch
Angle (°)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
-10
-10
-10
-10
-10
-10
-10
-10
+10
+10
+10
+10
+10
+10
Distance from
entry wall (in.)
1
1.5
3
6
11.25
12
22.5
33.75
44.5
1
1.5
3
6
11.25
12
22.5
33.75
44.5
1
2
6
22.5
1
2
6
22.5
1
2
6
12
22.5
38.8
Number of
yaw nulls
3
3
3
3
2
2
2
1
2
3
2
2
3
2
2
1
1
1
2
2
2
2
2
1
3
2
1
3
3
3
3
3
Yaw angle null
range (°)
-6to + 9
-7.5 to +10
-5 to +10
-5to + 7
-2.5 to + 7.5
- 2 to + 5
Oto + 5
- 1 to + 5
+174 to +188
+175 to +190
+175 to +190
+174 to +187.5
+175 to +185
+175 to +185
-6to + 5
-5to + 7.5
0 to + 7.5
-3to + 5
+174 to +185
+175 to +187.5
+174 to +185
-5.5 to + 8.5
-5.5to + 7
-6to + 6
-6.5 to + 6
-5.5 to + 6.5
Central
yaw null (°)
+4
0.00
0.00
+2
-2
182.5
182.5
176
176
177
190
181
+1.5
+3
+3
+2
+4
+3.5
8-5
-------
Wind Tunnel Experimental Results
Table 8-1. Summary Table of the Effect of Distance from Wall Entry
on the Number of Yaw Nulls Observed (continued)
Probe
Autoprobe Type S
(AS)b
Modified Kiel (MK2)
DAT (3D-1)
Spherical (MS502)
Pitch
Angle (°)
+10
+10
+10
+10
+10
+10
-10
-10
0.00
0.00
+10
+10
-10
+10
0.00
Distance from
entry wall (in.)
1
2
6
12
22.5
38.8
1
5
1
5
1
5
3
3
4
Number of
yaw nulls
1
3
3
3
3
3
1
1
1
1
1
1
1
1
1
Yaw angle null
range (°)
+174 to +190
+174to+187.5
+174 to +187.5
+174 to +187
+174 to +187.5
Central
yaw null (°)
182
182
183
181.5
182
180
3.50
+2.5
3.50
+2.5
3.50
+2.5
+2.5
+3
+1
Thermocouple facing upwind.
Thermocouple facing downwind.
8-6
-------
0.5
Q_
CO
VU f\
n 0
CD
Q
Figure 8-3.
Manual S-Type: Yaw/Wall Effect
60 fps,0°pitch, 1-42" from wall
-0.5
-1
-20
-10 o 10
Yaw Angle (degrees)
20
avg 1"
avg 3"
avg 6"
avg 9"
avg 12"
avg 42"
8-7
-------
0.5
Q_
(D n
±i 0
0>
Q
-0.5
-1
-20
Figure 8-4.
Manual S-Type: Yaw/Wall Effect
60 fps, 0°pitch, 1" from wall
-10 0 10
Yaw Angle (degrees)
20
8-8
-------
0.5
Q_
S o
-------
1 T
Figure 8-6.
Manual S-Type: Yaw/Wall Effect
60 fps, 0°pitch, 6" from wall
0.5
S o
-0.5
-1
-20 -10 0 10 20
Yaw Angle (degrees)
8-10
-------
0.5
a.
S o
-------
Figure 8-8.
Manual S-Type: Yaw/Wall Effect
60 fps, 0°pitch, 12" from wall
00
.0
Oc
.0
OA
.*f
Oo .
.£.
Q_
2 n -
— U
-------
0.5
Q_
CO n
±± 0
0
Q
-0.5
-1
-20
Figure 8-9.
Manual S-Type: Yaw/Wall Effect
60 fps, 0°pitch, 42" from wall
-10 0 10
Yaw Angle (degrees)
20
8-13
-------
0.6
Figure 8-10.
Autoprobe S-Type: Yaw/Wall Effect
60 fps, 0°pitch
Inches from Wall
0.4
0.2
Q_
ro
-0.2
-0.4
-10
0 10
Yaw Angle (degrees)
avg 1
avg 1.5
avg 3
avg 6
avg 11.25
avg 12
avg 22.5
avg 33.75
avg 44.5
8-14
-------
Figure 8-11.
