FoLro
T47
~
CE'2-ci•E,q-37
41'·2.
CALIBRATION OF FLOW NOZZLES
.USING WATER WITH EXTRAPOLATION TO STEAM
by
Albert G. Mercer
for
Westinghouse Elect
ric Corporation
Engineering Research Center Civil Engineering Department Colorado State University Fort Collins, Colorado 80521
CERB-69AGM37
f
1
lI
ENGINEERING
RESEARCH CENTER
Civil Engineering DepartmentColorado State University Fort Collins, Colorado 80521
CAL
IBRATION OF FLOW
NOZZLES USING WATER WITH EXTRAPOLATION TO STEN IPr
epared
by Albert G.M
ercerJune 1969
for
We
stinghouse Electric
C~rporation Plant
Apparatus DivisionBo
x
1047Pittsburgh, Pa. 15230
CER68-69AG~l37
CALIBRATION
OF
FLOW NOZZLESUSI1
G WATER WITHEXTR.A..POLATION TO STEAM
Introduction and
Summary
This
report presents
theresults
ofcalibration tests performed
on twosteam flow nozzles at the
EngineeringResearch Center of
ColoradoState University in
accordance with WestinghouseElectric
CorporationPurchase
Order56-IC-46472-G dated April 22, 1969. The
aim oft he calibration
wasto relate the stea m flow rate through the nozzles to
thedifference in pressure sensed at the nozzle taps. The
calibrationw as performe d ,vi th
waterin place of steam
andthe results
of thecalibration
werecorre ct ed for ste am analytically. The
primary
results for
waterare presented in Table II as a set of data
relatingdischarg e coefficient to Reynold's number.
The
Reynold's numbers for the
tests withwater
are lowerthan
thosethat will be experienced
withsteam.
Figure4 shows the
empiricalcurve determined for the extrapolation of the data to the
higherReynold's numbers
.The calibration data for the two nozzles were so close that one
curvefits both sets.
The
relationship between pressure differeilce and steam flow
. givenin Table III and Figure 5
were obtainedfrom
theempirical
curveof Figure 4
withanalytically
determinedcorrections for the
thermal propertiesof steam.
The · calibration
of
Nozzle lwas performed May 8, 1969 and the
calibrationof
Nozzle2 was done on
May17,
1969.The
calibration datawas collect ed by
Mike Wittington, andthe data reduction
wasdone
by IkramUl-1-iaque
, both graduatestud ents in Enginee ring.
Theor etical Conside r ations
The basic differential equation for th e flo w of a compressib l e fluid i n a pipe of varying cr oss section
(asfor example a flow nozzl e) . with negligib l e h eat lo st to the pipe wa lls, is
2
Jdu + d(
pv)
+d(~)
= 0 (1)2g
where: u is th e intern a l energy
pis th e absolute pressure
V
is th e sp ecific volume
V
is
the m ean velocity across the pip e
gis th e accelerati on of gravity and
J
is th e conversion fro m mechani ca l en ergy to heat
.The t erms of Equation
1r epre s ent ch anges in int erna l energy, flo w
workand kin etic energy in th 2.t order. The middle t erm of Equat ion 1
can
be divided into t w o parts to give
2Jd U + p
d
V + Vd
p +de~) 0
2g =The first t w o t erms of Equa tion 2 appea r in th e defi nition of entropy s
JTds =
Jdu
+pdv
(2)
(3)
The
two t erms on th e 1eft in Equation 3 are conside r ed t herma l t erms
and th e two t erms on th e r ight are conside red m echanical t erms .
Friction al ways causes a
loss of energy from th e me chanica l terms
to th e th
ermal t erms
.Th ere is no proc ess in a flow
-conduitthat
can cause energy t o flow in th e r everse dir ection.
