OCR error (C:\Conversion\JobRoot\00000AX0\tiff\2000XELB.tif): Unspecified error
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U S Environmental
Protection Agency, have been grouped into nine series These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields
The nine series are
1 Environmental Health Effects Research
2 Environmental Protection Technology
3 Ecological Research
4 Environmental Monitoring
5 Socioeconomic Environmental Studies
6 Scientific and Technical Assessment Reports (STAR)
7 Interagency Energy-Environment Research and Development
8 "Special" Reports
9 Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL MONITORING series
This series describes research conducted to develop new or improved methods
and instrumentation for the identification and quantification of environmental
pollutants at the lowest conceivably significant concentrations It also includes
studies to determine the ambient concentrations of pollutants in the environment
and/or the variance of pollutants as a function of time or meteorological factors.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161
-------
EPA-600/4-79-051
September 1979
THE EPA METEOROLOGICAL WIND TUNNEL
Its Design, Construction and Operating Characteristics
William H. Snyder
Meteorology and Assessment Division
Environmental Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences Research
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion. Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
AFFILIATION
William H. Snyder is a physical scientist in the Meteorology and
Assessment Division, Environmental Sciences Research Laboratory, U.S.
Environmental Protection Agency, Research Triangle Park, North Carolina.
He is on assignment from the National Oceanic and Atmospheric Administration,
U.S. Department of Commerce.
-------
PREFACE
The EPA Meteorological Wind Tunnel (MWT) is described in detail in
this report. The basic aerodynamic design and specifications were provided
by EPA personnel. The tunnel is quite similar in overall size, shape and
performance characteristics to the Environmental Wind Tunnel (EWT) at
Colorado State University (CSU). Plans for the EWT were purchased from
CSU and many ideas were incorporated into the EPA MWT. It was constructed
by Aerolab Supply Company of Laurel, Maryland. Professor A. W. Sherwood,
President, Aerolab Supply, provided numerous innovations that improved
its performance and simplified its construction. The tunnel was commissioned
in June, 1974.
The impetus for this report came primarily from a statement by Bradshaw
and Pankhurst (1964) that tunnel designers should "calibrate their tunnels
properly and publish the results, good or bad, so that wind tunnel design
may come to involve more science, less art, and no magic at all." It
will also, of course, serve numerous other purposes.
111
-------
ABSTRACT
The design philosophy, construction details, and operating characteris-
tics of the EPA Meteorological Wind Tunnel are described. Measurements in
the empty tunnel show that the mean velocity is uniform to within ± 2% at
any given cross section, at speeds as low as 1.5 m/s. The turbulence
intensity in the empty tunnel is typically 0.5%. A 2-meter-deep boundary
layer was obtained using elliptic wedge vortex generators and roughness
on the floor. Measurements are presented showing that this boundary layer
simulates, in both turbulence structure and dispersive characteristics,
a neutral atmospheric boundary layer over rural terrain.
iv
-------
CONTENTS
PREFACE iii
ABSTRACT iv
FIGURES vi
ACKNOWLEDGEMENTS viii
1. INTRODUCTION 1
2. CONCLUSIONS 5
3. DESIGN AND CONSTRUCTION DETAILS 6
3.1 Design Philosophy 6
3.2 Gross Characteristics 8
3.3 Entrance Section 8
3.4 Test Section 9
3.5 Acoustic Silencers 12
3.6 Power Section 12
3.7 Tail Section Assembly 13
3.8 Instrument Carriage 14
3.9 Operator's Console 15
4. THE LABORATORY AND INSTRUMENTATION 16
4.1 The Laboratory 16
4.2 The Minicomputer 16
4.3 Flow Measurement Apparatus 17
4.4 Concentration Measurement Apparatus 18
5. PERFORMANCE CHARACTERISTICS OF THE WIND TUNNEL 19
5.1 Calibration of Empty Tunnel 19
5.2 Development of Two-Meter-Thick Boundary Layer 21
REFERENCES 27
-------
FIGURES
Number Page
1 Overview of the EPA Meteorological Wind Tunnel 28
2 Overall plan and elevation of the EPA Meteorological
Wind Tunnel 29
3 Front, side, and top views of entrance section 30
4 Perspective and side views of entrance section 31
5 Test section details 32
6 Inside view of test section looking downstream 33
7 Turntable and details 34
8 Turntable mounted on dolly 35
9 Turntable dolly 36
10 Details of adjustable ceiling support 37
11 Overview of adjustable ceiling support 38
12 Gear motor, limit switches, and torque tube 39
13 Inlet silencer 40
14 Exhaust silencer 41
15 View of fan and motor from inside sound enclosure .... 42
16 Tail section assembly 43
17 Instrument carriage 44
18 Views of instrument carriage 45
19 Close-up view of cable drum 46
20 Operator's console 47
21 Floor plan of Fluid Modeling Facility 48
22 Schematic of minicomputer system 49
23 Typical calibration curve for hot-film probe 50
24 Effect of trip ring on velocity and velocity
fluctuations 51
25 Calibration of empty wind tunnel: air speed as a
function of fan speed 52
vi
-------
Number Page
26 Lateral profiles of mean velocity and turbulence
intensity; U = 1.5 m/s 53
27 Vertical profiles of mean velocity and turbulence
intensity; U = 1.5 m/s 54
28 Vertical profiles of mean velocity and turbulence
intensity; U = 3 m/s 55
29 Vertical profiles of mean velocity and turbulence
intensity; U = 6 m/s 56
30 Vortex generator system 57
31 Development of mean velocity profile along centerline
of wind tunnel; U = 3 m/s (except as noted) 58
32 Mean velocity profiles in log-law form 59
33 Development of longitudinal turbulence intensity
along centerline of wind tunnel; U = 3 m/s
(except as noted) 60
34 Development of vertical turbulence intensity along
centerline of wind tunnel; U = 3 m/s
(except as noted) 61
35 Development of Reynolds stress along centerline of
wind tunnel; U = 3 m/s (except as noted) 62
36 Lateral uniformity of mean velocity; U = 3 m/s 63
37 Lateral uniformity of longitudinal turbulence
Intensity; U = 3 m/s 64
38 Lateral uniformity of vertical turbulence
intensity; U = 3 m/s 65
39 Lateral uniformity of Reynolds stress; U = 3 m/s .... 66
40 Lateral concentration profiles taken through
plume centerline; U = 3 m/s 67
41 Vertical concentration profiles taken through
plume centerline; U = 3 m/s 68
42 Estimated wind-tunnel dispersion parameters compared
with standard Pasquill-Gifford values
(Turner, 1970) 69
vi 1
-------
ACKNOWLEDGEMENTS
The number of people who helped in establishing the Fluid Modeling
Facility is too large to mention individually. Special mention and thanks,
however, must be extended to Mr. Robert A. McCormick, who cleared the path
for the EPA Fluid Modeling Facility; Mr. Roger Thompson, who aided in the
aerodynamic design of the wind tunnel; Professor J.E. Cermak, who provided
several ideas and helpful suggestions for improving the tunnel; Mrs. Cooper
Atamanchuk, for her flexible and common-sense contracting abilities;
Mr. A. Wiley Sherwood, who with his engineering genius and pragmatic views,
constructed a functional and economical wind tunnel with high quality flow
characteristics; and Mr. Robert E. Lawson, Jr. for his enduring and pains-
taking efforts to keep the laboratory running smoothly.
viii
-------
1. INTRODUCTION
Under the Clean Air Act of 1967 (amended in 1970 and 1977), the U.S.
Environmental Protection Agency (EPA) is charged with the responsibility of
overseeing and enforcing air pollution control measures. Essential ingre-
dients of this responsibility are the identification of air pollutants and
the determination of their concentrations in the atmospheric boundary layer.
One of the primary responsibilities of the Meteorology and Assessment
Division (MD) is to develop and evaluate models to predict the transport and
diffusion of pollutants from source to receptor. Other responsibilities of
the MD include the evaluation of Environmental Impact Statements and the
provision of support to other federal agencies, Regional Offices of the EPA,
and state and local air pollution control agencies. To date, this support
has been provided primarily through the development and application of
mathematical models. To develop the models, EPA has relied almost exclusive-
ly on field programs.
In spite of the tremendous advances in computer sizes and numerical
techniques, mathematical models still require gross simplifications, both
because the fundamental fluid-dynamical processes involved in the dispersion
of materials are not well-understood and because computer memories are still
far too small to keep track of the detailed eddy motions in the atmospheric
boundary layer.
Field programs to obtain experimental data are extremely expensive and
time consuming. Data are required for various combinations of meteorological
conditions, various kinds of terrain, different source heights, various types
of sources, etc. Present mathematical models are suitable only where ade-
quate diffusion data exist or can be inferred, that is only for continuous
point sources fairly close to the ground over relatively flat terrain. They
are not suitable for calculating concentrations of contaminants in building
wakes, in hill-valley complexes, from high sources, within the domain of the
urban heat island, etc.
1
-------
In the past two decades, fluid (laboratory) models have shown great
potential for solving many of the complex fluid-dynamical problems that are
presently intractable through analytical/numerical techniques. Fluid models
appear to work best where mathematical models fail (i.e., where obstructions
such as buildings and hills block the flow). They show great promise for
simulating the urban heat island circulation, mountain-valley winds and even
transient phenomena such as the inversion break-up. In a few weeks in the
laboratory where atmospheric conditions may be controlled at will, answers
may be had that would require years of expensive field measurements.
A study was conducted to determine the feasibility of fluid modeling
and the usefulness of such a facility to the EPA (Snyder, 1972a). The
final report on that study included a critical review of the similarity
criteria necessary to model atmospheric motions (Snyder, 1972b), an extensive
review of the pertinent literature, descriptions of the facilities of
the major modeling establishments throughout the world (Snyder, 1974), and
recommendations concerning actions to be taken by EPA along the lines of
fluid modeling of air pollution meteorology. The conclusion of that study
was a recommendation that a complete fluid modeling facility be established
that would include: (1) an open-return, nonstratified wind tunnel (designed
such that stratification capabilities could be added at a later date), (2) a
water channel/towing tank with provisions for salt-water stratification, and
(3) a complete laboratory, including shops and staff.
