1/ AD—A049 050 NAVAL INTELLIGENCE SIPPORT CENTER WASHINGTON D C TM——ETC FIG 13/11

WA TER—JET pu. OcSiGn (RASCHET V000STRUYN060 EZHEKTORA) • (U)

N

S. I--
NOV 77 A S TSYGAM(OV
UNCLASSIFIED NISC—TRANS—397 7 .l lI

END
DATE
FILICO

2 —78
DAt
 (7
Pr — ;
~ ~ ~~~~~~~~
¶

~DEPAR~~~~~~O F T ~~~NAVY

CLASSIFICATION : UNCLASSIFIED
~
t ~~~~
ITLE :
~~~~~~~
7
E:
E EE ±0
~
9
7
.
.

~~~~~~~~~~~~~~~~~~~~~~~~ A. S.

>JAGES : ~~~—~ 11

J,,iI
S tstl 9, 1969
~~~ SOURCE~~~~~~~~~ ~~ ~~
IYeLNO.

C.3
( ~~S~R) ?~~ i$-17
7 ~
‘
•~~ ~~~ kJRIGINAL LANGUAGE : Ru ssi n

TRANSLATOR : LI)

APP OV~~) f. I f<

~
MCESSlO~I t ~ DATE
j j

•
~~~ £ ctIsu
“f D D C
•
~~~~~~~~~~~~~~~~~~~
4
U JAN 24 1918
SISThJBJTlO /AVA ,LUllITT C5OU
~
DISTRIBUTION STATEME ~~~.~~
• Approved for public releose
L I LU U L.S
Distribution Unlimited D

_ _ _ _

_ _ _ _ _ _ _ _ _ _ _ _ - - - -• . - -~~~~
.
.

. .
 - -
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -. -.

WATER—JET EJECTOR DES I GN

Water-Jet ejectors are coimnonly used aboard sh i ps as drainage systems.

This article proposes a fairly simple and reliable method of calculating
H highly efficient water-Jet ejectors.
The procedure described here makes use of energy balance formulas
cited by P.1I. Kamenev In[Gidr~elevatOry I Drug iye Stroynyye Apparaty
(Water—Jet Pumps and Other Jet—type systems), Nashstroyizdat Publisher ,
1950]. The formulas deal with the theory of anal ysis of similar high-
efficiency equi pment. The data derived from this theory show good agreement
with the results of bench tests conducted on a number of shipboard ejectors.
The calculation of a wate~-Jet ejector consists basi cally of determining
the necessary head of the operati ng water, both ahead and after the ejector,
the water flow rate, the ejection factor, as well as all principa l dimensions.
The calculations are based on the law of conservation of momentum due to
impact , Jet continuity equations and velocity/head relationships for liquid
discharge from the nozzles.
Shipboard water-Jet ejectors are generall y desi gned with inlet (suction)
pipes positioned at an angle 60°) to the ejector axis , this being the
~ j”
most appropriate arrangement in a drainage system . Sometimes, however, this
pipe is set at an angle of~ p90 . in the latter case, the entire Jet
kinetic energy is wasted , which makes this type of ejector rather Inefficient .
The basic energy losses In the ejector occur when the operating and
ejectable stream are mixed In the mixing chamber. An ejector will show
maximum efficiency only when Its design will assure th. east hydraulic

1
PT- —• -
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ _ _ _ _

• I

losses and , consequently, maximum pressure. Efficiency increases considerably

• when pressure Increases not only in the diffusor but along the length of the
m ixhg chamber as well.
In order to reduce hydraulic losses during inf low to the mixing chamber , it
Is desirable that the angle between the ejectable flow and the ejector
axis be m inim al ; furthermore, it Is important that the flow rate at the

• mixhg chamber inlet (outlet end of nozzle) be most advantageous, that the
diffusor be designed with a minimum aperture at the beginn i ng and that
this aperture be gradually expanded toward the diffusor outlet .
The velocity head ahead of the ejector nozzle depends on the position
of the operating water feed source (pump, force line) and the vacuum in
the m xing chamber. If the ejector Is positioned on a level different from
that of the source, then the head ahead of the nozzle must be determined
by taking into account the geometrical height equal to the length from the
• ejector axis to that of the qerating water source (plus or minus); here
one must subtract from the head the local resistances and friction losses
on the given section of the force line .
Design Formulas
To simplify the cal culation without significantly affecting Its fina l
results , the following assumptions were made : the specific weight of the

• operating and ejectable waters, as well as the mixture of both, are

2
assumed to be same and equal PPuuI l000kg/cIn ; in some formulas , specific weight
was not taken into account due to a decrease in ’~ or the expression of the head -

in meters of water column . in computing the operating water velocity at the

exit from the nozzle, the effect of its velocity head ahead of the nozzle
and the pressure of th.ejectable liquid was not taken into account. The
• friction losses and local resistances In both th. intake and delive ry lines

_ _ _

- - - -
 -
~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

were not specified ; therefore, both terms “suction head” and “delivery
head” are understood here to mean their total resistances , i.e., the
geometrical height plus total losses that are to be taken Into account
when calculating or selecting an ejector design.