3
For frictionl
essflow, Equation 2 becomes
2 d d(y_) = O
V p + 2g (4)
Under
conditions of incompressibility vis constant and Equ
ation 4 canbe integrated to give
(S)
Equation S can be expressed in ter
msof the fl01v rate
F,the cross sectiona l areas A
1 and A
2 , and a calibration coe fficient Kin the form
(6)
U
nder ideal conditions, the coefficient K would be unity but in practice it is somewhat less. The major cause for this difference is the bound
arylayer w hich has the effect of making the noz
zle seemsmaller than it is (displacemen t thickness). Since the boundary
layerthickness is less with higher flo
ws, the coefficient character- istically approaches unity as
.the flow is increased. The boundary layer thickness can be shown to be a function of the Reynold's number R
· definedby
e
VD
\) (7)
where D is th
epipe diameter and
vis the kinematic viscosity
.It foll
•::iws that
Kis basic
ally a function of
R •e
The pri
marycalibration usin
g wat
erobt
ainedrelat
edva lues of K and
R •e
As
long as
the expansion of steam is not great the
boundary layer thickness is no
tmaterially affe ct
edand the K - R
e rel
a tionship
obtained for water willbe app
licableto st eam as we
ll. Once the
ideal f l
owr ate pr essur e
difference
relati
onship is obtained for
steamthe
coeffici ent K
canbe
applied.
To obt
ain
the pr essure differenc e for ste
am,Equat ion
4mus
t be integ=ated with t he constraint that
the entropy of the steam
be~onstant
. Integration of t he
left hand t erm. cannot be
performed unlessthe p-v
relati
onshipis kno1m . However th e m ean value
the
oremfor int egra ls state s that there is a
constant k betwee n
zeroand
1such th at
v2 v2
1 2
z:g- - 2g
(9)w
here flv is
the change, in specific volume
. Then to get an expression
parallel
toEquation
6we
can writeEquation
9as
F = -
- - - - - - --- --
(10)- /
_ (v+6v)2Al A2
If 6v is
sma
llenough
the assumption can
be ma de that v
varieslinearly with p meaning that k
= 0.5.If flv is
too large
the nozz
lecan be
treated in steps with each step hav ing a sma ll enough
change in areathat the assump tion of
lineari
ty
is good.
To
solve Equat ion
10, the follo
l'lin g st eps can be performed·:
Step 1. Set K = 1
St
ep 2.
Compute
vcorre
spondingto p
1 and
the percent
dryness x
1 fro
m steamtabl es
Step
3.Set ~v
= 0j
I
I·1
i
6
Step
4.
ComputeP2
St
ep 5. Comput
eth
epercent dryness
x2for zero entropy
changef
romsteam tables.
Step 6. Compute
twfor P2 and
x2Step 7. Rep
eatsteps
4through 6 to convergence
Step 8.Correct
Ffor co
efficient
K.D
escription of Nozzles
The geo
metry of the noz
zlesare shown in Figure 1. The nozzles when received had identical m
arkings:
1 s. o. #3395-1
i Mil -S2319
4AC/S
HT
#213060
To
distin
guishbetween them th
eywere stamped 1 and 2 on the outside face of th
I epressure
manifold. ~licrometer r
eadingof ~1
ethro
atdiam- eter averaged
I4.516 inches at room t
emperature (about 20°C)for both
nozzles. The upstr
eampipe and t
apwas furnished by CSU and th
ediame ter at the t
apwas measured as 6
.081 inches
.Under 1000 psia and 544°F it is estimated that th
edi
~1eter would expand by 0.34 percent which would
meanan increase in flmv rat
eof 0.68 percent.
This expansion was
allo\'Jed for in the computations .
D
escription of th
eTest Facility
·
Th
etests were
performedat the
m.:iinflow calibration facilities
of th
eEnginccr ine Rese
arch Center whi
charc shown diagram
matically in
Figure 2. The
nozzles were insta ll ed
in the
12-inch lineas shown in
Figure 3.
