Raabe J. Hydro power - the design, use, and function of hydromechanical, hydraulic, and electrical еquipment

flacliine,

the

rot;iti~;g tappings can be conn*:cicd one after i~nother

by

n

sc;tn~ling

\:11\~

to

the station;iry ~nano~ncter. For the abbrcvirltio~~ of this procedure t\\*o

tliiligs

l~a\c to

be

provided. Firstly:

Tllc

hwitching

operation

nus st

be

possible

under

full

spcctl.

Second-

ly:

With

i111y

switching scveral tappings are connected wit11 the correspondin2 numbcr

of

stationary ~nanometers.

~cchatlicill design of a scanning valve: Fig.

9.4.9

shows n mcchanicul s\vitchi~i~ dcvicc. dcvclopcd

by

E.

Biir-

[5.17].

It

operatcs as follo\vs: By thc rcciproc;~ting niotinn of Icter

4.

parts

.:

illld

2

:iiove

up

and down, guidcd pcriphcr;tlly in the tccth

5,.

The friction

of

tlic seal dcccleratcs the rotntilig

part

2

ag:~inst pal-t

5

and

1:

By

this both parts arc rilovcd tan_eentially

by

nnc

:oath

pitch n~ainst cach

other. This is possiblc in tlic lif!cd position of thc !ever

4

due to thc clcarancc

of

one pitch in the

tangential

dircctioii (cffcctcd

by

!he

omission of each sccond tooth) and also in thc lo\vcred position

the lever due to similar

tangential

cleara~~cc (see Fig.

9.4.9).

This switching by moving the levcr 4 up and down turns part

2

about the axis against

part

1

in the

desired

manner,

so that thc pick

up

bore in

2

;~dvaiices along tllc seal

6

from one nlensuring port

to thc

next.

The up and down motion of part

2

against shaft

1

is also nccdzd

to

avoid ,II,~ ivcnr

and

friction during the s~itcliing ;tnd to provide tightness of the joint flange bet~vecn p;lrt

1

and

3

;ifter

any

switchirig by means of thc scal

6.

Since thc scanning valvc is fixed on thc frame of the stator

of

the torque reaction dynaninn~etcr (here

two oil-operated radial

12

cylinder positive

displacement

pumps are

used

to obtai~~

a

strons arid

steady

braki~:g torquc down

to

zero speed), the frictional torquc of its scal

7

is

elilninated as an

interllal force.

Fig.

9.4.10

shows the arrangement on a model Francis turbine with its pipes connecting

the tappings or ports

on the vane to the stationary manometer via the scanning valve

shown in Fig.

9.4.9.

9.4.5.

Measurement

d

the

torque

-

Fund:+mentals: The hydrodynamic torque exerted

by

the flo\\:ing fluid on

thc

rotor

may be measured by the twist of

the shaft between the runner and the adjnccnt jovrnal

bearing with the aid of strain gauges. As the shaft is rather stiff, very sensitive strain

gauges must be applied or their output

is

very s~nall and thcrcforz liable

ti,

crrors.

htlore-

over this doublful small electric \lalue must be transmitted to a stationary indicator.

Therefore

on

models, iorque measurement

by

reaction torque dynamometers is preferred.

Here the

rcaction torque, which is nearly equal to the hydrodynamic torque of the I-unner,

is

measured at the quasi-stationary, slightly swinging stator of the brake. This can be

done mechanically by weight or electrically via bending of a beam by means

of

strain

gauges, as used

e.g.

in

the lab of Hidroart, Milan

").

To

approximate the measured

reaction torque as far as possible to the desired hydrodynamic torque on the runner,

thc

friction of the support of the brake's stator must be made as small as possible.

This

can

be

done by incorporating the shaft bearings

in

a tube which is attached to the stator of

the

brake, see Fig.

9.4.6

arid

10.

This eliminates the bearing friction ;is an internal force

of

the system motor brake. The swinging quasi-stationary brake stator of this system

should be supported at least by means of roller bearings.

