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

Подождите немного. Документ загружается.

which may entail starting and stopping 10 times a day. With a reversible set

the following additional factors affect the generator design: hydraulic machine character-

istics, dual rotation, methods of starting the units in the pumping mode.

The present insulation design provides a built-in overload capacity, which can sometimes

be used by operating the turbine at a gate opening greater than the specified one. Pumped

storage involves frequent mode changes, starts and stops, and even reversal of rotation

and in aggregate these will amount to several thousand significant load changes per

nnnum. The epoxy-resin-based insulation system is designed to withstand this rapid

cyclic duty.

11.4.1.5. Special machines for low

head

The rim generator of Straflo turbines (Cap. 10.2) could prove to be the economic means

of generating electric power from the tides at selected sites, including several river estua-

ries, in the

UK.

Ratings of up to 50

MVA

at 60 rpm have been considered. Such generator

would have a rotor diameter of 12

This design combines excellent cooling by the

discharge passing the rotor bore with an easy accessibility of the alternor. Highest

possible natural fly-wheel mass ensures stable regulation.

11.4.1.6. Operating

regimes

In the past, very many hydroelectric installations were built as base load stations, but

with the changing pattern of generating capacity in an interconnected system, attention

is being given to re-assessing the role of these existing machines. In many cases it would

be more

ecor~omical to operate the sets as peak-load stations with the added security of

quickly available standby capacity.

achieve this, would mean increasing the installed

capacity by 25 to 50 per cent

and running at reduced load factor

cp;

see also [11.61].

Increasing the rating of the station could be achieved in a number of ways among which

uprating thc existing generators is immediately attractive. Replacing the existing bitumen

based stator windings by modern epoxy windings would permit a substantial increase in

permissible rating

both by increasing the volume of copper ir. the machine and by

providing an insulation system of much improved thermal conductivity.

11.4.1.7. Computerized design

Computer techniques for the electromagnetic and mechanical ciesign calculations are

increasingly being developed and these enable more detailed design studies to be carried

out before the final design is established

[11.60].

11.4.2.

Alternator, eIectric features

The alternator consists of

rotor as a carrier of

magnetic field with a magnetic flux

This is obtained by a certain numberp, of pairs of

excited poles uniformly distributed

around the circumference of

the rotor.

The rotor is driven by the hydroturbine. This has a speed

n[rpm], resulting from its head

and its essentially head-conditioned unit speed

n,,

(see Cap. 9.2) and a noininal

diameter

(from now on instead of

which henceforth denotes the diameter of the

alternator rotor) of the turbine runner. Thereby the magnetic field of the poles, uniformly

distiibuted around the circumference, induces in the

windings of they, pairs

poles

an effective internal voltage (in volts)

j),.@,

(11.4-1)

whcrc

C,,

is consLil~lt

9,517

atld

thc e~ectivc tn;~gnctic flux in

i~lor,,,,,,,

I-~;~rr~lonic;~ll

varying around the circumferunce.

'rhc

grid frequency

the number of pole pairsp,, and 11-~e spccd

or the angul

of set

(I,

are relatcd by

r velocity

In rarc c;tses \vith larse variations of head and thereby widely changing optimilm

speed

,I,

the speed a!ld the number of pole pairs either in the rotor or in the stator of alterllator

may

be changed at

\\ill

in two steps, especially when the ~naclline is a motor-generator

a pump-tarrbine uscci for turbine and pump operation, scc Cap. 11.4.7.

When

the alternator is loaded with its current

then thc internal voltage

differs from

the effective voltage

at the alternator termirlals by the following voltage drops to be

subtracted vectorialiy (Fig.

1.4.2): 1)

nf(X,

X,)

I, due to the synchronous reactance

and with

90-

lead against the current I, where

is the reactance due to the stray field

of the stator and

that due the armature reaction,

due

to ohmic resistance

alternator circuit (usually unimportantly small) in phase with the current

f).

Fig.