To find th
e flo w rate, water
from
the nozz
le is
divertedinto
volumetrictanks
fora sp ecified period of
time meas
ured and
controll
ed by a puls e coun ter op er a ting
off the AC power line
. The volume of w a ter collec t ed in
the tank is measured by a prec ision hook gauge . Th e w ater temp er a ture is m easur ed
on samp
les
colle cted from the flo
·1div ert er. The pres sur e difference was meas ured by two
separa t e sys
tems s e
t upespecia
lly for th ese test s (Mercury manometers usually
usedw ere not a
llowed und er the sp ecific ations). For high
flows, aset
ofthree differenti al manometers each 12
feet
long with acetyl ene t e trabromide
(S.G.2.95) as th e indicating
fluid were pro- vided. Th ese we r e con ne ct ed to th e no zz
le
taps in seri es with each
other
. The
tubeshad
large bor es
(9/16-
inch) for accurate meniscus
form
ati
on. · Forlow flows, an
inverted air-water manometer was
used
.
for
, larger manometer de
flections
.I
Thepropos al for this study st a t ed
that open wel
lhook
_guages would b e
used.for lowflo
ws but ci1e se proved
tob e impract ic al. The air-water manometer is
not quit e as accurate as th e
open well system was predicted
to be and
,as a
resu
lt, the data for
the lowest flo ws
showthe most scatte r .
D
es
cripti on
of the Test Proc edure
Each of
the flo w no zz l es were calibrat ed for 15 di s ch arges
, adhering as c
lose as possible to th e quantities outlined in the
pre-calibration data of Tab
le I. Th e only differen ce was
the
- - __
..._..___
"""'"_.,.._ ....Table I. Pre-calibration Data for Calibration of Nestinghouse Steam Flow Nozzles
Run No. 1 2 3 4 5
.6 7 8
910 11 12 13 14 15
Flow rate
(gpm) 450 675 900 1125 1350 1575 1800 2025 2250 2475 2700 2925 3150 3375 3600 Run time
(min) 10.0 7.0 5.0 4.0 3.5 3.0 2.5 2.5 2.0 2.0 1.75 1.75 1.5 1.5 1.5 Sample
(gals) 4500 4725 4500 4500 4725 4725 4500 5062 4500 4950 4725 5120 4725 5060 5400 Volume error
per rdg. (gals) 2 2 2 2 2 2 2 2 2 2 2 2 2 2
2Number of
readings 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
Percent volume
\error
.09 .08 .08 .09 .08 .08 .09 .08 .09 .08 .08 .08 .08 .08 .07 Percent time
error
. 0 0 .00 .oo .01 .01 .02 .03 .03 .04 .04 .05 .05 .06 .06 .06 (Ac curacy of 60 cps)
Manomet. er system I Hook gauges 1 manometer
.2 manometers 3 man.
Total
Man.
Def. (ft) 0.87 2.0 3.5 5.5 4.0 5.5 7. 2 · 9.1 11.2 14.1 16.9 19;0 22.1
.25. 2 28.8Error per
rdg. (
ft) .001 .001 .001 .001 .005 .005 .005 .005 .005 .005 .005 .005 .005 .005 .005 Number of
readings (ft) 2 2 2 2 2 2 2 2 2 4 4 4 4 6 6
Percent pres-
sure error
.22.10 .06 .04 .25 .20 .14 .11 .09 .14
.12.10 .09 .12 .10
substitution of an inverted air-\'/ater manometer for the hook guage
. .
arra_
ngement indicated for Runs
1 through4. The
reading error for
thisman:)meter nominally would be . 005 feet, as for the oth
ers·, but the long r n
time for these runs allm1•ed rep
eated readings for higher accuracy.