A

more favorable solution

is

to

Support it by oil- or air-lubricated 11ydrost;ltic bearings. This is done in Fig.

9.4.6

and

10

with radial bcarings. Another solution is a spring suspension

of

the brake stator in such

a

way, that either no spring force is exerted on theqsystem, when it is in its measuring

;;2-------

)

According to information givcn

to

the author

by

Mr.

Lrccl/i.

the

lab

of Ilydroart. Milan, appl~c's ~llis

hind

of

strain

g;illges

to

a

snlall loaded bcani for

all

kinds of Iaad and torque mrnsuremcnts.

In

reccnt research work.

the

author's lab at

TUM

has also used this vc~y sensitt~e device successfully.

Fig.

9.4.10.

Torquc reaction dynamometer with scanning valve. Brake: two multipiston radial posi-

tive displacement

pumps with stator on swinging frame suspended on torsion rod. Upper end

of

latter fixed on staticnary part. Zero output by throttling oil

flow.

Reaction torque on the

pump

casing twists

the

torsion rod, the twist of which is lneasured by the deflection of a light beam. From.

position, or that the tangential deformation of its suspension is used for the measurement

of the torque and the deformation due to the axial suspension is used for

the measurement

of an additional axial thrust

during operating of the machine, set:

Cap.

9.6.

-

Special brake design: Fig.

9.4.10

shows

a

special brake design in which the last men-

tioned principles are realised. At first the shaft of the turbine is radially

and axial1

supported in a tube attached to the brake stator, which eliminates the shaft friction as

internal torque.

This tube of the brake stator is radially supported by oil-lubricated hydrostatic bearings.

tube-like system is axially supported by means of the twisted rod. This rod is

tangentially clam

:it

its upper end by means of roller bearings, enabling only an axial motion of this rod end. At

end the rod is

fixed in the axial direction to a short axially elastic rod,

fixcri

at its upper

end

to

rigid frame. This si~ort rod is covered with strain gauges, to measure the variable axial thrust

und

operation of the turbine.

The reaction torque on the lower end

of

the elastic rod twists the latter. This twist

optically indicated by

a light bearn, which is emitted from

a

stationary lamp and

th

reflect~d

by

a mirror attached to the twistcd

rod.

This device

devclaped

nt

LeIi~.stuhl und

Labo~.

fur hydraulische Mnschincn und Anlagen der

TU

Miinchcn is c;ilibsated

by

n

,,,,ig1~t-induccd torque.

The

interllal torque can also be mcasured

by

a simpler- brake

if

the bearing torque is

,!iminated

by

n~casuremen t, see Fig.

9.4.1

1.

~i~

9.4.1

1.

Scheme

of

torque metering on

a

hirb pressure test rig in the Brunnenmiihlc

@:arch station of

J.

M.

Voith, kleidenhcim,

used

for modcls of turbines and pumps.

1

gene-

ntor motor;

2

cradled electric dyna~nomcter;

3

load cell for coarse rncasurcments;

4

coarse

balancing unit,

20x

10

kg

weights;

5

load cell for friction torque

It;

of turbine bear-

ing;

6

load cell

200

N

for precision measure-

ment of

FF;

7

constant preload

111,.

=

2000

kN

31

reversed direction of rotation;

8

normal

direction of rotation and torque;

9

model

machine.

See

0.

Eichler-,

G.

Hcir,lnilrl,

L.

Brur~dt

19.581.

Similar load cells in the fo~m of bending

cantilevers pasted with strain

gaugcs arc used

by

Hydroart, Milano. An cliniination of the friction

radial bearing load. This

can

be reached by putting

a

load on the bearing.

torquc of the guide bearing requires anilullcd

111

the torque in form of

a

couple wilhout radial

9.4.6.