1.4.2.

Vector diagram cf an alternator.

cffcctive magnetic

flux;

efTcctive voltage at the generator

terminals;

effective current;

reactance due to thc armature rcacrion;

X,,

reactance duc to stray

field of the stator;

phase shift angle;

effective inner voltage.

Qbviously

(Fi_e.

11.4.2)

phase shift

between

and

results from this. 'The internal

{oltage

depencis

the

magnetic

flux

and hence can be regulated by variation of the

elcitinz ctirr'cnt

I,,

This is

and supplied by the exciter, usually

stationary device,

but sometimes

rotary machine at the shzft's end [11.77; 11.781.

From the vector diagram

(Fig.

11.4.2) it is seen, that

the case of parallel operation

alternators. the ansle

between

and

depends on the current

is too large, the

alternator

may

fall out oT step, which must be avoided by regulation.

The phase angle

be:ween

and

and hence the power factor cos

depend on the

excitation current. Therefore the latter has to be regulated

function of tlie load.

Us~~ally three phase current is generated, occasionally also single phase current for

traction pot1-er-.

the case of a three phase alternator, the terminal output results fr~m

the carrcnt

and eit!ler

tile

star voltage

L',

or the delta voltage

3~,1ccs~

~~IU,COS~.

1.4-

Resides regulilr operation sometinies the idling generator is used to balance the reactance power on

the

line

so-cnl1c.d 'condenscr' or 'phase shiftcr' operation, which thc emptied hydraulic machine

has

to \vithstand

liO.1381.

"spinning reserve" also standby operation is practised.

The

terminal vo:tagc

has

kept constant

regulation of tlie exciter. The operating voltage

ranges

fro111

frequencies

16%,

Hz.

For smaller machincs standardized

voltages are recommended. Optimum magnetic and electric utilization of the machine is achieved.

when the voltage can be freely

selcctcd. According to experrence, the voltage increases with the

output

This can be approximated by a logarithmic law

PIPo

In(UiU,),

where

and

are the

output and

voltagc of a certain plant.

Excepting traction power with its grids of 16% and

Hz,

the frequency of AC grids is standardizcd

at 50, or 60

Hz.

Therefore regulation of an alternator to cover a certain load demand implies

constant terminal voltage under nearly constant speed according to

1.3-2) and the proportional

band (permanent speed droop) of the governor (see

11.2).

11.4.3.

Design features with respect

critical speed

General remarks: The design of alternator rotor, its shaft and bearings is influenced by

the effort, to shift the critical speed to the desired high level (usually above runaway) at

reduced cost. Therefore low head sets with low runaway speed show in general the two

bearing design with an overhung arrangement of turbine runner

and alternator rotor. In

order to reduce the distance of the rotor's centre of gravity from the guide bearing centre,

the umbrella or semi-umbrella design (Fig.

10.2.1.3)

is preferred.

The shortening of this distance also under higher head, when the large rotor mass then does not

allow this design, is one of the decisive needs for a high critical speed. Moreover such a speed requires

also a small mass of alternator rotor. As the latter is always larger than that of turbine runner, the

design has to be focussed on this small distance between rotor and its adjacent guide bearings.

The dominating mass of alternator rotor is a typical feature of hydroturbine sets especially in the

high head range. It requires corresponding machine tools. partly fabrication on site, a stress relief

oven there, or transportation facilities with sufficient loading gauges, also cranes at the shop and site

respectively, depending on the construction and the mode of rotor fabrication (in shop or on site)

as well as on the existence of a satisfactory erection bay.

11. Fly-wheel effects: The big rotor mass suits the large moment of inertia often required

for stable speed control and for limitation of overspeed after sudden load rejection. This

holds especially for smaller and high-speed units.

The fly-wheel effect of the rotor also effects the torsional vibration and thus the dimensioning of the

shaft in

some cases. The shaft diameter should then be selected so that no danger occurs in thccase

of short circuit or faulty synchronization and the critical torsion spccd range

avoided when thc

torque varies.