With the ejector in a steady-state operating mode, the water flow from
the nozzle is forced Into the mixing chamber , the vacuum In wh i ch Is deter-
mined (in water column)as

where H Is the suction head

2 of the ejectable flow ,
m wat .col;
V 2 is the velocity of ejectable
-
flow at the mixing chamber
inlet (outlet nozzle end), m/sec;
g”9.81 is gravity acceleration .
For an ejector operating in the suction mode, one should consider
that the sum of all resistances on the intake line added to the velocity

H head of the drawn—in flow at the beginning of the mixing chamber must be
much below atmospheric pressure: H <lOin wat.col .
For properly designed ejectors the geometric suction head Is as
high as H2—7m wat.col.
The pressure head grad ient of the operating water in the nozzle is
wrrtten as

i I, II,
~~ —.H, — ~~ _(H~_. p.), (
~)

where is the pressure head of operating water ahead of the nozzle ,

m wat.col.
The operating water flow rate, on leaving the nozzle (rn/eec) is written
as

________
~-•-~ ~~~~~~~~~~~~~~~~~~~~~~~
 •

•
e
—

v,~~~ 3 ’2g (H 14 H,) , (2)

•

where~~ Is the nozzle efficiency taken to be 0.95 to 0.97 for properly

machined nozzles.
For an ejector operating In immersed condition , the magnitude H2
(ignoring local losses) may be considered that of the water rai se height
that will aØpear with a minus SIgn in the expressions for Hvac and H1.
The most advantageous ejectable flow rate In the beg i nning of the mixing
chamber Is achieved when the specific weights of tI~te liqu i ds entering the
ejector (lyi*~~~ ) are equal , the cross section of the mixing chamber (of
cylindrica l shape) is constant and the inlet angles between the operati ng
water and the ejectable flow are equal to zero 21 —C(2—0), i.e., their
supply lines arranged along the ejector axis.

I ,

‘l’ Figure 1. Efficiency of Cast Diffusers

-

0,6 •
t1.4 -

-
-~~~~~~~ .
-
5g w
• . . • ~~~~~
‘SO 130
5g9
i •
•

For ejectors perforinlnq in the high suction mode (vacuum) generally

1~
________
- ~~~~~~~- - —- - -~~~~•.-
-- ~ -~-,~~--,,~~• --- - - - • • _
‘- • • -, ~~.
- - -,, - _ —~~ -,
~-~-. -r ---—-,w W
~
-
r~~. • ‘—w• — -
~~~

•
fitted with conical mixing chambers and ejectable water lines at an
angle to the ejector axis , the assumed flow rate will always be lower than
•

the most advantageous one.

In selecti ng the optimum ejectable flow rate, one must bear In mind that
increas i ng it will lower the impact losses due to mixing the flows and ,
to a certain extent , justify the energy expended for producing It; yet ,
the selected high flow rate may not satisfy the equation Hvac<lOm wat.col.,
and thus will have to be reduced .
In order to attain the optimum ejectable flow velocity in the beginn i ng
of the mixing chamber, ejectors are often designed with flap nozzles.
When the first model is bench-tested , the most efficient performance of the
of the ejector is used for determining the most advantageous distance from
the outlet end of the nozzle to the beginn i ng of the diffusor throat
which represents both a specific cross section of the mixi ng chamber at
the outlet end of the nozzle and the optimum velocity .
There are three cases of flows entering the mixing chamber:

;~~o3; v~ > v3; Us,

~
• where v’ -- is the averaged velocity of the mixture of both
3 flows In the beginning of the mixing chamber , (2)
determined from the momentum equation (2), rn/eec;

——
-F

• v Is the mIxture flow rate in the narrow cross section

3 (throat) of the diffuser,rn/sec
In the first case, the hydrodynamic pressure along the mixing chamber
-: renia Ins constant, In the second——Increases and In the third——decreases .
The momentum before and after mixing the flows with least mixing
losses may be expressed In the following form.

~ 5

• -
-• - •— • -

~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~ • • •~~• •.

 p-~ _ ..
• ~~~~~

• •

O,v,+ O,v,c a Os, , (3)

~
where G 1, G and G3 are the amounts of the operating and
ejectable w~ters and then mixture , m 3/
~
a--is the inlet angle of the ejectable flow into the mixing
chamber.
The ejection factor or the ratio of the amount of ejectable water to
• the consumption of operating w~ter:
0,
*

• Introducing ~ into equation (

3) and reducing it by G 1, we obtain

()
~
or -

Where -fl. is the ratio of the ejectable flow veloc i ty to the averaged flow
rate of the mixture at the begi nn i ng of mixing .
With the ejector performing In a steady-state mode, the energy balance
with allowance for losses, may be reprinted by the following equations:
for case 1 , when v ’ —v
3 3

a,u, — o , [~~~~~
i — (I I :+ ~~~T~) J G,Ff . +Gj1 +
~
v! (vi)’ ’ (v ) 2

for case 2, when v ’ )v

3 3
0,01,—G, ~~~~~~~~~
— (H,-f-~~ )] ,+ G,H3 -f
G,11

(v
I~~ t1+ a
~~~~~
j ~j
Tr o ’~ ~j
\ . O,(V; v 3 ) . . 0 t4
0S — ~~ T3/i j - -~~xj .--- r.* s j 1a.
~