The objectives of the fluid modeling program are, in order of priority:
(1) To establish the areas of applicability of fluid modeling
and to delineate the similarity criteria, i.e., to set the
standards for fluid modeling of atmospheric diffusion,
(2) To conduct basic, systematic studies of the flow and
diffusion of pollutants in the atmospheric boundary
layer, in particular, in complex flow situations
(around hills, buildings and highways), to aid in the
construction and development of mathematical models,
and to aid in the understanding of such flows, and
(3) To conduct applied model studies, thereby providing
direct support to other federal agencies and to EPA
Regional offices.
The first objective, setting the standards, is given top priority.
Neither proper rules nor areas of applicability of fluid modeling are well-
established. Various laboratories apply different, often conflicting,
-------
similarity criteria tc essentially the same modeling problems. Frequently,
reports of these studies are included as portions of Environmental Impact
Statements prepared by industry. Setting the standards is essential if EPA is
to: (1) make rational decisions on the Environmental Impact Statements, and
(2) conduct modeling studies of its own. EPA's role should be to determine
what types of problems can be solved through fluid models, what similarity
criteria need to be applied in particular cases, how much detail is required
in the models, etc. An essential part of this objective is the critical com-
parison of model results with full scale field data.
Basic, systematic studies are given higher priority than applied (speci-
fic) studies. Applied problems are useful, but in general, do not enhance
the understanding of the basic fluid mechanics. They do not enable one to
predict the effects of slight modifications to the model or to the flow.
Historically, this has been a major objection to wind tunnel studies. As an
example, many, many wind tunnel studies have been concerned with the problem
of downwash of effluents in the vicinity of industrial plants. The sponsors
have been, in almost all cases, the industries that are building or modifying
their own plants, and they have not been interested in broadening the scope
of the study so that the results could be generalized and made universally
applicable. Thus, in spite of up to 40 years of wind tunnel modeling of in-
dustrial plant downwash, we still cannot predict with any certainty what the
ground level concentrations will be for a new plant without running yet ano-
ther wind tunnel study. EPA's role here should be to conduct basic, general
and systematic studies so that, for the above example, a "building codes"
handbook could be compiled.
The first step in the establishment of the Fluid Modeling Facility was
the acquisition of the Meteorological Wind Tunnel, which is described in
detail in this report. Brief descriptions of the laboratory and equipment
are also given. The second major step is the acquisition of the Water
Channel/Towing Tank, which will be described in a separate report. The third
major step in the over-all plan is the closing of the return of the wind
tunnel and the addition of floor heating and cooling to provide for temper-
ature stratification of the flow. Because of the very large expense and the
"down-time" involved in this third step, its execution has been postponed
indefinitely.
-------
In Section 3, the design and construction details of the wind tunnel will
be discussed. Brief descriptions of the laboratory and its equipment are
given in Section 4. Calibration data on the performance of the empty wind
tunnel as well as measurements of the structure and dispersive characteristics
of a 2-meter thick simulated atmospheric boundary layer are presented in
Section 5.
4
-------
2. CONCLUSIONS
A meteorological wind tunnel has been constructed for the purpose
of studying the pollutant transport and dispersion characteristics of complex
flows such as those around hills and buildings. It is a very low speed,
open return tunnel with a long test section (2.1 x 3.7 x 18.3 m) that will
accomodate model sizes from individual buildings at a scale ratio of 1:300
to large terrain at 1:5000. Design and construction features of the tunnel
as well as the laboratory and instrumentation used for measurements in
the tunnels are described in detail.
The empty wind tunnel was calibrated and its performance characteristics
are described. A "surging" problem was discovered and corrected. Mean
velocities were found to be uniform to within ± 2% at any cross section,
and the turbulence intensity was found to be typically 0.5%. Both these
characteristics apply even at speeds as low as 1.5m/s and may be regarded
as excellent performance characteristics.
A 2-meter thick boundary layer was developed in the tunnel using a
vortex generator system to simulate the atmospheric boundary layer. The
properties of this boundary layer are reported in detail. It is a slowly
developing boundary layer that simulates the neutral atmospheric boundary
layer over rural terrain. It is reasonably two-dimensional and its dispersive
characteristics match "accepted" atmospheric data. This boundary layer
would be suitable for comparative types of studies dealing with flow and
diffusion patterns in the near-field of individual buildings, but further
testing would be required to determine its suitability for longer range
studies.
-------
3. DESIGN AND CONSTRUCTION DETAILS
3.1 Design Philosophy
A meteorological wind tunnel differs in two basic respects from a con-
ventional aeronautical wind tunnel. First, because the atmospheric boundary
layer usually extends far above the buildings immersed in it, the simulated
boundary layer in the meteorological wind tunnel must be quite deep in order
for the model buildings to be of reasonable size. In an aeronautical wind
tunnel, the reverse is required - great pains are taken to minimize the
thickness of the boundary layers, to make the flow uniform across the test
section, and to obtain as low a turbulence intensity as is feasible. A very
long test section is generally required in a meteorological wind tunnel in •
order to generate a thick boundary layer. Second, high wind speeds are
generated in aeronautical wind tunnels to compensate for the reduced size of
the models. However, because buoyancy effects are much more important in
atmospheric flows, wind speeds in meteorological wind tunnels are generally
reduced in order to accommodate these buoyancy effects. In aeronautical
terminology, then, the EPA tunnel would be classified as an "ultra-low speed"
tunnel. Because low turbulence is not a primary goal in the meteorological
wind tunnel, a large contraction ratio is not essential. Also, because of the
low speed, power consumption is minimal and the capital cost of closing the
return is not justified by the savings in operating costs. Thus, the EPA
Meteorological Wind Tunnel (MWT) was designed to be an ultra-low speed, open-
return type with a relatively small contraction ratio (3:1) and an unusually
long test section (18 m).
The overall size of the MWT was determined by the size of the test
section, which, in turn, was determined by the types of models to be used.
(Many factors, of course, enter into the decision on the size of the tunnel,
but only the primary ones are discussed here.) The spectrum of model types
to be studied in the tunnel ranged from individual buildings at a scale ratio
of, say, 1:300, to topographical models at scale ratios of up to 1:5000. If
-------
the atmospheric boundary layer is taken to be 600 m in depth, then the test
section height must be somewhat larger than 2 m in order to accommodate a
simulated atmospheric boundary layer for building studies. A 2.1 m test
section height was chosen for convenience, both to reduce blockage effects of
large models and to accomodate model scale ratios as large as 1:300. At a
scale ratio of 1:500, the model height of a modern power plant would be about
30 cm, and the model stack height would be about 60 cm. Allowing for
modeling some of the surrounding buildings and terrain irregularities, and
also for side wall boundary layers, the width of the test section was chosen
to be 3.7 m. Indications were that confidence could not be placed in model
predictions for downwind distances much larger than about 5 kilometers
(Snyder, 1972b). The required length of test section was thus 18.3 meters:
approximately 10 m for the downwind distance, 3 m for the model and upwind
terrain, and 5 m for boundary layer generators and development of the
appropriate boundary layer structure.
The design maximum speed was set at 8 m/s. Higher speeds are seldom
required; this is a matter of considerable economic importance, since power
consumption varies with the cube of the speed. The tunnel was designed
to be controllable over as wide a speed range as possible to allow for
possible future modifications. To permit modeling of buoyant plumes, the
tunnel was to be capable of operating at very low speeds as well (say, 30
cm/s), The size and speed of the tunnel are comparable with other recently
constructed wind tunnels for the study of atmospheric diffusion and meet the
design criteria put forth by Sundaram and Ludwig (1970).
Convenience of operation was a major factor influencing the design of
the tunnel. To this end, the test section was divided into five identical
subsections with interchangeable windows and floor sections. Transparent
windows are interchangeable with opaque (flat black) panels, thereby providing
the option of good visibility of the interior of the tunnel or an opaque
background for photography. One of the floor sections contains a removable
turntable, and because of the interchangeability of the floor sections,
the turntable may easily be placed in any of the five subsections. The
turntable rests on a four-wheeled dolly with an elevating mechanism; it
can easily be rotated to change the wind direction over the model or it
can be lowered from the tunnel and rolled to the shop; a different model
-------
mounted on a second turntable may be installed in the tunnel in a matter
of minutes. An instrument carriage is included to provide for three-
dimensional positioning of measuring probes with remote control and readout
of position. All controls and readout devices are centrally located on
an operator's console.
3.2 Gross Characteristics
Figure 1 shows an overview of the EPA Meteorological Wind Tunnel and
Figure 2 provides a sketch showing the major dimensions and labeling the
various sections. The overall length of the tunnel is 38.2 m and its maximum
height and width are 4.62 m and 6.43 m, respectively. Room air enters
the tunnel at the left through a honeycomb and a series of four screens
that remove the disturbances from the room air and produce a uniform velocity
profile. The flow then passes through a contraction that further reduces
the turbulence and also aids in producing a uniform velocity profile.
The flow then passes through the test section, an acoustic silencer, a
rectangular to round transition section, the fan, a diffuser, and finally
another acoustic silencer from which it exhausts back into the room. The
fan section is totally enclosed in a sound attenuating enclosure. The
ceiling height in the test section is adjustable to obtain a zero pressure
gradient and to compensate for blockage effects of the model.
3.3 Entrance Section
Details of the entrance section are shown in Figures 3 and 4. The
bellmouth is formed using a 15 cm diameter pipe, tangent to the inside sur-
face of the tunnel, around the entire perimeter of the entrance. The en-
trance dimensions are 4.0 m x 5.5 m. The floor of the entrance section is
48 cm from the building floor.
The flow straightener consists of a Verticel plastic impregnated
paper honeycomb with triangular shaped cells, cell "diameter" of 1.27 cm
and cell length of 15 cm. The cell length-to-diameter ratio of 12, sub-
stantially larger than the ratio of 6 to 8 recommended by Bradshaw and
Pankhurst (1964), was felt to be necessary at these especially low entrance
velocities for removing swirl and lateral mean velocity variations caused by
disturbances to the room air from draughts and obstructions.