Inches from Wall
Autoprobe S-Type:oYaw/Wall Effect
60 fps, 0°pitch
Q_
S
-------
-0.5
-50
Figure 8-12.
Autoprobe S-Type: Yaw/Wall Effect
60 fps, 0°pitch
o
50 100 150
Yaw Angle (degrees)
Inches from Wall
avg 1
avg6
avg 1.5
avg 3
avg 11.25 —^- avg 12
avg 22.5 -*- avg 33.75 -a- avg 44.5
200
8-16
-------
Figure 8-13.
Autoprobe S-Type: Yaw/Wall Effect
60fps,+10°pitch
"55
Q
-0.2
-10
0 10
Yaw Angle (degrees)
Inches from Wall
avg 1"
avg 2"
avg 6"
avg 12"
avg 22.5"
avg 38.8"
8-17
-------
0.6
Figure 8-14.
Autoprobe S-Type: Yaw/Wall Effect
60fps, +10 pitch °
0.4
0.2
CL
&
0)
Q
0
-0.2
-0.4
165
175 185
Yaw Angle (degrees)
Inches from Wall
avg 1"
avg 2"
avg 6"
avg 12"
avg 22.5"
avg 38.8"
8-18
-------
Figure 8-15.
-0.5
-50
Autoprobe S-Type: Yaw/Wall Effect
60 fps,+10 pitch0
I 50 100 150
Yaw Angle (degrees)
Inches from Wall
avgl" -^avg2" -»- avg 6"
200
avg 12"
avg 22.5" -*- avg 38.8"
8-19
-------
0.6
Figure 8-16.
Autoprobe S-Type: Yaw/Wall Effect
60 fps,-10 pitch °
0.4
0.2
Q_
£
-------
Figure 8-17.
Autoprobe S-Type: Yaw/Wall Effect
60 fps,-10 pitch0
-0.2
-0.4
165
175 185
Yaw Angle (degrees)
Inches from Wall
avg 1"
avg 2"
avg 6"
avg 22.5"
8-21
-------
-0.5
-50
Figure 8-18.
Autoprobe S-Type: Yaw/Wall Effect
60 fps,-10 pitch °
0 50 100
Yaw Angle (degrees)
150 200
Inches from Wall
avg 1" -*- avg 2" -•- avg 6" -s- avg 22.5"
8-22
-------
0.5
CL
e
Q_
-0.5
-1
-1.5
-20
Figure 8-19.
Mod Kiel w/ Therm.: Yaw/Wall Effect
60fps, 1" from wall, pitch=+10,0,-10 c
-10 o 10
Yaw Angle (degrees)
20
-g- 10 -B- 10 -B- 10 -e- o -e- o -e- o -^ -10 -^ -10 -v- -10
8-23
-------
1.1
0.5
0
-0.5
-1
-1.5
-20
Figure 8-20.
Mod Kiel w/ Therm.: Yaw/Wall Effect
60 fps, 5" from wall, pitch= +10,0,-10 °
10 0 10
Yaw Angle (degrees)
20
4^10
0 -V- -10 -9- -10 -9- -10
8-24
-------
0.5
0
Q_
-0.5
-1
-1.5
-20
Figure 8-21.
Spherical Probe: Yaw/Wall Effect
60 fps, 0°pitch, port #1, 4" from wall
-10 o 10
Yaw Angle (degrees)
20
8-25
-------
1.5
0.5
Q_
CO n
±i 0
Q)
G
-0.5
-1
-1.5
Figure 8-22.
DAT Probe: Yaw/Wall Effect
60 fps, 3" from wall, pitch= +10,-10
-20
-10 o 10
Yaw Angle (degrees)
-B-
10
-B-
10
10
-10
-10
-10
20
8-26
-------
9.0 CONCLUSIONS AND RECOMMENDED FOLLOW-UP
The conclusions found in previous sections of this report are presented below, along with Cadmus'
recommendations for follow-up actions that are technically justifiable based on the wind tunnel
findings. The recommendations relate to both the upcoming collaborative field tests and possible
revisions of EPA's test methods for volumetric flow.
MODIFIED KIEL PROBE EXPERIMENTS—SECTION 3
CONCLUSIONS
Wake ports are too insensitive to be used for
yaw nulling.
Fechheimer ports can be used to yaw null
with an accuracy of better than 1° at
velocities between 40 and 90 fbs and at all
pitch angles in the range ±30°.
