BENEDICT
Fellow ASME.
J. S. WYLER
Mem. ASME.
Westinghouse Electric Corp.,
Steam Turbine Division,
Lester, Pa. 19113
In this paper, we first present a state of the art review of the published work concerning
theoretical nozzle discharge coefficients.
Then, we develop a new nozzle discharge
coefficient based on an axisymmetric
boundary layer solution, which in turn is based
on a new axisymmetric potential flow solution. These solutions apply to plenum inlet
installations which offer major advantages over conventional ASME pipe inlet installations as to losses, and as to predictability of the discharge coefficient at the higher Reynolds
numbers encountered in industry. Next, we present new correlations for static pressure
tap errors and apply these to the theoretical (zero lap size) discharge coefficients.
Finally,
we present new experimental data showing how a laminar boundary layer is preserved
and tap error is accordingly minimized for the case of a plenum inlet with
ASME
nozzle contour.
Introduction
6.92 RD-o-
(1)
0.184 RD-o-s
(2)
2/3 D
UJJ1 LLIIIIIIIIUIIIIII
Ulllll
and
CDT
SEPTEMBER
<>v
= 0
By1
1
p
dV
dS
~~ p
dT
dS
0(2 + Hn)
d_Z
ds
kA
dU/dS
TJ
U
pU*
= 0
(3)
dUJdS
+ z
- u
{Fl}
Fl
(4)
dS
NomenclatureA area
CD = theoretical discharge coefficient
(zero t a p size)
COM = measured discharge coefficient
d - pressure tap diameter
1) nozzle throat diameter
e = error in static pressure measurement
H = shape parameter, &**/d
Hu = shape parameter, 8*/0
I = electric current intensity
k = thermal conductivity
m ~ mass flow rate
N characteristic exponent
q = dynamic pressure, rate of heat
transfer
R radius
266 /
Vol. 100, S E P T E M B E R
1978
Rd*
RD
Re
Re*
S
=
=
=
=
finite difference
momentum thickness
kinematic viscosity
fluid density, electrical resistivity
wall shear stress
potential
Subscripts
1 = inlet
2 = throat
/, = laminar
T = turbulent
Superscripts
i
= ideal or inviscid
\:i-
0.8
0.6^
2-D RESULTS
j
/
0.4
L^
/
/ /
AXISYMMETRIC
RESULTS
0 = 0.43
n?n
-1.6
~'"H
-1.2
-0.8
Fig. 2
-0.4
S /D
0
S/D
0.4
0.8
1.2
1.6
2.0
nized when r/q = 0. Transition, by the Walz method, is recognized when log R$ exceeds a critical value defined by
(log Re*) = 2.42 + 24.2 (H -
1.572)
(8)
V,
Fig. 3 Axisymmetric potential solutions to low $ ASME nozzle
A, +
~zA Hi
Energy Balance.
equation is:
(U
i ++
as
AS
B.Fi
1 +
"ft
R7
M
H
dU dS
'
~u~
(F)
(/a)
' - n
(5)
energy
(6)
~z -
Bn F- ,
AS
(7)
W)
(tv)
D%) /
(9)
\Z> a /_
which is valid in a rigorous sense only in the region where onedimensional flow prevails. This procedure yields a new potential flow velocity distribution from which is obtained a new
boundary layer solution. Such corrections, and such repeated
computer runs are only important at the lower throat Reynolds
numbers (say below 1Q6), as shown in Fig. 4. Even then, the
effect of this iteration is to increase Co by only a slight amount
(on the order of 0.1 percent at l i e = 105).
Obtaining the Discharge Coefficient. Solving the boundary layer equations yields: 8*/D, 6/D, 8**/D, R, and r/q (this latter
being the dimensionless wall shear stress), at every point of
interest along the nozzle contour. There are several formulations available to define the discharge coefficient from this
information. T h a t given by Rivas-Shapiro is:
CD
8*
4 D
+ \D)
y u
8U
J.
(10)
= 1
5*
(11)
"-0- S)t
l -
46V/ft
" - 4 52*/ft
4 8 2 **/A!
(12)
11 is clear from (12) that the discharge coefficient will be influenced by the energy thickness (5**) as well as by the displacement thickness (5*).
-"K
^^
Boundary Layer Transition. A boundary layer, which is initially laminar for a plenum inlet nozzle installation, may, at some
location in the nozzle and at some throat Reynolds number,
undergo transition and become turbulent. The boundary layer
solution is, of course, strongly dependent on the choice made
from among the transition criteria, and these include:
1. No transition. That is, the laminar boundary layer may
persist throughout the nozzle at all fiVs. In this regard, Schlichting [13] states: In a region of decreasing pressure (accelerating flow) the boundary layer remains, generally speaking,
laminar."