Tl~e

visui~.lization

of

flow

-

General rcmark: No measurement at a series of individual points can rcp!ace

the

visualization of the whole flow field. This holds true especially under off-design point

operation and moreover in

the cavitating regime. Up

to

the prescnt there is no device,

which enables the measurement of pressure and velocity in

a

cavitatirig flow.

-

Absolute flow:

A

siinple method of vis\lalization consists in fixing light thrcads

011

the

boundary walls of the flow

field.

For an air flow, eiderdown enables good observation of

wakes

and

the

boundary layer adjacent to the wall under stallcd conditions. The injection

of smoke in air, of coloured dyes or gas bubbles in water, or the electrolytic generation

of

bubbles there also enables an observation of flow effects especially close to

the

wall.

any rate the size of bubbles should be s~nall !o lirnit slip

by

pressure gradient. In this respect the

use

of

polymcric suspension instead of water according to

Fleischnpr

[9.11

I]

facilitates an approxi-

mate

measurement of flow

velocity.

The spark nlethocf, introduced by

IVeskr,

and later on developed

by

Fisto.

[9.113]

and

G~lllrs

[9.114],

shows the strain rate of a spark-ionized fluid path e.g., in rotor

channels and hencc the tcmporal change of velocity profiles in

badly

accessible ducts.

All

these methods need at least one observation window and one illurnination window,

'he

first with

a

plane external surface. For the observation

of

the flow in a rotor channel

'hey

require at

least

a

partly transparent shroud, hub or, in case of need, also vane.

Moreover

any

observation requires

a

high speed camera with a steady or speed-triggered

(*lroboscopic) light source.

-

Reliltivc

flow:

,\

perkct

c,bscrvation

of thc

flow

in

:I

r;~dial

rotor

is

ohiaincd

by

a

t1,lrlsparcrrt

htll.

and ;idj:~cent tr:1n5p;!rcnt

c:~sing.

This

neccls

cll.ive

from

tlie

side

of

tile

-

shroud.

A

conkcniont

~nct

hod

is

offered

by

the rotoscope.

1

lcrc

the

calner;i rests.

,I

light

beam

fro~n

the

ob5crvcd rotor

passe.;

a prism which rotatcs with

5OU/o

of the angular velocity

of thc 1-utor (DOIC pris111). Thus thc rcilcctcd beam of the prism remains stationary. Thereforc the

rccciving ciimera or human cye of

;t

resting observcr ciin also bc stationary.

-

Visui~lization of rclntivc flow dircction close to the vane: This method has bcen dcvclopcd by

(;.

Sc.l~lcilrrlle~-

[9.Y3]

for

a

1rrrnc.r.s

runner. Thc mean

flow

dircction

a

ccrtnin disti~ncc from the runner

\.allc Si~cc citn be indicaictl by micro flags. Thcsc light ~nct;~llic flaps of trinngulnr form arc arranged

so as to rotate eas~ly about thcir shafts. At thc end of cnch shaft

a

piston-likc disk is fixcd. This

lies

with a vcry small axial clcarancc

in

a specially adaptcd cylindrical chamber within the runner vane

surfklcc. This piston can be blocked by friction, when it is loaded by co~npressed air. This pressure

4

lasts itlso during thc rt~ndown oS the turbine and the following dismantling of the individual rotor

*

Lanes. \Vith the aid of pi~ttern making scctions and radial sections,

scratched

on the runner blade,

the "frozen" qilasistcady <lirection of the tirne-averaged relative flow can

be

obtained easily from

(his

4

flag position

(9.831,

(5.181.

*i

9.3.7.

Fluids

to

be

used

in

tests

5

+

'i

11.1

practice water and air dominate

as

test fluids.

The

advantage of air consists in the

easy

g

hanilli~?g of the light test equipment. This facilitates especially preliminary selective

testa'

Gas in

3

closed

tes~

rig

may also allow the Iieynolds number to be varied in a wide

range

for

efficiency scale effects. This can

be

attained, e.g. by changing

the

kinematic viscosity

by

means of heating

or

pressurizing the

fluid.

