Professional Documents
Culture Documents
Evaporation of Water As Related To Wind Barriers: NO. 2 Archive
Evaporation of Water As Related To Wind Barriers: NO. 2 Archive
Evaporation of Water As Related To Wind Barriers: NO. 2 Archive
.C6
NO.
25
copy 2
ARCHIVE
Completion Report
OWRR Project No. B-015 COLO
by
S. B. Verma
and
J. E. Cermak
Submitted to
The work upon which this report is based was supported in part by funds
provided by the United States Department of the Interior, Office of
Water Resources Research, as authorized under the Water Resources
Research Act of 1964, Public Law 88-379, Agreement Number 14-01-0001-
1436 (July 1, 1967 to June 30, 1971).
CER 7l-72SBV-JECll
1111111111111111
U18402 4845901
G8653
.C6
NO.
25
copy 2
ARCHIVE ABSTRACT
and mean velocity distributions were also measured over the wavy
boundary. In the region far downstream of the leading wave, where the
surface Reynolds number. Based on this result, the mass transfer data
from this study correlated well with the mass transfer data from surfaces
evaporation data from flat plate and sublimation data from surfaces
flow over equally spaced barriers has yielded valuable basic information
Chapter
3.3.2 Turbulence . . 46
iv
TABLE OF CONTENTS (Continued)
Chapter
3.8.1 Waves · 54
3.8.2 Fences 55
4.1.1 Waves 58
4.1.2 Fences · 61
4.2 Turbulence · 63
4~2.l Waves · 63
4.2.2 Fences 64
4.3 Mean Humidity Distributions . · 65
4.6.2 Fences . . · 74
4.7 Shear Stress Measurement on the Wavy Surface . . . 75
v
TABLE OF CONTENTS (Continued)
Chapter
v Conclusions . . 93
REFERENCES 95
APPENDIX A .105
APPENDIX B .106
Tables. . . . . . 109
Figures . .116
vi
LIST OF TABLES
Table
7 Values of f 115
s
vii
LIST OF FIGURES
Figure
11 Probes . . 127
viii
LIST OF FIGURES - (Continued)
Figure
ix
LIST OF FIGURES - (Continued)
Figure
x
LIST OF FIGURES - (Continued)
Figure
xi
LIST OF FIGURES - (Continued)
Figure
xii
LIST OF FIGURES - (Continued)
Figure
xiii
LIST OF SYMBOLS
Symbol Dimension
H Form factor
also heat flux
I Current amp
k K~rm~n constant
xiv
LIST OF SYMBOLS (Continued)
Symbol Dimension
Characteristic length L
m,n Exponents
p Pressure
Pr Prandtl number
(gm of water vapor)
q Specific humidity
gm of dry air
Ambient specific humidity
q' Fluctuations in q
qs -qoo
R Resistance of film or wire ohm
Reynolds number
Temperature e
t Time T
xv
LIST OF SYMBOLS (Continued)
Symbol Dimension
Friction velocity
z Roughness parameter L
o
z Hypothetical length parameter associated
om
with humidity profile L
6* Displacement thickness L
e Momentum thickness L
Wave length L
xvi
LIST OF SYMBOLS (Continued)
Symbol Dimension
Angular velocity T- l
Reference quantity
xvii
CHAPTER I
INTRODUCTION
flow and intended for protection against any ill effects of wind (43).
wind speed below the dangerous level for a particular kind of crop,
e.g., Bates (1930, 1945), Cheyney (1931), Denuyl (1936) etc., have
a field study and his results indicated that furrows if deep enough
evaporation and soil erosion from bare and vegetated currugated soil
required on how the local boundary conditions and adjacent wind barriers
affect the local flow. Systematic field studies on how various spacings
stand the transfer processes more completely, the study was conducted
in two parts. The first part concentrated on the flow over corrugated
areas (wavy surfaces) and the second dealt with the flow over equally
surfaces,
etc.),
CHAPTER II
This chapter deals with the basic concepts of mass and momentum
DU. a a
p 1 - ~ + ].I - - D•. (2.1)
~ - - ax. ax. 1J
1 J
au. au.
D.. = - 1 _J (2.2)
1Jax.- +
ax.1
J
au.
ax.1 = 0 (2.3)
1
3-direction is the y-axis which is lateral to the flow and projects out
of the plane.
5
(2.4)
_1 an +
~
w a2 u a --
(u'w') ~ (u' 2
p ax az 2 - az - ax
au
- +
aw = o (2.6)
ax az
and
(2.7)
The second and third terms of the right-hand side of Eq. (2.5) are the
au
-pT = \) -az --
- u'w' (2.8)
to turbulent transport, or
au « --
-pu'w'
lJ -
az (2.9)
-u'w' = KM az
au (2.10)
(2.11)
Thus, comparison between similar flows, for one with the KM known,
(2.12)
T is the wall shear stress. For values of zu*/v > 50, the universal
law becomes
7
(2.13)
(2.14)
involving a region of constant stress near the wall were used by van
which would describe the flow outside of the logarithmic portion, which
character of the turbulent boundary layer, since the wall shear supplies
0* = IC f /2 /). (2.16)
where
00 (U 00 - U)
~ 0 J u*
d (z/ 0) (2.18)
0
00 (U-U00)2
G =f u*
d (z/ 8) (2.19)
0
right in a U/u* vs. log plot. Roughness ~hanges the wall law
v
to
U
=A log v
(2.20)
where 8U/U* is called the roughness function, and measures the down-
ward shift of the logarithmic velocity profile. For fully rough flows,
where viscous effects have been removed, the inner law must be in-
8U
= A log v
+ D (2.21)
roughness function above has been used by Hama (1954) and Perry and
Joubert (1963). The form and orientation of the roughness elements used
9
by Hama (1954), Perry and Joubert, Betterman (1966) and Liu, et ala
(1966) have been screens or square rods perpendicular to the mean flow.
given by
6U 1 ctK+E
-= k log[(-v-) u*,] + C (2.22)
where ctK is the height above the roughness elements at which the inner
flow
U
-= fez c /K)
U 1 (zc +E)U*
k In - - v - - (2.23)
blend. For flow over a "K" type roughness, vortices of scale K are
shed from between the roughness elements. Flow above these elements
6U 1
= - log (2.24)
k
be larger than the smooth wall result since u* is much larger for a
rough wall.
this flow, (l would be small since the viscous layer would be near the
~u 1
-= k In v + (2.25)
tops.
u 1
(z
c + EK)u*
= - In (2.26)
k v
ment method and the momentum integral equation to obtain the values of
devised.
The slope of the log profile is 5.6 IC f /2, and using the momentum
element tops. When the data points give a straight line for distances
this type of approximate method does not always give consistent results
and use the moment balance over one roughness element. The details
of this method are given in Appendix (B). Using the value of C thus
f
determined, the slope of the logarithmic portion of the velocity profile
straight line with the correct slope of 5.6 IC f /2. When the slope of
are determined by this method, the velocity data points (u,z) can be
Jw,2/U 00
and the turbulent shear stress -pu'w' involve the complete
12
layers over rough surfaces have rarely been reported and empirical
(2.27)
for the outer flow results in an equation for the stress -pu'w' (107).
from the measurement of Klebanoff (1959) for a smooth wall, and is 1.77.
The applicability of the above analysis for rough walls has not been
established.
b = Uooodo
zKm dx (2.30)
13
-u'w' = K d (U - U) . (2.31)
m dZ 00
The above relation for the turbulent shear stress is expected to hold
G
for z/6 < 0.6, since the boundary layer is intermittent above this
level. For low values of z/6 (say, z/6 < 0.2), the additional
assumption of
U00 - U
«1 (2.32)
U00
does not hold. The analysis in this form for the turbulent shear
stress has not been examined for rough walls and the corrugated surfaces.
substance, e.g., the amount of water vapor associated with unit weight
~
at
+ U~
dX
+ V~
dy
+ W~
dZ
=~
aX
(K ~)
X dX
+ ~
dy
(K ~)
Y dy
+ ~
dZ
(K ~), (2.33)
Z aZ
u'q' = - Kx ~
ax'
v'q' = - K ~ and w'q' = - K ~
y ay z az
(2.34)
u ~ =a (K ~) (2.35)
ax az z az
lim q(x,z)
~n
= qs (a constant) (0 < x <
-
x )
0
Z-ru
coefficient R~
R~ (2.37)
ing K
z
= KM, Sutton used Taylor's theory of diffusion by continuous
K
= (O.25l)1-n
z l-n
(2.39)
Himus (1929), Hine (1924), Powell and Griffith (1935), and Thiesenhausen
similar to that first used by Blasius and later by Goldstein (1939) and
(1950) had obtained the same solution in an attempt to solve Eq. 2.35
heat transfer,
~ -- C
C 2 + 5(~)1/2
C
5 (a-I)]},
{a-l + In [1 + -6 (2.40)
H f f
Ce = E/pgU 00
6q (2.41)
-8/9 3.77 R -1
6.23 R
-1 x x
N = (2.42)
xi 4/45 xi 1/10
(-) (-)
Xl Xl
U2
in which T = Cf p
00
, Q = CH pC U 6 T, R = u*x , N = E'x , Xl =
2 P 00 X D D6q
distance downstream from the beginning of evaporation boundary, x'1 =
distance downstream from the leading edge of boundary, and E' =
evaporation weight/time x area.
17
Q.
1
= Zl-m Zom (2.43)
in which z
o
= roughness parameter and m = constant. The shear stress,
which is given by
T = (2.44)
then becomes
U = m1 (2.46)
power law
(2.47)
18
K = R,2 ~ (2.48)
z 1 az
and Eqs. 2.43 and 2.47, Frost solved the diffusion equation analytically
u= (2.49)
z
z
= 1. The limiting values in Eq. 2.49 are 0.75 ~ 8 < 1.25, where
o 1
and, except for a region very close to the surface, the distribution of
u 1 In u* z + 5.5 .