The step-by-step procedure adopted was as follows:
St
ep
1.With
the pumps operating and with water recirculating
throughthe nozzle but bypass ing the volume tric tank,
the valves were adjusted to obtain the desired flo
wrate as indicated by the precalcul
ated manomet
er deflections of Table I.St
ep 2. The follow ing data was record
ed:a. the tem
pera ture of flowing water
b. the tempe ratur
eof aiD
c. th
e temperatur
eof w
aterbled from m
anomet
erlines
d. the level of \'later in the volume tric tanks
e. the
manome t
er levels.
For
some runs the manomete r lines were bled to check for air accumulation and the manometers were short circuit
ed to checkthe
nullbalance.
Step
3. The timer was started which automatically div
erts t}:leflow into the volume tric
tanks,St
ep
4. Manometer
levels were measured repeatedly durin
g the run
to getas many readings as time allowed.
St
ep
5. The tim
er wasstopped which automa tically div
erts the flow to the sump .
Step
6. St
ep2 was
repeat
ed.10
St
ep 7. The volumetr ic tanks
were emptiedby pumpi_
nginto the
sump and pre li minary cal cul ations
were madeto catch
gross errors. Severa l r uns were r epeat ed on the bas is t hat th er e were gross
errors but only those for \vhich th e
source of the
errorcould be id en tified
.E
valuation of Syst emati c
ErrorsThe
calibrati on
equipment
hasbeen designed to m inimiz e or
correct for
most causes of s ystem at ic
error.Th e sourc es of
possible systemati c error
are:1.
calibr ati on of th e vol umetric tank
2.
vari ations i n th e 60 cycl e A C
current fo r th e timer
"3. non-uniformity
of th e flow
_div erting proc es s
4.
variations
ofth e densi ty
andvi scosi ty of th e flo wing
\-lat er
5.variation
oft he s pecific we i gh t
ofth e fl uid
int he
manomet
erand t
he connect ing
lin es
6.
ina ccuracy of the
manomet er sca l es
7.men iscus
effects8.
local vari at ion of the
accelerationof grav ity
9.
bouyancy of t he flu ids du e to loc a l barom etri c
pressure 10.tank leakage
andevaporation.
The evaluation
andtreatm
entof th ese error sourc es we r e
asfollo ws:
1.
Volu me tric t ank ca lib ration
Thet anks
haveb een ca l ibrated periodically by filling
th em with water in
240 poundincr ements
weigh ed on
asp ecia l scal e ca libra t ed b y
weights checkedby th e
National Burehu
of
Standardsin
Boulder, Colorado.The depth
volwe
_re
lationshipis
knownaccurate to
atle ast ±2
gallons overth
e6000 ga llon r ange
.2.
Timer variati ons The accuracy of the timer depends on the accuracy of the 60 cycle A.C. power supply. The power supply has been checked periodically against th e gov ernmen t time signal from Station WWV in Fort Collins. The average error experienced has varied from
.09 percent for a period of 1 minute to .0003 percent for 10 minutes.
3.
Div ersion proc
essThe
· time r equired to complete th e divers ion from th e sump to the volumetric tank, or vice-versa, is less than t wo second s . The equipment is designed so that the amount of flo w diverted vari es lin ear ly with time during th at int erval .· If it is perfectly linear, th ere is no error introduced by th e diversion process. How- ever, there is no way to evaluate this possible source of error except
•to .
see if there is a systematic error associated with sampling ti me and none h as been detected.
4.
Density and viscosity of the flo
wing water The density and viscosi ty of water was de t ermined fro m th e water t emperature using stand ard t ab l es* . Temperatures \v ere r ead with a precision thermome ter to 0.5° F. This perm itted density to be computed to
.02 percent and
I
viscosi ty to ~bout 1,0 percent. This accuracy for visco sity is acceptab le since th e discharge co e ffic ient was found to vary about .01 pe~c ent for a one perc ent change in viscosity. Incidently , th e mass density is the import an t prop erty of the flo wing fluid rather than the specific weight as f ar as th e flow no zz le is concerned .