Four stainless steel anti-turbulence screens of mesh size 7.9/cm
(20/in) and wire diameter 0.025 cm (0.010 inch) were specially woven by
8
-------
Tyler Manufacturing, Cleveland, OH, to cover the entire inlet area without
seams. This type of screen provides an open area ratio (B) of 64% and a
local Reynolds number (U^d/gv) of 80. These paramenters yield a pressure
drop coefficient of 0.71 (Davis, 1957), which theoretically provides a
reduction factor for mean velocity variations of 0.48 and a turbulence
intensity reduction factor of 0.76, for each screen (Pankhurst and Holder,
1952). For four screens the respective factors are 0.05 and 0.34. The
screens are spaced 20 cm (800 wire diameters) apart. This spacing is suffi-
cient for the turbulence generated in the wire wakes of one screen to decay
before the next screen is reached. It is also sufficient for insertion of a
long-handled vacuum wand for cleaning.
The screens are fastened to vertical aluminum pipes outside the two
sides; the pipes may be rotated to tension the screens. Downstream of the
last screen a settling section about 1 m long allows the turbulence in the
wakes of the screen wires to decay.
With an entrance area of 4.0 m x 5.5 m and a test section area of 2.1 m
x 3.7 m, the contraction ratio is 2.8:1. The length of the contraction is
3.3 m. The profile shape approximates two cubic arcs joined together with
equal slopes to form a smooth transition at the inflection point, which is
located approximately 1.2 m downstream from the entrance to the contraction.
The actual shape represents a compromise based on several theoretical papers
and previous designs and experiences of Aerolab Supply Co., Laurel, MD
(A. W. Sherwood, private communication). Table 1 lists the dimensions of the
contraction.
3.4 Test Section
Details of the test section are shown in Figure 5. A view of the inte-
rior is provided in Figure 6. The inside height, width and length are 2.1,
3.7, and 18.3 m, respectively. The sections are fabricated with 6.4 x 6.4 cm
(2 1/2 x 2 1/2 in) angle iron frames that retain 1.9 cm (3/4 in) plywood and
1.27 cm (1/2 in) acrylic plastic (plexiglas™) panels. One of the floor
panels is fitted with a 3.2 m diameter turntable that can be manually rotated
to any desired yaw angle with the air stream. Since all the floor panels are
of the same outside dimensions, the panel containing the turntable can be
placed in any of the five subsections. This involves the exchange of a solid
section for the turntable section through the use of a specially constructed
9
-------
TABLE 1. CONTRACTION DIMENSIONS
Distance from
Entrance (cm)
0
30
61
91
122
152
183
213
244
274
305
335
Inside Width
(cm)
549
540
529
505
477
449
424
403
387
374
368
366
Inside Height
(cm)
396
392
380
355
324
298
275
255
236
224
215
213
lifter. It consists of casters fitted on the ends of lengths of square
tubing which are bolted to the floor panels at the two sides through threaded
metal Inserts In the plywood floor. The floor panel to be moved is attached
to the lifter and rolled along tracks (floor supports at the sides of the
test section) to Its new position. All floor panels are sealed at the edges
by strips of foam rubber to minimize air leakage into the tunnel. To facili-
tate ease in movement of the floor panels, they were made as light and thin
as possible. To prevent sagging, adjustable jacks support the centers of the
floor panels.
Two circular turntables (Figures 6 and 7) are included so that one can
be fitted with models while the other is under test in the tunnel. A four-
wheeled dolly with an elevating mechanism (Figures 8 and 9) transports the
turntable from the shop to the tunnel, lifts the turntable to its position in
the tunnel, and allows rotation of the model in the tunnel. Two dollies are
provided: One is a hand-crank version; the other is motorized. The joint
between the turntable and tunnel floor is sealed with a gasket and toggle
clamps as shown in Figure 7.
10
-------
The sidewalls are 2.44 m in height, the entire lower halves of which are
fitted with acrylic plastic windows in three of the five subsections. The
other two subsections contain opaque (plywood) windows that are interchange-
able with the transparent ones for photographic purposes. Since the windows
are heavy, a trolley system running on rails at the top of each side is pro-
vided to facilitate handling. A fluorescent light fixture is attached to the
top of each transparent window to provide light inside the tunnel. Conven-
ience outlets and switches are provided at each section for the lights and
miscellaneous equipment. The windows are sealed at the seams with foam rubber
that is compressed by the action of the toggle clamps retaining the windows.
Covered, 2.5 cm portholes provide access for electrical cables inside the
tunnel.
The ceiling (see Figure 5) is fabricated of 0.95 cm (3/8 in) marine
plywood to provide greater flexibility as compared to the 1.9 cm (3/4 in)
floor and sidewall panels. The individual plywood sheets are edge-glued and
laterally stiffened at 61 cm intervals with cross battens of 2.54 cm (1 inch)
square steel tubing. The ceiling is supported at 112 points by whiffle trees
attached to roller chain (see Figures 10 and 11). The chain is carried by
sprockets rotated by 14 torque tubes oriented transversely across the top of
the tunnel at 1.2 m intervals. Separate gearmotor drives (Dayton Model 6K303,
30 RPM, 1/20 KW, 115 V, Ratio 57.5:1) with speed reducers (Dayton Model
2Z421AWK, Ratio 60:1) on each of the torque tubes enable an operator to put
a smooth contour in the ceiling. Limit switches prevent the movement of the
ceiling more than 30 cm in either direction. Switches at the operator's
console control the gearmotors. A pointer attached to a bracket near each
torque tube provides a measurement of the ceiling height to a person at the
operator's console. Electromagnetic disc brakes (Dayton Model 3M342, 115V,
60 Hz, 0.1 kg-m torque) on the gear motors prevent the torque tubes from
rotating when the power is turned off. Figure 12 presents a close-up view of
the ceiling lift mechanism. A hinged (sliding) flap over the inlet silencer
at the downstream end of the test section connects the flexible ceiling to
the transition section. Windows (57 x 122 cm) spaced every 1.2 m in the
ceiling provide for overhead lighting and photographing of models from above.
Clothesline retractors take up the slack on the roller chains. A catwalk
to facilitate personnel access to the top of the tunnel has been added since
11
-------
the photograph in Figure 11 was taken.
The edges of the ceiling are sealed within the sidewalls by strips of
linear polyethylene that are guided by aluminum strips at the edges of the
ceiling (Figure 10). An inflatable gum rubber tubing (amber surgical) forces
the plastic sealing strips against the tunnel sidewalls when air pressure is
applied within the tubing. The tubing is deflated to adjust the ceiling
height and inflated to 70 to 100 kPa (10 to 15 psi) to make the seal after
the appropriate height is obtained.
3.5 Acoustic Silencers
The mechanical and aerodynamic noise emanating from the fan is reduced
by acoustic silencers. The inlet and outlet sections of the fan are fitted
with interior partitions forming a large honeycomb of plywood. All surfaces
of the cells thus formed are covered with a 2.5 cm thick sound-absorbent
material (Lina-Coustic fiberglass duct liner) that is cemented to the plywood
and protected at the leading and trailing edges with formed and perforated
sheet steel. The inlet silencer is 2.1 m x 2.7 m in cross section, 2.4 m in
length, and is divided into 12 cells. A photograph of the inlet silencer is
shown in Figure 13. The outlet silencer, shown in Figure 14, is of similar
construction. It is 3 m square, 2.4 m in length, and consists of 12 cells.
The exterior of the fan section, from inlet silencer to outlet silencer, is
completely housed by a double-walled enclosure of gypsum sound deadening
board. The interior and exterior wall coverings are independently supported
by separate frameworks. A fiberglass blanket is placed in the space separat-
ing the walls. The enclosure contains a door for access to the power section
and two windows to provide cooling air to the motor and eddy-current coupler.
The noise level in the room when the tunnel is run at maximum speed is
approximately 60dbA.
3.6 Power Section
The fan is a 1.8 m (6 ft) diameter axial flow fan with 5 adjustable
pitch blades manufactured by Buffalo Forge Company (Size 72D5, design II,
type S, Adjustax Vaneaxial, Arr. 9). It is driven through a V-belt drive by
an Eaton Dynamatic Eddy Current Coupling (Model No. ACM 9143, 75 KW, Max.
Speed 1685 RPM) to provide variable rotational speed. The coupling is driven
by a constant speed AC motor (Westinghouse, 75 KW, 1770 RPM, 3 ph, 60 Hz,
230/460 V) with a reduced voltage starter. This power system is very easy to
12
-------
operate, provides a quite steady air flow rate and a wide range of speed, and
produces a minimum of noise. To start the tunnel, the operator need only
push the motor start button, then the coupling start button, then turn the
speed adjustment knob to obtain the desired tunnel speed (the one-turn poten-
tiometer supplied by the manufacturer was replaced by a ten-turn meter for
finer resolution of speed setting). The fan rotational speed is continuously
monitored using a photomechanical light chopper (40 hole disc mounted on the
fan shaft and an electronic counter). The maximum fan speed is 900 RPM
(~ 8 m/s). It is quite steady as low as 20 RPM (~ 0.15 m/s). This type of
system was chosen over a variable pitch fan with a two-speed motor for two
reasons: (1) a variable pitch fan generally functions efficiently for the
high tunnel speeds, but at low air speeds the pitch must be reduced to the
point where the blade sections near the hub produce an axial velocity distri-
bution affecting the flow in the tunnel. The use of a two-speed motor on a
variable pitch fan tends to alleviate this condition, but it remains a
serious defect. (2) also, because the noise generated by a variable pitch
fan is primarily a function of the blade tip speed, the noise level is
essentially the same (high) at all tunnel speeds. Figure 15 shows a view of
the fan and drive motor from inside the sound enclosure.
3.7 Tail Section Assembly
Details of the tail section assembly are shown in Figure 16. A 2.4 m
(8 ft) long transition section of 0.48 cm (3/16 in) thick steel was developed
by triangulation to connect the rectangular acoustic silencer to the round
fan. A safety screen was placed at the entrance to the transition section to
prevent debris from entering the fan. A flexible rubber joint connects the
transition section to the fan to prevent vibrations of the fan from being
transmitted to the remainder of the tunnel. A 3.7 m long round diffuser with
inlet diameter of 1.83 m and a divergence angle of 13.5° (total angle) saves
energy by converting some kinetic energy into pressure energy. This rather
large divergence angle is considerably larger than that (5°) recommended by
Pankhurst and Holder (1952), but, because of the low energy consumption, an
efficiency approaching 80% was felt to be adequate when compared with the
resulting savings in capital costs and floor space from a shorter diffuser.