-k
v
(2.51)
u
-= 1. In (z-d) for z > z0 + d (2.52)
k z -
o
-u = (2.53)
and
u
-= q'" (2.54)
in place of Eq. 2.51 and 2.52, respectively. In Eq. 2.53 q" and a
with Eq. 2.51, and similarly, q'" and a also will depend on the
range of (z-d)/z such that Eq. 2.54 agrees with the more accurate
o
Eq. 2.52. Also, in the turbulent layer over a smooth or a rough surface,
there is a region not too far removed from the surface in which the
(2.56)
surface. From Eqs. 2.53 and 2.54, one can easily obtain
2 z
T = p _U_ (....£)20. (2.58)
o q ",2 z
for rough surfaces. Equations 2.57 and 2.58 are special cases of a
(2.59)
in which
1
s = 1+0. <5 = 'V , and £ = (_1_)
q' ,
2/1+0. (2.60)
'
and
s = 1, <5 = z,
o
and £ = 1/q",2
saturated, was constant over a plane at a distance from the datum plane
Thus, the boundary conditions (2.36) were assumed valid for a rough
z = 0, o < x < x
o
, as an aggregate of elemental, continuous, infinite,
q(x,z) and the evaporation rate. His expression for vapor concentration
above a smooth saturated strip is identical with the one first obtained
by O. G. Sutton (1934).
problem by Sutton (1934) and Calder (1949). The analyses by Sutton and
22
was taken into account, the surface temperature of the downwind wet area
by
F = pw's' = - as (2.62)
pK az '
in which w' and s' are the fluctuations in the vertical component
T = pu'w' =p ~.
au
az (2.63)
23
(2.64)
wind speed U,
au (2.65)
az - kz
U ! In z (2.66)
-= (z > z )
k zo o
roughness rather than from the level of the zero mean wind which may be
U 1
z -d
T (2.67)
- = - In
k
24
1 1
k In Zo = - In (2.68)
k v
diffusivity and eddy conductivity are determined by the first and second
Z
q-q = q* In (2.69)
s Z
om
and
T - T (2. 70)
s
from the base of the roughness. The quantities q* and T* are known
E = pw'q' = - pK ~ (2.71)
Z dZ
25
Combining Eq. 2.71 with Eqs. 2.63, 2.64, 2.65 and 2.67, one obtains
(2.72)
H =p C W't' =- P CP Kh az
aT =- P CP 8 'ku* T* (2.73)
P
obtained, both in field and laboratory, are contradictory and the ques-
tion is far from settled. The work of Swinbank (1955) suggests a value
of 0.7 for Kh/K , while that of Rider (1954) suggests about 1.3, for
M
neutral conditions. Pasqui11 (1949) found that in neutral and stable
z
0.8 for -0.03 < L + 0.10
~
j(-
- 3.2 z/L 0.35 for -0.8
z
< - < -0.03, (2.74)
M L
z
3.0 < -0.8
L
Eqs. 2.63, 2.64, and 2.71 and assuming Kz = KM ' one arrives at the
equation
2
u* a
E =- p ~ (2.75)
(au) az'
az
E = (2.76)
E = (2.77)
E = pk 2 (U2-Ul)(ql-~2) (2.78)
(In z2/zl)
The above formula was first derived by Thornthwaite and Holzman (1939).
E = (2.79)
This equation was first suggested by Pasquill (1949 a and b) and later
neutral conditions, Eqs. 2.78 and 2.79 are also subject to this
restriction. However, Pasquill (1949) has shown that the Deacon (1948)
for the modification of Eq. 2.79 for unstable conditions. The modified
equation is
2 2 2(1-S )
pk (l-Sl) (ql-q2)(U 2 -U l )zo 1
E = (2.80)
(Z2- d )1-Sl -(zl-d)l-Sl 2
In stable conditions, Rider (1954) has shown that Eq. (2.49) does not
hold and its use would lead. to an underestimation of the eddy diffusivity
equation:
E = p(O + K ) ~ (2.81)
z az
to obtain:
pku*(qs-q) D
E = 1 ku*z
for z » --
ku*
(2.82)
n D
K
z = ku*(z+z 0 ). Thus, on integration of Eq. 2.81, he obtained:
described below:
the water;
value u*R (u*R > u*R)· By requiring continuity in wind speed, a re-
(2.84)
o= (2.85)
E pk u*R Z
~
dz (2.86)
o
(2.87)
we get
(2.89)
the outer turbulent layer was concerned. Norris pointed out that by
vapor, except for the small difference in D and v , one uses the
distribution of water vapor and velocity must obey similar laws. Based
(2.90)
in which
(2.91)
r =
transfer rate varied with ReO. 6 ; Arvia et al. (1964) found that it was
.
proport10na 1 to Re 0.5 . Smolsky and Sergeyev (1962) investigated heat
benzol, and butanol) from a free surface into a turbulent air stream
Ta -Tw
Gu = T
(2.94)
a
32
with T and T the dry and wet bulb temperatures of the free stream
a w
respectively.
Nunner (1956), Smith and Epstein (1957), and Dipprey and Serbersky (1963)
found that the presence of roughness increased the heat (or mass) trans-
fer rates. On the other hand, Pohl (1933), and Kolar (1965) found the
greater than the thickness of the viscous sub-layer, but less than that
modified in such a way that a decrease of the mean mass transfer would
layer -- a viscous sub layer near the solid interface in which momentum
over individual protrusions will also occur in the case where the
but less than that of the turbulent boundary layer. Levich proposed a
33
c
= (-.i) 1/4
2
or
(2.96)
Re = ~U /v .
00
cases which satisfy the condition Sc» 1. Equation 2.96 agrees with
cathodic sites on the corroding pipe. On the other hand, Lai's (1969)
evaporation data, from his experiments on small water waves (Sc: 0.60),
34
did not correlate well with Levich's equation (2.96). Lai obtained a
account the fact that protrusions themselves extending into the boundary
fusion. Therefore, Eq. 2.96 gives somewhat low values for the mass
differences.
and mass transfer in the turbulent flow of fluids through smooth and
3 1/4
A = (~) (2.97)
o £
A-
o
T = (2.99)
e UA-O
surface of the vortex element is in contact with the medium in which the
heat transfer occurs. Thus, Eq. 2.98 indicates that the rate of heat
contact with the surface. Combining Eqs. 2.97, 2.98, and 2.99, one
obtains
(2.100)
or
(2.101)
36
mass and heat transfer in flow over rough surfaces, such that
u*h l
--v-- > 100 (where hI = 30z o = the equivalent sand roughness), Owen and
(2.102)
where
_pU_(_z)--,.(_q_s-_q_) = U( z) [U (z) +
(2.103)
E u* u*
(J = D(orv (1) , m and n are constants to which values 0.45 and 0.8
were assigned. In Eq. 2.102, <11 was suggested to be a constant for each
to and from some rough surfaces. He analyzed his results in the light
the entire atmospheric boundary layer in an "exact" sense does not appear
dimensionless parameters which should be nearly equal for both the model
~e of Similarity Requirement
for only two of the requirements. These are equality of Rossby numbers
and equality of Reynolds numbers for the model and the atmosphere.
38
patterns.
UL (2.104)
=
KM
because the eddy diffusivities are not known accurately in either the
Nemoto (1961) expresses the Reynolds number criterion for modeling wind
velocity as
U LM 1/3
cxM
= (-J (2.105)
U
cop
L
p
This is much more plausible form of Reynolds number similarity and has
be increased for the model as was the case with ordinary Reynolds
each other. Thus if the flow over the model is rough, the Reynolds
flow over the model is not rough, its turbulence characteristics would
criterion pointed out by Jensen (1958) that the roughness parameter for
i. e. ,
z L
~ = -E. (2.106)
z L
om m
mass and momentum transfer from smooth and rough surfaces. The information
40
encountered in field research and because of the great amount of time and
CHAPTER III
tunnel and the U.S. Army Meteorological wind tunnel. Both tunnels are
in the CSU wind tunnel. The measurements were taken on sinusoidal waves
cut in three sizes (A = 4.2 in. and h = 1.70 in., 1.0 in. and 0.50 in.)l.
2
These waves were cut from styrofoam blocks with a single shaper blade
and the shaper blades used to cut these waves. The styrofoam waves
were laid in aluminum pans (Figs. 3 and 5), which were of sizes convenient
for moving. Aluminum channel supports were spot welded to the bottom
of these pans; the waves were later screwed down in sets to these channel
supports to keep them from floating when the pans were filled with water.
An asseumbly of six pans was used (Figs. 3 and 5). The first pan con-
tained a set of 4 1/2 waves, the next four contained 5 waves each, and
These pans were placed adjacent to each other in the wind tunnel (as
shown in Fig. 3), and a sheet of felt was positioned very carefully over
the waves, following their contour. Small stainless steel pins were
used to tack the felt to the waves. The ends of the waves toward the
walls of the wind tunnel were sealed. When positioning the individual
styrofoam waves in the pans, care was taken to leave a small (about
1/32 in. wide) crack between each wave to allow water to flow up from
the bottom of the pans (Fig. 5); thus the continuity of the wavy surface
was not disturbed. Once the pans containing the waves were assembled,
bottom of each pan. The water level was maintained so that water came
to the top of the crack in the troughs of the waves (Fig. 5). Before
beginning a run, the felt surface was saturated with water; it was
observed that the felt remained wet everywhere as long as the desired
section 3.7). No water was supplied to the pans so that all the waves
were dry.
false floor (approximately 4 3/16 in. high). Also, just upstream of the
transition from the contraction section to the false floor (Fig. 3).
entire width of the wind tunnel between the aluminum sheet and the false
floor at a distance of about half an inch from the leading edge; the
brushes were used for tripping the turbulent flow and thickening the
a smooth transition from the end of the waves to the wind-tunnel floor
(Fig. 3).
and the sharp edged side stood vertically as a fence, the other side
served as the foot of the fence (refer to Fig. 6). All fences were cut
the full tunnel width. The front edge of the first fence was located
40 ft from the wind tunnel entrance. From the first fence downstream,
all fences were fixed on a 1/2 in. thick plywood floor. The spacing
between each fence was chosen at a constant value L for each run. The
44
plywood floor was made to cover the total test section. Double stick
scotch tape was inserted between the fence foot and the plywood floor
and 3/8 in. wood screws were used to fasten fences on the floor. Half
inch wooden strips were glued to the bottom of the plywood to elevate
it one inch above the original aluminum floor. In order to measure the
pressure taps were drilled in front of each fence through the plywood
floor. One pressure tap was drilled at the wind-tunnel center line,
the other two were drilled 12 in. to the right and to the left of the
center line. Plastic tubes were connected to the pressure taps and let
outside to the tunnel through the space between the plywood floor and
shown in Fig. 6.