If the synchronous machine can be dimensioned for optimum electrical and magnetic utilization,

without mechanical overstressing, the rotor has the so-called natural fly-wheel effect.

If a fly-wheel effect greater than

the natural one is required, it

necessary to increase the bore

diameter so far as mechanical stresses permit (so-called fly-wheel type rotor).

II'

is desired to retain the borc diameter corresponding to the natural fly-wheel effect, the required

can be achieved by an additional fly-whcel with a diametcr larger than that of the stator bore.

For reasons of accessibility in vertical sets, this fly-wheel is installed underneath the bore. It is then

no longer possible to install and

dismantlc the complete rotor

fly-wheel) through the stator bore.

The physical origin of the

largc rotor mass in relation to that of the turbine runner, especially striking

at high-head machines, results on the one hand from the modest strength of the magnetic field in

iron even under saturation. On the other hand it results from the limited admissible density of

current in

the conductor, to avoid heat stresses and damage of insulation by waste heat. As a result

the force between rotor and stator is combined with a rather low pressure

bar) of the

conductor bars on the radial walls of the notches. This could be overcome by cryogenics.

Against that the hydraulic reaction turbine at a

head

has a

differential

pressure about

the rotor vanes, yielding up to 70 bars. Moreover the propelling force on the

runner is extended over

the-whole

vane-fillcd space of the rotor, whereas in the alternator rotor it is limited to the surface

near thc gap between rotor and stator.

111.

Shaft ditlmetcr duc to critics1

speed

nnil

rotor mass:

The

shaft di:lmclcr is

usual,

dctermincd

the nccd, to set tlie lowest critic:ll

speed

u),,

due

to flcxural vibrations

somc~vh:lt

~I!>OVC

tlie ruIl:I\il:Iy speed cora (scc also Cap.

10.5).

According

the similarity

Ia~vs (Cap.

9.2)

or,

z3;'

K~~(~~IIJ)'~~/D~,

(1 1-4-4)

where D, is from now

the runner diameter of the turbine, Ktr its rated blade speed

coefficient, and

its runa;Nay to rated speed ratio (a type-conditioned figure).

The critical speed

(I),,

mainly determined by the imaginary deflection f.of the horizon,

tally assumed shaft under the weight nt, of the alternator rotor, according to

The mass of the rotor, density

Q,,

axial length

diameter D (from now on), with

LID

as a design feature, may be expressed by

n1,

D~(L/D).

(1 1.4-6)

With a distance betweeti the centre of the bearing and the rotor proportional to

an area

moment of inertia of the shaft, diameter

cl,,,

proportional to

d:,,

the shaft deflection

becomes

(~/&,h)~

L2!E, (1 1.4-7)

where

is Young's modulus of the shaft. Inserting this in (1 1.4

5),

gives the lowest order

critical speed

GI,,

const (d,,,/~)~ (Dl

(E/Q,)'/~/L.

(1 1.4-8)

Introducing the mass of the turbine runner, diameter

DT,

axial depth L,, density Q,, in

the same

way

as that of thc alternator rotor after (1 1.4-6), the ratio of the

rotor masses

reads

nl,/nlT

(LID) (LT/DT)-

DID^)^.

(11.4-9)

Safe operation requires

wCr/ora

const (in general)

(11.4- 10)

Inserting

1.4-4) and (1 1.4-

1.4- 10) and expressing

DID,

from (1 1.4-9) brings

the shaft to rotor diameter ratio

The upper limit of

d,,/D

depends on the rotor type (Cap. 11.4.7).

11.4.4.

Dimensioning the alternator rotor

Diameter of the alternator rotor: Calculating the diameter

has to start from the

turbine-conditioned runaway speed

o,,

being

times the rated speed

(hence in terms

of angular velocity).

Thus

o,,

io.