5. Specific weight of manometer fluid The specific weight of the manometer fluid was dete rmined from t
emper:i turer eadings of th e
* Meyer, C.A.
,~!cClintock, R. 13
., Siverstri, G .J. and Spencer, R.G. JR.,
Thermodyn
amic and Tr
ansport Propertiesof Steam,
AS~IEResearch Committee
on the Pro perties of Steam.
AS~lE,Un ited Engineering Center, New York,
1967.12
s
urrounding air
.The r el ationship of
the sp
ecific weightof the
acetylene tetrabromide to t empera ture was evaluated from pycnometer
determinations.6.
Manomete rs scales The m anometer scales ,-,ere
standard st ee l
meas
uringtap es accurate to at
least
.002 feet over ten f eet
.7.
Meniscus effects
Thecapillar ity associat
ed withthe meniscus is idea
lly balanc ed out in a
differentia
lm
anometer.Never-
the
less,large
tub
edi ameters (9/16-lnch) were used and the tubes were
thoroughly cleaned and fill ed with fresh
manometer fluid.
8.
Acceleration of gravity
The var iation of g
over th e
United St~tes is 0.2 percent
. InColorado the value differs about
.04 percent
fro m a value of 980.0 cm /se c
2:
9.
Bouyancy of air
Thebouy ancy of th
ea
tmosphere affects
the specific weightof ivater by about 0
.13 percen t. The difference
in this
bouyancy atFort
Collins (Elev 5000) and at sea level is about
.02 percent
.10.
Tank le akage and evapora tion Th e amounts of water lost due
to tank leakag~ and evaporation are
complet e ly negligible.
Evalu
ation of R andom Errors
R
andom errors are
those tha
tcause scatter of the data. They arise mainly fro
m thedetermination of the volume of
the sample and the
deflectionof th
emanometer.
The
volume of w
at
er collect ed in t he vo
lume tric tanks is
determi ned by me
asuring
the height of th e wat
ersurface with a hook
guagc befor
eand aft
er th e run
.Th e hook
·guagc has a
vernier whichreads t o
0.001feet, equiva l
entto about
0.8ga llo ns of t ank
capacity. However , it is di fficul t to repeat r eadings to this accura cy and
±2gallons is a reasonabl
eestima t
eof th e maximum r andom error for each meas urement. T ab le I gives f or each run the r esulting percen t error in volume expect
edfr om each of t he two mea
surementsr equi r
ed.Th
e manomet
er deflection is determined by
measuring the eleva t ion of
the top of both menis cuses r e lative t o the t ape scale. Th e tap es are gr aduat ed in hundr
edths of a foot andr eadings are estimated to thousandths of a foot. However
,flu ctuation of th e
meniscu s level and error s of para ll ax make
.005f ee t a r easonab le maximum error for
manometer readings. The l ast r ow of Table I shows th
eexpected
.
percent error i n pressure f or this order of accuracy. Thes
efigures are r easonab l e exc
eptfor th e first fou r run 6 which are based o"n ho ok guages readable to
.001i nch
.Without hook guages , the p
ercent
pressu re error can be expected to b e about five ti
mesthe v2
.lue
given. Itshoul d be remembered t hat th
esquare root of pressure is
onlyone ha lf as much as fo r pressure itself
.Furthermore, the mano meter 4eflections r ecorded were th
eaverage of a number of readings for th ese first runs , which should i
mprovethe accuracy
..
Results for Wat er
The va lue of K and R obt ain
ed forth e
15runs for each fl ow
e
nozzl e are giv en in Tab l e II. The va l ues are also shown plotted in Figure 4. The scatter of t he data is emphasized by th e larg
escale chosen for K. The data for th e t wo no
zzl es were s o clo se tha t on
ecurve could be fi tt
eclt o both sets . Th e curve chosen was th
eexponc1:tial
14
curve discussed earlier. The coefficients A and B were chosen to produce a least squares fit. With A = 0. 35 and B = 0 .45, the data of Nozzle 1 fitted with a root-mean-square error of 0.34 percent and Nozzle 2 with an error of 0.32 percent.
