Theoretical and Experimental Study on Fluid Flow in Valve Channel

8/2/2019 Theoretical and Experimental Study on Fluid Flow in Valve Channel

1/17

Purdue University

Purdue e-Pubs

International Compressor EngineeringConference

School of Mechanical Engineering

1982

Theoretical and Experimental Study on Fluid Flowin Valve Channels,Parts I and II

L. Boswirth

This document has been made available through Purdue e-Pubs, a service of th e Purdue University Libraries. Please contact [email protected] for

additional information.

Complete proceedings may be acquired in print and on CD-ROM directly from the Ray W. Herrick Laboratories at https://engineering.purdue.edu/

Herrick/Events/orderlit.html

Boswirth, L., "Theoretical and Experimental Study on Fluid Flow in Valve Channels,Parts I and II" (1982).International CompressorEngineering Conference. Paper 371.http://docs.lib.purdue.edu/icec/371
http://docs.lib.purdue.edu/http://docs.lib.purdue.edu/icechttp://docs.lib.purdue.edu/icechttp://docs.lib.purdue.edu/mehttps://engineering.purdue.edu/Herrick/Events/orderlit.htmlhttps://engineering.purdue.edu/Herrick/Events/orderlit.htmlhttps://engineering.purdue.edu/Herrick/Events/orderlit.htmlhttps://engineering.purdue.edu/Herrick/Events/orderlit.htmlhttp://docs.lib.purdue.edu/mehttp://docs.lib.purdue.edu/icechttp://docs.lib.purdue.edu/icechttp://docs.lib.purdue.edu/


2/17

TIIEORETICAL AND EXPERII"'ENTAL STUDY ON !


3/17


4/17

TABU!: 1 Valve channel entrance contours with constant pressure and velocity

({' . . parameter,tan r=:-y 1

b) fo r a single s lo t :Y"' 2:JT( irrttan( ,JT/4- lf/2 ) +in lf] j x- 2+; (co sT - 0 .5 )

t G+-t__n___,.,

2b/t=0. 5

s

0 .2

X

0 0.56 10

611a ; : ..parameter'( tantr"' 'fo r a circular_holeL3]-

y/Ro 'f:'/.L( )I)" ' n 1.0 000.05 0.9450.10 0.9110.20 0.8760.30 0.8410.40 0.8?.40.50 0.["1?0.70 (\.7971.00 l'.?f,f ,1.40 0.7832.00 0.7 82

.. ..-s=b

1 1 .5 Xb

Fig.4 a Poten tia l flow veloc it ies along wallsb Velocities along valve pla te (e::oo)and laminar boundary layers .

40

2


5/17


6/17


7/17


8/17

@

0 -1rfJ.7

0.50.4

0.2

0.1

J1

o.s....-L v/ " /_IJ___ -- -- -- - r r - . - -

as -1 sf'f:> 1.5_l_ .l. 1 l iH - / - -

:us/

v V-1 ./_,.......... ....... 1ifrs-D.6//6/

.. ._ 5 .\ .. ---D-{ .5 \

{ --....."4--- \" ...,._...iift s,o.6mm ' _\

0 .2 1

.....

.._- ........-.__ '4 .. ,........... ,.

/ -

rnm

2. 5 7 1o'* 2. s 7F ig .9 Experimental '( see eq(5)) fo r a mu lt i - r in g -p la te va lve (5 circu la l 's lo t s , f r e e a rea i n the se a t p la te A,2 6 em; 2b=3,5 mm).For th i s valve Adepends on s ra th e r s l ig h t ly and th e re fo re was rep laced by itsmean value

0 .706 .- a) e-=tJ-( s) fo r va lues p 2/p1 = 0-i-0,8 b) c )used(2) v iscous e f fe c t s are im portant .Equations fo r incom pressib le flow with lo ss coe ff i c i e n t s o r e f fec t ive flow a reas depending on Reynolds number can be used .G )E ffec ts of v isco s i ty as well as e ffec ts of compress ib i li ty are to be expected.Equations fo r com pressib le f lu id s with flow coe ff ic ien ts depending on Reynoldsnum ber must be used .

DISCUSSION OF EXPERIMENTAL RESULTSExtensive experiments on flow lo s se s of ava lve were performed byH.H .Weiss and the au thor. This i s described in d e ta i l elsewhere i n th ese P roceed ings[? ] .