Rotor dimensiolling may start from the admissible

stress

o,,

critically stressed element, usually on the outerrilost diameter

Accounting

for fatigue and notch effects by a factor

then in a limiting design

where

the

mean rotor density.

Then the diameter

of the turbine runner follo~vs from the head

and the

flow

the

hence given type number

(Cap.

9.2),

the resulting ;lngular \docit)-

nrld blade speed

KLI

by means of

facilitate, in the case of a vertical set, dismantling and assembly of the runner (for

reasons of submergence at the bottom of the set's pit) by lifting

through the bore oT the

alternator stator, the following condition must be satisfied

(if

isnoring step up gears.)

where

DT, is the maximum runner diameter. This is usually satisfied

dcsizns of very

low

head are exempted. On the other hand

and with

the outmost stator diameter

allow a convenient placing of two adjacent alternators

\vithin a distance glven

the lateral width required by the spiral chamber or its equivalent (see Cap.

4.2).

11.

The iron length

of the alternator rotor: On the one hand this results from the rotor's

required moment of inertia

Expressing this by

I;,

LDJ.

\~.ith

design-

linked parameter, accounting for the start up time of set

required (Cap. 11.3), with the

parameter

k3 due to the torque vs speed graph and the shaft efficiency

11,-

of the turbine.

the equation of motion of the starting set gives for the iron length

rotor, which has

the so-called natural fly-wheel effect, the following expression

With the requirements of magnetic flux and current coverage of the alternator. the latter's

output

PTqG, ils speed

30w/n, the pole pair numberp,,, the rotor length ho~vevsr

becomes

P,/(C~~D~~)

nP,/(30 p,D20C).

1.4- 16)

The parameter

W/(m3 rpm)) depends on such features as the permissible temperat lire

rise

AT,

the termillal voltage and an iron fillifig factor for bundle lamination

the stator

1.831. In the case of water cooling C may be raised by about 60

9.b.

w,l

and

in (11.4-2),

in (11.4-15j and

1.416) and

(1)

1.3-12).

(11.4-13).

(11.4- 15), (1 1.4- 16) must be compatible with each other in

limiting desi2n.

11.4.5.

Alternator

output as a

function

working

data

The output: By means of the similarity laws

(Cap.

9.2).

the flow

reads:

Q1,

D+(g~)112, where Q,, is the unit flow and D, the runner diameter of the turb~ne.

Introducing tl~is in (1 1.4-

15)

gives the iron length

which the turbine nceds for

certain

start up time

of the set due to

stable regulation. Putting this in

1.4-

16).

and

using

for the number of pole pairs

p, the relation (1

1.4-21,

the generator output becomcs

should be noted that

has the dimension kg/(m/s2) (in practice

mirl,'m3).

Accounting for

1.4- 8) and

1.4-4), the required distance of the lowest critical speed

due to flexural vibrations of the

set from its runaway spced according to

1.4-

10)

yields

This in

1.4- 17) gives the alternator output the form (Q,, due to rated flow)

!tic 4iniiInrit; I;i\vs

(3.2)

the ripccific hcnd

can

bc rcduccd to lhc runner blad

cuclfir.ic11

(rn

1),'2)'1(?

~11').

Tlic strength of thc gcncr:ltor rotor under runawal

requisc\ t~ccordlng

(

1.4-

I?)

tile

followit~g linkepe of the circunlfcrclltial rotor

tip

spccd

lliZ

;tntl

the admissible stress

o,,,,:

((0

D/2)'

o,,,/(kp,

i2),

where

is the

rotor

clenhity

and

the runaway to rated specd ratio. Both the last relations give

Putting this in

1.4.-

19)

the alternator output becomes

Ir:

a limiting design, now considered,

Did,,

and

LID

are more or less constant. For a

reaction turbine the speed coeficient

and the runaway to rated speed ratio

rise

slightly with the type number

(Cap.

9.2).

