ML20101M584

From kanterella
Jump to navigation Jump to search
Mut Vortexing Evaluation
ML20101M584
Person / Time
Site: Crystal River 
Issue date: 04/04/1996
From:
FLORIDA POWER CORP.
To:
Shared Package
ML20101M505 List:
References
M-96-0005, M-96-0005-R00, M-96-5, M-96-5-R, NUDOCS 9604050448
Download: ML20101M584 (15)
v • d • e


Text

~

3 3 Fo4% -oz.

)

Florld2 e

Power

V ANALYSIS / CALCULATION

SUMMARY

DSCIPUNE CON *ROL No.

REYSION LEVEL DOCUMENT IDENTIFICATION NUMBER M

96-0005 0

)

TITLE CLASSIFICATION # CHECK ONE)

MUT VORTEXING EVALUATION

@ safety netated 0 won s v.tv n.i.i.d MAR /SP/COWR/PEERE NUMBER vENoOR oOCuuENT NuuSER i

EU ITEMS REVISED APPROV,ALS 1

Design Engineer

[MT[Md Date

)

Verification Engineer

((dg[

Date/ Method

  • R Supervisor i

Date

  • VERIFICATION METHODS: R - Design Review; A - Altemate Calculation; T - Quaification Testing DESCRIBE BELOW IF METHOD OF VERIFICATION WAS OTHER THAN DESIGN REVIEW t

a L

i e:

PURPOSE

SUMMARY

The purpose of this calculation is to determine the submeraence depth necessary in the MUT to orevent the entrainment of the overaas innto the outflow liauid.

r

^da?B

-4 jV(gesc

jQi,

.g n,,

9qyp ?v n.. w. f$

. s. 7l, v

~p is g

<g

.. RESULTS

SUMMARY

f

'iThe results of the analysis show that a submergence depth of 3.63 inches is required for a flow of 325 GPM;oh of the tank.

4 Other submergence criteria are provided on page 4 of the calculation for different flow rates.

9604050448 960404 PDR ADOCK 05000302 R efe5 P

PDR n a T. ve. or P=ni a s se: Noc.., E,.n e.,

Florida DESIGN ANALYSISICALCULATION-Power Crystal River Unit 3

- ro.

Page 1 of 5 AEvts40N DOCUMENT l(.Mr.NTf GTION NO M-96-0005 0

TABLE OF CONTENTS SECTION DESCRIPTION PAGE 1.0 P u rpos e...............

2 2.0 Design inputs...

2 3.0 Assumptions..

2 4.0 References.................

2 5.0 Cal.culations.............

2 6.0 R esu lts.....................................

5 ATTACHMENTS NUMBER DESCRIPTION NO. OF PAGES

)

anvt Weak Vortices.at Vertical intakes, Journal of Hydraulic g

1.0 i Engineering / Vol.113, No. 9, Sept.,1987.

9

.q 2.0 iSir.ing Piping for' Process Plants, Chemical Engineering, p

June 17,1968, pages 205 & 206..............

2

-g u r u..., n.e.

m N...,

eye....,

, os

Florida DESIGN ANALYSIS / CALCULATION TU Crystal River Unit 3 Page 2 of 5 cocuwsr ewora su susa M-96-0005 0

1.0 Purpose

The purpose of this calculation is to determine the submergence depth necessary in j

the MUT tank to prevent the entrainment of the tank's overgas in the outflow stream.

This type of hydraulic abnormality can lead loss of pump efficiency, possible cavitation, vibration, surging, etc. and damage to the pump.

2.0 Design inputs:

2.1 Buffalo Tank Dwg. M-6057 2.2 Attachments 1.0 & 2.0

3.0 Assumptions

This analysis does not contain any assumptions requiring later confirmation or preliminary data.

4.0

References:

4.1 Weak Vortices at Vertical Intakes, Journal of Hydraulic Engineering, Vol.113; No. 9, Sept.,1987 (see Attachment 1) 4.2 Sizing Piping for Process Plants, Chemical Engineering, June 17,1968,pages 205 & 206 (see Attachment 2)

?

5.0 Calculation

Reference 4.1 gives the results of studies of weak vortices in intake structures that have vertical outlets that convey the water to the pump suction. This paper develops an equation that can'be used tol estimate if weak vortices can form on the

~

free surface of the water in tri5~idtake]2s configuration is very similar to the fre water surface within the MUT. tank.

an However, the tank does' not have the mechanism within the internal flow path of the tank to mechanically induce circulation flow. All water inputs to the tank are directed to the surface by spray nozzles. There are no side entrances which would allow a tangential flow velocity to develop to create a circulation flow around the suction nozzle at r' bottom of the tank.

Therefore, the equation developed in the reference 4.1 can be used to estimate the submergence necessary to preclude the formation of weak surface vortices which may damage the pump by allowing the tank's overgas to be entrained in the outflowing liquid.

RET LAe of P.ard R E 5P. %Catar Engmeefrog gg

DESIGN ANALYSIS / CALCULATION Mglorida rd Crystal River Unit 3 Page 3 of _b_

CDGvMENT OENideCATsON NO.

RE WS4)N M-96-0005 0

The equation from reference 4.1 is S/D = 2.5 + [ 4/3 ]{ Fo }" + 40 { N }"'

r where: S is the submergence, inches D is the diameter of the outflow pipe, inches diameter of the outflow pipe, V/(gD)"y in and Fo is the Froude No. based on velocit Nr is the circulation number For the MUT tank, the circulation number is zero. Therefore, this equation reduces to the following S/D = 2.5 + [ 4/3 ]{ Fo }"')

The required submergence can be evaluated for various outflow rates in GPM. The results of this evaluation are given below.

l GPM Velocity, ft/sec Froude No.

Submergence, in 25.0 0.63 0.192 11.85 50.0 1.26 0.384 12.9 75.0 1.89 d.575 13.78 100.0 2.52 0.767 14.56 125.0 3.15 0.957 15.29 150.0 3.78 1.151 15.96 175.0 4.41 i

1.343 16.6 200.0 5.04

~%1.534 17.2 I

225.0 5.67

-- =1.726 17.79 250.0 6.3

'1.918 18.35 i

275.0 6.93 2.11 18.9 300.0 7.56 2.302 19.42 325.0 8.19 2.493 19.94 350.0 8.82 2.685 20.43 5 98 R E T ufe of F ent RE 5 P. Nxtear Eng neenng

DESIGN ANALYSIS / CALCULATION Mglorida N

Crystal River Unit 3 i

Page 4 of 5 DOCUMENTDENT4:GATKJN No.