• for case 3, when v ’3(

v3 it is the same as in case 1 plu s C is.

~~~~~~ _.•~ ••_ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~ ••-- -~ ...- - - • —•—~•~r-~ -•— ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

 —
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~

•
I
—

In these expressions H and H3 are given as H~ ink g4’m 2 with specific

2
weights introduced in kg/rn3 for a general case of an ejector using various
liquids . The addend values enclosed in parentheses are the total losses
caused by mixing the flows in the ejector.
Since frict i onal losses in the mIx ing chamber cannot be determined
prior to i dentifying the dimensions of both the chamber and the diffusor
throat , they are arbitrarily categorized as losses In the diffusor , as
•
-

a result of wh i ch the latter ’s efficiency Is lower than usual.

The total head m wat.col . developed by the ejector for case 1 (v’3.’v3)
will be found from the energy balance
,‘ 2 _.2
~
~‘2 (6)
•
~~~~~~

for case 2
(7)
for case 3 (v (
v ):
3 3

(8)
• H$+H
(v ~~~~~ v~
*4

where H2-—is the suction head

(with allowance for losses), m wat.col ;
H3——i s the delivery head of
the ejector, m wat.col ;

I~ ——I s the diffusor ’s coefficient of

local resistances ;

~ --is the coefficient of local resistances when the

~~ flow of the mixture entering the diffusor
from the mixing chamber becomes constricted.
It Is well known that the ejector’s efficiency b (
*) is determIned as the
ratio of useful work to that spent , i.e.,
~ ~~~ ~~
a _ i uo_ t ux.
Z : .
The interrelationship between efficiency (it) ,velocity factor (?)

• •• ~ • •~~~~~~~ •
~~~~~~~~~~~~~~~~~~~~~~~~~~~ - -
 : ,
-

and local resistances (C) is —c”-L—t .

~
,
•
- The diffusor ’s efficiency as a function of its conicity and the
duration and quality of machining may vary within — O.7 to 0..
8
~A
For diffusors with properly mach i ned surfaces that are in contact
with the liquid s ‘i~~O.8 to 0.85. -

The efficiency for unfinished cast diffusors should be taken according

to Fig. 1 , as a function of the operating water head ahead of the nozzle.
• The coefficients of loca l resistances for a mixture whose flow is
constricted when entering the diffusor from the mixing chamber are deter-
mined from curves in FIgs. 2 and 3.

• 0
_ _ _

20 JO 4~ 30 00 70
~ S
Fig. 2. The local resistance coefficient for a flow which
is constricted when entering the diffusor from a
• coni cal chamber.

ii •

v— ——
___ s
i ii i
~~~~~~~~~~~ ‘

S
‘

• Fig. 3. The local resistance coefficient for a flow which

is constritted when entering the diffusor from
• a rectangular chamber or from that of a different
shape (other than conical).
The equality of the balance and the total heads (6) to (8) shows that
the least total losses due to mixing the flows in the chamber as well as