It is seen, that

is proportional to the line

frequency

j' and

the design factor

The latter can be raised by

60%

in the case of

water

cooling. Hence the output of an alternator, keeping cverytlling else constant, is mainly

inversely proportional to the squared speed

1n.a real desip. this holds only from

certain mitlimum speed on. Below the latter, the

output depcnds otherli ise on the speed

of the set. Strictly speaking in a li~niting design,

Pfr*)),

(see

Fig.

1.4.1).

Hence by means

t!~e sin~ilarity laws, the type number

can

expr:ssed as follows

function of the angular velocity

and

the head

In the

cas5 of a Iarsc reaction turbine. the submergence links a given head

to a certain

type

number

,I;

(Cap.

9.2).

given

)IT,

and

the last relation can be considered

11mitir.g relation for thc sped

The latter must be adapted to the line frequency

mcans of

1.4-

2).

Thcn the output follows by means

F(o)

according to

1.4-21).

The

output

furlctio~l of the mass ratio

1.4-9)

and cooling: Inserting

1.4-20)

ira

1.4-

gives

(tl,,,'D)'

(o,,/(k

E))

(IJ/D)10'3

(L=/D~)~/~

2'3.

(11.4-23)

This in

1.4-21)

shows that the output is inverse to

The latter reiation is valid only for an alternator whose outermost parts operate with

stress-lin1i:ed peripherai velocity

rmaway. Below a certain speed

and for a rotor

hose tip diameter

limited by manufacture and portability, the limit of alternator

output docs not depend in inverse proportion on the speed of the set (see Fig.

11.4.1).

\Vith the specific loss

It;,

the heat flow

h;P,

has to cross a surface

means

heat transfer coefficient

a temperature difference

'7:

Hence the squared diameter of

the alternator rotor

const

[It;

c/(a

T)]

DIL.

(11.4-25)

The

generatorlturbine rotor mass ratio

p*,

(11.4-9),

can be reduced to

LID,

D3,

LT/DT

and

D:.

Expressing

(11.4-25),

(11.4-13),

the mass ratio

reads

(")'I2

(4;")"'

const

L7-

aAT

~u~(gH)~"'

Inserting this in the relation

1.4-24), the alternator output of a limiting design. at given

head

gH,

reads as follows

const

(h~11k3)112(("314(;)"2(U~9~a~~

)

(5)

1.4

27)

Thus the alternator output increases by raising a) the overheating

(limited by insulation

material), b) the resulting heat transfer coefficient

Hence, 60% increase in output by water cooling

reflects 150% growth in

Since the heat transfer coefficient between coolant and metal for water

is about 100 times that of air, the above demonstrates a predominantly diminishing influence of the

heat conduction term in the iron and copper on the value

General consideration: From (1 1.4-

16)

and

1.4-2) the output of an alternator with a grid frequen-

can be expressed by

const

CD2

Lf.

For homologous machines

-.D

and at

given grid frequency, the generator output becomes

proportional to its volume, namely

D3.

The alternator's power loss is mainly proportional to

its output and mainly converted into heat flow

Q:,.

Hence

Q:,

D3.

a given permissible temperature difference

between the alternator and the coolant entering,

and at a given heat transfer coefficient

this heat flow has to pass the alternator's surface

Hence

Q:,

With the proportionalities

and

Q,+,

D3,

the resulting heat transfer coefficient

required becomes

D/A

Thus cooling of an alternator beconlcs more difficult with increasing physical dimension of the

machine.

11.4.6.

Cooling

practice

General survey: The purpose of cooling is temperature control of zones with heat loss

(conductor, magnetic materials), and its components which may be damaged by excess.

heat

[11.75].

Overheating is established by a series of temperature rises. In the following this may be

shown

in the case of air cooling. The temperature of stator winding can be divided into

the following factors:

air from coolers to air entering the ventilation ducts,

air entering the ventilation ducts to the iron,

iron stator copper winding.

The first is due to the heat carried in

tlie air before reaching the stator ventilation. The

second is

function of the heat transfer coefficient from iron to air, that is on the velocity

of the coolant and the shape of the ventilation ducts. The third cause of temperature rise

results from construction techniques.