I

-

34casurements carried out

by

Klrir~

at Lehrs!uhl und Labor fiir hydraulische h4aschincn und Anla-

"

pcn at Tcchnicnl

University

Munich ~lth air, Argon,

CO,

operating

a

1;T

have rcvealed that

the

efficiency charactcnstic.

5.2.

the hill dingram, of a hydraulic machine does not only vary (as always

silspcctcd) with the Reynolds numbcr but in

a

much stronger degree with some molecular properties

of

a

fluid.

3

Tt~cse arc e.g. the number of atoms in a rnolecule and henc:: the adiabatic

exponent.

These changes

concern as well the level as also the form of the characteristic.

Thise invzsiiga:ions msy

be

considered

as

a

proof, that a rcliable prediction of the characteristic

o

a

hydraulic inachine

tail

be made only

by

tests

in

water and thcse under

a

Keynolds number as

c10

as possiblc

to

that of the prototype

(9.1151.

Changing the fluid results also in

a

change oft

calibration curves for the

instruments use,

e.g.

Venrlrri

nozzle etc.

9.5.

Measurement

of

unsteady relative

and

absolute

flow

in

a

Kaplan

Turbine

by

a

vectorial probe

of

quick

response

from

D.

Castorph

9.5.1.

Introduction

For stress ca1culatio:ls in the runner vanes of

a

Kaplan turbine, the dynamic meas

ment of the real unsteady flow field, around the runner vanc with periodically

randomly varying fluctuations of

pressure.and velocity,

is

of special interest.

A

nulnber of references described by

Lotz

[6.27;

6.281

and

Lienharr

[6.32]

show that, in

theoretical investigations, approximate computations of the fluctuations in the

veloa

field of two cascaties

in

serizs moving relative to each other,

have

been

made,

where

360

bc,-ausc

of

the

cfl'c~~ivvc work involved, experiniental invcstig:rtions of thc unstc;~d]r

no,v.irldi~ccd forces espccioll: acting

on

Kopl:ln

rllllncr vnlics arc rase.

~l,.asurcme~~ts oftk stresses

in

the

vane surfilcc by strain

gaugcs

;,re givc~i hy

(;i.c-itr

illid

Brli-p[

10.37).

iince

fluct~~atio~ls of hydrodynamic forces are caused by fluctuntions of pressure and vclocit>.

li~i

kno,,t~cdgc of the

real

unsteady flow field as the source of unsteady lo:~d

is

of

iniportnnce.

"sual method of dynamic measurr~nent in water arc based either on the nicasur.crnc.!lt of

unstcndy flow velocity by means of hot film probes.

as

has been c;lrricd o11t by ~\lollc~trkr~~~/:

[9.1(,].

or

on

the

measurement of unstcady tofa1 pressure

by

nieans

ofa

Pitnr

probc

\:

ith

built-in trans~l~.~~~.

,,

by

Ger.ic.11

19.451.

Both methods s~lffer from tlic fact. that tlic sin~ult~ncous nleusurcrncnt

of

prc.ssure and velocity at a definitc statio~i is impossible

illid

that the \clocity tiirection

can

be

mc,1-

s,,rcd

only

in

a limited range.

9.5.2.

Measurement

of

the

absolute velocity

field

of

the

runner

In

order

to

avoid the above-mentioned disadvantages,

a

new conical vector probe, siniilar

10

that of

Conrotl

(Fig.

9.5.1)

but

of

quick response

has

been developed

by

Cnsrorpll

at

~~hrstuhl und Labor fiir hydraulische Mascllincn und Anlagen at Technical Universitv

h.!unich,

West

Germany. This was for

the

measurement

of

the

uiistzady flow field

\~,ithi!l

a

KT

operating

in

water

[9.116;

6.291.