RgvissoN M-96-0005 0

Reference 4.2 has information on the flow conditions that will develop inside the tank near the exit point when the liquid level is approaching the pipe exit or the bottom of the tank. These flow conditions are has follows. A liquid circular weir will form when the Froude number is less than roughly 0.3 or the ratio of submergence height to i

diameter of the outlet pipe is less than 0.25. With this flow configuration, the vapor / gas core that is formed in the pipe and tank is not appreciably pulled into the downflowing liquid. For this flow condition, the flow is self venting.

When the Froude number is greater than 0.3, vapor / gas will be entrained into the downflowing liquid unless sufficient liquid height is maintained in the tank. This entrainment height can be evaluated by Harleman's equation presented in reference 4.2. This equation is presented below.

S/D = [ VI( 3.24 {gD/12}")]

Where: S is the submergence, inches D is the diameter of the outflow pipe, inches V is the velocity of the liquid in the outflow pipe, ft/sec 2

g is 32.174 ft/sec Using this equation, the following table presents the submergerice to prevent gas entrainment vs. outflow rates in gpm.

GPM Velocity. ft/sec Froude No.

Submergence,in 25 0.63 0.192 1.3 50 1.26 0.384 1.72 4.5.

jg'.[

75 1.89 0.575 2.02 100 2.52 0.767 2.65 j

125 3.15 0 957 2.48 150 3.78 1.151 2.66 175 4 41 1.343 2 83 200 5 04 1.534 2 99 225 5 67 1.726 3.13 250 6.3 1.918 3 27 275 6 93 2.11 3 39 300 7 56 2 302 3 52 325 8 19 2 493 3 63 6 e5 RET de of F ant RE5P Nwear Eng<neev,3

Florida DESIGN ANALYSIS / CALCULATION Power Crystal River Unit 3 cxmmra Page 5 of 5 i

CDG4 % TNT OENTIFOATON No.

FtE vision M-96-0005 0

6.0 Results

The submergence given by the equation of reference 4.2 is considered the proper method to evaluate for gas / vapor entrainment into the downflowing liquid exiting from the MUT into its outflow pipe. The results of the reference 4.1 equation are not considered viable since the tank configuration and flow conditions into the tank are not the same as those used in the reference 4.1 to develop the equation.

w i

4' nGk hk cws-E eg.

' [,

$k.

.(x I

- a f.3c t.'

.1(L E*

  • j ',, -1 ; '

scy

's

-; t d 5

' *E a 5 s g:

p ;;:-

f

,w

.w 1

m e

7 ;.

S.h

.h R E T v'e of P, ant RE5P Nucea' Eng met'mg gg

WEAK VORTICES AT VERTICAL INTAKES 3

and Simulation of Low Flow flydraulics. Ihr-

. Miller and Harty G. Wenni. By fcffery G. Whit-

.,,,, w w

losure by authors.................

,,,,,,,,,,,,,w,dyees= epen dyee Distribution Coeffidents for Grase-Lined Chan-hrtel M. Temple. By Nicholds Kemisert. Closure by mien d yeemed ami ew =m et a ys='8aa *** as-.e= =d P'"h'"" d".Tlie _"*he k.

" susehe hdeste lhat e

  • 1221 en amelyste ef,

^

' " dese I

to eveed week woudens et wrHemi b

Insgo dhuensteudene ashungsgenee to wqease wow type of onesesWu h-beehn.Bdem m eethiedme wlE i l 1227

,i,, u.sen, unseen ese to ne seended, n..,es,ed m e.n.sessaan een gence le else he, ply 2, ^ -myon oppseeds new engte end headsam lengeh/

h

.. m a <

-fh~'1

, kt lumecoucnost

.ed

    • M$( f (MML*j,~ -

- Intake vortices are a resenet of angular momentum conservation et the 9.g g*.*

, z g-ANALYSIS / CALCULATION flow constriction where engester veloci increases with the decrease in

%e.

i et free surfe= sows into do,ed w ca,.ui.rea. nwy o=ur m conduits, such as sinks or aethtub drains. In large closed conduit in-b d,NfM'T R(

tekee. however, vortices ese often a severe problem to be avoided. They l

DOC 10etui

-1 ATTa have been found to cause flow reductions, vibrations, structural dem-oge, surging due to vertex formation and dissipation, and a loss of ef-l flEV jr O St(EET _l___ Old

  • ficiency in turbines or pasmpo. In certain instances they have also been pi a safety hazard.

I Pump and turbine perfonnance is highly sensitive to swirling flow.

Hydraulic pumps and turbines are designed sumning that the fkiw intc-l

.)

the machine will be axial and uniform. An intaks: vortex can (nuw.t swirling flow to enter the machine, resulting in off operation, a i7 tion. An air-

[

loss of efficiency, and poselbly cavitation, surging, and entraining vonex can also reduce the discharge into the intake. Swec-

.g, ney, et al. (1982) state that at pump Intakes no organized or subsurface i,

vortices equal to or greater than that vimmeDy represented by a coherent Ij swirt into the intake (dye core vertices) can be ellowed. Tresh-pulling i

-I and air core vortices, therefore, should also be avoided. A similar cri-terion is appropriate for hydreturbine intakes, since the flow through i

the hydromachine is simBar. One difference from a pump, however, is

-that a turbine has guide venes upstreesn of the runner that may elimi-r l

note a smou swirl. Another difference is that wet friction In a long pen-stock may euminate swirt befose it reaches the turbine (Baker and Seyre I

1974; Hecker and tarson 1983).

to the results of an,,."__.^ " invcJgellon de-

'Ihis peper the formstlon of weak, free eiarface vortices et vertical

' (l signed to Intakes with a heedroce channel. A weak vortex to desned as e coherent.

'Aseec. Prof., St. Anthony FaEs Hydr. Imb Dept. of Ch. and Mt.wret Engtg..

i Univ. of Minnesota, " ~

t MN 35414.

7. "$. that. of Ov. and Mineral

^

t^'_.,.'^.

^^

. Aa=a ta

' Grad. Res.7^

_ _^. Minneapolis. led SS414.

Urde.