At pitch and yaw angles equal to zero, a
wake port placed between 1-3/8 inches and
3-5/8 inches from the Kiel head can be used
to measure static pressure. For velocities
between 40 and 90 fps, each port can be used
to derive a constant calibration coefficient.
For a wake port located 3-5/8 inches from
the Kiel port, pressure drop measured by the
probe is independent of pitch angle in the
range ±30°; for wake ports placed closer to
the Kiel head, measured pressure drop is
dependent on pitch angle.
The modified Kiel probe using wake port 4
can be used to determine total stack flow and
yaw angle in the velocity range 40 to 90 fps
for pitch angles hi the range ±30°. The
probe cannot be used to determine pitch angle.
RECOMMENDATIONS
The wind tunnel test results provide a sufficient technical basis for including the modified Kiel probe
in EPA's planned collaborative study. Before including this probe in the collaborative study,
however, EPA should consider performing a pre-test of the probe in a power plant stack to confirm
the probe's field readiness based on a comparison of its velocity measurements with those obtained
using DAT and Type S probes. If the modified Kiel probe gives plausible results relative to the DAT
and Type S probes (e.g., velocity measurements lower than those of the non-nulled Type S, but
higher than the DAT's), it should be included in the collaborative field study. In that case, EPA
should review the results of the collaborative study to determine whether revising Test Method 2 to
allow use of this probe is technically justified.
9-1
-------
Wind Tunnel Experimental Results
VARIABILITY OF 2D PROBE CALIBRATION COEFFICIENTS
WITH VELOCITY CHANGES—SECTION 4
CONCLUSIONS
For all probes except the manual modified S,
variability in the calibration coefficients
derived at individual velocity settings was
low. The variability was especially low for
the Autoprobes.
As reflected in the percentage change in
mean Cp and in the low coefficients of
variation, the calibration coefficients for all
the tested 2D probes, except for the modified
S probes, are stable in the 30 to 90 fps range
(40 to 90 fps for the modified Kiel probe).
Across all velocities, the greatest variation in
calibration coefficients was exhibited by the
manual modified S probes, reaching as high
as 7.15% across the tested range. It may be
advisable to require this probe to be
calibrated at a wind tunnel velocity close to
the prevailing velocity where the probe will
be used.
The modified Kiel probe should not be used
at velocities less than 40 fps because its
calibration coefficient appears to be very
sensitive to velocity changes below this level.
This specific velocity cutoff applies only to
the modified Kiel probe dimensions tested
here; modified Kiel probes of the same
design but with a different diameter and port
sizes may have different velocity cutoffs.
RECOMMENDATIONS
EPA should consider revising the calibration requirements currently found in Method 2 (40 CFR Part
60, Appendix A) §4.1.2.3 to define conditions under which a probe would be required to use
calibration coefficients obtained at multiple velocities. For example, such a requirement might take
the following form:
Step 1: Testers would be asked to declare the
Minimum velocity (V^ where the probe will be used:
Maximum velocity (V^ where the probe will be used:
.fps
, fps
Step 2: Following the procedures in Method 2, §4.1, the tester would derive Cp1™" and Cp1™", the
calibration coefficient at the declared minimum and maximum velocities respectively.
Step 3: Users would then calculate the percent change in calibration coefficient (PCCC),
where:
9-2
-------
Wind Tunnel Experimental Results
xt tnox
PCCC = -i — — p— * 100%
Step 4: Users would apply the following criteria to determine the required number of cali-bration
coefficients to be used:
PCCCi 3.0% =» Use a single calibration factor determined by averaging Cp™" and C,,""" .
PCCC > 3.0% =» Determine Cpl at the velocity (VJ mid-way between V^ and Vn
nun •
Determine the percent change in calibration (PCCC1) between Cpl and C™. If PCCC* 3.0%,
record Cpl and Cfm. If PCCC1 > 3.0%, determine Cp2 for the velocity V2 mid-way between
the VIBK and Vj. Determine the percent change in calibration (PCCC") between Cp2 and C,,1""
and PCCC"' between Cp2 and Cpl. If either percent change is <. 3.0%, record the calibration
coefficients and velocities. If either percent change is > 3.0%, continue generating the
calibration curve until the percent change in calibration coefficients between adjacent velocities
is <, 3.0%. Repeat these steps for the velocity interval between V, and \^. Once the
calibration table is established, the tester should use the calibration coefficient for the velocity
closest to the velocity measured in the field.