44

With re spec t to compress ib i l i ty valves behave s im ila r to nozzles in f lowinc lud ing choked flow condition .Chokedf lo w condi tion i s es tab lish ed a t lowerp ressu re r a t io s th an th eo re t ic a l ly ca lc u la ted fo r no zz le s (fo r th e t e s te d valve chokedfl9w cond ition occured a t Thls phenomenon i s qua l i ta t fveLy unders toodan d quan t i ta t iv e ly expressed by 2 ex p e r i menta lly gained co e f f ic ien ts ,A:.m= 2k P 0 [p2/k_ pCk+1)/k]} c:-- -1 1}1 ( 5)

w ith P = 1- ':A(1-P/P1 )f!-= \:( s); ) \=A (s ) o r constan t ro mass flow rate k r a t io o f sp e c if ic hea t


9/17


10/17


11/17

15% 3 . 1...J..

smm

0.611 522.53

s/b

0.343 0.5710.8571.143 1. 429

theory.fs fs3.1 3.3

2.833.0 4.4 3.4 6.73.9 10.24-5 14.3 5.0 20

5s-eq( 9

4.1 35-93 8.4 811 315.1

85% of to ta l area

0nrseq(7

0.4920.4110.343 0.297 0.257

experim ent

0.290 0.4010.492 .5450.57 70.60 3

0.8540.7 020.5 74() .11770.4 040.3521) re la ted to free area in sea t p la te A=26cm 22) CD I 8 = CD! b/s

betw een valve p la te and seat pla te . Theouter r ing of the valve corresponds to caseS .1 , the remainin g s.lots to case S.3. Thetab le of makes c lear the theo re t ica lcalcu lation of the discharge coeff ic ien t

no pressure recovery2e 5.5mm 2b 3.5mme/b 1.57 f/b=0.8 57

1.431.4 01.391.361.37

1 _ _ , . . ~ , - ~ . - - - . - - - - - - - - - ; 0Dis

0 2 Jmriil i f t s

Fig.12 Theoretical calcu lat ion of discharge coeff ic ien tfor a m ulti- r ing -p la te va.lv,"Jan d compar ison with experimental resu lts-A roundin g effec t of the boundar y layermay increase the e ffec t iv flo w area.-Pressure recovery may occur in so mes lo ts .0Dis

On the other hand one can calcu la te an in -compressible discharge coefficien t CDIfrom the experimentalI t ca n be seen tha t besides a constantfac tor the j e t fl ow theory can expressvalve loss behaviour adequately.

For small pressure differences eq(5)in Part I yie lds (-to)

For the calcu la tion of the experimentalvalue C and A are us ed from the rangewhe re effec ts are not predominant(p 2/p 1 =0 to 0 .8) . The re su l ts are comparedin the diagram in Figj2 . The experimentalvalues are la rger fo r a l l l i f t s s by afactor1-4+0.03.This may be a ttr ibu ted to thefo llowing-facts:-Geometrical differences betw een a lo ngrec tangular slot(cases 8 .1 ;8 .3 ) and them ulti-r ing -p la te valve.

- While the s lo ts in the sea t p la te are in -terrupted by s tru ts the area bet ween theseat p la te and the valve i s free fo r.throughout(area blo cked byfree area in sea t p la te : 26 em ) .

47

With some experience and feed back the resu lts of je t flo w theory may be veryusefu l fo r the design process,-especia lly fo r optimization procedures(where constantfactors are not im portant).


12/17

PRESSURE LOSS DATA FOR LONG RECTANGULAR SLOTS 1 x 90-deflection flow.,.....---------------- ---------- - - - - - - - - - - - - - - - - --- --

.- _ / " ) . . e F 20.1

f=Acp(p,,-P:z.)Cp -- -+ - - -- -- - - - , ~ i .. .. . -.. .! CX)1.5 2.

J! '1.51 I I t 1.25---- -- ---: =1- II

05 I I0 0.5 1 1.5 2 bS.2. Th Mo

'\. fsr - - - - " " \

/,,,.'"-

- "' . /

3

2

10.1

-

48

0.2.

ol..o .20

10

0

- I --i

.------11-----

0.2

i0.6 1 2 3b

1 j1.25I

---- 1.5

2..f: ,__=Oob

0.5 -'1 s 15 2.'b

-' II 'i I i - - - - - - - --T, .

04

I ' :1/.I '

./. ,, .It ' 'I , ' I I' I-' ' . 1f-- 1h1I I l

06 0.&;

I

: ! I: I;-j ; fI - , i I:l;!':s'b

!..