.h open until Febnsory 1,1988. To entend the closing date one Y

of Journale. The snonth, a wvtteen segment sniest be Aled with the ASCE

.T_ -_.;r; 8er '8d8 Pe wee subsenetted for review and blication on I

n,f eracri,,g.

i 18.1986. Th peper le part of the femnuel of I

February

)

Vol.113. No. 9. ".r....M,1987. CASCE. ISSN 073M429/87/0009-ll01/Solbl.

/

Paper No. 21775.

I 1101-k i

--__,,____[~~~~------___.

[a( egag _ [(_4M 83, [Sr.

8I ent dye core ent: ring the intakr, o class of vortices is be evuineutlon of the resulte is gj 39 aga

( g j 4,a np :nd turbine intakes. The marya intakes for i.phpower d

84 _ 8'* # %, y [I.................. (2) preliminary dreign of now le, intalw sub.

[#

a r

e es. The emperiments emphaels[e awidth raiso7 the heedroce., g, ge + g, g s,g g q

.nce, intake veiodty, and the g er a, ar es ' e8r er. {

rDi' sy " " " " ' " * " ' * " g3)

(52_l ' ',

+

w e h tu. c va.num ro.onario s 8C N NM Q.,k-N

=-

M g,,,,

g,,,q,

g,,,,$ = l g -.I. #PiSD'l -Q'

. flow lictd in which intehe vortices occur is highly three-dimen-

--- - Y 8C k + I h h8C L

p 884

l. s.Ilowing only minimel elmplification of the equations of motion Y 8C D 8

Jimensionless parameters that should be used to describe vorteu 8

r y,sg -

lbl the

,3 ' g,sg + g gag gg--l

.................. (4) 9' 8Yg - - M* h I

) 9 48C*

stion have been in debate for two decades, one examp e e eg

+-

of Jrin. et al. (1978), and following discussions (Amphlett 1979; L4 Q.

9 velocity cornpo-Jell H79; Hebaus 1979). Recently, Odgaard (1986) has used a Ran-where ab, el,, and sq, = the radial, antal, and and angular coordt-vortra and an assumption on the definition of an alt-entraining vor-nents, respecdeely; r, s, and 9:= the redlel, o theoretically develop an equation for the su'._.o-a required g - the seculeretton of gravity;

(\\

notes: P = presount; p = Said density; J -,-a; D = Intake throet old such a vortex. The applicebuity of Odgaard's work, especieHy

, = kinemette viscosity of the fluid: S =

N 1,

diameter: Q = How sete; and F. d he farnekt values of 2e egr (circu-ye core vortices, has not been tested, however.

>f t

.e most desirable dimensional analyste involves normaltration of the lation). Using S, rather then D; in the expression for ( results in a rel-atively simple dienensionless Q. J =; for circulation, as wul be ohnwn i

itions of motion, as was performed by Anwar (t966). The authorstf difflerent set of later. Bage. 2,3, and 4 Identify sin dimensionless peresneters which de.

l

? repetted Apwar's analyele, resulting in a slight asiMe flow with axlal symmetry (Fig.1) may be"es s ssyiIikfin' terms

+b.*t 'N scribe the How: Mr = Sr./Q, a circulation nuinber; R = Q/es, a Reynolds ensionle'sfpsrameters. The equations ofilR60ost' ford 46dy incom-i number; $/D, a skuesneloniens _ '_ -

,. g:(r /p)(#P/ar)(5'/Q');(1/pX8P/

l 8

he dimensionless stream fusketion,4, first defined by Lewellen (1962)

!]k> b

^_.. will be converted

'C 8tXSD*/Q'); and gSD'/Q'. Mietter three l

r to dimensionless vedebles in Eqs. 9 and 10.) Dh4 ding Eq. 2 by (S/D)'

would isolete the circulation number identified by Anwar (1966). Dr./

f M

(18)

'hq

['

Q. Thee le at Hrot appearance a snore logicet cholee than Nr = Sr./Q, 3

Qq ph)s l

a which is frequently a dependent parameter, is-m am since the i.

replaced by the N,diesneter. The retics F./Q, however, may be re. subm j

  • sr e

,,,; y; h duced to a form that is inversely dependent (16) g4 i.;

8' t

e, 8'

I"8

"'T hh

!MLh d the dimensionless vertables 3 = r/D; ( - r/S; and r = 2=(v.r/r.)] ()

h Q8g far-fleid circulation is the Ene Integral of angular velocity times radius j

at r - R:

f D

s o a-

.s.

Reedt = 2seh R = 2=Rv, ten =............................ (5) l T. =

~

where e = the ande between the approach velocity vector Y and the vector normal to the control surface. In adeBelon, the discharge through a

the cytndelcel control surface of redlus A and height 5 + oh is g

f

Q = 2*Re. ($ + AA)......................................

zy,,,,,, ;

,g,,,g, y p,,

unun,. uu/na,na,u u, nuunusuunan S_r., 2 0,R$ i.n e,

t.r,.

}

' ^ " ' "

  • Vertes Formonen at Q

2sRp,(5 + A4) f Ahi l

( l + Si 10.1.-Definmon Shech 9er Vortettles and Peremetwo of

\\

.<tical Intene t103 t102

Friet relationehlp Ca S/D cdd Froude number, however, can be i

seen from Eq.10 to be 5/D ~ F oriS/D ~ Fil". The writers' c1tperience

'I

~~N.

N le that when.._

_;d values of $/D at whide o given i of vertra Nj occure are plotted werene the Froude number, Fe = F g, and a cir.

I

/

f culation parameter, Nr as ten e, al far better resoluffon of the dela is e

\\

achieved. A

of 5/D vedsue F = V/yQ and T.D/Q = D tan

~

L' 8 * ***

\\

dent e/S could resultIn a deudy resolup of the d&ea because a deIof the I

e t

vertable la the.,-. mate, aut_.

ca, le contained in esc

,8

. ;;; errore in out..-.p.w would affect each of (s

three terme. R the three vedebles.

f h

a f

Raoucnoes ce CotCULAMoM alone Humanece Ceuessa s

I

% % g "'

A number of.m _J m have emperbnentaBy in ted free surface i

N h

\\

vortex formation at an outlet in the center of a teak with a given j

drculation (Anwee 1966: Deggett and Keulegen 1974: Jean, et el.1978).

i t

-Plan sketch of Ver11oelInteho ter Peremeter and Vertelde DefinflIon The drculation wee controlled using venee or jets lesping from the stranged hydrotur-

]