—
 ~~~
___
~~~~~~~~~~~~~~~~~~~~~ - —

• in the diffusor occur with v?v3, i.e., when the mixing of flows is
• concurrent with an Increase in pressure .
f
•
• It is , therefore, recommended that water—jet ejectors be designed only
•

-
on the basis of case 2 of mixin g the flows.
Basic Dimensions 2L EJectors
The basic ejector dimensions include the diameter and length of the
• nozzle, the 4nixing chamber , the throat and diffusor , as well as the diameters
of the inlet pipes for the operating and drawn-in water and the mixture ’s
outlet.
The diameters of these elements are determined on the basis of a Jet
continuity equation wh i ch , for a general case , may be expressed by the
following relationships :
-•

0)0 010
J;
3,~f — ~~~~ f ’..0,785d ; d— 1,13}’
—
U—

where v is the flow veloci ty , m/sec;

• G Is the flow rate of the l iquid , m 3/’(

f is the flow cross sectional area (of the pipe), cm 2
d is the diameter of the flow (pipe), cm .
The diameters of the inlet and outlet pipes are determ i ned on the basis
of permissible flow rates assuring normal ejector operation , while those
having connecting flanges or half-nuts are accepted on the basis of GOST
-
or other standards. -
1
.

for Inlet and Outlet Pjp~s, rn/sec

Flow Rates Recommended —
I •- —

-
Operating water inlet v0 3 - 6
Ejectable water inlet V~ 2 1 -
3
Mixture outlet v4 2 -

Higher mixture flow rates in the outlet pipe after the ejector diffusor
• are permissible only when the pipes are short and the number of areas of
local resistances with low values of their coefficients Is small.

I.
.

‘
I
 -r ? ‘• • -
y wvr .r ~~~~~~~~~~~~~~ ~~~~~~ •~~~
• ~~~~•
~~~~t~~~~ ~~~~~~~ ~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~
~~~~ ’ ~~~~~~~~~~~~~~
_—
\
• •

The nozzle aperture is generally assumed to be within 20-300 , that of

the diffusor - 8 - 12 °, their length being defi ned by the formula

2tg -j--
where D , d are the major end minor diameters of the nozzle
or diffusor , mm;
ais the dlffusor aperture.
The nozzle’s outlet edge (the cylindrical portion) must be as short as
possible (nó~more than one halt of the diameter of that cross section) in
order to reduce friction losses which may be quite consi derable at a high
flow rate.
The length of the mixing chamber together with the diffusor throat for
complete mixing of the flow is assumed as

3 is the throa t diame ter of the diffusor , mm .

where d

The length of the mixing chamber from the outlet end of the nozzle

to the beginning of the diffusor throat is recommended as follows :

1, (1.5— 2,5) 43.

It may be well to point Out that wi th greater lengths of mixing chambers

and d iffusors in then (and higher flow rates in the diffusor throat)

there w il l be rather considerable friction losses whi le shorter lengths

will resul t in poor mixing and lower productivity .

-
• Using the formulas cited in this procedure and having available a

calcula ted ejector (or its basic dimensions) , one could recalcula te it to

other operating modes for increasing the suction head , productivity or

the del ivery head by raising either the pressure head or the consumptIon
• of the operating water (in the latter case, i t will be necessary to increase
the diameter of the outlet end of the nozzle.

10
 - — _ _ _

t .
.‘ .
. — 4

References -

-. I
A rons, G. A. Ejectors (Struynyye Apparaty) , Gosenergolzdat Publishers , 1 948
Baulin , K. K. Ejector Design. Otoplen iye i Ventilyatsiya , 1 938 , No. 6
• Berman, D. D. Theory and Design of Water—to—Water Jet Pumps. I zvestiya
VT I , 1935, No. 3
Kop’yev , S. P. Auxiliary Equipment of Power Station Machine Shops
(Vspomogatel ’noye Oborudovaniye Mash,Inn ykh Tsekhov Elekt rostantsiy) .

Rzhanitsyn , N. A. Water—Jet Pumps. Onti NKPT , 1 938

Sokolov , Ye. Ya and Zinger , N. I. Ejectors (Struynyye Apparaty) .
Gosener golzda t , 1960

11
I

~~ iLt
i ~~~~~~~~~~~~~~~~~~~~~~~~~ -• • •