Wind tunnel tests have demonstrated that calibration coefficients for the manual modified Type S
probe have significantly higher variation with velocity than other 2D probes. For this reason, we
recommend against including the manual modified Type S probe in the field tests. We do, however,
recommend including the Autoprobe modified S probe in the field tests.
If the results of the preliminary field tests conducted with the modified Kiel probe lead to the inclusion
of this probe in the planned collaborative tests, the four modified Kiel probes that will be used should
be calibrated in the wind tunnel at velocities between 10 and 90 fps, to determine if there should be
a low-velocity cutoff for use of this probe.
9-3
-------
Wind Tunnel Experimental Results
PITCH ANGLE EFFECTS ON VELOCITY MEASUREMENTS
PERFORMED WITH 2D PROBES—SECTION 5
CONCLUSIONS
Calibration Coefficients
• The calibration coefficients of all probes,
except the Type S and French, were stable to
within 3.5% in the -10° to 20° pitch range.
• None of the probes, except the modified
Kiel, maintained a stable calibration
coefficient in the -30° to -10° pitch range.
Velocity Determinations
• In the ±10° pitch-angle range, no single
probe excelled under all velocity conditions.
Maximum deviations from axial velocity
ranged from +3.5% for the Autoprobe
operated with a modified S pitot head to
+7.3% for the French probe.
• In the ±10° pitch-angle range at the higher
tested velocities (60 and 90 fps), all probes
except the French measured velocities that
were less than or equal to 4.9% of the axial
velocity. The modified Kiel had the lowest
deviation at -2.2%.
In the ±20° pitch angle range, across all
velocities tested, none of the tested 2D
probes kept deviations from actual axial flow
within ±5%.
In the ±20° pitch angle range over all
velocities, the manual Type S had the
smallest maximum deviation (-6.7%) and the
smallest maximum overestimation (+5.2%).
At 60 and 90 ftps, the maximum deviation for
the manual Type S probe (-5.5%) was also
lowest of all tested probes.
At negative pitch angles, the manual Type S
probe always underestimated true axial
velocity—the more negative the pitch, the
larger the underestimate.
RECOMMENDATIONS
In revisions to Method 2, EPA should consider limiting the use of 2D probes to specific pitch
ranges, even if yaw nulling procedures for 2D probes are included in the revised test method.
Pitch-angle range restrictions would require sources either to (a) explicitly determine the extent of
pitch in a stack (e.g., by performing a pre-test 3D traverse) or (b) demonstrate by some other
means that flow is within stipulated pitch-angle limits.
The wind-tunnel tests suggest that, for all 2D probes except the French probe, the acceptable
pitch-angle range should be restricted to 0° to +10°. The manual Type S and modified Kiel
probes gave consistently low velocity measurements between 0° and -10° pitch. At pitch angles
of 10° and greater and velocities of 60 and 90 fps, all probes gave consistently high
measurements. If allowed at all, use of the French probe should be restricted to situations where
pitch angles are essentially 0°.
The collaborative field tests should be designed to provide additional data on how pitch angle affects
the accuracy of velocity measurements.
9-4
-------
Wind Tunnel Experimental Results
ACCURACY OF YAW ANGLE MEASUREMENTS AND EFFECT
ON VELOCITY MEASUREMENT—SECTION 6
CONCLUSIONS
Use of Fechheimer ports produces consist-
ently better yaw angle measurements than
other procedures for measuring yaw angles.
In the wind tunnel, the manually operated
Type S probe gave yaw null and velocity
accuracy results comparable to those of the
probes employing Fechheimer ports.
The maximum yaw angle error using the
Autoprobe1 was higher than the maximum
errors produced by either the manually
operated Type S probe or any of the
manually operated probes that employ
Fechheimer ports.
Among the manually operated probes, the
largest error in yaw angle measurements was
produced by the modified S probe: 6.6°.
Similarly, the largest error with the Auto-
probe was produced using the modified S
probe head: 5.4°.