Pressure tapped fences, which were used to measure the form drag
of fences, have the same shape and dimensions as the other fences except
some pressure holes were drilled. Two tapped fences were made--one for
a fence. Seven pressure holes were drilled on the first fence, with
each hole spaced 1/8 in. in the vertical direction. And since the
separation pressure was constant over the height of the fence (see
Plate, 1964, and Nagabhushanaiah, 1961), only three pressure holes were
drilled on the second fence (Fig. 6). Three sets of such pressure holes
45
were drilled on both tapped fences, one set was at the center line and
the other two were located 12 in. right and left from the center line.
Pressure gradients over that part of the test section which was
is given below.
Physical Set Up Primary Measurements
Wave Assembly (Wet) Distribution of mean velocity, mean
humidity, mean temperature, and
local mass transfer coefficient.
Wave Assembly (Dry) Distribution of mean velocity,
turbulence, wall pressure and local
heat transfer coefficient. Drag
measurements and visualization.
Fence Assembly (Dry) Distribution of mean velocity
turbulence, wall pressure and local
heat transfer coefficient. Drag
measurements.
measurements.
pitot tube must involve an error because of the presence of the above
static taps were put on two different arms held along the y-axis ((56)
between the two taps. Both the total and static heads were measured by
with a D.C. amplifier and a digital voltmeter, was used to obtain the
The moisture content of the air flow was measured by sampling the
gas stream through a specially built sampling probe (see Figs. 11 and 12),
very slowly, with time because the recirculating air gradually absorbed
both the local moisture content in the boundary layer and the free-
probes to the dew point hy~rometer consisted of 1/4 0.0. teflon tubing
valve. The surface air humidity was determined by measuring the surface
in the sample gas. This is the dew point temperature of the gas; the
moisture content of the gas can be determined from this using some
optical sensing bridge for sensing the dew formation. The amplifier
the mirror. The temperature of the mirror, representing the dew point
(110).
surface air temperature, respectively. The free stream and surface air
station.
after the experimental runs in the wind tunnel. This was done to check
for any suspected drift in the stability of the batteries or the standard
upstream wave (Fig. 3). This special wave facilitated the direct
with the aid of an "automatic feed and metering system". Both the
49
"evaporation metering wave" and the "automatic feed and metering system"
the wave (of wavelength = 4.2 in.) was divided into eight sections of
equal width with thin (0.020 in.) copper baffles comprising both the
sides and the bottom of each section. The copper channels (which con-
block with the top edges of the channels forming the desired sinusoidal
shaped wave of desired amplitude and wavelength (Fig. 7 and 8). Strips
of blotter paper (0.022 in. thick) were cut to fit vertically in these
channels; the tops of the blotter papers were carefully sanded to form
automatic feed and metering system described later, was supplied to the
top of the blotters by capillary action. Thus, the blotter surface was
kept completely wet as long as the water supply was maintained. Brass
screens (18 x 14 mesh, 0.011 in. diameter wire) were placed at the bottom
channels, etc., in the wave. The presence of the thin copper walls
wave to another. From a top view, the measuring wave was 2 ft long. Two
50
additional waves (called the buffer waves), each also 2 ft long, were
provided on each end of the measuring wave (Fig. 3). These buffer waves
were also kept wet to minimize lateral diffusion of water vapor. Care
was taken to seal the ends and sides of the metering wave to avoid any
thermocouple beads were placed in the center of the top surface of the
blotters; their leads ran 1/4 in. downward from the surface and then
along the horizontal length of the wave between two blotter layers, to
Metering waves of three sizes, namely A = 4.2 in. and a = 0.85 in.,
0.5 in., and 0.25 in., were built separately. Figure 7 shows the
maintains and measures the supply of water to each of the eight sections
of the metering wave; it is similar to the one used by Cermak and Lin
(1955), and Nelson (1957). Figure 8 shows the arrangement needed for
one typical wave section. Each wave section was linked to a graduated
burette, which in turn was connected to its own water bottle. These
burettes were closed by rubber stoppers at the top; two brass tubes
(0.092 in. 0.0.) were inserted in each through holes in the rubber
stoppers. The bottom end of tube A (Fig. 10) was used to control the
created by the water loss from evaporation at the top of the wave section,
This volume of water drawn from the burette was replaced by the ambient
51
air coming through tube A. Thus, the evaporation of water from each
wave section took place at ambient pressure and the amount of water that
evaporated from each section in a given time was measured from the drop
were corrected to account for the volume occupied by the brass tube A
in each burette.
heat (Fig. 15). The heated element was a Oisa 55A90 flush-mounted film
relation
and R are the hot and cold resistances, respectively, of the film
c
and a' is the temperature-resistance coefficient, equal to 0.3%/oC
The wall-heat flux in the form I2R/~T can be related to the shear
stress T through a calibration of the form
(3.2)
52
Brown (1967). The calibration, Eq. 3.2, describes the behavior of the
must be represented by
T
U = w z +
11
and (3.4)
W= 0
uniform stream using the values of T obtained from many sources, ((1),
(28), (32), (95)). Two typical calibrations are shown in Fig. 15 and
16. Note should be ade of the following results: First, the calibra-
tion
~~R = M T l / 3 + P (3.5)
53
(3.6)
2 3
A = -3.9 x 10- B = 5.6 x 10-
5 2
C = -8.0 x 10- D = 7.0 x 10-
4 5
E = 5.7 x 10- F = 1.5 x 10-
Using these values for the constants, T can be obtained from measure-
ments of 12R and /::"T. The response Eq. 3.6 was used to "predict" the
shear stresses for the calibrations shown in Fig. 16. The "predicted"
and calibration shear stresses are shown in Fig. 17. In general, the
when the measured calibration 12R's and ~T's are used in the
(3.7 )
where WI and L are the width and length of the film. The effective
length, L ,of the film can be determined from Eq. 3.7, and for
e
54
The goal was to determine areas of erosion and strong shear, which
heat loss and wall pressure. An attempt to correlate the heat losses
to shear stress was also made. The results of this part of the measure-
3.8.1 Waves - The drag force on the waves was determined in two
This pressure tapped wave has the same shape and dimensions as other
styrofoam waves except that 17 equispaced holes were drilled across the
wall pressure at that point on the wave and the static pressure at a
reference height in the free stream. Fig. (69) shows the wall pressure
surfaces, Hsi and Nath (1968) had previously designed a special shear
cm x 59.6 cm x 59.0 cm. This plate was separated from the foundation
55
One end of the arm was attached to the shear plate and the other to the
of the restoring arm, were installed at 1.27 cm from the end of the
restoring arm which was attached to the foundation plate. The shear
plate moved back and forth in a horizontal plane only; it had a natural
arms and the magnitude of the applied force. The shear plate would
return to its original position after the applied force was removed.
plate readout.
tape to the top of the shear plate. The shear plate was put in the wind
th th
tunnel so that the above wave pieces would occupy the 26 , 27 , and
th
28 wave positions; other styrofoam waves were put on each side and at
each end of these pieces to maintain the continuity of the wavy surface.
Care was taken to properly seal the ends and sides of these pieces. The
force to the plate and recording the bridge output. These measurements
3.8.2 Fences - The drag force on the fences was measured by means
of two tapped fences; one was the "stagnation fence" and another was the
56
the upwind surface of a fence was measured through seven pressure taps
of the fence was measured through the pressure tap drilled through the
z = 7/8", 6/8", 5/8", 4/8", 3/8", 2/8", 1/8", and z = 0 were recorded.
By substituting the separation fence for the stagnation fence, the
separation pressure on the lee side of the fence was measured. Separa-
h
Df = f (Pf - Pb) dz (3.8)
o
where
each fence.
the waves, a visualization study was made. Air motions were traced by
releasing smoke near the waves, and visible records were obtained by
means of pictures taken from a 35 rom still camera. The smoke was pro-
located outside the wind tunnel floor by means of a tygon tube terminating
CHAPTER IV
effort to maximize the available moisture and heat for plant growth.
discussed.
= 0.99 U 00
Figures 18, 19 and 20 clearly show the changes in the
means that the boundary layer reached its equilibrium condition for
59
thickness 8 and form factor H , are shown in Figs. 21, 22 and 23.
profile
U 1 R,n
z
z > z (4.1)
u* =k zo o
for nearly neutral conditions (see page 61). In Eq. 4.1, z is the
and £ =h - d
c
is the height measured above the respective wave crest,
d is the zero plane displacement, and h the wave height. The value
then calculated using values from the drag measurement results. Values
from the bottom of the waves, the effective zero plane is raised by a
half and a quarter of the wave height for the steepest and the inter-
mediate waves, but remains unchanged for the shallowest one, which seems
reasonable.
Considering the above and using Eq. 4.1, values of u* were com-
for all three wave situations. The results indicate that the friction
velocity initially undergoes a rapid change when the air first encounters
the change of roughness, and then adjusts itself in the far downstream
region. This type of response in the wall shear stress has been pre-
Yeh (1970).
lowest few inches (above the wave crests) was large, and a typical
aT
value of the Richardson number = g az for 1 in. < z < 4 in.
f (au) 2 c
az
3
was found of the order of 10- . The Monin-Obukhov parameter
L = H
(with H - L E, see Section 4.4), was of the order
o e
k K (~)
T pC
P
3
of 10 cm. This means that an error of about 2 to 3 percent is involved
from the initial profile (at station x = -4 ft) when it encounters the
collapsed into a single curve. This means that the boundary layer
and Eq. 3.7 was used to determine the form drag where
h
Df = J (Pf - Pb)dz . (3.7)
o
(4.2)
62
If the distance between fence L is not large, the shear stress acting
form drag acting on the fences. It is possible to assume that the aver-
T
o
= L1 D (4.3)
T
o
(4.4)
1
- pU 2
2 00
(4.5)
Values of measured cD' Cf , and u* for all six runs are shown in
are the shear velocities obtained from the slopes of the semi-log plots,
U vs log z .
R
x
is high, Rx = U x/v ,
00
is independent of R
x
and is a function
distance far downstream from the leading fence, the important length
held constant in the present study, the length scale must be L. The
measured C for each run at the largest x is plotted in Fig. 28, and
f
the relation between C and RL is found to be
f
63
(4.6)
= 1.05
4.2 Turbulence
for the last wave position for a typical case are shown in Fig. 30.