However, as will be explained later, this large divergence angle created
problems of surging. The problem was eventually alleviated by installing a
13
-------
trip ring as suggested by Pope and Harper (1966). The trip ring was a 2.54 cm
wide piece of thin metal installed circumferentially around the entrance to
the diffuser. The resulting loss in top speed of the tunnel was insignificant
(less than 5%).
The fan, diffuser, motor and coupler are bolted rigidly to a framework
that is supported by coil springs to permit some movement (see Figure 16).
The downstream end of the diffuser is connected to another transition section
through a second flexible rubber joint. The second transition is 1.83 m long
and has an octagonal entrance and a square exit. The flow exiting the dif-
fuser thus sees an abrupt shape change from round to octagonal.
3.8 Instrument Carriage
Figures 17 and 18 show the details of the instrument carriage which
provides the capability of positioning a probe anywhere within a 17 m x 3 m
x 1.5 m parallelepiped to an accuracy of ± 1 mm. The positioning, control,
and readout are effected remotely from the operator's console. The slewing
rate in all three coordinate directions is adjustable to provide fast
positioning for long traverses and slow rates for accurate settings in the
vicinity of the set point.
The main support frame is a 5.1 x 15.2 cm (2 x 6 in) tubular steel
member. Longitudinal positioning (x-direction) is provided by a gearmotor
(Dayton Model 3M234, 1/20 KW, 2.8 RPM) driving a sprocket-chain arrangement,
which in turn drives wheels that roll on tracks on each side of the test sec-
tion. The tracks are 5 x 10 cm (2 x 4 in) tubular steel running the length
of the test section 1.4 m above the floor. They are slotted to house power
and signal cables inside. A pneumatic cylinder (Schraeder Model 3452A) ac-
tuates a two-speed transmission that provides high and low speed ranges.
Maximum speeds are 5000 and 700 mm/min on the high and low ranges respective-
ly.
A block driven by a roller chain on a vertical bar provides for vertical
positioning (z-direction) of the probe. The bar is fastened to a yoke that
slides transversely on the mainframe. Vertical drive is effected through a
gearmotor (Dayton Model 3M234, 1/20 KW, 2.8 RPM) that drives a square shaft
(parallel to the mainframe), that, in turn, drives a mating sprocket fastened
to the vertical bar. The sprocket slides axially on the square shaft. Maxi-
mum speed in the vertical direction is 400 mm/min.
14
-------
Transverse positioning (y-direction) is provided by a gearmotor (Dayton
Model 3M235, 1/20 KW, 6.7 RPM) that slides the yoke on the mainframe via a
looped roller chain. Maximum speed in the transverse direction is 1000
mm/min.
Theta Instruments, Inc. optical encoders (Model Nos. 05-100-31-2 and
05-250-31-2) sense the probe motion through drive mechanisms similar to those
of the gearmotors, except that special care has been taken to minimize back-
lash. The gearmotors are mounted on one side of the mainframe, with power
cables run through the slotted track on the test section sidewall. The opti-
cal encoders are mounted on the opposite side of the mainframe, with signal
cables run through the other slotted track on the test section sidewall, in
order to minimize electrical noise and interference from the gearmotors and
power cables. The cables are wound on large drums at the downstream end of
the test section (See Figure 19). Hanging weights on the drums provide con-
stant tension on the cables to take up slack when the carriage is moved down-
stream. The cables then wind and unwind from smaller drums on the outside of
the tunnel, from which they are routed to the operator's console. Blockage
of the cross-sectional area by the carriage is approximately 3.5%.
On/off reversing switches and silicon-controlled rectifiers at the
operator's console provide power to the gearmotors. Monsanto electronic
reversible counters (Model 106A) provide direct readout in millimeters of the
x, y, and z coordinate positions of the probe. The coordinates can be
"zeroed" at any position to establish an origin.
3.9 Operator's Console
The operator's console is shown in Figure 20. All tunnel operational
controls and readouts are provided at the console. These include control and
readout of fan speed, control and readout of probe position, ceiling height
controls and limit switch indicator lights, and pressure controls for the
ceiling seal. For convenience, approximately 6 m of extra cable makes the
console somewhat portable.
15
-------
4. THE LABORATORY AND INSTRUMENTATION
4.1 The Laboratory
The floor plan of the Fluid Modeling Facility is shown in Figure 21.
22
The laboratory occupies approximately 1860 m (20,000 ft ) of floor space.
The main floor area contains the Meteorological Wind Tunnel, the Water
Channel/Towing Tank (to be described in a separate report) and an Instrument
Calibration Wind Tunnel used for small scale experiments and calibration
of meteorological wind instruments.
A wood and metal working shop contains tools and equipment for the
construction of detailed models from wood, metal, and plastics. Minor
modifications, repairs and additions to the facility are also performed
1n-house. An electronics shop is used for the repair and maintenance
of instruments and other electronic equipment and for the development
of new instrumentation. A photographic darkroom is available for proces-
sing black and white films and making prints.
4.2 The Minicomputer
A Digital Equipment Corporation POP 11/40 minicomputer is located
within the facility to sample, process and store all laboratory data.
A block diagram of this system is shown in Figure 22. Its main components
include a 16-channel, 12 bit analog-to-digital converter, a POP 11/40
central processing unit with 80K words of memory, 3 disk drives, 3 mag-
netic tapes drives, several terminals (including a refresh-graphics ter-
minal), and an electrostatic printer/plotter. Its operating system is
RSX-11D which permits multiple tasks and multiple users to access the
system simultaneously. Real-time analysis of the outputs of electronic
data gathering instruments provides instant feedback to the experimenter
on the results of data being taken. The refresh-graphics terminal allows
the operator to display data graphically as it is acquired and to make
decisions based upon data taken to date; e.g., is the velocity profile
smooth or do additional points need to be taken?
16
-------
The system as configured permits sampling rates up to a rraximum of
1000 samples/s. It is programmed to carry running sums and sums of squares
during sampling (in triple-precision arithmetic) so that, for example,
the mean velocity, turbulence intensity, and Reynolds stress, are available
immediately after the last sample has been acquired. The magnetic tape
drives provide for storage of digitized data for later analysis, such
as the computation of spectra or correlations.
4.3 Flow Measurement Apparatus
Velocity measurements in the wind tunnel are made with TSI constant
temperature anemometers (Models 1054A and B are available). Hot-film
(as opposed to hot-wire) sensors are used exclusively, as the higher
frequency response of hot-wires is not necessary and the hot-films are
much less fragile. Since hot-film anemometers do not directly measure
the air speed, they must be calibrated in a known air stream. This is
done by placing the hot-film sensor next to a pitot tube in the free stream
flow in the wind tunnel and running the computer program HCALX. The air
stream velocity is calculated from the differential pressure indicated
by the pressure measuring device connected to the pitot tube, in general,
an MKS Baratron model 170M electronics unit connected to a model 310BH-10
sensor head. This velocity is input to HCALX through the operator's ter-
minal and the A/D converter samples the output voltage of the anemometer,
typically at 500 samples/s for 15 s. After a sufficient number of calibra-
tion points (typically 6) are established over the velocity range of interest,
HCALX then computes the best fit to King's law E2=A+BUa (E is the output
voltage of the anemometer, V is the mean velocity, and A, B, and a are
constants to be determined). It steps a from 0.30 to 0.60 (in steps of
0,01), each time computing the best-fit A and B and each time retaining
the least-squares error. When finished, it then searches for the best a,
i.e., that which minimizes the least-squares error. A typical calibra-
tion curve is shown in Figure 23. Note that the zero-flow voltage is
recorded, but is not used in the calculation of the best fit curve.
To make turbulence measurements in the wind tunnel, then, the best
fit A, B, and a values from HCALX are input to another program called
HOT, which generates one or more tables of digitized voltage versus velocity.
During the sampling period, HOT samples the anemometer output, uses a
17
-------
table look-up procedure for linearization (much more efficient than calcu-
lations), and carries the running sums of (1) the velocity indicated by
each film, (2) the square of the velocity indicated by each film, and,
if cross films, (3) the cross-product of the velocities indicated by each
film pair. After the sampling is finished, HOT uses these sums to compute
the mean velocity, the angle ot the mean flow, the two turbulence intensity
components, end the Reynolds Stress.
Because the hot-film is sensitive to ambient temperature changes,
the rocrr, temperature is carefully monitored. If it changes by more than
0.5°C, the operator inputs the new temperature into HOT, in which case,
f\
a new look-up table is generated according to E =(A+Ba) (Tf-T )/(Tf-T ,),
T new T Co I
where T- is the film temperature, T is the new room temperature,xand
l"cal is the room temperature at which the sensor was calibrated.
4.4 Concentration Measurement Apparatus
Pollutant dispersion is studied by releasing a hydrocarbon tracer
mixture from the model source, collecting samples through a sampling tube,
and measuring the concentration of tracer in the sample. Tracers that
have been used are methane and ethylene, being chosen for their buoyant
and neutrally buoyant properties, respectively. Four Beckman model 400
Hydrocarbon Analyzers (flame ionization detectors, FID's) are used in
the continuous operating mode to determine the concentration of tracer
in the sample. The "zero" and "span" are adjusted on the FID using "zero"
air (less than 1 ppm total h>drocarbcns) and a 1% mixture of methane in
air (or equivalent ethylene standards), respectively. The zero and span
are checked after each concentration profile is obtained to be certain
that the electronics has not drifted. The output of the FIC is also proces-
sed by the minicomputer through the program HCA. The response time of
this instrument (typically 0.5 s), however, is tco large to obtain useful
statistics on concentration fluctuations.
Because the wind tunnel is open-return, background hydrocarbons in
the room air tend to increase during a testing period. To account for
this shifting, background measurements are taken periodically during an
experiment. The HCA program subtracts the background hydrocarbon concen-
tration from the sample value by assuming a linear change of background
with time between readings.
18
-------
5. PERFORMANCE CHARACTERISTICS OF THE WIND TUNNEL
5.1 Calibration of Empty Tunnel
The basic performance characteristics of the empty wind tunnel, with
the ceiling adjusted parallel to the floor, are presented in this section.