[j side of the tase @ng 1976). The typicalshowet in Fig.3,le far from a drcular

/5 is often small. the drculation parameter identifled herein may, Y

(

l-bine intake structure, schematice9yoach flow drculation (engle) may be i

n ctses. be empressed as f.'

b, tank, however. Althou the appr the etructure near the intake on circulation is I,

l estimated, the effect ttn a

.............. (8) generally unknown and difficult to determine (Brochard 1983). Most hy-ai t a n a...................

i,9

(

i.3 D l.;

droturbine intakes also have an approach channel (heedrece) that will i

at the up.

i I*7

'.'i j i(n algnificantly redeste the circulation m the approach flow. Nrs t

irculation parameter given by Eq. 8 is olmilar to that used byO(N 2

both reduced by the factor 8/(2wR), where B = the heedrace width.

iiI st. (1978) m the analyels of their experimer.tal data because it

,' g ;

The primary season a heedrece will reduce circulation is the lateral remeters of i in less auto. correlation between the independent nto the cir-grade line (4 + V'/(2g)) caused by the flow sepa-g difference in gression; e g.

submergence was not incorporate edge of the channel weEs. Higher velocities l

p.

g n partmeter. The equations developed herein give a theoretica ration around the and poselbly even a higher water surface elevation wEl occur on the leit j

5 war (Fig. 3), laabing toward the lateke. The energy grade-line on the 9>

lin-3r thi2 dedsion.

two pressure gradient terms resulting froen Eqs.;2 and:4 wilthe Froude"riumbers that govern the intake i

Q

[,

y right won wSI be less than that at the left wou, redadng th j

e that the pressure distribution le approximately hy'drostatic, then angle.

f

- pg. end JP/ar = pgar/#r, phere z = the distance from the water Intake geometry (t/8) and flow a by side weg friction. The process I

i

e. The pressure gradient and body force terms in Eq. 4 then be-Circulation will be reduced furt le slow, however,..r_k.g much larger L/8 rettoe than normally en-I

. a. W. W..... (9)

E U '...........................Ac6 u a pp h

t JP SDh

~

f y.

i Q'

a Jr

/

ig as the pressure distribution is close to hydrostatic in the verticality and

,a l

h i

son, which is a good assumption for weak vortices, t e grav

,8 are gradient terms in Eq. 4 cancei. In addition, the pressure gre-V[

t and body force terms in Eq. 2 become 29 i)

)'

(;)'

JC 16(;

,3C t

g s"

(10)

f. g.

S' g S't)'

, at 16

= 7 p, n,-=7 3

=

- vi l

t 4Q/

e F = V/\\/g$. Fn = V/\\/gD; and V = mean inlet velocity = i t Eq.10 indicates that the choice of a proper Froude number, g ven FIG. 3. -Oedinblon Stateh of Vertleellntake vigt fteedrese Channel irculttion parameters of Eq. 8. le not straightforward. The appro-tt05 i

It04 '

~

m.sn

-~

I Emmnimeenn Faceury Q,

i t formation in An ex m_ rol fhsme was buBt la invest gate wor en I

~....r in.m.ql;

. i ippo verskal Easkes weih a W channel. ne nuaw, shown in rige. s l

m and 6, is 7 m (23 ft) long,1.4 m (S ft) wide, and 3.2 m (4 ft) deep, i

~.-

consisting of a etirin6 eein, a trenettion section, and a test section. Flow

@b d' '

b t

was suppued froen the Missiselppi Stiver in a once through mode with l

a 20cn (8-in.) supply Ene to the stilung basin. InSow was measured ll j

with an ortace meter coldwated in-line and connected to a manometer ll with either mescury or Medesn blue se indicator Gulde, ne staling beein I

=

.e..

wee designed to af straight uniforrn Sow out of this sec.

with a (8-in.) lateral diffuser pipe shown l

tion. This wee in Fige. 5 and 6 with 5 cm (2 in.) diameter holes detBed to stagger at 4.-contros volume ter seemenown theneesa Applicensen

+45* and -2if off hosteental, and directed at the rear wet. The flow i

wee further smoothed by Gowing through a 1Sen (6 in.) thick rock crfb, i the field, since the wall boundary layer increases gradustly which consisted of socks coesser then a 2cn (3/4in.) slew. Finally, a any large scale eddles.

ee. This phenomenon will be neglected here.

transition mone of 2 m (7 ft) wee used to dem Icsted a relatively uniform l

rformed on an intake very simuar to that of Fig.

Velocity..u_ __.

2 after the rock crGP t

the iAh of clatulation along the headrece velocity over the crose necelon (Rindels and Cultiver 1983).

nts were x

e 7 nstallation of two 1.2 m to descri number et dimensionlese parameters using tan e L/8. etc.

Vlevel observation was made poseitde 1 I

without success. Finauy, a very approximate momentum f um 1.4 m x 2 cm (4 ft x 3 ft x 3/4 in.),44 walls at the side and end l

i is developed that proved successful. If we assume: (1) That e _, 4 I

on in circulation is proportional to the reduction of the y-i of momentum in the channel:(2) that the w Jr' mean y

"' ' 3 I

~~~~

nt of velocity V, may be used in the momentum theorem,

~

~

_. g

~~~ ~

0. V, = V tan a; and (3) that the pressure difference between

\\ k 2T

.b

! right wtlls is proportional to the y-component velocity head, sj 3;;

3

  • tc q.4 l

' M

f. $

m i

- ftp V......................................... (1 1)

F 8

I a

  • J b

Eso w

um a -

nam am*

    • t".

= pressure on the right wall; P = pressure on the left wall;

- ms.

c i,

t

.',U,b r 5 e-- -.-

i constznt of proportionality on the order of one or two, then ntum theorem may be Integrated between x = 0 and r = L, t

t-b 3 lll FIlk 5.-Plan vise et _', _ i ^ 'FeeAny 3

.t

'3

.r

'[

M

'd-a

_a S

I

................................(12) g l

= ; itn a UE 5 v -= = a.

I

  • i g '*"
  • g,,, g. n.

AQh'4 q

O 5 + M'.hhd 5 >> M in most applicQ" ^ q

.p'

.7y-s we w:'4 g g,. u..

u,

o..