The intensity profiles are similar for a wave for positions 15, 20, 24,
and 25, and are also similar between the three waves. The measurements
of Liu, et al. (1966) and Corrsin and Kistler (1954) are also shown for
comparison.
00 dU 00
~x
T 00
-
p
=
f U(U oo - U)dz + dx J (U 00
- U)dz - W(U 00
- U). (4.7)
z z
The values for the terms were evaluated using the mean velocity
term. From the mean velocity profiles, the vertical velocity W, given
by
z
d
W= - dx J U dz (4.8)
z
c
stresses using the momentum integral approach give values of the wall
method. The profile shape for all three cases is :imilar. The measure-
ments over square rods perpendicular to the mean flow by Lin et ale
(1966) and over water waves by Karaki and Hsi (1968) are also shown for
comparison.
stress layer above the boundary. Although the stresses for the cor-
law. The same conclusion was reached by Perry and Joubert (1963). In
profile law gave reasonable values for the origin shift; the effective
typical run are shown in Fig. 32. "Over shoots" are found to exist in
65
when flow encountered the fences. But one can see from Fig. 32 that at
station x = 0, u
,2 almost retained its initial quantities and the
plotted in Fig. 38. The shear velocity u* determined from mean velo-
Unlike the smooth wall case, the turbulent shear stress -u'w'
near a rough wall does not have a uniform region. This can be found
1(11), 1(9), and 1(7). The corresponding for each station is also
stream, and surface air humidities. It has been shown (e.g., Chamberlain
66
evaporation from a free liquid surface; so, the air humidity at the
and more water vapor effluxing from the wave in the boundary layer. The
Fig. 39.
41, and 42. Except for large heights, most of the data points, taken
Montgomery (1940), Pasquill (1949), and Lai (1968). The humidity pro-
q-qs
files in the far downstream region were then plotted with vs
q*
z
axes (Figs. 43, 4, and 45), in which z is again the height
z
om
67
measured from the effective zero plane, which is the same as for
the line
z
= q* ~n
z (4.9)
om
46, 47, and48 show these humidity distributions. Shown in Figs.49, 50,
and 51 are the enlarged views of the lower portion of the corresponding
increasing until just before the next crest. This result facilitates
typical results are plotted in Fig. 55. As mentioned, free stream and
along with the local temperature T(z) to account for changes in ambient
the equation
T - T
s
(4.10)
in which z is the height above the effective zero plane, with the same
H= (4.11)
was conducted upwards through the floor of the wind tunnel and this was
vaporization.
inside the troughs of the waves in the far downstream region. Figures
behavior. The mass transfer data reported in this study reveal some
the surface of the specially designed wave located in the far downstream
were taken with all upstream waves completely wet, at various ambient
air speeds for waves of all three sizes. The ambient and surface air
The result for all cases are plotted in Figs. 59, 60, and 61. As
59, 60, and 61 have been projected to the next consecutive wave to
for each of the three wave sets, is essentially the same, indicating
that the flow pattern for each wave remains unchanged in the correspond-
ing range of Reynolds numbers. In each case, the evaporation rate shows
from 3.6 for wave A to 2.40 and 2.10 for waves Band C, respectively.
After reaching the minimum, the evaporation rates show relatively little
Photographs taken for visual studies are shown in Figs. 62 and 63.
The most noticeable feature of the flow is the vortex (or roller) seen
vortex region 'v', one can also see that the region marked's' is the
place where the flow separates. It is also noticed that the region 'a'
is the approximate vicinity where the airflow contacts the wave surface.
As seen in the photographs (Figs. 62 and 63), the flow on the upstream
near the place where the flow contacts the wave surface.
wave sets, are shown in Fig. 65. A general feature noticed is a maximum
'a' where the flow contacts the surface and a minimum near region's'
(8 and C) .
(Fig. 65). Downstream from this minimum, in region 'v', the evaporation
rate for steeper waves is found to be smaller than for shallower waves.
Also, in the case of waves A and B, the evaporation rates are relatively
(A) is much better defined than the vortex in the shallower wave (C).
picking up more and more moisture, thus reducing the evaporation rates
noticed, especially near the place where air flow contacts the wave
surface (Figs. 62, 63 and 64). The evaporation rates become significantly
in the vicinity of region 'a'. In this region, the incoming dry air
direction. The corresponding shift in the wall pressure maxima can also
be noted (Fig. 65). This indicates that the region where the dryer air
comes in contact with the wave surface is reached sooner for the shallower
waves than the steeper ones, which is reasonable. After this maximum, the
73
also to the fact that the low moisture-content air coming in contact
with the wave surface in region 'a' picks up relatively more moisture
as it flows along the wave before reaching the region's' and so is left
as discussed in Section 4.3 (see Figs. 52, 53, and 54), a typical set
is fairly low in region 'v', but later it increases toward region 'a',
wave" were also taken when all upstream waves were kept completely dry.
Typical results are shown in Fig. 68. For this dry condition also, the
The above measurements were repeated with the measuring wave turned
sections is negligible.
4.6.1 Waves - Local heat loss from a point source of heat on the
between the heat flux and the wall shear stress will be discussed later.
The heat flux profiles, in the form I2R are given in Fig. 69 and
0
70 for a small (7 rom x .2 mm) element approximately 35 C above ambient
temperature. The three dimensional vortex between the waves removes heat
from the element and produces large values of I2R in the troughs (Fig.
transfer using an area source of water vapor (Section 4.5) can be used
of the heat flux agree with the mass flux measurements. The separated
flow after the crest removes less heat from the wave surface than the
the crest. As the next crest is approached along the wave surface, the
4.6.2 Fences - Local heat flux profiles for fences are shown in
are definitely not met. However, assuming the wall shear stress is
is most noticable for wave A, as wave C has a very small vortex flow
between the crests. After the trough region the shear stresses increase
to the large value at the next crests. Shear stresses on wave C follow
a sinusoidal pattern along the wave profile. Kendall (1970) has observed
76
the same sinusoidal pattern for waves with small h/A values. However,
his sinusoidal waves did not have a separation region in the troughs.
Section 4.5). Knowing the velocity and humidity profiles, two other
flow was obtained. The reliability of these two methods was cross-
developed region.
The first method, which may be called the mass balance method,
d
o
E = J pU( z) [q ( z ) - qoJ d z . (4.12)
o
height z. Curves for a typical case are shown in Fig. 76. An optical
c 0
planimeter was used to obtain the value of the integral f pU[q-q ]dz. QO
o
The value of evaporation rates were obtained by using Eq. 4.12.
77
o
This method requires measuring slopes of the curve pU[q-qoo]dzJ
o
vs x; therefore, measurements at a number of closely spaced longitudin-
view of this, this method was used to evaluate the evaporation rates in
the fully developed (far downstream from the leading wave) region only.
used.
sections. The vapor flux at the surface can be obtained from Eq. 2.72
fetch for all of the three cases. In general, the evaporation rate, in
crests and then decreases, leveling off in the far downstream region.
When the air first encounters the presence of waves, there is a sudden
Also, when the relatively dry air contacts the first few waves, it
absorbs much more moisture than it does in the downstream region. The
evaporation rate later decreases and becomes contant in the far down-
In the far downstream region, the results obtained from the first
direct measurements; the first method agreed within 10 percent and the
second, for most cases, within 20 percent (Table 4). Considering that
encouraging.
ambient air,
For a fully developed boundary layer under near neutral conditions, with
problem leads to
E 5l., v
= f[ , IT] , (4. 13)
boundary layer was obtained near the evaporation metering wave. The
located at both ends of the metering wave, completely wet during the
79
the present study can be correlated on the basis of Eq. 4.12. As the
v
where D (=Sc) is included in the constant 0.10. In connection with
the above relation there have been numerous values proposed for the
exponent to the Reynolds number for the transport of mass or heat from
for flow through rough tubes, Kolar (1965) and Nunner (1956) found values
of 0.986 and 1.0, respectively, for the exponent to the Reynolds number
formed from the friction velocity and diameter of the tube. On the
Eh
pD(q -q )
(4.15)
s 00
results are in agreement with Mahato and Shemilt's (1968) results (with
80
waves (Sc: 0.6) with Levich's correlation (2.63), Lai (1969) reports
(Sh)
o
= 0.055 (Re*)O.90
0
(4.16)
gives the best fit correlation for all the different sets of data. The
from any type of surface, some examples of which are considered in the
4.8.3 Comparison with evaporation data from water waves - Lai (1969)
U 1 tn ~
= k (4.17)
z
o
~l
with the data obtained in the present study, the following considera-
constant speed or celerity c. The wind blows over these waves with a
constant with time and direction. The flow is unsteady and the boundary
same speed as the waves is introduced. When viewed from the moving
kc
(z ) moving coordinate
~_o~~__~____~_____ = e u* (4.18)
(zo) fixed coordinate
Ez
Using this modification, values of ( Sh) o -- pD(q 0 -q ) and
s 00
were computed from the set of data, (taken in the approximately fully
Fig. 82. As shown, the evaporation data from the water waves and those
from the fixed waves are reasonably well correlated by Eq. 4.15.
smooth flat plate. To correlate their data with that of the present
study, one needs the values of the friction velocity u* and the
relation
zu*
~n --- + constant. (4.19)
v
shown in Fig. 83. The line representing Eq. (4.15), which correlated
the data from the solid and water waves, can be extrapolated to correlate
camphor from surfaces roughened by two types of glass: "one was mottled
by irregular pyramids in relief and the other reeded, with the generators
their sublimation data with the evaporation data, one must account for
(4.20)
Ez
can be attempted, in which ( Sh) o -- 0
pD(~q)
where E is the mass
transfer rate per unit area and ~q is the difference of the surface
and Sherwood (1934). Levich (1962) and Mahato and Schmilt (1968) used
and Colburn (1934), Reynolds et ale (1958), and many others report a
PrO. 33 dependence, whereas Kolar (1965) obtained a PrO. 5 dependence
on the Nusselt number. In this study, n = 0.33 has been chosen for
line
(Sh)
___~~o~ = 0.064 (Re*)0.90 (4.21)
(Sc)0.33 0
(Re*) and the Schmidt number Sc. This implies that the roughness of
o
the surface can influence (Sh) only insofar as u* and z are
o 0
by knowing the velocity profile above a surface, one can predict the
relation
(4.22)
in which
b u*k Alvk
rb = [R.n y + (-0- + R.n (4.23)
u*R
Equations 4.21 and 4.22 have previously been presented as Eqs. 2.57
r = [~n b ] -1 , (4.24)
b az
o
u*k
in which ~na = 1.76
with and being the friction
u*R
velocities inthe intermediate and turbulent layers of the three-layer
that the value of mean humidity measured at 10 meters above the water
The roughness length z of the sea surface has been found to be about
o
0.6 cm, regardless of wind velocity. On the other hand, for hydro-
dynamically rough lake surface, z was found to vary from 0.55 to 1.55
o
cm for the wind velocity ranging from 1 m/sec to 15 m/sec measured at
tion
86
Ez
o
(4.25)
It turns out that Norris' results are not very sensitive to the
and ambient humidity, using relation (4.24), one can reliably estimate
= pu*[qs-q] _ U(z)
(4.26)
E u*
has been used [Chamberlain (1966), Chamberlain (1968) and Owen and
(4.27)
where m and n were assigned the values 0.45 and 0.8, respectively,
-1
Values of B and were computed from the experimental
for u*z
o
Iv > 40, along with the data points from Chamberlain's wind-
(19). The results show that, for the range of u*z oIv considered,
-1
B is always positive, indicating that the resistance to momentum
u*z o Iv the value of B-1 will depend upon the type of roughness
elements, the B-1 values obtained in this study are of the same order
a time.