RECOMMENDATIONS
In revising Method 2, EPA should consider allowing yaw nulling of 2D probes for determining axial
velocities. The method should stipulate that a probe to be used in the field with yaw nulling must be
calibrated in a wind tunnel. Such a calibration procedure should involve rotating the probe to find
a yaw null (AP = 0) and recording angle relative to true yaw = 0°. The absolute difference between
the rotational position reached after yaw nulling and true 0° yaw would have to be less than or equal
to some standard (e.g., 5°). This standard would have to be met over the range of allowable pitch
angles in 5° increments. If the absolute difference exceeded the standard (e.g., 5°) at any pitch
setting, the yaw nulling procedure would not be usable with the tested probe.
EPA should consider re-testing the yaw nulling capabilities of the Autoprobe Type S and Autoprobe
Modified S with the curve-fitting algorithm currently employed by the manufacturer.
1 As noted in Section 6, the Autoprobe in this study employed a "sweep" procedure to determine the yaw null rather than
the curve-fitting algorithm subsequently implemented by the manufacturer.
9-5
-------
Wind Tunnel Experimental Results
CALIBRATION CURVES AND ACCURACY OF VELOCITY
DETERMINATIONS FOR 3D PROBES— SECTION 7
CONCLUSIONS
Over the pitch angle ranges ±30° and ±20°,
the velocity accuracy of all tested 3D probes
was within 5.5% of the true axial velocity.
In the ±10° pitch angle range, the velocity
accuracy of all tested 3D probes was within
3.8% of the true axial velocity.
No significant gain in accuracy is attained by
developing calibration curves at three veloci-
ties rather than at the two specified in draft
Method 2F.
In order to minimize error in the flue gas
velocity measurement, calibration curves for
3D probes should be developed at velocity
settings close to, or possibly bracketing, the
prevailing velocity where the probe will be
used.
The largest error in measuring pitch angle for
all four probes was 4.3°, which occurred at
90 fps and -20° pitch and produced a veloc-
ity error of 0.3%.
RECOMMENDATIONS
To investigate the third conclusion (above) further, EPA should consider deriving two alternative sets
of calibration curves for each DAT and spherical probe used in the collaborative field tests. The first
set of calibration curves would be derived for the velocity ranges specified in draft Test Method 2F;
the second set would be derived at wind tunnel velocities that bracket the velocity(ies) found in the
field. (EPA may want to define bracketing ranges to be within ± 20 fps of the prevailing velocity.)
The resulting velocities and volumetric flow rates using each set of curves would then be compared.
A previous study2 noted that scratches on the head of a spherical probe can significantly alter the
pitch angle calibration characteristics of that probe. EPA should consider designing and conducting
wind-tunnel tests to assess the impact of scratches on spherical probe heads on the accuracy of pitch
angle and velocity measurements.
2 Kbnge.MD. 1987. Detailed Near-Wake Flowfield Surveys and Comparison to an Euler Method of an Aspect Ratio
4 Rectangular Wing. M.S. Thesis, Mechanical and Aerospace Engineering Department, North Carolina State
University, Raleigh..
9-6
-------
Wind Tunnel Experimental Results
ENTRY WALL EFFECTS ON YAW NULLING—SECTION 8
CONCLUSIONS
In a laminar-flow wind tunnel, three yaw • Multiple yaw nulls are found with the
nulls are observed at wall entry distances of Autoprobe Type S at all distances from the
up to about 9" using a Type S probe. The wind tunnel center line. A "pseudo-entry
center yaw null is found within 1 ° of the true wall effect" appears to be associated with the
yaw zero. Autoprobe.
At entry wall distances greater than about 9", • At distances near the entry wall, unique yaw
unique yaw nulls are found with the manual nulls are found with all probes that use
Type S probe. Fechheimer-type ports to determine the yaw
angle: the modified Kiel, spherical, and DAT.
RECOMMENDATIONS
In should be noted that probes that detect multiple yaw nulls near the entry wall in a controlled
environment such as a wind tunnel may not exhibit the same behavior under field conditions, where
the increased variability of the gas stream may prevent the occurrence of this behavior. Field test data
should be observed to see whether this is the case. Therefore, we recommend the following:
• In the collaborative field tests, EPA should observe whether multiple yaw nulls occur near the
entry wall when using 2D probes.
• In the collaborative field tests, EPA should observe whether the Autoprobe Type S and
Autoprobe modified S produce multiple yaw nulls away from the entry wall due to the
postulated "pseudo-entry wall effect."
9-7
------- |