Table II
Discharge Coefficients and Reynold's Numbers for the Flow Nozzles
Nozzle 1 Nozzle 2
K R K R
e e
.9681 1.94(105
) .9778 2.51(105
)
.9686 3.33 .9656 3.64
.9761 4.38 .9746 4.86
. 9.712 5.41 .9736 6.15
.9670 6.37 .9742 7.28
.9740 7.37 .9770 8.47
.9696 8.11 .9770 9. 79
.9700 9.08 .9761 11. 23
.9736 9.91 .9773 12.42
.9770 11. 83 .9788 14.08
.9809 13.10 .9802 15.83
.9788 14.06 .9816 16.95
.9810 15.03 .9831 18 .10
.9778 15.99 .9824 19.15
.9801 16. 89 .9818 19.97
~
Results for Steam
Pressure differences and flow rates are related in
Table
IIIfor
dry saturatedst
eam.Th
et
able covers
flowrates from
30,000to
300,000 pounds of
st
eamper hour
,with
pipe linepressures
varying from 500 to 1000psia.
Thepressure
differ
ence
values weredeterminc.d
by computer
usingth
estep-by-st
epproc
edureoutlin
ed earlier.Pressure differences assuming incompressibility arc
shown
inbrackets
in Table III forcomparison.
Thesteam prop
erti
es includingviscosity were
obtainedfrom
tablesin
theASME
publicat1onr
eferr
edto
earlier. Key values of Table IIIarc plo_tt
ed inFigure
5 to provide a more convenient form for use.Of course,
th
e steamin
thepip
e line maynot be in
adry
saturationcondition.
Table IVsh0\•1 s
1 forcomp aris
on,pres sur
e dif- ferencesfor steam ,-1h ich
is 99 percentdry
.. .
C)LOAADn STATE UNIVfRS!TY
STEA14 QJAL1TY 100 PERCENT ORY - - - -- - - --'-T_A~RULATEO VALJES ARE PRESSURE OIFFI tN PSI FOR ~Tr.AM
B~ACKETEn VALJES ARE PRESSURE I..,--?ST ASSU14I~G t~r.o.,.PREss-i~tLlrv PIPE PAF:SSURE
PSIA
- -- - _ -2.0J) - -·. -
(
560
_ _ _ ...:,6(10
FLOw RATEi TN ll'OUNOS Pl:R ~nUR
30000 60000 9ooon 120000 l'iOOOO 180000 210000 2i.onno 300000
,i.o5 1,617 ;1~-5~ -.,57?. 10,419 15 294 21,327 2A,7n6 37,7n• 48,736 --- -· ·--
;i.0111 1,6011 3,604> 6,396T. -9,979T 14,355> 19,5211 1 25,477> 32,2n1 1 39,759;·
• 389 1,553 3,512 6, 30? 9,979 14,621; 20. 351 27, 1l 4• 35, 7111 45,956 ,388) 1,544) 3,463) 6,145) 9,58A) 13,792) ( 1 ~. 7<;5) ( 24,478) 30,9<;91 18,200) ,3111 1,49ft 3,376 6,05?. 9,57?. 11;.001 19,1tc;4 2-..050 33,9A1 43,521 ,373) t l,41'!51 I 3,332) ( S,912) I 9,2251 (13,270) ( \R,045) I 23,"i5ll ( 29,781'!) 36,754) .. • J.60_ _ _ l_._4..3_0 _ ~-3..Zll _ ___
s---~- °--- -
..9_...il L _ _ u .. 4.!t.O . _l 13 ._~J 6 ___ 24_190 0 _ 32, )Q J 41 •. 3 '._l9_,360) ( 1,4311). 3,210) ( 5,b9t,) ( A,81Hl ( 12,7R4) ( \7,3A4) ( 22,,r,A9) ( 2~,&97) 35.408)