Six

micro pl.essure-transducers

of

the type Kulitz

~QL080-25

(freqt~ency

70

kilo

cps)

enable the simultaneous indication of six fluctuating

pressures in

the six ports in

the

head

of

the probe. T11is is done by application

of

DC

amplifiers acd

a

light beam oscillograph with moving

coil

galvanometer

of

4

kilo

cps.

Qs:at(z=Ll

Qtot

,,

pa

1

Fig.

9.5.1.

Vectorial probe of quick response for simultaneous measurement

of

instantaneous

prcs-

Sure

and velocity vector in water operated turbines, developed by

Casrorph

[6.29].

a)

Sect~onal vicnr

0f6-hole probe. b) Pressure

transducer,

stations of pressures

pbl,

pI2

due to pitch angle

/i,

P,,

.

pO2

due

to yaw angle

u.

1-our of the prcssrlre tr;~nsduccrs arc

used

for thc indication of the sp;lti;ll local vclocity clircction

Oi>c

;lt

tllc

;II>CY

of tlic cone intlici~tcs the total prcssrlrc,

slid

tllc 1:1st 011c :~vcr;~gi~~g

1.11~

values offou;

ports uniformly distributctl

l lo rig

the pcriphcry of thc cylinclric;~l pi~rt, i~ltlic;itcs thc pcripllcrnll

Y

avcragcd slatic prcssurc. Thc diflkrcncc of thcsc 1;1st two figures

gives

thc lociil dynamic prcssurc and

hcncc thc vclocity.

13cc;lusc of its comp;ict constr~~ction, the hcad of the probc has n di;ln~ctcr of only

8

mm, which

is

fairly sni;~ll compared with thc runner riiamctcr uscd of ci12 mrn. Ilccausc of the mcasurcmcnt

in

cold

wntcr and thc frcqiicntly m;rilc cotltrol of zero load point undcr ntmosphcric prcssurc uscd

as

a

rcfcrcncc figure, the

influence

of tcrnper;~iilre and tin~c cli~pscd on thc calibration characteristic could

bc practically clirninatcd.

The prcssurc

transducer

1b1Q

L

has

a

diaphragm of stainless stccl and

a

full Wheatstone bridgc on

its

back. The rcarsidc morcovcr is connected with thc atmosphere. Thus thc pressure i~ldicated coin-

cides always with the

gauge prcssure. The simultaneous i~ldication of this figure by all the six

tr:insduccrs in the hcad of the probe has bccn icsted undcr unsteady conditions. The wholc probe

has been calibrated dynamically

in

thc pressure conditions of the prototypc turbinc, c.g.

in

a

pulsating pipe flow.

Since in unstcady flow the probe cannot be turned for a permanent compensation of its angle ofyaw

-1

lbetwccn the vclocity vector and the probe hcad axis in

a

planc normal to thc probc shaft),

the

calibration curves depend on the

yaw

angle

x

and the pitch anglc

/?

(between velocity vector and

probe head axis in

a

planc containing both).

9.5.2.2.

Measuring planes, results upstrear11 and downstreani

of

the

rotor

The turbine, with a vertical shaft fed by a semi-spiral casing and equipped with a straight

4

cortical draft tube, was operated

in

an open test circuit under the two heads of

fI

=

1,8

m

and

2,8

in.

The location of the measuring planes and the positions of the probe relative

;

to

the runner are shown in

Fig.

9.5.2.

The following measurements were carried out at the

$-

bep with

n,,

=

127.5

rp~n,

Q,,

=

1,39

m3/s.

'4

-

Measurements between the gate and runner: Fig.

9.5.3a

shows the pressure

fluctua-

tions indicated

by

the six ports

of

the probe for

a

typical position. The components of

the

Fig,

9

5.3.

Mc;trt~rcnlc~~ts.

G)

hc.~d

II

=

2.8

ni. poirit

I.