A series of tests was undertaken to examine the steadiness, uniformity
and turbulence intensity of the core flow and the growth of the sidewall
boundary layers. The first tests investigated the steadiness of the flow.
Very low frequency fluctuations in flow speed, i.e., over the order
of a minute or longer, are referred to as drift. Drift can result, for
example, from changes in line voltage, gradual changes in air temperature,
from "hunting" in the fan speed controller, etc., but has never been
a problem with the MWT. Velocity fluctuations faster than a few hertz
and generally associated with rotational motion are called turbulence
and will be discussed later. Surging, as defined here, refers to fluctu-
ations in flow speed in a middle-frequency band, in the range of 2 to
60 cycles per minute. It is generally associated with separation somewhere
in the wind tunnel circuit and is most easily identified as irrotational
motion, i.e., two velocity sensors placed in the test section will show
nearly perfect correlations in fluctuations regardless of their separation
distance or direction with respect to one another.
Initial tests in the MWT showed surging to be a problem, annoying and
bothersome when attempting to obtain time averages of velocity, etc. The
surging amounted roughly to sinusoidal variations in speed, with peak
amplitudes of approximately ± 3% of the mean (1% rms) and frequencies
roughly proportional to velocity. An example of the surging is shown
in Figure 24.
Probing various parts of the circuit with smoke wands and attaching
wool tufts to the sidewalls revealed separation in the diffuser downstream
of the fan. The separation was undoubtedly caused by the large divergence
angle of the diffuser. The problem was easily solved as mentioned in
19
-------
Section 3, by installing a trip ring that forced separation to occur at
the entrance to the diffuser. The trip ring reduced variations in speed
to under 1% peak-to-peak values (under 1/2% rms). It reduced the top
speed of the tunnel by approximately 3% (see Figure 24).
The tunnel was then calibrated, i.e., flow speed was determined as a
function of fan speed. Figure 25 shows the results. Speeds above 1.2 m/s
were determined using a standard Pitot tube and micromanometer. Lower
speeds were obtained by timing smoke puffs. The data in the figure show
the wind speed, for all practical purposes, to be directly proportional
to fan speed. Although the design speed of 8 m/s was not obtained with
the present blade pitch angle, it could easily be exceeded in a matter
of a few minutes by increasing the pitch angle, since the motor is taxed
to only 70% of its capacity with the present adjustment. The flow is
quite steady at all speeds, including those below 1 m/s.
Velocity surveys were made with a TSI Model 1054A anemometer and a
1210-20 hot-film probe. Signals from the anemometry were processed in real
time on the minicomputer, as discussed in Section 4, except that during
this period of time, the operating system was RT-11, a single task, single
user system allowing sampling rates up to 6000 hertz. A suppression was
applied to the bridge voltage fed to the minicomputer to produce a signal
within the range of the A/D converter; the turbulence intensities presented
here thus represent total fluctuations - including drift and surging -
since low frequency fluctuations were not filtered out as is conventionally
done. Generally, one minute samples were used to form time averages and
sampling rates were 2000 samples/s, which is far beyond the Kolmogorov
frequency. Surveys were made at 3 wind speeds and at 3 downstream loca-
tions, representing the entrance, middle and downstream end of the test
section. Turbulence intensities were normalized with the local mean velo-
city.
Lateral (cross-stream horizontal) profiles of mean velocity and
turbulence intensity at a nominal velocity of 1.5 m/s are presented in
Figure 26. Figure 27 presents the vertical profiles of mean velocity
and turbulence intensity. The mean velocity in the core, i.e., outside
the sidewall boundary layers, is uniform to within ± 2% of the mean at
any given section. The core flow accelerates by approximately 7% over
20
-------
the length of the test section because of the growth of the wall boundary
layers. The core flow can be made to be nonaccelerating by adjustment
of the celling height, but it was not attempted in this case. The turbulence
intensity, nominally 0.5%, would not be considered low for a high-perfor-
mance aeronautical wind tunnel, but may be considered excellent for an
ultra-low-speed meteorological wind tunnel.
Figures 28 and 29 show the development of the boundary layers along
the floor of the test section at 3 and 6 m/s respectively. It may be
of interest to note that the boundary layer thicknesses are well-predicted
by
. 0.376 x
= n V5
Rex
where 6 is the boundary layer thickness (U = 0.99 U^), x is the distance
from the entrance of the test section, and Re is the local Reynolds number,
A
U^x/v. The above formula is, of course, the standard textbook formula
for the growth of a turbulent boundary layer over a flat plate.
5.2 Development of Two-Meter-Thick Boundary Layer
The simulation of atmospheric dispersion in a wind tunnel is especially
difficult. In order to model buoyant plumes, it is desirable to use the
largest convenient model scale (Snyder, 1972b), and hence, the thickest
possible boundary layer. It is also desirable to develop this thick boundary
layer in as short a fetch as possible and to maintain this boundary layer
in an equilibrium state over the region of interest downstream of the
source. Finally, it is important that the otherwise undisturbed boundary
layer be two-dimensional.
The aim in this first development of a simulated atmospheric boundary
layer was to obtain something representative of neutrally stable flow over
rural terrain, say, somewhere between open country and woodland forest.
The boundary layer to be simulated was to have the following approximate
properties (Davenport, 1963; Counihan, 1975):
Boundary layer thickness, 6 = 600 m
Profile Shape, U/U = (z/6)1/6
21
-------
Roughness length, ZQ = 30 cm
Scale Factor = 300
The generator system chosen was a slight modification of that developed
by Counihan (1969). The elliptic wedge vortex generators and castellated
barrier were scaled versions of Counihan's: the height, H, of the generators
was 1.83 m; their length in the streamwise direction was H/2; their spacing
was H/2; the distance from the castellated barrier to the front edge of the
generators was 5H/6; the barrier height between castellations was H/7; the
height of the top of the castellations was H/5; the castellations were aligned
with the center lines of the generators; the tunnel height was 7H/6; and the
distance from the test section entrance to the barrier was H/3. Four genera-
tors were spaced across the test section such that the centerline of the wind
tunnel was mid-span between generators. However, the roughness elements dif-
fered from those of Counihan. Two-dimensional roughness elements were used
to maintain the boundary layer in equilibrium. They consisted of wooden
strips H/100 (0.019 m) in height and H/36 (0.051 m) wide and were spaced H/4
(0.475 m) apart along the floor of the test section perpendicular to the flow
direction. Figure 30 provides a sketch of this vortex generator system.
Measurements described here were taken as described in Section 3; the
sampling rate was 800 Hz and the averaging time was 1 min. Unless otherwise
noted, all measurements were made at a freestream velocity of 3 m/s.
The development of the mean velocity profile along the centerline of the
wind tunnel is shown in Figure 31. These measurements were made over a period
of approximately six weeks using both standard straight film probes (TSI Model
1210-20) and x-film probes (TSI Model 1243-20). The agreement between the
data taken with the two types of probes is regarded as excellent. The mean
velocity profile is seen to be unchanging over the range of 4.5H < x < 7.5H,
and to be approximated very closely by a l/6th power law, as desired. The
data at x = 6H at twice the wind speed show the flow to be essentially
Reynolds number independent over the range 3m/s < U^ < 6m/s. Some less exten-
sive measurements reported earlier (Snyder and Lawson, 1976) showed a l/5th
power law at l^ = 1.5m/s, but this difference is not regarded as significant.
Finally, the measurements at x = 8H are slightly different from those at
x = 7.5H. The former are probably influenced by the proximity to the acoustic
22
-------
silencers at the downstream end of the test section, since x = 8H is only a
few centimeters from the silencers.
The mean velocity profiles plotted in the log-law form in Figure 32 were
used to define the aerodynamic characteristics of the surface:
where u* is the friction velocity, d is the displacement height, and k is the
von Karman constant (0.4). The displacement height was found to be negligi-
ble. Within the precision of the measurements, the roughness length is found
to be 0.65 mm (0.00036H), and invariant with respect to position or wind
speed. Note that the depth of the logarithmic portion of the boundary layer
is approximately 0.15H. The roughness length of the prototype boundary layer
would be 20 cm, slightly less than the desired value. According to Counihan
(1975), this roughness length would correspond to moderately rough terrain,
say, agricultural/rural, and would yield a power law index of approximately
l/6th, which fits the data quite well.
The development of the longitudinal and vertical turbulence intensities
and Reynolds stresses are shown in Figures 33, 34, and 35, respectively.
These turbulence quantities have been normalized by the free-stream velocity
U^. Again, these data were obtained over a period of approximately 6 weeks
using both straight-film and x-film probes. The repeatability of the measure-
ments is regarded as excellent. The figures show the streamwise development
to consist of a slight decay above z/H = 0.15. Evidently, the flow close to
the surface is nearly in equilibrium with the surface, i.e., the roughness is
matched reasonably well to the barrier height (Robins, 1975). Notice that
the turbulence intensity and nondimensional stress profiles are only very
slightly affected by changing the wind speed.
Counihan 's (1969) data at x = 4.5H are reproduced on the figures for
comparison. The agreement is reasonably good considering that Counihan's
estimated roughness length was proportionately twice as large. Whereas sur-
face values of turbulence intensities and Reynolds stress are matched very
closely, the present measurements show a more rapid decrease with height and
hence lower values up to approximately z/H = 0.4. These differences are
probably due to the different character of the roughness (two-dimensional)
23
-------
and to the lower value of roughness length.
The constant stress layer is at most 10% of the boundary layer depth.
2
The surface stress, approximately 0.0022 U^ , is essentially independent of
position; its magnitude is consistent with that obtained from the log-linear
plot of the mean velocity profile in Figure 32. It is also very close to the
proposed curve of Counihan (1975) for ZQ = 20cm (uV = 0.0023 Uro ). The
longitudinal turbulence intensity at an equivalent full-scale height of 30 m,
however, is somewhat lower than that proposed by Counihan (1975). The pro-
posed value is 17%; the present value is approximately 14%. Finally, in the
surface layer, the ratio of vertical to longitudinal turbulence intensities
is 0.50, compatible with previously reported values (Counihan, 1975).