Q j

ra u-

,i

...........3......................s....(13) g =1 -

)

tzn o i

....ii.,.

o O*"

- tsn o A

i

, frk,.

l I

l ill be usedian analysis of the experimental date persented in g.

section. Assumptions 1 and 2 are an extremelfMods approxi-

.I reality. The other tion, however, is to attempt various com-at random, which certainly would not pts. s sessen visse et %--a.ieness Yose recany s of the relevant nura wiuced the relation for NF given in Eq.13.

11o?

1108 8

n

=

]

For a given expedmental run, inGow discharge end e wm 4 ev: tion were d i ;ru,,,, gg,, w;g,,,,,,g,c,,,,,,,,,,

h ov.

lume. The interior components of the test section were t e m h olde-wee also seeorded on a strip chart. Discharge late the be5 mouth Intsk2 are prod $ced using four plexiglass panels for eachtsdrace walls and the beumouth int wee found by adang hedow discharge to the pendece of water surface wee.

and water surface area. Water surface elevadon was (3 il)gscejalle ic,,,_

until a F.2r> dye core vertex formed, es de-h time role of W

edlarge'vadationsinlength.Thelength't the oo

,g.$ m16 Yt),

ellowed to termined byinjecting thsough a into the water. Water 09

. k will used in the q-%ent were the..m

' 9.m 19.5-It) long combinations. The 1.8-m (6.ft) counbination le vehse and 2 x 10'8 surface rate of varied a

l In Figs. 5 and 6. The bellmouth wee centered 0.4 m (16 in.) from m/s(0.6 x 10-s ft/sec) with no discernible impact on the measured value of criticalsub,ca foe ore vortices (Itindele and CuUlver 1983)

.3

ir sidewalls.ns shown in Fige. 5 and 6.

l rder to contral and predetermine the now approach angle to theice, guide vanes were surface rete of dwnge, were the approach now aqle, dkhg e Other potential sources aperimental errors, aside from the water l.dewalls. The 11 vanes were 21 cm (91n.)In lengthMrith a thicknese h

had pivote euremente, and sneemmeesnent of waeer/ surface elevation. As stated pre-it was documented that guide vene angle le a good represen-cm (0.12 in.) and a specing of 11 cm (4.5 in.). Each vene between e top and bottom in order to very approach now anglel (1978), the vene en6 e vlously,f approach Sow angle, within 2%. Vane angle could be set to 1

within 20.5 degrees; thus, the uncertainty in the vene angle le 27% at tation o

. IJsing a technique described by Jain, et a.l of the Dow leaving the g

, hickness were used to compute the ang e g o the 7.5* and 22% at 30".

= i. The maximum adjustment for vene thicknese in comput ag An orince eneter, co86 rated in place, was used to sneesure discharge.

l t d through y h, g!q The accuracy of the calibration is 0.5%. In addition, the discharge was

. of the now was 2%. Culde vene performance wee eva ua e l es of t

j ;4 -

unstead for a number of reasons (Padmanabhan and Hecker 198 l ographs of dye strenklines taken at two depths and two va u i

d i e discharge (Rindles and Gulliver 1983). The streaklines indicate

, imateady inSow, and the drop in water surface eleve-L d that guide vane

p l There wee no
,y a]

[j tion over time These, the uncertainty in discharge measurement was s

ciuding the guide vanes performed their function well an 20.006 cme 10.21 cf",), mresponding to an eencertainty in Froude num-of 10.27. This le a significent source et error and could

! e is a good representation of approach now ang e.

g id vis-,g ber,V/

ence ci vorticity in the flow caused by the vanes.

m n int 2ke throat diameter of 0.15 m (6 in.) was chosen to avo l

account for some of the acetter ht the data.

g q%g l dtles,

' i impedtnce of vortex formation exceP at very low intake ve oThere were three sources of eror la the stage measurements (water 1

t diameter surface elevation). The point gesq tseed was accurate to 20.6 mm (20.002

' ough there is still uncertainty on the criteria. The throat I

.i

  • /* > 5 x 10'. Dagget and It). In addition, the svood jn the fhsene w~rld swe5 and contract in con-eed d Jzin. et alls (1978) criterion of g*Da VD/* s 3.2 x 10' to avoid vlecous

,3 junction with successivt periode d A ing and wetting. This swelling.

,legan's (1974) cdterion of Rocts indicate that viscous impedance of vortex formation le pose yg lble f

and contraction caused the level of the top of the beamouth to be slightly d

o i V s 0 2 m/s (0.68 ft/sec) or at F s 0.17. This criterion is surposee' (1978) critedon of R = Q/(v5) off, cetteneted to be 10.3 mun (20.001 ft). There was also a human error r>

e the water surface elevation, estimated to be t1.2 mm l

virtually all of the data taken. Anwer s five to avoid in caused the total tancertainty in sneesurement of stage i x 10' hulicates that the ratio (5/D)/F must beless thanC. Anwar*e critedon le cous effects at water temperatures near 20' d in Tullis, et al. (1986) to be less then *I.4 mm (20.015 ft), which corresponds to an error of (t0.004 ft),

i l

o turpassed by most of the date. Data presente diameter, D = 5 in.,

20.01 in the disnensionless M 7 ydye core worten is somewhat arbitrary s,

I

.hcr.tr a critenon of R = 4 x 10' and a throatassed by the data presented herein. PadmanaR = 1.5 x 10'.

and developed the pesernt of time a vortes is present me a bhan The definition of a ** for the convenience of the capedments. Hecker b

.d Hecker (1964) ound no siptficant scale effects a ovei s withlarge (1981) has studies where tierbine and pump Intekes hich tre also su

.ch of these cdteda wee L.A for air. entraining vort ce,k vortices wouldcriterte for h case wortiere present less then 50% of the time. The

.locities nur the air core. A similar criterion for weat (1986), for example, should have ellerton somewhat leer, arbitrary but difAcult to in-J l6 t

.gictity be somewhat less restrictive. Tul s, e a,und no scale effects for dye core vortices down to an R =

Q/(*S)of wrbre found Co'Porste into the sneesurements suported herein.The dif that wortex forsnellon de a transition

~-- from a l radial pproximately 5.000.

flow to a swirEng now. The point oftkneition le not we[re be-cause there le a range of now conditions over which either a purely re-ka:unement Tecmeous hich The measurements were designed to identify the submergence at w dial or e swirlIng 30W con occur, elmilar to the wet-known transition ii l b

. coherent. persistent dye core vortex forms, herein called cr t ca su.

from laminar to turbulent boundary-leyer flow. Just as a disturbance in i

f at least ten etc wee the free streem over the boundary layer can cause a turbulent spot that nergence. A coherent dye core vortex present or t

Jelined as persistent. Ten see was believed'to be sufficiently long oh the flow' field.The dye-later dioelpetes, a estisebence in the intake oppseech flow can cause a avoid the Deeting vortices that will pass t roug of voetensthet should be fleeting dye core vertex, which appears infresquently and will not sig-h f

f core vortex was chosen because it le the ty igne.