13/17

PRESSURE LOSS DATA FOR LONG RECTANGULAR SLOTS "2 x 90-deflection flow,- survey

. ------- -----. AiR)-- --- - ' +l ' -i "_[_________ '

\ v

values(T h,Mo)- -E xper im en ta l values derived from I .E .Ide l 'ch ik : Handbook of HydraulikR esistance.Isreal Program for S cien tif ic Translations IPST Cat .No.1505Springfield Va.1966----:Ex trapo lated curve

-----------------------------------------------------------------------------'49


14/17

-- - - - ' "" '

PRESSURE LOSS DATA FOR LONG RECTANGULAR SLOTS 2 x 90-deflection flow

5.3 Th*M TIT

---T71iI

20 III

15 II

107 I

If; I i

.5 t-i1I I:' 4 ~ - - ~

3

2 0.5 0.7 1 s 45b5.5 Mo

b0.811.2.

poten tia l flow solution available in l i te ra tu re (to the knowledge of the author). An approach wasused to calculate fs. In principle the 2x90-gef lec tion flow was composed by 2 independant 90 -de-'f lec tion flows. A certa in correction was used to -account for the fac t that the second go 0 -deflec tion flow occurs Without a stagnation poin t.

S.4I f 1?>2f reattachment of flow and pressure recoveryoccure. For 1 2:>4f pressure recovery i s fin ished.Application ofmomentum theorem re su lts in the fo lloW ing

[ T.(Vfss.s' - 1)]Js ,S .4 = j s ,s .3 1- fs ,S .3 The amoun t of pressure recovery depends on widthofs lo t 2f. Optimum width: two times jet width 2d(according to case So3; 2d= f S 3) In th is specia lcase the formull to'

S.6Th*MoNo reattachment and pressure recoveryoccurs for 1, 2 ::1 5f. Pre:; ;sure recgveryin the f i r s t goo-aeflect1on flow 1s

Pressure recovery in both go 0 -deflectionflowsfin ished at l1 .t: l -1 1.2.

8. .

6

4---1-------+

50

10-- -'

6" --- ----t' --1 '-- --1 --' '. - --- --

4+ - - - - -fs

,


15/17

PRESSURE LOSS DATA FOR CIRCULAR CONFIGURATIONS Survey

c.z

-


16/17

PRESSURE LOSS DATA FOR CIRCULAR CONFIGURATIONSLoss data fo r c ircu la r configurations are derived from experimental re su l ts pub lished ing rea t de ta i l by Schrenk 5 .Exp-e riments were .ca rr ied out w ith la rge models (. diamet-ers ofhole about 50mm) and water as workin g f lu id .D a ta as lo s s coeff ic ien ts . ,fo rce coeffic ien tReynolds numbe r boundary laye r th ickness e tc . were derived by the autho.r fromdata . Reynold s number e f: f :ec ts seem to be weak in the -c oncerned range(o

Conditions of transitio -n

_1 r ct1 11 _ l (._2rnea,n VwLIe-s 1Dr d. :.? Ql' 'lc l T-..;:;:(for ; >2 :pressure c..3!)

52

0 ? I.I.... 1d01

!c:Lt' - ij -... __ _____ -II 'It;..4ai-- --- - - --- -- --wcz)

OJ1 2 D 3

d'=tQd ,t-

() (,

for =1.2


17/17

- - - - - - - - - - - - ---PRESSURE LOSS DATA FOR CIRCULAR CONFIGURATIONS

jc.31Exlfs

'fs,c,2 roughly Unfar decrease. due to pressurerecovery1

0 2 ' ,4 . 6 t s

Ic.Lt!Ex [5]]..f.s .5 = 1.2D-;r 1/i-:!:J

t C.5/C .6 I:EX [5] 6

-

'1

2

ti!I

:

'II- ...

I

Red . oo------1-----l.--1......__,_I H k

5 I -1c s--t-- -

/()

-- ..-.. _.... __1-- - - -; II ---+-----I

d1-- r)rJO t .d r 'd;l, = 12od Ird =qf5

f- ---- ==-- ..- - . - . ----- - - -- +-1-- ---..-- , I ! i IIO,g 0.03 0.05 0.1 .s 0.2ci--- I_ _ _ _ j

Theoretical and Experimental Study on Fluid Flow in Valve Channel

Documents