The Reynolds stress measurements (Figure 35) show considerable scatter,
which is due at least in part to an insufficient averaging time. It is
difficult to discern any trends amidst the scatter except possibly a decrease
in the stress with downstream distance at mid-elevations.
The lateral homogeneity of the flow is examined in Figures 36 through 39.
The mean velocity (Figure 36) is seen to be within ± 5% of the mean for any
given horizontal plane. The turbulence intensities (Figures 37 and 38) show
wider deviations. Aside from the decay of turbulent energy with downstream
distance, the turbulence intensities may be judged as reasonably homogeneous
at any given section, i.e., deviations from the mean turbulence intensity at
any given plane are within ± 10% in the middle and lower portions of the boun-
dary layer and within ± 20% at the upper levels (where the "mean" turbulence
intensity is typically only 2 to 3%).
The Reynolds stress measurements (Figure 39) show very large scatter,
typically ± 50%. The lateral deviations in stress appear to be vestiges of
the fins at the highest elevations, but do not appear to be correlated direct-
ly with the existence of the fins in the middle and lower portions of the
boundary layer. The deviations from the mean appear to diminish and the
stress appears to become more laterally homogeneous as the flow develops
(although very slowly).
On the whole, the measurements of the boundary layer structure at
x/H = 4.5 compare quite favorably with those measured by Counihan (1969).
The boundary layer is seen to undergo a very slow streamwise development as
was also noted by Castro et al. (1975). The most disturbing aspect of this
24
-------
boundary layer is the lateral nonuniformity of the Reynolds stress. Figure 39
shows much wider deviations from the mean than have been reported previously.
However, the previous measurements have not examined variations across the
entire width of the wind tunnel; instead, they have typically measured only
three vertical profiles: one on the generator centerline, one at midpoint
between two generators (midspan), and one halfway between those two (quarter-
span). It is perhaps only fortuitous that previous investigators have found
such a high degree of lateral uniformity in their Reynolds stress measure-
ments.
The streamwise development of the boundary layer is similar to that
observed by Castro et al. (1975); it is mainly a decay in the turbulence
levels in the regions z/H > 0.15. Contrary to the results of Castro et al.,
however, slight differences in vertical turbulence intensity and Reynolds
stress were observed at different wind speeds. Hence, this boundary layer is
not entirely Reynolds number independent. The roughness used by Castro et al.
was much larger than the present roughness; they mentioned that initial work
had been done with smaller roughness and that a Reynolds number dependent
flow had been obtained.
Measurements of plume dispersion were made in this boundary layer to
determine whether its dispersive characteristics were similar to accepted
atmospheric dispersion characteristics. A model stack of height 0.2H emitted
a neutrally buoyant effluent of a 1% mixture of methane in air. The ratio of
effluent speed to wind speed was maintained at 1.5 to avoid downwash in the
wake of the stack. Lateral and vertical concentration profiles through the
plume centerline were measured at three positions downwind, as shown in
Figures 40 and 41, along with "best fit" Gaussian curves. The lateral and
vertical standard deviations of plume spread are plotted along with Pasquill-
Gifford curves for stabilities B, C, D, and E (Turner, 1967) in Figure 42.
Vertical dispersion within the wind tunnel boundary layer is closest to
stability C (slightly unstable), whereas the horizontal dispersion is closest
to stability D (neutral). The comparison with "accepted" atmospheric data is
regarded as good. These results are fairly typical of other comparisons of
wind tunnel measurements with Pasquill-Gifford curves, cf. Robins (1978) or
Wilson and Netterville (1978). The lower slopes of the o versus x curves
from the wind tunnel are also typical of other wind tunnel results (ibid,
25
-------
above references) and result from the slowly decaying wind tunnel turbulence
and possibly, in the case of the lateral spread, from the inability of the
wind tunnel to simulate the low frequency fluctuations in wind direction and
wind direction shear. Even though the turbulence structure is developing
somewhat with downstream distance and the Reynolds stress measurements show
rather gross nonuniformities, this boundary layer may be acceptable for a
number of purposes. It may safely be used for short range diffusion studies,
for example, for comparison types of studies where the object is to determine
building downwash effects from a short, nearby stack. It should obviously
not be used for longer range studies, where the plume width would be a sub-
stantial fraction of the tunnel width, without further testing.
More recent experiments with a trip fence and gravel roughness have
shown a much more laterally uniform boundary layer. A report on this new
boundary layer with much more extensive measurements and analysis is forth-
coming.
26
-------
REFERENCES
Bradshaw, P. and Pankhurst, R., 1964: The Design of_ Low-Speed Wind Tunnels.
Prog. Aero. Sci., v. 5, p. 1-69.
Castro, I.P., Jackson, N.A. and Robins, A.G., 1975: The Structure and Devel-
opment of a 2m Simulated Suburban Boundary Layer, Central Electricity
Generating Board, Res. Dept., Marchwood Engrg. Lab., R/M/N800, March.
Counihan, J., 1975: Adiabatic Atmospheric Boundary Layers: A Review and
Analysis of Data from the Period 1880-1972, Atmos. Envir., v. 9, no. 10,
p. 871-905, Oct.
Davenport, A.G., 1963: The Relationship of Wind Structure to Wind Loading,
Paper 2, Proc. Conf. on Wind Effects on Bldgs. and Structures, Nat. Phys.
Lab., June, HMSO, London, 1965, p. 54-102.
Davis, G., 1957: Non-uniform Flow through Wire Screens, Ph.D. Dissertation,
Univ. of Cambridge.
Pankhurst, R.C. and Holder, D.W., 1952: Wind-Tunnel Technique, Pitman and
Sons, London.
Pope, A. and Harper, J.J., 1966: Low-Speed Wind Tunnel Testing, John Wiley
and Sons, Inc., NY., 457p.
Robins, A.G., 1975: Wind Tunnel Modeling of Plume Dispersal, Central Electri-
city Generating Board, Research Dept., Marchwood Engrg. Labs., Mech.
Research Memo. No. 236, Dec.
Robins, A.G., 1978: Plume Dispersion from Ground Level Sources in Simulated
Atmospheric Boundary Layers, Atmos. Envir., v. 12, no. 5, p. 1033-44.
Snyder, W.H., 1972: Fluid Models for the Study of Air Pollution Meteorology:
Similarity Criteria, Facilities, Review of Literature, and Recommenda-
tions, Unpublished.
Snyder, W.H., 1972: Similarity Criteria for the Application of Fluid Models
to the Study of Air Pollution Meteorology, Bound. Layer Meteorol., v. 3,
no. 2, p. 113-34.
Snyder, W.H., 1974: Fluid Modeling Facilities for the Study of Air Pollution
Meteorology, Unpublished.
Snyder, W.H. and Lawson, R.E., Jr., 1976: Determination of a Necessary Height
for a Stack Close to a Building - A Wind Tunnel Study, Atmos. Envir.,
v. 10, no. 9, p. 683-91.
Sundaram, T.R. and Ludwig, G.R., 1970: Simulation of Small-Scale Atmospheric
Turbulence in a Laboratory Facility, Pres. at Subsonic Aerodyn. Testing
Assoc. Mtg., Ottawa, Canada, May 14-15.
Turner, D.B., 1967: Workbook of Atmospheric Dispersion Estimates, PHS Pub.
no. 999-AP-26.
Wilson, D.J. and Netterville, D.D.J., 1978: Interaction of a Roof-Level
Plume with a Downwind Building, Atmos. Envir., v. 12, no. 5, p. 1051-59.
27
-------
i
0)
c
3
h-
c
I
8
O)
_0
o
o
9)
Q-
LU
0)
O
0)
'£
0)
a
£
O)
IT
28
-------
REFERENCES
Bradshaw, P. and Pankhurst, R., 1964: The Design of Low-Speed Wind Tunnels,
Prog. Aero. Sci., v. 5, p. 1-69.
Castro, I.P., Jackson, N.A. and Robins, A.G., 1975: The Structure and Devel-
opment of a 2m Simulated Suburban Boundary Layer, Central Electricity
Generating Board, Res. Dept., Marchwood Engrg. Lab., R/M/N800, March.
Counihan, J., 1975: Adiabatic Atmospheric Boundary Layers: A Review and
Analysis of Data from the Period 1880-1972, Atmos. Envir., v. 9, no. 10,
p. 871-905, Oct.
Davenport, A.G., 1963: The Relationship of Wind Structure to Wind Loading,
Paper 2, Proc. Conf. on Wind Effects on Bldgs. and Structures, Nat. Phys.
Lab., June, HMSO, London, 1965, p. 54-102.
Davis, G., 1957: Non-uniform Flow through Wire Screens, Ph.D. Dissertation,
Univ. of Cambridge.
Pankhurst, R.C. and Holder, D.W., 1952: Wind-Tunnel Technique, Pitman and
Sons, London.
Pope, A. and Harper, J.J., 1966: Low-Speed Wind Tunnel Testing, John Wiley
and Sons, Inc., NY., 457p.
Robins, A.G., 1975: Wind Tunnel Modeling of Plume Dispersal, Central Electri-
city Generating Board, Research Dept., Marchwood Engrg. Labs., Mech.
Research Memo. No. 236, Dec.
Robins, A.G., 1978: Plume Dispersion from Ground Level Sources in Simulated
Atmospheric Boundary Layers, Atmos. Envir., v. 12, no. 5, p. 1033-44.
Snyder, W.H., 1972: Fluid Models for the Study of Air Pollution Meteorology:
Similarity Criteria, Facilities, Review of Literature, and Recommenda-
tions, Unpublished.
Snyder, W.H., 1972: Similarity Criteria for the Application of Fluid Models
to the Study of Air Pollution Meteorology, Bound. Layer Meteorol., v. 3,
no. 2, p. 113-34.
Snyder, W.H., 1974: Fluid Modeling Facilities for the Study of Air Pollution
Meteorology, Unpublished.
Snyder, W.H. and Lawson, R.E., Jr., 1976: Determination of a Necessary Height
for a Stack Close to a Building - A Wind Tunnel Study, Atmos. Envir.,
v. 10, no. 9, p. 683-91.