TRANSITION BY SEVERE
LAMINAR SEPARATION
(BUBBLE FORMED)
lO 5
THROAT
lO6
REYNOLDS NO.,
10?
3. Transition at the Indifference Point. This earliest of transitions occurs when Re exceeds Rg*, and is identified as the Tollmien-Schlichting indifference point of (8). To his comment
t h a t the boundary layer in accelerating flow is generally laminar, Schlichting adds:
"even a small increase in pressure
causes almost immediate transition." The basic indifference
point is known to move towards higher Ro*'s as the wall roughness and/or free stream turbulence decreases. Walz [10] indicates that Re* increases by as much as 750 at very low turbulence levels. T h e turbulent boundary layer t h a t results from
transition at the indifference point will cause a CDT which
exceeds CDL for Rj> < 10 because such a boundary layer grows
at a smaller rate and is thus thinner than the corresponding
laminar layer at the nozzle throat.
3. Transition Because of Laminar Separation. Whenever r/q
~ 0, the laminar boundary layer can no longer prevail.
In its mildest form, the laminar separation will trigger transition to a turbulent boundary layer in much the same manner as
indicated for the indifference point transition. However, transition by laminar separation usually takes place later than by
indifference. In its most severe form, a separation bubble may
grow and cause a thickening of the resulting turbulent boundary
layer. This has the immediate effect of decreasing CDT with
respect to CDL4. Flat Plate Transition. According to flat plate theory, transition occurs when R = 470. However, this zero pressure
gradient value has little to recommend it for the sharp pressure
gradient flow found in ASME nozzles.
I t is t h e existence of these various transition possibilities
that introduces such large uncertainties in theoretical discharge
coefficients. Fig. 5 describes, by schematic curves, the effect
of the various transition criteria on the discharge coefficient.
One must rely on experimental data to determine which of the
transition criteria applies for a particular nozzle profile.
s /o
Fig. 6 ASME nozzle displacement thickness versus contour position
Vol. 100, S E P T E M B E R
1978
RIVAS-SHAPIRO
LAMINAR
S/ D
Fig. 7 ASME nozzle momentum thickness versus contour position
AUTHORS'
TURBULENT-2-;^--~~~"~\^ s ^^AUTH0RS' LAMINAR
/"^
^-^'f^
/
RIVAS^ ^ ^ ^ " b
,,-SHAPIRO ^ = ^
/
^.--^
COTTON-WESTCOTT
A//X
/r^
/
/
- H A L L ( F L A T PLATE)
LEUTHEUSER'S DATA
J
/?
THROAT
I06
REYNOLDS NO.,
107
^TURBULENT
_FLAT_ PLATE
"TURBULENT
Fig. 9
= L
5.25
RD~- ;
(13)
Authors
S/ D
A S M E nozzle shape factor versus contour position
(14)
0.106 R D --w
(15)
Cotton-Westcott
I05
THROAT
I06
REYNOLDS NO.,
I07
I0 B
R0
1.00
1
AUTHORS' LAMINAR
THEORY ^-___
_____
~
99
LLT
UJ
o
o
UJ
5.98X
CO
Q
SOUNDRANAYAGAM
2 " NOZZLE
0 4"NOZZLE
#
.97-
I05
THROAT REYNOLDS NO., Rp
10
= / (A*)
y.
(17)
/(R_*)
(18)
where
R_* =
V*d/v
etf/pv*
= ( - jCR,.*)2
and
=
(16)
T<22/>2 =
(R<i*)2
(19)
&
A,
A
-X
D
.04
.08
.12
HOLE
RAYLE(I949),L/D=4.2
"
, L / D = 15.6
"
,L/D=270
RAYLEII959)
SHAW ( I 9 6 0 )
FRANKLIN a WALLACE (1970)
.16
SIZE, J
.20
.24
.28
(INCHES)
400
129 ( p * ) 0.555
800
1200
600
2800
Fig. 14 Static tap error as a function of wail shear stress and tap
Reynolds number
'
7.0AUTHORS ,(21)
to describe static tap error in the range, 0 < Rd* < 385. In
the fully turbulent region, we have correlated all the available
data by the new expression.
j g
6.5-
2000 2400
j *
6.0-
lc
5.5-
A
-
(21)
0.269 ( R d * ) o-sss
g^4)'
CD +
CD3
1/
RAYLE
X SHAW
O RAINBIRO
D
4.0-
FRANKLIN a WALLACE _
3.5-
5.0
5.5
LOG T
Fig. 15 Static tap error as a function of wall shear stress and tap
Reynolds number
Here, CD and (r/q\ are specified by the boundary layer analyses, (13), (14), and Fig. 10 respectively, and the quantity
(e/rh is defined by the correlations of (20) and (21), depending
on the tap Reynolds number,
=
jr/q
'P
\ RB
(24)
(22)
For most practical cases, et/qi, < <C 2 /<J 2 and can be neglectedFor a plenum inlet, (22) reduces to
CDM =
4.5-
Rd*
Co
1 - /3<
5.0-
RAJARATNAM,
(20)
,
(23)
1978,
Vol.