model of a field furrow is shown in Fig. 87. For the furrow orientation,
the wavy surfaces of this study is shown in Fig. 88. As expected, the
establishment.
less than the evaporation loss from the bottom of the shallowest wave C.
can be defined as
where U[z] is the local wind speed measured after placing fences, and
U [z] represents the local wind speed measured in clear tunnel. The
o
values of U
o
used in this analysis are measured at x =- 4 ft.
Figure 89,90, and 91 are plots of constant R's for L =6 in., 12 in.
by the ambient velocity. Furthermore, all six runs show that a constant
Figure 92 shows the wind reduction pattern between fences for Case II
h L
~=~ (4.29)
hm Lm
However, when applying this result to design the wind breaks, a smallest
91
when the wind reduction R > 70% is needed for y < h , fences must
(b) Waves: Wind speed is reduced more at any height above the
wavy surface than for a comparable height above a smooth surface. The
R1 = (1 - ~)
Uo x 100
Figs. 94, 95, and 96 for the corrugated surface. The wind speed is
reduced the most at any height by the largest wave (Wave A), as would
from the change in wind speed between the start of the waves (A = 0)
th After the 15 th crest, the reduc-
and approximately the 15 wave crest.
tion in wind speed changes slowly at any height above the wave crests.
factor
R
f
=~ (1 - ~) x 100
u*o Uo
92
a surface did not increase the surface shear, the above two wind reduc-
tion factors would be the same. The values of R are shown in Fig.
f
th
97 for the 25 wave crests.
The largest wave gives the best windbreak effect for two reasons:
1) The wind speed is reduced the greatest for this wave, and 2) vortices
are created between the wave crests. Both effects provide sheltering
for seeds in the troughs. The vortices have a small velocity (49) and
do not remove the seeds. This sheltering also prevents moisture loss
from the wavy surface (110), providing further growth potential for
seeds and small plants.
93
CHAPTER V
CONCLUSIONS
concluded:
Ez
o
(5.1)
Evaporation rates given by Eq. (5.1) represent an upper bound for eva-
follow a logarithmic distribution and the results reveal that one can
z
q - q = q* R,n
z
(5.2)
s
om
94
surface are intimately related to the flow structure inside the wave-
Also the form drag coefficients of the leading fences were found to be in
height were found for the smallest fence spacings. However, it was not
REFERENCES
96
REFERENCES
24. Coles, D. (1956), ''The law of the wake in the turbulent boundary
layer," J. of Fluid Mech., 1, p. 191.
35. Goff, J. A. and S. Gratch (1948), "The humidity ratio (mixing ratio)
of moist air at saturation," Special Report, University of
Pennsylvania Thermodynamics Research Laboratory, Navy Contract
NOBS-2477.
46. von Karman, T. (1939), "The analogy between fluid friction and
heat transfer," Trans. American Society of Mechanical
Engineers, 61.
52. Kohler, H. (1950), "On evaporation from snow surfaces," Arkiv fUr
Geofysik. Bandl, nr. 8.
76. Nunner, W. (1956), "Heat transfer and pressure drop in rough tubes,"
V.D.I. Forsch. Series B, 22.
79. Pasquill, F. (1949a), "Eddy diffusion of water vapor and heat near
ground," Proceedings Royal Society, Vol. 198A.
84. Plate, E. J., (1965), "The drag on a smooth flat plate with a
fence immersed in its turbulent boundary layer," ASME, Paper
No. 64-FE-17.
102
93. Rossby, C. G. (1936), "On the momentum transfer at the sea surface,"
Papers in Physical Oceanography and Meteorology, !, No.3.
APPENDICES
105
APPENDIX A
(A-I)
the weight of water vapor associated with unit weight of dry air, and
fs Ps
qs = 0.62197 p-f P (A-2)
s s
spaced pressures p (in. Hg); the part of the chart that is pertinent
APPENDIX B
),,/2
D =f (-p w) sine ds , (B-1)
-),,/2
Fig. B-1.
Flow ... I
Wave Surface ~e /PW
~
Figure B-1
2'1TX
Y a cos -)..- (B-2)
ds = 1 + (2a7T)2 . 2 27TX dx
). SIn ).
(B-3)
sine =
),/2 0
2a7T
Pw sin 2~x dx] ,
27TX d
D =- 12), [ f
0
Pw sin -).- x + J
-),/2
(B-4)
in which ds is in inches.
Flow
To
A
1
Figure B-2
Perry et al. (82) used for evaluating wall shear stress in case of flow
shear stress T
0
is assumed to act along a surface BC positioned at a
preceding trough by the amount the free stream static pressure has
T
o
A
D
= (B-5)
2a 2a
Using Eq. (8-4) in Eq. (B-5), the effective skin friction coefficient
T
o
C (= 1 2) was determined. A difference of 2 to 3 percent was
f
"2 pU oo
cross checked with the shear plate measurements, again by using the
TABLES
Table 1
Performance characteristics of the Army Meteorological
and the Colorado State University Wind Tunnels
motor ~
o
Speed control: fine pitch control pitch control
3. Temperatures
Ambient air temperature soC to 9s°C. not control leg
Temp. of controlled boundary SoC to 205°C ambient to 95 C
4. Velocities
Mean velocities approx. 0 mps to 3"/ mps approx. 1 mps to
27 mps
Boundary layers up to SO cm up to 20 em
Turbulence level low (about 0.1 percent) low (about 0.5 percent)
S. Pressures adjustable gradients not controlled
6. Humidity controlled from approx. 20% to not controlled
80% relative humidity under
average ambient conditions.
III
Table 2
Effective average skin friction coefficient C (waves)
f
(Far downstream region)
U00
Wall Pressure Shear
(ft/sec) Integration Plate
Table 4
Comparison of Average Evaporation rates E/(q s -q) * obtained00
from Different Methods
(Far Downstream Region) (Upstream Waves Wet)
First Method Second Method
(Mass Balance Method) (Profile Method)
E = :x[!OpU(q-~)dZJ E = -pku* q*
u00
Direct
Waves ern/sec Measurements
Table 5
Data for Correlation (Re·)h vs (Sh)h
Table 6
Data for Correlation (Re.) 0 vs (Sh)o
----_._-_.
Table 7
Values of f [Goff and Gratch (1948)]
s
Atmospheric Pressure
TOp 23 in. Hg 24 in. Hg 25 in. Hg
FIGURES
1 12
'1
r=~dTurns r----
i 01.
Direction of Air Flow
26'
14'
, ,fil Test Section
16" Blade Axial Fan
Driven by 75 HP Motor
20'
~ II
~
~
II Two ~'":..-'" ~
~ II
#~
'-l
, ....... Screens "'"""
.',I
-
J ~ "'....
"'....
~,~'------- -------
6'
~./~ ::
I'
6'S"R
'-~ ,I
a'S"R
14'
L
,- 35' _l
N~
Plan View
7"
67 8 Evaporation Metering Wove
~ (Detail: see fig. 5)
~ Plank
jPans7\
, , ,
'\ ' \ Buffer waves
!
i
I
I
7 I"
'\ ~
I
I \
\
, ~18if121,,121,,121"12i'j~8~126L~,J
-·----20· II"-+-· 9' II~"
8
.I..;_~
to Automatic feed
arrangement
Plan View
Wind Tunnel
Floor
Tank
Tubing
Reservior
.. P-'
N
I-Il
Plasticine Filler
Felt
Water
Level
.1. 10 I • I• 1
16 Test Section Downstream - - - - - - - - - -
u I-"
N
N
>e I • I • L -
[
Bross tubing
.....
1--- __
--24 11 N
VI
I 42"
I.
~Od ...J~ screenS~:'!'\lo1?~~~I_
Thermocouples
Thermocouple lead ------ ~
. <i~:L .:
.~~-
Thermocouple ____ Copper channels : •••• __ .
- \ Side
piece
q
?- \
I
Mochine screws ~ ~
1
Aluminum block) Aluminum block--
(Dimensions for Wove A)
Elevation End View
Tube A
':>Clamps
,-- Clamp
Clamp
\ Burette
"Tubing
Thermocouple
~
N
0'.
Fig. 10. Part of automatic feed and metering system (typical for each wave section).
"LO.
Stainless
steel
10"
~
N
......:J
";'
z'
I"
(scale: 1;1)
Rectangular Angle Moisture Probe
Moisture Pro be
a b c d e
Up to
250'
Sample
•
In
115 Vae
D.C. Power In
Up to
Sample
•
In
115 Vae
D.C. Power In
Thermal
Boundary
Viscous ~-Layer , z =aT
Sublayer
x
______-1____L~~~____~>~>~>~>~>~j~. . . .~. .~,~>~>~>~.~>~>~>~------------------------------- ____
(Le> L)
.21
.19
~
-
LL
en
a. .17
-
~
...
• 15
.13
.1 1
2.9 3.0 3.1 ~2
12R
~ T x I 0 3, wo t t s /0 C
0.22
c:
....oo
::J
0-
w
0.20
CD
tn
c:
o
Q.
tn
•
CD
cr 0.18
•
•
."
....uCD •
•
:0
...