Sundaram, T.R. and Ludwig, G.R., 1970: Simulation of Small-Scale Atmospheric
Turbulence in a Laboratory Facility, Pres. at Subsonic Aerodyn. Testing
Assoc. Mtg., Ottawa, Canada, May 14-15.
Turner, D.B., 1967: Workbook of Atmospheric Dispersion Estimates, PHS Pub.
no. 999-AP-26.
Wilson, D.J. and Netterville, D.D.J., 1978: Interaction of a Roof-Level
Plume with a Downwind Building, Atmos. Envir., v. 12, no. 5, p. 1051-59.
27
-------
0)
c
3
I
re
o
'01
O
O
03
0.
LLJ
V
O
o>
'E
0)
6
28
-------
rfrt
UJ
>
a.
O
0)
c
3
CO
_0
o
"o
o
CD
0.
LU
0>
g
<->
CO
I
0)
•o
-------
SCREEN SUPPORT, TOP
CLEAR LUMBER, 1 in x 8 in 4 REQD
INLET BELLMOUTH
STEEL PIPE, 6 in DIA
SCREEN, FLOW STRAIGHTENING
WS TYLER INC STAINLESS STEEL
20 MESH x 010 in WIRE
INLET REAR FRAME
STEEL TUBING4 in SQUARE
CONTRACTION SECTION SKIN
STEEL SHEET 3/16 in.
s SCREEN SUPPORT, BOTTOM
CLEAR LUMBER. 1 in x 8 in 4 REQD
7 ' ' ' ' ' GROUND LEVEL
v "•"•„;-"••; , ,, THK ENTRANCE STRAIGHTENER
STEEL PLATE 7,n x 8 ,n x /, ,n THK VERTICELCO PAPER HONEYCOMB
Bin CELL LENGTH
FILLER ADJUSTER MOUNT -
STEELTUBE,4m SQUARE
SCREEN MOUNT TUBE—"^ f
STEEL PIPE 3 in DIA 3 REQD
21 0123456789
FRONT FRAME LEG
STEEL TUBING 6 in SQUARE
REDWOOD FILLER,SIDE
/ CLEAR LUMBER, 1 in. x 8 in. 8 REQD
=3—1=
SCALE, ft
1
3
-I I
CONTRACTION REAR FRAME
STEEL ANGLE 2% in x 2'/. in
SKIN REINFORCEMENT
. STEEL STRIP 4m x % in THK TOP & BOTTOM
v CONTRACTION FRONT FRAME
STEEL ANGLE 2V, in x 2V, in.
SCREEN TUBE LOCK
STEEL PLATE 2 in THK 4 REQO
Figure 3. Front, side, and top views of entrance section.
30
-------
31
-------
SCALE, ft
0 1
SCALE, m.
r
CEILING TRAVEL LIMIT
CEILING STIFFENED
STEEL TUBE 1 in. x 1 in.
SECTION B B
WINDOW
PIEXIGLAS '/.in.THK
V
•CABLE CHANNEL -
TOGGLE LATCH
WINDOW FRAME
STEEL TUBE 2 in x 2 in
SEAL SLIDING STRIP
POLYETHYLENE, 3/8 in x I in
, SEAL, INFLATABLE HOSE
NEOPRENE3/. in 0 D
.CEILING PANEL
SEAL PLATE "YWOOD.3«»
ALUMINUM STRIP I1/, in x 1/8 m
(ONE EACH. TOP AND BOTTOM)
CONTROL CABLE AND SENSOR WIRE CHANNEL,
STEEL T CHANNELS in x 2 in ONE INCH OPEN
r— — H- r. r
t
.,
!
h
L
TIE BAR STRUT
STEEL PLATE 2V, m x 7 in x 'A in. THK
WELD (ALL CONTACT CORNERS)
WINDOW CHANNEL,
ALUMINUM. % in x % in
WINDOW
PLEXIGLASS, '/. in THK-
WINDOW FRAME-
SEAL FLANGE
STEEL TUBE 1 in x 1 in
I BEAM
STEEL Th in. x 3 in
FLOOR REINFORCEMENT
STEEL TUBE 2 in x 2 in
FLOOR STIFFENER
STEEL TUBE 1 in x 1 in
SEAL
SPONGE RUBBERY, in xl in
SEAL
- SPONGE RUBBER1/, in x1 in
.ANGLE
STEEL 2% in x 2% in
rt--:^3
2101234567
r^r L-l 1 J <• ' ' ' ' '
SCALE, m
404812 16
SCALE, cm
FLOOR PANEL
PLYWOOD,'/, m
TIE BAR
STEEL ROD % in DIA.
BOTTOM VIEW
SECTION A.A Figure 5_ Jest section details.
32
-------
Figure 6. Inside view of test section looking downstream.
33
-------
DETAIL'A'
A A
210, 2 4 , 6 , 8 | 10 t 12 , 14
TURNTABLE
MASONITE, 1/8 in. PLYWOOD, V, in.
SEAL
SPONGE RUBBER V. in. x >/. in.
SCALE, cm
Figure 7. Turntable and details.
34
-------
o
T3
O
O
E
c
3
00
0)
35
-------
—-ft
o
T3
c
D
eri
«
36
-------
o
Q.
a
'
u
.a
CO
t+-
o
0)
O
0)
h.
3
37
-------
o
a
a
c
01
o
.Q
CO
T>
ca
£
-------
o
a
a
3
O)
c
a>
o
.a
to
4-
o
Q
o
37
-------
o
a
a
w>
O>
c
'5
u
.n
CO
•a
CO
£
OJ
eu
I
-------
Figure 12. Gear motor, limit switches, and torque tube.
-------
0)
u
09
CO
o>
40
-------
Figure 14. Exhaust silencer.
41
-------
£
(/>
_o
u
-------
8
£
8,
iZ
43
-------
LONGITUDINAL POSITION GEAR MOTOR
DAYTON 3M234, 1/15 HP. 2.8 RPM
PNEUMATIC GEAR CHANCE CYLINDER^)
SCHRAEDER 3452A
TRANSVERSE POSITION GEAR MOTOR
DAYTON 3M 235 1/15 HP 67 RPM
SPROCKET CHAIN
'.. m PITCH x% in WIDE, SIDE GUIDE
TEFLON SLIDE ALL 4 SIDES OPTICAL ENCODERS
THETA INSTRUMENTS, INC.
05 250 31 2 05 100 31 2
VERTICAL POSITION GEAR MOTOR
DAYTON 3 M 234 1/15 HP, 28 RPM
CARRIAGE SUPPORT ROLLERS
HARD RUBBER 3 in DIA
CONTROL CABLE & SENSOR WIRE CHANNEL
STEEL X' CHANNEL 2 in x 5 in 1 inch OPEN
PROBE HOLDER GUIDE BAR
STEEL BAR Kin x 2 in.
IN
TEST SECTION WALL
FRONT VIEW
'J&
Trf
CARRIAGE FRAME
~ STEEL TUBE 2m x 6 in
VERTICAL POSITION DRIVE SHAFT
SQUARE STEEL BAR'/. in
TOP VIEW
12 6 0 12 24 36 48 60 72
VIEW'A'
SCALE, in
30 15 0 30 60 90 120
SCALE, cm
Figure 17. Instrument carriage.
44
-------
00
cc
UJ _
Q in
O Q
O
o_
O
cc
5=
_l uj
< I-
O 2
^o
cc H
UJ
J
(O
O
o>
V)
c
0)
i>
00
t~
£
i,
59
-------
E
3
_Q
ca
o
*f-
O
0)
a
3
_0
o
46
-------
O
O
8
(B
l.
-------
u
«5
LL
D>
"55
"8
o
c
JS
Q.
O
O
§,
48
-------
<
IU
oc
4
ui
o.
>
z <
Z UJ
< 5
x <
o
o
cc
oc
o
u
2 a,
I <
Ul
a.
a
«- I- Ul
0. Z U
O LU O
O. O CC
o
uip
ZtJ
aoiu
U.
r
E 4
OC _l
01
I
I
a
o
u
'E
•s
(D
^> UJ U)
gag
£
I
I
.a
s
O CO
ol
L
!li
4< I
O H I
— «O i
SS I
2lu I
1
49
-------
CO
_l
O
PROBE
SLOPE
INT
ALPHA
CALIB. TEMP.
WIRE TEMP.
8807-2
6.916
9.118
0.470
= 18.500
219.80
O CALIBRATION POINTS
O ZERO FLOW VOLTAGE
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Figure 23. Typical calibration curve for hot-film probe.
50
-------
<0
O
Z
.E
E
of
n
CM
I/)
O
'^
(0
O
CM
<£>
CM
0)
(0
oas/iu '
(O
O
O)
C
in
uu
o
z
oc
JL
cc
I-
o
O> O T-
II II II
X >. N
LU
5
£
f,
il
CM
CM
oo
-------
a
ui 4
ui
a.
CO
a
z
5 3
x = 6.1 m
y = 0
2= 1067 mm
100 200 300 400 500 600 700 800 900 1000
FAN SPEED, rev/min
Figure 25. Calibration of empty wind tunnel; air speed as a function of fan speed.
52
-------
12
10.5
7.5
-f
4.5 -r
z = 800mm
A 1.83m
x= Q 9.15m
* 14.6m
-2000 -1500
-1000 -500
0 500 1000 1500 2000
y, mm
1
E
O
1.4
'l '• i i i i i i i i i i i i i i i i . i i i i i i i i i , i -
||^«-*— A — A_A^i_A_^_A_^_A_-A— ^—- ^ .
: to ^l1 i
1.2- :p ^ -
t 9 f -
1 n 4- w r, -'"
1'° t 7 * :i
i * i -i
•8 ••- I j I ;|
-i s ! :i
.6 -
.4 -
5 uj i
S — ' -
LU "* i -
Q _J ' -
35 g] -
d Z ' "
^ Pi.
•21 Tr
o ."" , i !
-2000 -1500 -1000 -500 0 500 1000 1500 201
y, mm
Figure 26. Lateral mean velocity and turbulence intensity profiles; U = 1.5m/sec.