H00 avoided in most pump and turbine intake i

N

. Ske

' ' - - - - - ' - - - - - - - ~

7

,, {,

nely impact turbine or pump performance. This disturbance could se. for o:mple, to s temporery surge in the flow or a rare com-

=*

o gep*I c-n of tip vortices shedding from part of the hydrouBe structure.

j in any transition phenomenon, the results of the expedments will

, p, g eF t

some scatter. Since the critical out.%;.ae is found by reducing

    • 8 l

e

' ergence until a persistent vortex forme, measurement errors w'll

,(*

-a bias towards lower submergence levels. The conservative ep-h to submergence guidelines, therefore, le to place an envelope g

r over the measured values of critical submergence, as suggested t

e/s

  • umphreys, et al. (1970).

l f

t:3

\\b e critical submergence at which persistent dye core vertices form M

. merstred over a range of intake Froude numbers from 0.25-2.2 at

/,

.. f

  • e rangiments of headrace espect retto end approach flow angle. In

~

. e, 4 e.

3*

~, g# 4 j

91 individual measuremente of critical sub..a.

.ae'were made. The u.

g I

i 0

  • g

.idual dits are available es an addendum to Irindele and Gulliver j, i

N'C.

l j

1). The distance between the intake end the olde walls was always j

f P

i

-r c

swirl t

imes the throat diameter. The welle did not greatly seen se

.,8 i

-lopment over the intake. The data, therefore, should ig a merimum value of submergence that will assum ro dye core r,,

g d W"% H W j

e mhsurements are plotted in Fige. 7,8, and 9 se S/D versus F, j

,,,,,-f,,, g,,g, gj,

,7 the thirteen combinations of approsth angle and aspect ratio. AE i

h!

weeuwwie were snede with en lateke throat diameter of 0.15 m (0.5 y

s ft) and a heedrece width of 0.81 m (2-2/3 ft). The fotowing observatione Ti r.}

,f

[,A

. o s l

may be mede:

o

.k I

o

-(

O

1. All of the.___ _ a.."a a significant dimenelonless sub-m u-

.,,9 y -(*.

wegence to eveld dye core greetee then 2.5.

.s 2.

. 7 shows the ateleal _ _:-.s et the shortest heedroce length, with L 8 = 0.43. A great increase in the These eh;.as wee op.

i heedrace Perent as the inflow angle changed frorn 1 l

r+r..

r+r.

end ePProach flow angles see elegee no those of anony hyd.m-;. fa-cGhees. -b d flow angle le thus an important parameter foe these s,

Intekee

[

3. The Incieese with approach flow angle le not as significent for an L/B retto of 1.75, shown in Fig. 8, whm en Anaense in a from 15-v

,,uhed in a seletively eme'l (~ 0.7) Increase he S/D.

f = *..

a q g4

4. This inavese in S/D with 1,,_ - 's flow to reduced furtlwr p

whh the L/8 retto of 3.14. In fact, even the extreme flow

,,., y at t

    • y*s approach angle increened the required dimenoloniess submer-e 4

gence by only 0.8.

2 4*:

5. The rapsired _ i__.;;.ae fee.

Er and I./8 = 3.14 is appront.

weely the some as foe a - 30' sad L/8 = 1.75. Thus, the effect of the w/g heedrece length / width reato upon reducing circulotton is obvious.

l

6. The date are aE rnughly the same for the three lerigth/ width ratios t 7.-critices svenneegenee cleasuremente and Emelope Curwoe et Ets.13 for when a s IS*. This indicotes that the approach angle into the channel l

eerese Loaem et e.st a terem Enerense to intene cener unel: rsa - e.as; sney not be of,__g_^

as if it le less then 15*.

- Ny j

inst tito,

- --. -. ~. _,

)

ysis with p = 2 gave acceptable muks. Wee resuhe will give en in.

dication of the maximum submergence required to cvoid dye core vor-

.L tices at an intehe with a heedrece channel. A _r -i produced flow a

net (Coulver, et al.1996) sney often be used to determine the angle of f

the approach flow to the antake. Poettioning the side weIIe closer to the d

intf e and the installation of antidorten devkes wNl reduce the required k

J',. Frement y, a hydraulic enodet study to the best jf f

l

-- -.v. 4,_F (and perhope the only) meene of inceeporating the vertoue errengemente i nto the Intake design.

i f

_ 1 to the results of Jean, et al. (1978), who

1
h",*

{

-g Eq.14 le most m, ud the submergence at which an air entraining vortex would form I

K at a vertical intake. Join, et al. foijmd that the relotton:

,,,, ~

g s

g = 5.6N)

  • F%"...............'.................. l.......

y 8

g.

O 1.,

a gave a good description of theit data when viscoue effects were ex-I 3

(,

5/e 3

g ciuded. The prknery differences L a.; Eq.15 and those developed

  • ,s 8..

.c n herein is that Eq.14 hee an Intercept at S/D = 2.5 and a much stronger y.

'W dependence espen NF. Both differenese are probably due to the oppos-

n

- o,

.:g,

+ " "

b g

ing cderte for cdkal _ _i,_6 e.g., dye core wereue alretraining f' I tu j vertices. Dye case vertices are chosen as a criterion herein because they 8

ehould be avoided in snoet turbine and pump intakes (Sweeney, et af.

6 x

7J

,M

.k,.D q 1982).