100 /
271
TAPS
R
1/8" THROAT
ASME
PLENUM
INLET
TAPS
NOZZLE
PERFORATED PLATE
5 0 % SOLIDITY
AUTHORS'
-
THROAT
REYNOLDS
NO.,
a plenum
The /3 Effect
Actually, there are two general problems of interest in applying boundary layer solutions to flow nozzles. The one, plenum
inlet installations, has been dealt with at some length in this
paper. The pipe inlet installation, on the other hand, has been
all but ignored in this presentation, even though it is the most
conventional of applications. This is because no rigorous boundary layer analysis is available at this time. For one thing, the
inlet flow is rotational, and this is counter to t h e boundary
layer concept of an irrotational core flow. For another, and
possibly more important reason, there is a flow separation in
the vicinity of the corner formed between the inlet pipe and the
nozzle face. This separation is to be expected, but it has so
far precluded further boundary layer solution. T h e existence
of this separation strongly suggests the use of a cubic or double
cubic nozzle in place of the elliptical ASME nozzle, as has
often been suggested in the literature, [26, 27, 28, 29].
Several supposed analytical /3 solutions are available in the
literature, [8, 30, 31, 32], but these are in direct opposition to
each other. Although we believe t h a t our CD equation (12A) is
applicable to both pipe and plenum installations, the pertinent
values of S* and 8** must be available before theoretical predictions can bo made. As we have seen, boundary layer solutions
are available for plenum inlets, but we know of no such solutions for (3 flows.
CI
(%)
1.75
.9888
.0007
4.303
.17
3.01
.9914
4.6
.9920
.0018
2.770
.23
5.2
.9912
.0013
4.303
.32
5.6
.9934
.0016
2.447
.15
6.6
.9955
.0008
4.303
.20
7.0
.9958
.0002
4.303
.05
7.4
.9944
.0005
12.706
.45
7.9
.9946
.0012
12.706
1.08
8.4
.9948
.0006
12.706
.54
8.9
.9940
.0014
2.571
.15
9.6
,9950
Vol. 100, S E P T E M B E R
N-l,p
DATA
BOUNDARY LAYER
THEORY
THEORY ADJUSTED
FOR STATIC TAP
EFFECTS(d/D=0.06)
272/
fc
(xio-S)
I I N C H ^ 2 . 5 4 CM
Fig. IS
A2
2g pAp Vi
1 - /3*
(25)
where the significant pressure drop across the nozzle was determined from about 25 readings taken during each weighing. A
special 120 inch Statham U-tube manometer, with digital readout to 0.001 inch_mercury under water, was used to provide
Ap. A typical Ap had a standard deviation of 0.023 and a 2
sigma confidence interval of 0 . 0 1 inch Hg under H 2 0 .
Fig. 17 presents our experimental results in terms of CD,
the best value of CD at a given R j . The confidence interval
defined by
CI
(26)
Summary
1. We have given potential flow velocity distributions for
ASME nozzles for /3's of 0, 0.43, and 0.6, which all show an overacceleration and then a diffusion near the intersection of the
contraction section and the cylindrical throat section of the
nozzle.
2. We have obtained and described in some detail new
boundary layer solutions to an ASME nozzle with a plenum
inlet. These show the possibility of wide variations in Ci> depending on whether the boundary layer remains laminar, undergoes an indifference point transition to turbulent, or develops
a laminar separation bubble.
3. We have derived a new Co formulation which accounts
for energy thickness in the boundary layer as well as the conventional displacement thickness. Significant differences in Co
are to lie expected when using the more complete analysis as
compared with t h e conventional.
4. We have obtained new static tap error correlations-for
several ranges of t a p Reynolds numbers, and used these in
conjunction with our boundary layer solutions of r/q to predict
these effects on measured discharge coellicients.
5. Our work has been compared favorably with that of
others, where possible, concerning: potential flow velocities,
boundary layer parameters, and static t a p errors.
6. Finally, we have presented results of now experimental
studies of a plenum inlet ASMF. nozzle in water, and these
confirm our theoretical predictions.