2-

Pal

/--,-.-r--

..-&

-

-,,--d

---A-

-.--

m,:,s,lring pln~ic

(VL)

be-

"'

,,,eel,

guidc vancs

and

I

2

Pa2

--d-

-----/----..--

ru,ll,cr,

position

of probe

J/

g

,

0,17.

a)

Prcssurc

re-

m

so

cords

of Ihc

SIX

pressurc

Rtat

+--L,4--4--,-*-\~*.C~.\z"

\.--*--

,,,,l15duccrs

\.'s

timc clilpscd.

plot

c/~--c/-~-,-

bl

~bsolute mcridioual ve-

locily

con~po*ient

c,,,.

;~nd

I

*],irl

vclocity colnponent

5

p~l

~~---+------~-*A~-~--~

-+---A%

0.

vs

rime

elapsed.

aftcr

clim-

,

r.

n

inntiol~

of

minor distur-

PI32

--.J--*.,.-----d--

,--.

---------./'1-4.rp

-.,.

I-.\-

bilnces. After

Casiorph

16

291.

i

-

.-----

,

*---.----.--r

&3>

T-!

.

;

i\--------

-?\

I-

-I

L-O,OIS

-

---

-

Trvonsr

---

--

-

time

scale

4

runner

blades

'I

revolution

--d

b)

+--

vclocity resulting from this and the calibration chart are presented in

Fis.

9.5.3

b.

From

the ccrves for the probc position

s/s,

=

0,17

the upstream influence of the four runner

vanes on the absolute flow can be

recognized

(Fig,

9.5.3

b).

For

the cyli~ldrical scctivn near to

the

hub the influence of the runncr and the gates is

reduced lo almost nil. This may be due to the large distance of the measuring station

~II

the strcamwise

direction

either from the runner or the gates.

All

the measurements have

hen

repealed by a conventional

Conrad

probe with its time-averaged indications.

See

19.1091.

-

Measurements past the runner: The

Fig.

9.5.4a shows the pressure fluctuations of the

probe for a typical position. The evaluated

compone~lts

of

the

velocity are

-

401s

-

Trunner

-.-

t,me

stole

Lrunnu

Modes

I:ig.

9.5.3.

b1c;\surcrncnts

at

I!

-1

2.8

m,

point

I

in

pI:~nc

H

L

dnwnstreain of

thc runner, station

s/s,

=

0.32.

rt)

Prcssurc

rccorcis

of

six prcssurc

1

:S

transducers

vs

time

i*,

elrtpsed. b) Meridionll]

3

cornpollent

c,,

whirl

coniponc~it

C,,,

and

radial

conlponcnt

c,

of

absolute

velocity

from

a).

Numbcr

of

inden-

tations and thus number

of

wakes reflects

the

nunibcr

of

guide vanes.

After

Custorplt

I6.291.

bl

-

-95

4

i

C

1

i!

8

4

F

---0)

-1.

i

l

I

Y

1

---'I2

e

1

revolution

-20

t

LO

80

120

160

ms

'LOO

t

shown in Fig. 9.5.Jb. Note the wakes of the four runner vanes.

The

presence of

tip

clearance vortices could be recognized by the large fluctuations in a measuring station

close to the runner tip.

At the outer cylindrical section

s/s,

=

0,32

the pressure fluctuations were found to

have

a frequency equal to that created by the gates. This means that the viscous wakes due

t

the gates extend even past the runner, naturally thcn sliced into pieces.

Fig.

9.5.5

shows the comparison of the time-averaged (for one runner revolution) resul

of the dynamically measuring probe with those of the

Conrad

probe for measuring plan

VIA.

364

F;~,

0.5.5.

Co~np;trison of

timc-:lvcr-

jfL,~

1

L,l"~-il!.

cnti1;wnculs

;]I

11:c

(11-aft

illlct

(s).

h~;ld

If

:=

2.8

rn.

point

1,

in

,,]allc

S.