4h Esamour Ascutances i

The Rapidan hydroplant le a retrofit of an existing dem, with two 2.5-

~

O MW turbines, half the nurrber originally Installed with a greater total

  1. y:

8$

.,. Ig g ** '

y cepetity. Intake vertices were one of a number of concerns with thie

,d'

. S = 24.5 ft (7.47 en): D = 9.75 ft (2.97 l>

9 ?*'

l i

ei retrofit. The followtng date p/s); Y = 8.3 ft/sec (2.53 m/s); L/8 - 1.33; 4*

m); Q = 620 cu ft/sec (17.6 m F - 0.47; and S/D = 2.5.

var An electric analog potentlet flow analysis wee carried out on the intske

^^ -

2 and enweepe cuene of Eas.13 ser (Gulliver, et al.1986) and could pe used to determine the engle c*f ap-e.-cenicat -Aas. =

,,ece tsaeth of 2.49 m: t/s = 3.14; N = NT proech flow. The Ave streamlinee flowing into the right headrace were at the following angles (one heedrece width away from the entrance):

40',52*,65*, ylr, and 90'.1hese are very poor approech conditione. The wo plotted in Figs. 7,8, and 9 le a best. fit envelope curve developed everage of these stroemline anglee, Sr. wee uead in determining that onsidiring all of the data almultaneously in lineer regreselone on NT si 0.$7. Eq.14 Indkstes that with tlwee r.onditione an S/D value of

14. successively changing the power on the NT term, and attempting 10.7 would be required to evold free surface wortkee. The intake was d

e values of p to = 1,2. and 3). The regression weighted the equare very close to the hee 4 rete wolle, had the required submergence le prob-I

.lusta at ove the envelope curve ten times the equered residu,le be-ably lese, but tMe gives an indlantion that a hydraulle model study le the envelope curve. The resulting equation le es follows:

necessary. er, whh en S/D of 2.5.

^

al e..':.T..cd that there The hydrM %odet sh:dy that wee em_, __ _ " _

...............W) wee a vortes, sobleen af the intake. A strong alt eore vortex formed that i

-2.54 F T + 40 NF'....

f 3

was difficult to elimirwee, rUy because of the poor approach of the V/V5D; V = velocity in the intake throat; Np.

now to the heedwe (Gu ver, et al.1966).

[I 4 IL/B) tan of. (i.e.,8 = 2); and L = distance from the headrace F

=

F Sunsaany Asas h 4

rance to intake center line.

s 2.5. The Fil' dependence Week, free surface vertices, defined by a coherent, persistent dye core bcse equations are applicable for 0.2 s Fo hat developed in Eq.10. The ability of Eq.14 to describe a range ofsubsequent to dye injection, have been studied for vertical intakes with l ;

theorem anal-iriewesione indicates that the approximate momentum I h 1913 e

titt lI

-- - - - ~ ~ ~ - ~

- ' ~ ^

~ ~

~

,Z - nt1b w '

& W[*

.sw 1

9

" eor.,.idbed to be the best experirr.cntal re< ult; it is t

u r sn!'itions. It seems likely that his equation could

. f.

' p

,r, = 0 "U yi, Dp. Thne results agree uith thw for aim be applied to two immi<cilde liquid phas.es niecan.

a downtion%r liquid around a stationary labb:e.

t+r de.=ign e. Iluu mer, the toemcient of 3.21 gis en in i

I. In sammary, large ( d E; 1 in.) Inng hubblu become Fig.13 shoahl be changed in his experimental valac of g

J

'tr&pped in vertically downtlowing liquids when : <

2.0. In addition the lighter phase density shnald re.

g 100 ep. and:

place pc in Eq. (14).

,f Vt = 0.31 Y pt D 2

(19 When the velocity is less than predicted by Eq. (10), Y hirlpools in Process Vessels si

' bubbles will rise. And at higher velocities, bubbles will be sucpt donnward and remo"ed from the pipe. If a Compared with iriotational flow, whirlpool forma.

~

a-continuous warce of vapor is available. 0.31 G l't/

tion is usually very unpredictable because the forces ttf (psD)6 <1 can be expected to produce pressure pulsa.

I 2.

ion and vibration.

ci Fig.12 has been prepared to aid in the solution of
17..

Eq. (IG). The flowrate is converted to gym., the diam.

History of a Downflow Problem i

en eter is changed to in., and pn is assumed to be much As less than pr. Two additional lines are plotted on Fig.

Although the problems with dow;n: low are many 1

12. The first is for (Nr,)s = 1. near which pressure and varied, ne particular pr blem is interesting be.
the, palsation amplitudes will be high and siphons will cause it illustrates a recurrme design problem.

i The piping configuration shown in the photograph l

na,,

fcrm readily. The second line is encountered when was sized without any downflow technology. The t

nd (Nr,) t is increased further; frictional force offsets 30.in. cooline. water return line, roughly 40 f t. lone, At, gravitational force, and no pressure gradient will be was desiened f r appr ximately 13,000 gpm. of water.

ach present in vertical downflow. This latter line depends Bottom of the 30-in. line was near atmospheric pres.

sure.

en the Reyr.o!ds number and pipe roughness. The line Af ter startup, the equipment near this large water

D given is for high turbulence conditions (e.g., water) line vibrated severely. A study revealed that vibra.

r th e =. 0.00015 f t.

ti n originated in the 30.in. pipe. Furthermore, the 2

N pressure at the top of the pipe was near one psia.

I k *,, ^,..

otational Downflow From Process Vessels Also water temperature of 40 C. was uncomfortably close to the.'m c. boiling point of water at one psia.

Thus, the continuous vapor source needed for pres.

Irrotational downflow is often confused with the sure pulsation and vibration was apparently coming l

krge.bubb!c phenomenon inside a vertical pipe. Flow I' ",' ""jj,'; 'h[ t ;i "

h Fiir.13 been available at the patterns at tne exit of a process vessel, however, are g

more complex than the simple bubble dynamics. Liquid time, the designer could have predicted the niphon in the desien stace, and made an appropriate change J

height and entrance geometry become important vari.

to climinate the possibility of pressure pulsatien, i

t ables that complicate the analysis.

The 1ine could have been made larger to that it was i

I 3

.'With the geometry shown in Fig.13, liquid forms a ny.,

d, w smaller s that friction woull raise the pressure at the top of the pipe. The prob-s crcular weir in the vessel when the Froude number tem could also have been corrected with a restrie.

},

(pt. D)4 is less than roughly 0.3 or H/D is less than tion orifice near the base of the 30.in. line.

) !

t25. Liquid flows down the pipe in a falling film. The

.j.7 j [

ter of the pipe and ve.ssel contain.s a vapor core that g;*

  • g:

p ;

not appreciably sucked into the downflowing liquid.

1 igg

.p g.

is self venting flow is not compl.etely predictable y

i I i q e q

g l; - T,

' T

, as a result, may occur at Froude numbers some.