0.16 • Run
Run A
CD
a. • B
(I)
U)
.......
CD
CJ)
0.14
...
0
CD
.s::.
CJ)
••
", ..
~ 0.12
...
0.10
OJO 0.12 0.14 0.16 0.18 0.20
1/3
T ,Shear Stress Used for Calibration
1.3
1.2
1.1
1.0
WAVE A
Crests
0.9 -
o I
\l 3
0.8 o 6
6 12
x 18
0.7 o 24
Ze o 25
-
8 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 0.8 0.9 1.0
U
Uoo
1.4
1.3
1.2
1.1
1.0
WAVE B
0.9 Crests
t::. I
0.8 o 3
'V 6
0 12
0.7 x 18
Ze 0 23
o 25
8 0.6
0.5
0.4
0.3
0.2
0.1
1.4
1.3
1.2
1.1
1.0
0.9 WAVE C
Crests
D. I
0.8 o 6
'V 12
0.7 <> 18
Ze
o 23
o 25
8 0.6
0.5
0.4
0.3
0.2
0.1
00 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
U
UQ)
~2.2
~
<..)
I H
E
14 ~
~
91-
8t-
i ~
2
0r y-- 0
0
8
1.3 ~ 1.8
0
-~
Q)
,--...
E
I- 7
1.2 ~
<..) -
- 1.6
c:
E Q)
1.1 l-
-~ 6
~1.4I//
CIO
8*
I
Q)
::t
......
~
91- VI
10 (} 0"-
W •
8t- 0.8
:t ~t
.-'
0.4
0.2 a 20 40 60 80 100
False Roor • I ., Waves x (in.)
1.4 ~
':[ ~ 20~ 0
H
tJ')
0
Q)
r
1.3 l-
-
_8
tJ')
Q)
~
~
0
E
1.8 ~ 8
0
1.2 ~ o 7 ~
1.6
't-
E "0
0
~ Q)
6 ~
I.I~ 't-
"0
:l
tJ') 1.4 I 8*
Q) c
~ Q)
::s
H 1.0 ~ ~ 5 E 1.2
Q)
E -c::
1.0 ~ ~
~
09 ~ ,-- 4 Q) 0 8 tH
'-J
t -
c:: ~
0.5 L
2
IL
Oo6t-v---
I I
0.4 0 20 40 60 80 100
Fal se Floor • I • Waves x (i n.)
1.3~
-
---
~ 8 18 ~ 0
H
--- °8
~
u if 0
E U)
e
-
Q)
1.2~ 7 ~
u 1.6
\:J E
-
Q)
0
1.1 J- 5 6
U)
~
1.4
c -c
Q)
Q)
"-
H 1.0 ~ .§. 5 :3
U)
1.2
c
Q)
8*
0.9~ ~ 4~ - E
1.0
-
~ ~
8 .....
08~ 3~ .: 0.8
Cl>
• (.N
00
~
0.7 J- 21- ~ 0.6 ~
---- Wave C
~t
I~~
0.4
06t
05 0.2
0 20 40 60 80 100
False Floor • , • Waves x (in.)
-.004
-.008
I
Wave B
I
t:J) -.012
r
~
E dR
E -.016 -S!l 0.25x 10-5 ~Sl
dx In.
~
(.M
t -.020 I- 1..0
a..~
I
~ -.024
Fig. 24. Free stream static pressure distribution (waves) - typical case.
0.2
o.lsl I
/
"
.. Uoo
0.16 ~
,
I
//~
Wave
0 A } X = Floor
0.14 Jo- I 0
/:).
B
c Uoo ~ 20 ·fps
I
0.12 ~
I I
~
uoo ~ I 9
I II
0.1
~
+::-
0.08 ~ I I /:).
."",-- /:). A A
0
II n
0.06 .... II /
/
'"
0.04 .... ::::--.3//
" ---
0.02 r "
''--..dI
I I I I
~20
I I
0 20 40 60 80 100 120
False Floor · I- Waves x (in. )
0.9
0.8
0.7
u
Ua o I(-4)*
~ 1(0)
o I( I )
o I( 2)
~
'V I( 5) ~
~
0.4
• I(9)
• 1(i2)
• 1(13)
0.3
0.1
01 I I I I ! I
o 0.1 0.2 0.3 0.4 0.5 0.6 1.1
y/8
20~ _3
en
1
Q)
s::. I
u /
.s: 2
ISr-
~
~~ Cl)
--tr--------'" " ~
~
u .. .,,'"
c *
(.()
,-0-- - - --",....-
~
I ~----------- 1.5 ~
N
(.()
10 1.4
I 0
.....0---------- ..., 1.3
\
,
\
1.2
J:
51- \
"
h = I in. • I L; 12in. 10
O.
-6 -4 -2
II I
14
0.015 R- O. 258
L
....
u
0.007 ....~:-,-_-'-......Ii.....-""'--_ _ _ _ _ __ I
~
0.2 I I , I I I I I ,
002 004 O.Q6 008 Ql 0.2 0.4 0.6
Dimensionless Fence Height t
1.4
1.2
•\ • n/uco}
P. 25
Th
Crest. Wave C
• ;m/uoo
1.0 \.'-\
.
L i u , e t. a I. (I 966 )
a
\. \
- . _ . _ Corrsin Kistler (1954)
ze
08 ~\
8
'\\
- \\
06
.. \\\\ .\
.
0.4 .. ...\ '\\\
.. : \
0.2
.: :.\.\
0
0 004
, ~\, 008 012 016 020
n/Uc» , P/Uoo
1.2
08
ze/S
0.6
""
0.4
""
0.2
"\ •
0
0 0.2 0.4 0.6 Q8 1.0
T/Tw
18
16
2
O~~~~ ____~__~____~__- L__~L-__~__~ ____~__- i____i -__-L____L-__-L
2 3 45678 9 10 II 12 13
It • Dlstonce from Leading Fence (tt)
W,2{ ft~/secZ)
18
A'~'!'."'~
16
14
12
j 10
:: 8
f6
4
2
0
-4 6 7 8 9 10 12 13
ll. Distance ftom LeodinQ Fence (ft ) "
U'W'( ft~sec2)
18 l.....L...L.hJ
0 -2 -4
16 -
14
12
ilO
u
-~ 8
1:
2'6
f
4
2
0
-4 8 II 12 13
It • Dlstonce from Leodmg Fence (fI)
~
~, ,
"... ........
A
..... ,
,,
'~ "
002 " ''0..
" "
", '" ....
"~ ........ .......... -0..
........... _-*
0
0 02 0.4 0.6 0.8 1.0 1.2 1.4
z
-8-
6
o
0
4
en
Q)
,r:.
u
.S I
N I
3 I
o ~ I
~I
~I
2 vi
0
-;i
"'tl
o I
1
oI
01
J
0
0.5 1.0 1.5
- u'w' (ft. 2/ sec. 2 )
8
• ••
7
•
•
•
6 (\J I
-
~ (X)
cD • I
•
H
I • •
•
-
."
CD
5
-
m
H
(X)
~
cD
4
I
I
.s::.
0
c I • •
N
4 •
H <D
10
rt)
• • I
3
....
c
0
0
N..
" F~::sI N ::s..
::s \..--.-.J
•
••
l :.
•
(J) (5
.a
E
2
~
(J) •
~
•
•• •
•
12 3
UT
2
,
-.-.
UW
4
(ft 2 / sec 2)
5 6
9
• WAVE A
Crests
8 m\\
01
~ :~~
•o 6
3
• 18
6,
12
• 23
\l 24
~ 5. to "" " • 28
en
w
a:::
u
~~ ~~~
~
4 \ CJ1
~
~
0
a::: 3
LL.
~
(!)
2
w
::r:
oI ! ~! ! •• ~o! l'
.02 .04 .06 .08 .10 .12 .14 .16 .18 .20 .22 .24 .26 .28 .30
q -qoo
q. -qoo
~ 8~~
WAVE 8
Crests
o I
• 3
.. 7 o 6
..... • 12
en 6 18
w
a:
u 6 o• 23
24
~ • 25
oa: 5 \J 28
lL.
~ r\ ~~~
..-
'"
4 VI
N
W
:J: 3
00 .02 04 .06 .08 .10 .12 .14 .16 .18 .20 .22 .24 .26 28 .30 .32 .34
q -qoo
qs -qoo
9
• WAVE C
S Crests
(J)
w 01
o
z
7
fj.
o 12
6
. • 23
~ 6 • 24
w • 25
a:::
u 5
~
0
....
f!: 4 ~\ ~~
U'1
~
t-
~ 3
jjj
I
°0 .02 .04 .06 .OS .10 .12 .14 .16 .IS .20 .22 .24 .26 .2S .30 .32 .34 .36
q-qoo
q -q
s 00
Fig. 38. Mean humidity profiles at various longitudinal positions, wave C.
-- ·9
en
b81 AOO 8
E
0
~
- 7
lsl ~
a
Q)
~8T
. ~ 8 m
E
en
Q)
5
~
0
c:
4
o
~ 3~
Momentum B.L.
6 Temp. B.L. ......
U1
o Mass B.L. .J;:I.
..... 2
-1
cD
o" I I I I I I
o 20 40 60 80 100 120
dry false I _ wet waves x (inches)
floor - - , - -
Fig. 39. Development of mass, momentum &temperature boundary layers over waves
(typical case).
10
8
6~ WAVE
CRESTS
A
01 6 23
4tL ~ ~~ • 3 • 24
z o 6 \l 25
r-~ ~~
(in)
2
t, ~ ~
'" ~ ~~
• 18 • 28
~
U1
U1
o .02 .04 .06 .08 .10 .12 .14 .16 .18 .20 .22 .24 .26 .28
(q -~) I(qs -qCl))
{~)4~~ ~
0 12
•A
23
24
• 25
I-'
V1
-..J
I
.8
.6
o .02 .04 .06 08 .10 .12 .14 .16 .18 .20 .22 .24 .26 .28
(q- ~)/(~ -qoo)
24th -3
9J- 0 Crest, Zorn:: 0.89x 10 in.
Z
q-q =q ln zom
8
S •
-
( J1
00
6 ~'----------------------------~--------~--------~
1000 Z /zom 10,000
q-qs z
Fig. 43. Plots of vs (far downstream region), Wave A.
q* zorn
q-q
_s
q* 8
Wave B
th 2
o 25 Crest, zorn = 0.45 X 10- in.