,347 l,3A7 3,131 5,605 A,849 12,917 t7,BR4 23,A'i2 ]0,9c:;6 39,384
,3471 l,3!10) 3,096) 5,4941 8,572> 12,3'301 16,767) I 21,RA3) 27,6781 I 34,151>
,335 1,339 3,021 5,405 R,526 12,433 17.190 _.?2,_f:!!38 __ 2..? .. ~4...3 ___ 3.L .6.17 t ,3351 I 1,3331 .( 2,9901 I 5,3051 I l'l,277) I 11,90!',·l ( \6,191) I 21,1311 26,72,r,) I 32,9771 - - - -~c.""'2"'"'0,___ _ _ __ _.32~ ____ l..291t __ . .2.21.a ___ s...z...Le. .... ___ 8L2.26 ______ l.L.9..B.3 ___ -1.b...S.SJ ___ 22.o.01t-_ 2~.1;41 36~o'J.5
I ,3241 t 1,2~8) I 2,890) ( 5,1281 ( l'l,000) I 11,501'!1 I JS,649) I 20,425) I 25,8]3) 31,R74l
,314 1,251 2,822 S,043 7,945 ll,564 15,955 21,lAS 27,344 3/t,553
,3131 t 1,246) I 2.796) 4,9611 7,7411 I 11,1141 I 15,1411 19,71,ll I 24,9Q4) 30,8391 ,304 1,211 2,731 4,879 7.682 11,174 15,402 2n,4?7 2&,3;>9 33,211 l't60
.3031 I 1,207) I 2,708) 4,804) ( 7,496) I 10,782) ( 14,662) I 19,136) I 24,204) I 29,864)
- - - - -"'~B-'-n _ _ __ _ ____,,,..,2,_9'--'--4 ___ _..!..!.ll..4 _ __ 2 646 .~ .! 7?_5 ___ ___ 7.~:3~-- __ 1_9.._8QJ_ ___ 4 8':J_~-- _l9..t22l ____ ]!.?-'38~ _ __3l ,9!7 ________ _ ,29,.> I l.170) I 2,624) 4,6St,l I 7,265) I 10.450) I 14,211> I 18,5471 I 23,4S8) I ?8,943)
_ _ _ _ 1=0=0 _ _ _ ___ ,285 1. JB 2,s&s ,~579_ 1,201 101461 j,1391 19,051 z4,So6 30,024
~ 851 1,135) 2,545) 4,516) 7,047) 10,116) 13,784) 17,990) 2'-2.753) 28,0741
- -_ !. 2 7 7 _ ·-- ! • 1 0 5 _ 2 • 4_ 8 9 4, 4 4 I 6, 9 !:12 l O , 1 3 6 l 3, 9 3 9 l 8, 4 15 2 3 • & f!J 2 9 , 7 4 8 I ,277> l,l'lll 2,471) 4.3841 6,840) 9.819! \3.JAO) 17,463) 22,0Abl ( 27,252) _ _ _ _ _ ,~4._c.ll _ _ _ __ _ ,_~6.'!_ . . . 1.on 2.417 4,311 6,774 9,830 1:t.So9 l7,A"i2 22.._90~ _28J45 __
.269> 1,070) 2,400) 4,259) 6,&451 9,5581 12.9971 ln,9631 21,4S51 I 2&,4721
,262 1,043 2,348 4,188 ,r,,57A 9,541 13,104 17,104• 22,l'l7 27,811
,2611 1,0401 2,333> 4,1401 6,459> 9,2QJ> 12.614> 16,4A9> 20,sss, ?5,732>
,254 1,014 2,283 4,071 6,392 9,267 12,721 ll't,7R8 21,509 2&.937
- _7]JO ___ _
,254) t 1,011 l 2,269) 4,026) 6,282) 9.036) 12,288) 16,018) . ( 20,284) 25,028) _ _ _ _ ...,_,A_,..n..,,o _ __ _____ .~48___ L91}7 2.221 3~9~0 !a216 9_.Qo~ 12,)59 l"',300 _ 2.0.~10 .. 26.,ll3
1 .2471 ( ,9A4) 2,208) 3,919) ,r,.114) R.7951 11,959) 15,,r,09) 19,74?) 24,3581 -!l..2.._l'i __ - ---
.. _ ri_4 f)
. 880,_
91)0
'140 .