Acknowledgments
We would like to express appreciation to our colleagues:
Kddie Phelts, for help in resolving problems encountered in
developing t h e boundary layer computer program; Dr. Steve
Bennett, for providing the WECAN potential solutions; and to
Norm Deming and Les Southall, for suggesting and supporting
this study. Vic Head, Chairman of Fluid Meters Research
Committee SCS-WG1, offered many helpful suggestions during
the review of this paper.
16 Ray, A. K., "On the Effect of Orifice Size on Static Pressure Reading a t Different Reynolds Numbers," Translation in
ARC Rep. T P 498, Nov. 1956. (See also ARC F M 2479).
17 Shaw, R., "The Influence of Hole Dimensions on Static
Pressure Measurements," / . Fluid Mech., Vol. 7, P a r t 4, Apr.
1960, p. 550.
18 Jackson, J. D., "A Note on the Relationship Between
Static Hole Error and Velocity Distribution in the Boundary
Layer," App. Sci. Res., Section A, Vol. I I , 1962, p . 218.
19 Rajaratnain, N., "A Note on the Static Hole Error Problems," J. 'Roy. Aero. Soc, Vol. 70, Feb. 1960, p . 370.
20 Rainbird, W. J., "Errors in Measurement of Mean Static
Pressure of a Moving Fluid Due to Pressure Holes," Quart. Bull.,
Div. Mech. Engrg., Nat, Aero. Est., N R C Rept. D M E / N A E ,
1907 (3).
21 Duffy, J., and Norbury, J. F., "The Measurement of Skin
Friction in Pressure Gradients using a Static Hole Pair," Proc.
Inst. Mech. Enqrg., Vol. 182, Part 3H, 1967-1968, p . 76.
22 Franklin, It. E., and Wallace, J. M., "Absolute Measurement of Static-Hole Error using Flush Tranducers," / . Fluid
Mech., Vol. 42, Part 1, 1970, p . 33.
23 Zogg, H., and Thomann, H., "Errors in Static Pressure
Measurements Due to Protruding Pressure Taps," J. Fluid
Mech., Vol. 54, Part 3, 1972, p. 489.
24 Preston, J. II., "The Determination of Turbulent Skin
Friction by Means of Pitot Tubes," / . Roy. Aero. Soc, Vol. 58,
1954 p. 109.
25 Wyler, J. S., and Benedict, R. P., "Comparisons Between
Throat and Pipe Wall T a p Nozzles," ASME ./. Engrg. Power,
Oct, 1975, p. 569.
26 Rouse, II., and Hassan, M. M., "Cavitation-Free Inlets
and Contractions," Mechanical Engineering, Mar. 1949, p . 213.
27 Redding, T. H., "Flow Characteristics of Metering Nozzles," The Engineer, July 1963, p . 129.
28 -Morel, T., "Comprehensive Design of Axisymmetric Wind
Tunnel Contractions," ASME J. FLUIDS ENQRG., June 1975, p .
225.
29 Hussain, A. K. M. F., and Ramjee, V., "Effects of t h e
Axisymmetric Contraction Shape of Incompressible Turbulent
Flow," A S M E JOURNAL OF F L U I D S ENGINEERING, M a r . 1976, p .
References
1 Fluid Meters - Their Theory and Application, Report of
the ASME Research Committee on Fluid Meters, o'th edition,
1971, Bean, II. S., ed.
2 Shapiro, A. H., and Smith, R. 1)., "Friction Coellicients
in the Length of Smooth, Round Tubes," NACA T N 1785, Nov.
1948
3 Rivas, M. A., and Shapiro, A. II., "On the Theory of
Discharge Coellicients for Rounded-Entrance Flowmeters and
Venturis," TRANS. ASME, Apr. 1956, p . 489.
4 Hall, G. W., "Application of Boundary Layer Theory to
Explain Some Nozzle and Venturi Flow Peculiarities," Proc.
Instn. Mech. Engrs., Vol. 173, No. 36, 1959, p. 837.
5 Cotton, K. C., and Westcott, J. C , "Throat T a p Nozzles
Used for Accurate Flow Measurements," ASME / . of Engrq.
or Power, Oct, 1960, p . 247.
6 Leutheuser, II. J., "Flow Nozzles with Zero Beta R a t i o , "
ASME J. Basic Enqrg., Sept, 1964, p . 538.
7 Soundranayagam, S., "An Investigation into the Performance of Two ISA Metering Nozzles of Finite and Zero Area
Ratio," ASME J. Basic Enqrg., June 1965, p . 525.
8 Cotton, K. C , Carcich, J. A., and Schofield, P., "Experience with Throat-Tap Nozzles for Accurate Flow Measurement,"
ASME J. Engrg. Power, Apr. 1972, p . 133.