----

timc-;lvcr;~gccl

vnlucs

of

dy~i:~n~ic

mc;~.;ul-emcnt

with

thc

dcvelopcd

hy

Curlo~pR

16.291.

-

_

timc-avcr;~gcd ~l~t;:si-ht:l~ion;~ry

,,,,,~~~1-~1ncnt

I

usrtal

6-whole

,,,,,ic;ll

17rol)c:

ant1

rccortfs

by

mercury

llln,,~~netc~~s.

After

('(I.sIo~~)/I

[6.?9].

For the c!etes~~iin;ttion of the influence

of

the gi1tt.s' pitch and the non-uniform inflow from

the

semi-spiral casing

on

the rclativc vclocitj; ficlcl,

a

PI-arldrl

tube. Fig. 9.5.6a, also

equipped

\\lit11

two

Kulitc p!-essul.c tsunsducers. has been fixed

011

thc runner face. The

supply

voltage

its \\.ell

as

the signills of the transdl~cers have been transmitted from the

rotating system

through

311

eight channel ~nercury-opcrated ring chamber.

For cr~librntion the timc-:tvcr-ased valucs of the ports within the relative flow field have

been

compared

\\:it11

thc v;tli~cs indicated by respective ports under air injection.

9.5.3.2.

Espcri~~icntal rcslilts

from

tl~c

rotating probe

Fig.

9.5.6b

shvws the pr-essure distribution for two different positi~ns given

by

the

tiansduccrs of the

Prtrr~tltl

probc.

At

thc

outer section

s/s,

=

0,25 it can be recognized,

that

the frequency of

!hc

fluctui~tions of total pressure ecluals that of the gate number.

Addi~ioiially thc i:lck of a~isynimetry in the flow due to the semi-spiral casing may be

noled.

The pressure fl~ictu:itions due to the gate pitch. have been fo~111d

ai

the outer sectior~ to

bc

eq~ii~l to velocity floctu:ltions of

F

I

to

3

%

of

the

average value. The computation of

--

----.

Fig.

9.5.6.

Measurement with a quick re.

Rtat

-*

sponse probc

on

the runner blade of

a

KT-

a)

~ranhtl tube on

thc

runncr vane face

in

the

longit~~dinal section and cross section, a air

pro,

~a~,hp,.~-vdh+~ci,~'i,l.c.-S~~+-

ejection for calibration of plot on site,

b

air

/

I

ejection for calibration ofp,,,,, on site;

c

Kulite

2-

191

transducer;

d

runner vanc face. b) Przssure

i

o

T&+/---

7

:,

Test point

I,

head

I1

=

2,8

m;

s/s,

=

0,25;

records

of

the

two

transducers

of Prandtl tube.

0

UI

-

-401~

-

gu~de

vane pltch

above pr:ssurt: face; below suction face. After

-

24

gu~de

vanes

TIOM~I

Castorph

[6.29].

-

wakes according

to

Kelitps

and

Sears

[6.30]

lends to velocity fluctuations depending

on

the gate pitch, which agree approximately with the values measured

I*).

9.6.

Ir~fluence

of

runner vane number on relative

and

absolute

flow

in

axial turbines according to air tests and comparative-predictions

carried

out

by

W.

Kiifznel

9.6.1.

Formulation of the problem

The

main

aim of the following treatise

on

thc one hand is to obtain an experimental proof

of the real relative and absolute flow field

in

an axial reaction wheel with the typical

design

features

of

an hydraulic

axial

turbine.

j

")

This seems

to

be

slnall.

Rut

in addition to this. there also exist velocity fluctuations up to

4%

which

independent of the guide vane pitch. This results in \elocity fluctuations of up

to

f

7%

on

the

pressurc

ad

suction face

of

the blade. Thereby the tluctuntinn

,,I'

the differentia! pressure may reach up to 14% of the

average value.

366

Raabe J. Hydro power - the design, use, and function of hydromechanical, hydraulic, and electrical еquipment

Подождите немного. Документ загружается.