~.L\ 0 0
o 28th Crest, zom=0.497 x 10- in.
2
7'
.I.
zorn
-
V1
1.0
5' <
100 / 1000 3000
Z zorn
q-qs z
Fig. 44. Plots of vs (far downstream region), Wave B.
q* zom
q -q
~9 Wave C
q. 3
o 23rd Crest, zom =0.58xI0- in. o
th -3
o 24 Crest, Zom =058x 10 in.
Z
q -qs :: q. in-
zom
......
7 0"-
o
I I
6 .~~~~~-------------~------
700 1000 10,000
Z/Zom
q-q
s vs z
Fig. 45. Plots of
z
(far downstream region), Wave C.
q* om
UCD
5.0
4.5 2in3
2ll ~
424 ~
, ~
: ' ._ _ L~
Positions
Wave A
3.5
3.0
en
C»
-5 2.5
E
1;; 2.0
C»
~
U
1.5
E
-
~
~
01
0i)
1.0
.....
0\
.....
:r 0.5
o
-0.5
-1.0
-1.5
20 0 5 10 20 30 o 5 10 20 30 0 5 10 20 30 0 5 10 20 o 5 10 20 30
q- qCID ....2
--Xlv
q,- qoe
Fig. 46. Vertical humidity profiles at various positions inside a wave trough, Wave A.,
Uoo
50
231 2 3 4 5 24
45
r
--23 irest
~ Positions
40 Wave 8
3.5
3.0
en
Q,)
-5 2.5
~
en 2.0
Q,)
~
u
E 1.5
.....
0\
-
0
~
.:E 1.0
0\
N
Q,)
I
0.5
-0.5
rd
23 crest list position 24th crest
-1.0 0 5 10 15 20 6 5 10 15 0 5 10 15 20 o 5 10 15 20 30 0 5 10 15 0 5 10 15 20 0 5 10 15 20 30
q-~ X 102
qs-qoo
Fig. 47. Vertical humidity profiles at various positions inside a wave trough, Wave B.
Uoo
5.0
4.5
4.0 23 I 2 3 24
~
LL_J.~J
Positions
3.5
Wave C
3.0
en
Q)
.r;
.~ 2.5
.-
II)
Q)
U
E .....
-
e
~
01
1.5 0"
(,M
.;u
l:
0.5
0
rd st 3,d position 24th crest
23 crest 1 position
-0.5
o 5 10 20 30 0 5 10 20 30 o 5 10 20 30 40 0 5 10 20 o 5 10 20 30
q-q__
CD X 102
q.-qoo
Fig. 48. Vertical humidity profiles at various positions inside a wave trough, wave C.
16~ Uco
1.4
\~\
1.2
1.0
23 I 2 3 424
08~
~Al
t::.
0.6 i I I I I
Positions
-;n
Q)
.&.
(J
£
0.4 t Wave A
Wave A
in 0.2
~ 0 23'd crest
u
st
E 0
0 1 position
~ (">
2 nd position
{:; 3'd position ~
E01 -0.2 0"-
iii t::. 4th position +:>.
J:
D 24th crest
-0.4
-0.6
-0.8
-1.0
-1.2
Fig. 49. Enlarged view of lower portion of humidity distributions in Fig. 46.
2.0
Uoo
1.8
1.6
1.4
1.2
1.0
0.8
UI
Q>
"5c: 06 231 2 3 4 5 24
~ 0.4
~ Positions ~
U Wave B 0\
VI
~
-
~
02
Wave B
~ 0
ii rd
J:
0
23 crest
1st
-0.2 r 0
0 nd
position
2 position
j rd
-04
6. 3 position
~ 4th position
D th
-0.6 5 position
0 24th crest
-0.8
Fig. 50. Enlarged view of lower portion of humidity distributions in Fig. 47.
2.6
UOD
2.4
2.2
2.0 Wave C
0 23rc1 crest
1.8 st
0 1 position
6- 2M position
16 3,d
0 position
\l 24th crest
1.4
Ii)
Q)
..c:
(J
.: 1.2
-
11'1 23 I 2 3 24
ct±:n
Q)
U 1.0
E ......
0 Positions 0\
.!: 0.8 Wave C 0\
.:E
CI
'c; 0.6
:I:
0.4
0.2
-0.2
-0.4'_1 _ _ ~ __ ~ _ _ '. _ _ '.... _'. _ _ '.-. -.~-. _!...-. -.!.. .. _.!.._ -.:.-. -..!.-. _.!. __ ~_ -...!. __ 2.-. ~-.
_ I.. ...
q-q.
q.- q.
Fig. 51. Enlarged view of lower portion of humidity distributions in Fig. 48.
167
- - - - - - Uoo
) x
-
u
en
Q)
~
8
E
0
~
lL.. 2
"'C
Q)
~
:3
en
0
Q)
~
en
Q)
~
0
c:
-
~
00)
0'
- - - - - . UQ)
q- qQ)) 2 Isoconcentration
( q _q x 10 Lines
S Q)
4 5
-....
en
Q)
3
~
u
12
E
0
~
lL.
"0
Q)
~
2
::3
en
0
Q) 15
:E
-cJ)
Q)
.s=
0
c: 20
....
.s=
22
.~
Q)
:I:
-I rd th
23 Crest 24 Crest
4----------~--------~~------------~----------~
Isoconcentration ---t-Ux,
Lines q -q.
2
( q _q ) X 10
s 00
~--------~-- 5
--
fn
Q)
~
u
E
0
~
~ 2
~
...
Q)
::J
fn
2
-
~
fn
Q)
.s::
0
.!:
..
-
.s::
C'
'Q)
I
rd 24th Crest
23 Crest
-3 • 25
z 4
(in)
I o s v 28
-18
2 ~ - ~
-
-...J
0
I •
.70 .74 .78 .82
(T -Ts )/(Too-Ts )
8.6
Uco
-
8.2
7.8
7.4
7
WAVE A 0
6.6 0 23 rd crest
6.2
•
0
lit
2nd
position
1\
•
• 3 rd
4th
1\
-
1\
5.8 6-
f/)
CI,)
.s::.
()
5.4
• 24th crest
.s
~
en 5
w
a::
u 4.6
~ Positions
a:: 4.2
0
u..
~ 3.8
J:
(!)
w 3.4
J:
2.6
2.2
1.8
1.4
'80 82 84 86 88 90 92 94 96 98
---~-,
Fig. 56. Temperature profiles (upper portion) -at various positions
inside a wave trough (Typical case).
1.6
Uco
1.4
1.2
1.0 WAVE A
rd
I:::. 23 crest
0 1st position 23 12 3 424
0.8
0.6 ••
0
2
3
nd
rd
4th
II
II
II
INJ1l Positions
I-
CJ) o ~
W '-l
a::: N
u -0.2
~
o
a::: -0.4
u..
I- -0.6
I
(9
W -0.8
I
-1.0
-1.2
- 1.4 I- ."
.56 .58 .60 .62 64 .66 .68 .70 .72 .74 .76 78 .80 .82 .84 .86 .88 .90
(T.-T )1 (Too-T.)
Fig. 57. Enlarged view of lower portion of temperature profiles in Fig. 56.
173
Isotherms
96
4
-
u
( /)
Q)
~
3
E
0
lL.
~
90
"'C
Q)
~
:::J
(/) 2
cQ)
~
(/)
Q) 85
.c::.
()
c::
-
.c::.
C'
Q)
I
-I
Wave A
rd th
23 Crest 24 Crest
If)
0
5
)(
~I~: 4. ~
-...,J
~
3
I / A \\ / / A. \\ /
2
11
1.7
I" 4.211 .1
6.1
UQ)
5.9
<> '" 10 ft./sec.
5.7
A-20
5.5
5.3
5.1
4.9
4.7
4.5
4.3
Q 4.1
)( 3.9 ~
'-I
wl~
3.7 V1
3.5
3.3
3.1
2.9
2.7
2.5
2.3
2.1
1.9
1.7
1.5
1.3
1.1
0.9
4.2"
0 4.1
)( 3.9
lLI
1y.8 3.5
3.7
cr 3.3 ~
'-J
3.1 0'1
2.9
2.7
2.5
2.3
2.1
1.9
1.7
1.5
1.3
1.1
0.9
~
~ ..-----
~
~
\~
( ( ..-
--..J
r:fJ
...... ~
-.20
~1~8
0.....
Q"
I N
~I,
_1-
-.30 ~
IT"'J7B"""""C~,,,,'),,,
-......J
1.0
/(:::~<:"
1.-(11 /t".
e~~:~[;~:~
.. 11
I.Iv
-.40
4.211 -I
0 3.8
)(
3.6
1 8 3.4
wf 3.2 ......
00
01" 3.0 o
2.8
2.6
2.4
2.2
2.0
1.8
1.6 ~1.7"
1.4
1.2
1.0
0.8
,. 4.2" --l
Uoo
q -qoox 102
-qoo
S
~
00
~
Crest -I Crest
/- u I
5.4
5.2
5.0
Uoo
4.8
4.6 E_ gm
4.4 cm 2 /sec
gm water
4.2 q -q -
s 00
gm dry air
4.0
3.8
If)
U ~ 20 fps } 0 wet upstream
3.6 oo
0 L 0 dry upstream
)(
3.4
Wl!~ 3.2 ~
00
3.0 N
2.8
2.6
2.4
2.2
2.0
4.2"
Fig. 68. Comparison of local mass coefficient distributions for wet and dry
upstream cases, (Typical case).
(
• •• • • •••
•• • •• •
1.2 J- • ••
• •
I
• • • •
• •••
••• •
• • • ••
• • • • •
• •••• •
• •• • •
•
• • •• • • • •
-
~
~
•
•••• • • ••• •
• •• • • • • • ••
ci
- • ~
)(
00
a:: tN
'"
H
UCD
• u... = 20 fPS} Wave A
• UCD = 40 fps
LJ Crest Crlest
1 -I I I
10 0.8 0.6 0.4 0.2 .0 0.2 0.4 06 08
x / ).
0.9
en
::0 • Uao = 20 fps Wave C
~
ci • • ......
"
a:: • •
00
+:.
•
H
M
•• • • •••
• • •• • • • • •••• • •
I • • • •
08
Criest
o 0.2 Q4 0.6
x/)..