9 Benedict, R. P., and Wyler, J. S., "A Generalized Discharge Coefficient for Differential Pressure Type Fluid Meters,"
ASME J. Engrg. Power, Oct. 1974, p . 440.
10 Walz, A., Boundary Layers of Flow and Temperature,
(translated from the German by H. J. Oser), the M.I.T. Press,
1969.
11 Benedict, R. P., "Analog Simulation," Electro-Technology
Science and Engineering Series No. 60, Dec. 1963, p . 73.
12 Benedict, R. P., and Meyer, C. A., "Electrolytic Tank
Analog for Studying Fluid Flow Fields Within Turbomachinery,"
ASME Paper 57-A-120, Dec. 1957.
13 Sehlichting, II., Boundary Layer Theory (English translation by J. Kestin) McGraw-Hill, N. Y., 1955.
14 Hall, G. W., "Analytical Determination of the Discharge
Characteristics of Cvlindrical-Tube Orifices," /. Mech. Engrq.
Science, Vol. 5, No. 1, 1963, p. 91.
15 Itayle, It. E., "An Investigation of the Influence of Orifice
Geometry on Static Pressure Measurement," MS thesis, M.I.T.,
1949. See also, "Influence of Orifice Geometry on Static Pressure
Measurement," ASME Paper 59-A-234, Dec! 1959.
58.
30 Hall, G. W., Discussion of "An Experimental Investigation of the Flow in a Classical Venturimeter," by D. Lindley,
Proc. Inst. Mech. Engrs., Vol. 184, Part 1, No. 8, 1969-70, p. 147.
31 Au, S. B., "The Prediction of Axisymmetric Turbulent
Boundary Layer in Conical Nozzles," ASME J. App. Mech.,
Mar. 1974, p. 20.
32 "Steam Turbines," A N S I / A S M E P T C 6-1976, An ASME
Performance Test Code, 1976, p . 30.
33 Benedict, It, P., Fundamentals of Temperature, Pressure,
and Flow Measurements, 2nd Edition, Wiley-Interscience, April
1977, p. 191.
APPENDIX
= (Ui/Ui +
iy\
Bz
An = (Ui/U + i)F3
Bn
Fi = 2 + Ar + (1 + Ar) ff
F, = (1 + A) a
F3 = 1 -
Ha
NNT
- aH
= 1
= 1
= 0.2317ff - 0.2644 - 87000 (2 - ff)*>
NT
= 0.268
aL
= 1.7261 (H -
aT
= 0.03894 (H - 1.515)0-7
1.51.5) 0 - 7158
1.515)-
where
fl>7
2 = 0 2?^
.symmetric
"" 1/(1 + A')
i? =
/UK)
Tidr-
D \RD
5*
PTTRHP
ir# =
Pl
/'Pi - p A
\
1 - 2
/ pTrIWUjs/2g,
Actual
u)
= KE,
KE,
52*
1 - 2 52*/%
APPENDIX
Discharge Coefficient
Discharge Coefficient
ideally
CD
-.
(i -
2 -1 - 2 ^ - 2 ^
/?*) ( 1 - 2
s.yii,)
I * ( i - 2 ^ v
i l - 2
(~-~&^_-
(12a)
Si*/BJ
1 - 2
h*
1 - 2 d2*/Ri
1 - 2 5 2 */2 -
(12)
2 52**/-R
Plenum
and
0 = (fl,/i) =
U\/U\
/Pi - pA
\
/idea!
2(7=
-(1 - 0)
Aetually
d,*/B2)U2
/32
'1 - 2 W '
i - 2 5i*//ei
r/i
and
o a:
1 _ ._- j ; . . _ d .
DISCUSSION
D, R. Keyset
The authors are to be congratulated on taking yet another
step forward in improving the understanding of the flow coefficients of flow nozzles generally. Especially enlightening is their
inclusion and treatment of the effect of the pressure taps on the
flow measurement. I t has been a recurring weakness in the field
of head class fluid metering to concentrate on the nozzle or orifice
itself, and ignore or circumvent the effects of other parts -of the
measurement system, such as the pressure taps. In this paper
J R-S
KE
( (KE)dm
(KE)drh
J AA
2(7c
= 27r
f* tihdr
2
9' J 0
^9c
CD
f ->+ r
[//
:- dr
i?
V o l . 100, S E P T E M B E R
1978
\i-2hvru)
274 /
If it is in-
'
R-i
/2TrpRm\
\
'Note that the loss term has been neglected in this formulation.
cluded, the resulting discharge coefficient becomes