,241 ,961 2,163 3,854 l't,048 8,761 12,016 .l5,A40 20,2(18 25,341
,2411 I ,9591 ?.,1511 3,81,;1 r;.954) 8,5Ml ll,t.46) 15,1991 19,2241 23,719) ,235 ,91& 2,106 3,753 'i,B!:IR 8.5?.7 ll ,,r,R9 15,402 19,695 24,608 ,235) t ,9341 2,095) 3,718) 5,BO!l B,3441 11,346) .( 14,AOBl 18,710) 23.1101 ___ _ ,Z2J ___ ,912 2,053 3,b57 5,73£, B.303 11.379 · 14,9~5 19,152 23,914 ( ,22'i) ( ,910) 2,042) 3.b24) c;.655) 11,113) 11.060) ( 14,435) }8.25A) 22,5271
,223 ,890 2,002 3,565 5,590 A,OQ(l \l,0112 14,51'19 }8,&35 23.25S ,223) ,BAl3l 1,99?) 3,534) S,515> 7.9121 l0,7A7l l4,n7Bl 17,801,) 21,970)
,218 ,868 1,953 3,477 5,451 7,887 \0,800 14,?ll !B,145 22,630
1 .2101 ,8661 1,9441 3,4491 1 s,301, 7,7401 10,525> 11.1111 17,31111 21.,H>
____ _._2_1_3 ,847 1,906 _3,393 __ 5,318 _ 7,693 10,531 l1,A52 17,&7B 22,036
I ,2131 ( ,846) ( l,A97) 3,367) ( 5,253) 7,555) \0.274) 11,41'191 lb,96nl 20,926)
.,.?QB __ .827 __ 1,861 3,313 ·- 'i.19! . 7,5n7 10,273 13,"iOB 17,2)2 21,471
, 2 O 8 l I -;a 26 1· I l • B 5 3 l 3, 2 8 B I I c; , l 3 O l 7 • 3 7Q l I O , 014 l l 1 , n 9 & l \ti, 5 6 1 l 2 0 • 4 3 7 l
96p _ ,203_ _ ,808 1,IH8 3,235 . 'i,069 7,32Q 10,!l27 13,\AO !6,Bn7 20.932
,203) I ,807> I 1,810) I 3,212) I 5,0121 I 7,2091 9,'l03l 1?,7941 I lb,l'l?l 19,9661 - - - ·- - '9'-'8"-'0'--- - - - - __ . .JJ~. ____ _..19 0 ____ l...1I.6. ____ 3 -!. 61 _ ____ 4 • 952 _ _7 • l 5A _ 9 1.791 . l? • A~S . 1bt400 20 • 41 7 . ( , l 9 8 l I , 7 A 9) ( I • 7 6 9) I 3, 13 9 l I 4 , 8 9 Bl I - 7 • 0 A f,) 9 , 5 Fl l l l?, c:; O 4·1 ( I 5 • B I ~ l \ 9, 514) ,l94 _ ___ L77~ .. _ l .. 736 _ _ 3,09p ___ 4,839 t,,994 9,564 12,564· 16,010 19,923 ,1941 I ,771) 1,730) I 3,069) I 4,789) 6,8AQ) 9.3671 l;>,;>;>61 15,111,3) 19,n79)
_t
ooo
Table III