53
-------
0)
in
r-
II
JB
LULU 'Z
8
oj
JD
CM
o
0)
~E
z>"
T3
(0
§
c
CD
-------
I" 1 1 1 1 1 1 1 1 1 1 I—I" 1 1 1 1 1 1 1 1 1 1 |— •rr'"'— -, r- i — f-'<"T-T-*T'T~T~*t-*T*^ 1 1 1
1
; ' |
i III 1
00 *"" •+ -t-'
• • " Ttri
^ %_ j
; <3D* ft
- ° /f
i > x J?
* ^
: ** D ""
: / Q
: /=' ;
; • / i i
& • A £1 fi tfl m *» . H H Fhdfitfl^
* D V St B "•"•B^" ™ 1-T'»J*'1P
3 CM v r- «- t- r-
UIUI 'Z
iii_iu — * — gj — ^ — *r— ffi — ft **£?^hn
^Kt^OTlfilr
: «.,_,_a_,;-,-,-,-^
* 1
: ^
- -
-
§'g§8§§§§§§S<
N O 00 S ^ M O « 5 "*
L m ai
" 5 ^
r 3 m
D
1
Q.
'M
c
K) m
1
U
4- O c
0 •
3
•e
t «* i
! M ^>
a
o
1 9 1
r co >
i
u> E
" 1
O CD
"" 1 >
E co
« 5 «
^ D g
g,
ii.
q
r-
3
UllU 'Z
55
-------
-+—IT—r—t—r
,==r^r-»^trM1^ f.*
CM
i'f •
E E E fih
= 8 -.5 t'J
«- en r- jji .-t
: OD* j'f
?fi F
0 /ll
: u u /at
. >. x ^ .
: !/•:
: / B
IE -j •
; ^-S-l—ft—^—ffi—Q—^—Si-^jT**1 .
^ ' * ' ' ' '" ' ' ' ' ' ' ' •*• ' ' -J~^-J~- '--'-•< >—*—»—' '— 1 1— t— < 1— ' f— 1— I tj
3OOQOO OOOOOC
3OOOOOOOOOO
MO09(O: ge «
•"tj I
^> D
-" jg
*0
« a
. «l 1
c
— . 0)
3 <=
1
3
-i—r-r-4-v-r-r
(O
CO
8
S
O
O
CO
o
o
to
0
o
o
CM
0
o
0
0
o
08
0
o
«0
§
tuui 'z
CM
.,? _ o
f I
-w«i o
o
o>
CM
0)
^
g,
56
-------
57
-------
^ ' ' ' ' ! i — i — i — 1 — i — i — i — t — I — i — i — i — t— 1 — n — r — i — 1 — r — i — j — i — 1 — i — i — i — i — t — i — j — i — i — 1 — i — r 1 — 1 — i — r-i — I — r— i 1 — r-
: * ''i "-If -__ ,, <, '•
* *~*if~i~D<' *W fl < '
< (0 ^"^ D * a'-ij? , < <,
^ ^~ T ° * •
< •£ Q t <3
• Q 3 ^1
||X£ "8
^ H H H 3 **
| x x x x D
; d " '* I
q
r~
en
00
(V
(O
"! 3
01
^
o
r^qo>oqrvco in^- co CN*-O
— r- ••-.... ^
r> ,, ;;| -
r ' * 1 o c
-J E * L, 4. ,
^"Iio
: 1 1
* X X X
: j < C*
^
q
^
en
ra.
u>
8
"? 5
3
CO
r-
t- q en oq r; N|x to in ^ to PJ <-; o
»-'«-'''"'' ,_
: ff '! * ': ^ • * . .
: "****?.•£'•->
.: "o N':'^-':;»f
: * ^_ " * * *
• • * "E i "" * 9 >
555(005 *s
• d — — " < ^ 5
(0 I I I — iE""
: ,| x x x jjj
: X < D * • r
q
en
00
^
5
.
CO
CN
(—1
rqqoD^qin«trofsr-o
^ *~ *^
.: ^ *Q -0 ^Q
: * S » •-' E '' „
: * "^^'E*3^,
L. ^ 1 ^
: S 505 * ,
: *T 1 H *
: II X X(/)
: X < D*
•-
q
*~
en
CO
r^
ID
in 3
^ 3
CO
CM
r- q en oo r~ N|x
o
X
• —
"E
n
ii „
n
D
"53
c
c
3
?
i
»^-
o
0)
1
c
u
O)
"Jo
J?
hfl
o
a.
'o
o
0}
>
c
co
CD
£
4-*
C
1
a
o
0)
CD
Q
,_:
CO
0]
8>
LL
58
-------
1 UUUU
1000
100
E
N"
10
_— _ — . r — r 1 . r — __ — ! — ,__,__ , , , 1_
D x = 4.5H
O x = 6.0H
A x = 8.0H
: O,
diV4
,/L
ili Vy
/ ***
Jb 63r-
Jrf3?"'
iJ^
1.0 1.5 2.0
VELOCITY, U m/sec
2.5
3.0
Figure 32. Mean velocity profiles in log-law form.
3.5
59
-------
"8
a
8
X
w
J 1^0000
U. II II II
I X X X
O CD 00
o
in
E
gSS
c
3
-a
o
03
U
01
s
=
CO
c
01
o
CO
a.
O
q en oo
60
-------
in X in
*t
n r>j
s
o> 08
to in
n 04 «-. o
O
c
v>
(D
8
CO
II.
I
I
1
5
8
o
1
1
3
"5
o
+3
CD
oq r» ,^3; to u>
61
-------
s
I I
in x "3
q ~" o ^ *
a '
(IS 1
§0
X
03
CO - — •
g
W 1
i ?
0 3
q N8 if
I 1 I
r *^
§ 1
in &
o >
8 o
O ^
OJoqp-. (oin^fON'^b ^
o «;
I q iS
»*
^
E
r r
r lJ ^
cc" *c <:
*
Vj »^ »., ^ ^
c
n 03
O sj
o
en
B c
§ -2
• re
CM <"
§^8 |
in "j «
§3 (X)
-a
0
o c
S tr
, ^ lB s -5
i. ?.?... , . . , , i0 *-
»oqr»
m 9 S,
>- 3 CO
O U-
O
O
in
o
8
01 oo r*« Nl^ to in ** n CM ^ o
-------
1.1
1.0 -
IL
£ —
"7 . —
.6
.4
.2
.1
0
A
n
*
e
a
+
X
x/H
4.5
6.0
7.5
4.5
6.0
7.5
4.5
6.0
7.5
z/H
.083
.333
.833
•i A .flr
*
.A.
a
. .1
..A
-1.0 -0.8 -0.6 -0.4
-0.2
0
y/H
.2
.4
.6
.8
1.0
Figure 36. Lateral uniformity of mean velocity; U00=3m/s.
63
-------
x/H
z/H
A
n
•"K
e
85
+
X
C
#
4.5
6.0
7.5
4.5
6.0
7.5
4.5
6.0
7.5
.083
.333
.833
.175
.150
.125
.100
.075
.050
.025
*
A
tr
• EB'
D
?*-
1
1
A
-1.0 -0.8
-0.6
-0.4 -0.2
.2
.4
.6
.8
1.0
y/H
Figure 37. Lateral uniformity of longitudinal turbulence intensity; 1)^= 3m/s.
64
-------
.175
.150
.125
.100
.075
.050
.025
A
a
*
®
ffl
+
X
<:
#
x/H
4.5
6.0
7.5
4.5
6.0
7.5
4.5
6.0
7.5
z/H
.083
.333
.833
x
-
. J
l\
-1.0 -0.8 -0.6 -0.4 -0.2
0
y/H
.4 .6 .8 1.0
Figure 38. Lateral uniformity of vertical turbulence intensity; \J00= 3m/s.
65
-------
x/H
z/H
-uw
U,,
.0022
A
a
ffl
+
X
C
#
4.5
6.0
7.5
4.5
6.0
7.5
4.5
6.0
7.5
.083
.333
.833
-0.8 -0.6
-0.4
-0.2
.2
.4
.8 1.0
y/H
Figure 39. Lateral uniformity of Reynolds stress; 11^= 3m/s.
66
-------
2.0
CUHj
1.8
1.6
1.4
1.2
1.0 -
.8
.6
H$ = 400mm
x/H
^ 3.3
H 6.7
'I- 10
-750 -600 -450 -300 -150
150 300 450 600 750
Y,mm
Figure 40. Lateral concentration profiles taken through plume centerline; 11^= 3m/s.
67
-------
1000
900
800
700
600 -
500
z, mm
400
300 -r
200
100
.50 .75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75
Figure 41. Vertical concentration profiles taken through plume centerlme; 11^= 3m/s.
68
-------
1000
100 —
I I 1
1
x, km
1000
100
10
10
Figure 42. Estimated wind tunnel dispersion parameters compared with standard Pasquill-
Gifford values (Turner, 1967).
69
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before complei ng)
i. REPORT NO.
EPA-600/4-79-051
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
5. REP.ORT DATE
THE EPA METEOROLOGICAL WIND TUNNEL
Its Design, Construction and Operating Characteristics
September 3979
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
W.H. Snyder
8. PERFORMING ORGANIZATION REPORT NO.
Fluid Modeling Report No. 6
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Sciences Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
10. PROGRAM ELEMENT NO.
1AA603 AB-20 (FY-78)
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
Environmental Sciences Research Laboratory - RTP, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
in-house 6/75 - 12/78
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The design philosophy, construction details, and operating characteristics
of the EPA Meteorological Wind Tunnel are described. Measurements in the empty
tunnel show that the mean velocity is uniform to within ± 2% at any given cross
section, at speeds as low as 1.5 m/s. The turbulence intensity in the empty
tunnel is typically 0.5%. A 2-meter-deep boundary layer was obtained using elliptic
wedge vortex generators and roughness on the floor. Measurements are presented
showing that this boundary layer simulates, in both turbulence structure and
dispersive characteristics, a neutral atmosphere boundary layer over rural terrain.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COS AT I Field/Group
Air pollution
* Wind (meteorology)
* Wind tunnels
* Atmospheric diffusion
* Design
13B
04B
04A
13M
13. DISTRIBUTION STATEMENT
RELEASE TO °UBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES
78
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
70
-------
"> "D O
co n; ^
O 3 —
O 5i C
— a;
< —
— CD
C
O
I
cr o IT
V) I
w m
=: ZD
.O 3
s!
O)
00
t/1 -1 —
------- |