'I ;{

I l hat greater than 0.3. D. S. Ullock,85 for example.

p i i k iented that the transition occurs at (Nr,)r. = 0.55.

ca h

l When the Froude number exceeds 0.3, vapor will be

' Q h f((

4 j

ained into the downflowing liquid unless sufpcient s

g j

strainment point or critical liquid height has been QQ-jw) lf mei 3

id height is maintained in the process vessel. The d

- tr 9.

q I

ll rmined experimentally by Kalinske.:: and theo-

-4 y

q i,

J cal!y by liarleman and others.:2 liar!cman's equa.

i f I shown in Fig.13, should be a conservative desigr.

I l

y it; vapor will not be sucked into the downpipe at g

f above the liquid height predicted by this equation.

j; j p..

[ g The critical liquid height at which entrainment first I

3 7%

y g

i alm rs is frequently unknown. To test for vapor en.

f.

l s

'ement, the critical height is compared with that

p. !

l h would exist if entrainment were ignored. If the

{

-p l

pr. Y cal height is greater, vapor will l'e rucked into j

downilowing liquid.

Q Ri l

4 illstleman studied selective withdrawal of saline ua.

.7 I

el Engineering-hne 17,1968 205 i

)

/

^

.f

, FROCEs5 PIPING,

((

J 2

  • j* l that initiate uhirlpools are generally weak. Yet the The shaded area in this figure indicates uncertainty in'.

i effect of a whirlpool can be dramatic. The rotating the transition point. Actually, the uncertainty is much

l. quid can open a vapor core in a vessel that will prop.

greates because the observation as to what constitutes d

arate through the outlet piping. and perhaps into a fioth or slug flow is somewhat arbitrary. The reader '

f pu mp.

is ieferrrd to Gosier's first articien for his distinc.

9 Although tangential inlets most easily initiate whirl-tion betacen slug and froth patterns.

i1 pmis, centrifugal pump < can also induce a rotating

,T jl finw" in an upstream sessel that in turn opens a vapor Case History: Vibration from Upflow

!)[

core ancl feeds vapor into the pamp suction. Whatever y\\

the cause, whirlpools can be easily eliminated with a One.ipplication of Fig 14 involved vibration in a

.M straightening cross that is installed at the vessel outlet long 30 in. pipe in which two phases were flowing. A B

nozzle.

sertical riser at the end of the horizontal section vi.

brated at a low frequency and an intolerably his Liq Two Phase Upflow amplitude.

Three modifications were made to eliminate the sus-Sizing of piping is particularly diflicult with vapor pected slug-flow pattern. The first change was a red

'i,%

.nd liquid in coeurrent upflow. If a line i.s sized for tion in pipe size from 30 in. to 24 in. Only the ri

, i' g

low pressure drop, it may well cause slug flow, with and some horir,ontal piping immediately upstream of resultant pressure pulsation and vibration. A primary the riser were reduced. The second modification was f

design objective must, therefore, be to avoid a slug-to install more gradual horizontal to-vertical tra 3 {

flow pattern.

tion piping. Finally, connections were installed for.

g At least four distinct flow patterns can be observed

.,g in upflow. In order of increasing vapor rate, these are: vapor injection into the 24 in. transition piping.

injection nozzles were sized to supply momentum tha

] : J.

bubble, slug, froth, and annular. Only the slug to froth was thought to be lost by the liquid at the transiti

f[

transition is of interest here.

This last change was later found to be unnec

!Fj Govier and others. ".:7 have run extensive testy on When the new installation was placed in servicell u

h l

.tertical upflow of air-water mixtures. They experi-was found to pass a higher vapor rate than bei mented with tube diameters from 0.G2 to 2.5 in. One

- g series of tests studied the effect of liquid rate on flow without appreciable vibration. As shown by the poia in Fig.14, the suspected flow patterns were predi g

patterns, while a second series concerned the effect of correctly.

t

,ji s ariation in vapor density on flow patterns.

5,j in the slug to-froth transition, we can study the isothermal Two. Phase Pressure Drop

'.i

' *?.

ip; flow pattern in terms of the Froude numbers (Nr.h.

, d, 8 and (N,,)c. Such a correlation is shown in Fig.14.

As with other aspects of two-phase flow, press I

drop estimates have a high uncertainty associsted wi N,1 them. The designer should expect his pressur 1, ;,

f!

estimrte to be typically m40G of the true value.

f

. p 8

cause of (Fis uncertainty, he should try at least j

i l j j 'j l l different pressure-drop correlations if he desires,

- - ]

II tional confidence in his estimates.

-;,,.- y% ~ -

_t

/ /

be specific for a particular flow regime. The reaso 7

ideally, pressure-drop correlations should prob i

o

/ /

that a dispersed flow pattern, for example, would l,,

d6 n,

ff expected to behave different from a alug-flow pa _

!{

' h y

.,y',,_

g This approach has not met with great success, r

,, 5 incipient entroinment point, }p'/

/'

because the available flow-pattern maps are not yet J

defined" i

If *00'h'd d8, C', #C' "5 '

/

/

kg agggg.ggggh I

l j

M85' mum H/0 and i i I

/ A failed to differentiate between frictional pressure j

-v / /g{o for c,rcuier.e,r /j and momentum. based pressure drop. This latter g

h3 ltsoveersD) f sure drop is that associated with expansion cf the' p

phase as pressure is reduced. It is particularly l

4

'Self venting f e.

1 j2 y /g[).

/D to hrevent tant with the high mass velocities and the low 2 4 (H/D)'.

/

[I gos entro.nment sures that are used to generate data for correls y/

23 %[

Twn correlations that wdl be considered here I

p)

C [/3/l V /f DN 3 24 (H/0)2'5I the enrrelation of Leckhart and.Tlartinelli,as and r

t c

homngeneous model of Dukler and others." These

'k 0}

! ! I I ! I I relatiot.s are easy to use, and are more accurate' g

0 0.2 04 0.6 08 t.0 t2 t4 g

most other correlations? " Of the two, Dulder's

^

g<

Lieu.d he.cht to d ometer ratio, H/D relatinn will be slightly better for tr.ost appl.ica' Ik.th.T!artinelli's and Dukler's correlations IRROTATIONAL co*ntio* rnay occur en vessels-Fig 13 expected to gne better accuracy for horirectal r

1.

a 206 u

June 17,196 s-Gemkalla o

i-i 4

J

Retrieved from "https://kanterella.com/w/index.php?title=ML20101M584&oldid=4305658"