1.8
L
r
.... 6P= P-Po
Jr·~
1.6
P = r2R 16T
1.4
1.2
~ ~T = Tf - To
L=6h Po = 15.45 mw I °C
1.0
0.8 6P= 2.16 mw 1°C
0.6 at x= -IOh
1.8
1.6 L=12h
u
0
..........
1.4
~ 1.2
E
1.0 ~
00
a.. 0.8 U1
<J _L __ l_~L
06
1.8
L=18h
x/h
Fig. 71. Wall heat-flux profiles (fences).
(
~p
o
L=6h Cp = t pUoo2
-02
~/J.../../ at x= 12 h
Cp= 0
O~8
Q.=>
IQ.. .....
00
Q. -IN
0\
II
CL
U
Q2
L = 18h
o
J~
xl h
Heated Area
Wave Surface
• •
10
If)
Q
)(
N
8 I-'
::::l 00
00
Q"
,
-IN
too
••
••
• • 20 fps ••• •
2 - U... •{.40 fps
Wave A
•
o,Crest
1.0 0.8 0.6 0.4 o Q2 0.4 06 0.8 1.0
xlA
• •
•
8
1ft
Q
)( I F Uao :: 20 f ps Wove C
N
~ 6 ....
_~
.......
I '. J \ / 00
!,O
...
4
2
.~ /. •
OICrest
1.0 0.8 0.6 0.4 0.2
x/>..
16
14
I~ WAVE A
12 1\\ .~ Crests
01
• 3
X 6
N 10 6 12
0
)(
+ 18
\l 23
8 o 24
=>I=>
-.
8 8 8 *A 25
28
C7
I 1 C7
I
--
C7 C7
6
I·~... ~~ \. 'l. ~
'D
0
0
0 2 3 4 5 6 7 8 9 10
ze ( Inches)
q-qco u
Fig. 76. Variation of - vs z over waves (typical case).
(q -q ) u c
s co
(
7
I
Uoo !:::: 20 fps WAVE A
6J- gm
I E-
"" cm 2 /sec
rt')
0 5 ~ I ~ q -q -
gm water
s 00 gm dry air
I "'-
)(
1
8 4
WIC1'
I
fn
C1'
3
2
I: -
.....
\.D
,,
O·
0 20 40 60 80 100 120
false
floor
---t- waves
x (inches)
6
Uoo ~ 20 fps
rt)
~J4~1
WAVE B
gm
E -
I en I cm 2 /sec o
CF I
gm water .......
q -q - ~
21-1 s 00 gm dry air N
o I~----~------~------
o 20 40 60 80 100 120
false
floor
~ waves x ( inches)
6. WAVE C
Uoo ~ 20 fps
gm
E-
cm2 /sec
gm water
rt)
0
><
4i q -q -
s 00 gm dry air
8
CT
......
WII (/)
CT I ,f ~
6 o 0 9
1.0
r."J
false
--+--
o I _ _~_-----=_--.JL-_
o waves 20 40 60 =-____
80 1-_ _ _- .
100 120
x (inches)
floor
100
o Wave A Jl r-..J
0.41
A
6 Wove B lL f'j
0.24
A
h ,...,
o Wave C
-X 0.12
J
I
en
oS
J:.
lJ.J 0
Q...
10
II
_..c.
J:.
en
h ~ 0.41
0 Wave A T
6 Wave B h ~ 0.26
A
10 0 Wave C l1.. ~ 0.12
A
'8
CT
I
o (/)
N .!:!
W 0
Q...
II
0.1
10
10
o h
Wave A ~ ~ 0.41
6 Wave 8 t ~ 0.24
o Wave C ~ ~ 0.12
• Water waves (Lai - 1969)
N1~
W~
o
<:a...
" .1
-ti
.s:::.
Cf)
.01
.001
10 100
Fig. 82. Correlation of water waves data with the data of this study.
197
100
o
-8
cr
I
• Flat plate (Cermak a Lin - 1955)
.... en
w S
0 1
Q..
•
.I
.01
.I 10 100 1000
(R)
u.zo
e. 0 = &I
Fig. 83. Correlation of flat plate and water waves data with the
data of this study.
198
10
• Wave A ~ ~0.41
• Wave B t~0.24
• Wave C t~ 0.12
•
(Sh )0
(SC)1/3
10 100
( Re. ) 0 = ~
11
h
0 Wave A ~ 0.41
1:
h
~ Wave B -X
~
0.24
0 Wave C h ~ 0.12
1-
10
- - - - Norris ( 1948)
- - - Kondo (1962)
-.0
N
cE
w?,
(/)
c-
o
Cl.
E Z0 U Zo 0.90
=0.055 ( + )
p D (qs-qamb)
10 100 1000
{Re.b = U v* Zo
Fig. 85. Comparison with Norris' (1948) and Kondo's (1962) results.
200
1
. . (T his stud y )
•0
Wave B, A = 42In., h = 1.0 m.
Cylinders, I in. dia. , 2 in. apart
rJ " ,0.31 in. dia., 1.88 in. apart (Chamberlain
~ Half Cylinder, 2 in. dia. , generators touching 1968)
•
6
A static wave like profile, A = 10 in. , h =I in .
Artificial Grass (3.15in.xO.2in.xo.OI in.) -(Chamberlain 1966)
30
20 0
~e
• • 0
0 0
8- 1 e e
.~
-
o 0
e
10
•
e
- !J
- - --
rl
-l'l
8 !I o 0
6 66
6
6
6 6 6 6
4
2 ~----~------~--~--~~------~----~~--~~~
10 20 40 60 80100 200 400 600 1000
-1
Fig. 86. Plot B vs for this study along with
v
Chamberlain's (1966 and 1968) data.
(
Flow Direction
N
o
~
0, I
'0"",
""'O/s ture Furrow Mode I
~ --+
A o A
-2" +"2
65.4 r T. (OF)
65.2
65.0 I - U ao
64.8
64.6
"
64.4
642
o
RUN A I
o
RUN B ~ N
o
64.0 6. RUN C ) Uao ~ 20fps N
63.8
636
!
634 lO"
WAVE B
63.2
630 ~------~~--------~--------~----------~-------~
~ - - - - - - 4.2" ---.------------.---.---
Fig. 88. Variation of surface temperature on a saturated wavy surface (Typical case).
(
8
Case n: • 30 fPS} _ It
Case m: • 45 fps L- 6
. ",. ,.,- __ h-.--.... 20%
,k''''''
6 /'
/ /
.. /
,//
z/h J I .,/-t
//e
/ .........-
-------...-
~ ~."., ~ /~ ___-------..-..-.----A. N
•~/.-.t--'"
.................. _,.....---'......
_..-_. _-+_--~--e---A- 50%
2
... ..--..-
_.Jt:.:_~_~~..-----e.--- .............. -~ 60%
- - . - --"--_.- - --- . - .~-.---- --- . .,.-
°0 ~ 8 IL2 x/L 16 1
20 . ---2'4-
I I -------. I -----. ----------r I T- --··r----r r-----r--,----,
o 20 40 60 80 100 120 140
x/h
Sr Case I: • 30 fPS} L = 12 II
.
I
z/h 4~ .J&
/"" y / ...
_ ---
~_L--'"
•
- ..... 30°1
I,
y
./
-r~""'-
-----' /0
N
o
2~I .....--+--
&-_.10. _..-~ . . . . - .-,-_-1Ir------- ....... -----.~--.-40%
..
+::0-
......... --=-- •
r-- __
I&'-'-
- 4 _ J . - - ..... -~
I ......:: ..... ----...-- . - ....... --.-.-.A--..---t- ...... 50%
o 20 40 60
I 80
I
xl h
100
I I
120
-,---,---r--
i40 160
8
Case E:
Case 3L:
30 f
45 fps
PS} L = 18 u
z/h 4L
,//.
//
I 30%
-------- -
//
.a--/¥ & - - . - - . - .....- - - - - N
o
.
U1
~----
2
_~_
• .._-----A------t-------
------.-
40°Yo
0'
_
_ _ .&-_
_
---1-- . .
, F e r c e Height I .
'
•
----=-=-==--==--=--~
In.
_I __------1-
50°";
0
- • -I I S------.J.
9
I I - T- - T ~-T -~T~-T -- I • - -,-~ --~--l
r- T T -.-----,
a 20 40 60 80 100 120 140 160
x/h
3 ----------____________________________________ 400/0------------
z/h 2~ 50%
N
0
CJ\
60%
70%
80%
0' II I I I I II
o 0.2 0.4 0.6 0.8 1.0
x/L
Fig. 92. Wind reduction factor between x = 10 ft and x = 10.5 ft fences and
between x = 10.5 ft and x = 11 ft fences.
(
3
- - - - - - - - - - - - - - - - - - - - - - - - 40%
z/h
2
tv
o
---------------------------------------50%------------- -....J
1 I
01 II ----1-_-----'.-2- 0.4
o 0 1.0
x/L
10%
3
(/)
Q)
.s::::;
..
u
c::
~
....
f/)
Q)
U
~ 21 ./- ~ ---A 20%
Q)
>
I~
0
~ : 25% N
0
00
Q)
>
0
.0
... • 30%
<l:
~ ~
-os
~
C'
&~ ~
:z:: 0
.J) 5 10 15 20 25 30
(
Wavelengths, A
Start of Waves
Fig. 94. Wind reduction factors (waves), wave A.
(
...
fJ)
fJ)
Q.)
3
~ _____ 10 %
U
Q)
>
0
~ 2
15 %
Q.)
fJ)
>
o Q)
~.s=
<{ 0
c: / ~ 20%
... t- ~
N
0
.s=
(7)
Q)
::I:
/
" !"O
~
10 15 20 25 30
Start of Waves
Wavelengths, A
(/)
+-
(/)
Q)
~
2
u 10%
Q,)
>
o N
~
~ ~ o
.s=
Q,) (,J
> c:: 15 %
0-
..Q
<{
-
. &;
0'
°cv
_ - - -•• 20%
::I: o Mean Surface Leve 1
to
~
15 20 25 30
Start of Waves
Wavelengths, ~
2.5
2.25
• Wave A
2.0
• Wave B
-( /)
Q)
1.75 • Wave C
lo-
u 1.5
Q)
>
0
..0
1.25
<!
-
.s:::.
C'
"0;
1.0
::I: 0.75
0.5
0.25
0
0 0.2 0.4 0.6 0.8 1.0 1.2
R : U.
f u. o
U)
- ( 1 - - X 100
U
o