To establish as accurately as possible the intensity

these types

heat exchange,

complete knowledge

coolant flow is necessary, i.e.

fluid flow calculation to evaluate the pressure required for coolant circulation.

Heat flow calculation.

Fan or impeller pump system design.

II.

Ti~c coolilnt circuit for air cooling: The circuit clesi~n depcncls or1 1-nrinus fc;l[1lrcs

ths

rilnc:lillc.

\Vitll

air ;is

coolant nltcrilators can be divided inlo

tllc

fo!lowing groups

ligh spoed ~e~lcriltors (small

diameter,

low pole numbers).

Slow

gcilcrators (I:lr~c di:lincter, large pole numbers).

Reversible

motor-gcncrntors for pump-turbines and tlon reversible rl~otor-gencrat~~~

for ternary pumped storage scts.

In air-cooled machines

distinction is also made between axial arid radial ventilation,

where the first is preferably employed in the group 1) and the latter in group 2). Both

methods of cooling can be used in an open or closed system.

With

axial ventilation

(Fig,

11.4.3.11, the air is supplied predominantly by a radial

fan.

The bore and parts of the outer surface of the active iron are cooled directly. This alsd

applies to the pole winding

the gaps between the poles. Heat is removed from the stator

winding indirectly

over

the length of the axial slot, whereas the end windings are cooled

directly but have different temperatures corresponding to the temperature rise of the air.

within the axial slot.

With radial cooling (Fig.

11.4.3.2), the air is blown symnletrically into the gaps between

the poles from both sides

axial fans. 'The centrifugal effect of the rotating poles diverts

the air in the radial direction

and into the cooling air ducts of the stator iron. The stator

end windings lie directly in the cooling flow from the fan.

With larger bore diameters the air can also flow radially through the rotor (group 2).

Additional vanes may

bc providcd in the spaces between the rotor rings to assist the flow.

111.

~ooli'n~ systems:

Open

systerns

for smaller ratings withdraw the cooling air from the

turbine

hall

(Fig.

11.4.3.3)

and discharge

into the maclline pit. In the case of greater

Fig. 11.4.3.

Air

coolin_p s!steinj:

Cooling with axial ventilation by a radial flow fan;

cooling with

radial vtntil;ltion by

axial fans;

open system, cooli~lg air is drawn in from the turbine hall and

expelled into the

generator

hall; 4 cooling air supplied from and discharged to

the

atmosphere;

re-coolin_c of ventilating air by water/air heat exchangers;

air supplied by a separate motor fan

unit for reversible units. (Drawing from Brown

Boveri and Cie: Synchronous machines for hydro-

electric power plants.

Spccial issue publication No

CH-T

130082

E.)

power loss the cooling air is preferably drawn from the :~tmosphere via filters and ducts

and

subsequently discharged

(Fig.

11.4.3.4).

The air can be discharged through openings

the warm air duct and used for he;~ting the power house if necessary.

The closed system is used when the atmosphere is polluted and for high output machines.

The air circulates in a closed circuit and is re-cooled by waterlair heat exchangers

(Fig.

11.4.3.5).

It is possible to maintain an almost constant operating temperature by

metering

the quantity of cooling water or mixing warm water with the cold water

according to the load on the machine. The heat exchangers are either flanged

the stator

frame or let into the foundation cooling ducts.

synchronous machines for reversible operation (group

and pole changing two-speed

machines

are ventilated separately. The cooling air is supplied by motor driven fan units

(Fig.

11.4.3.6).

~urther efforts to increase the specific output and improve the utilization of the machines

have led to the application of liquid cooling. The economic advantage of liquid cooling

dependent on the evaluation

losses. Its application is particularly justified for high

outputs and high speeds. Moderate values of fly-wheel effect (bulb generators) permit

full

utilization of the advantages of liquid cooling.

IV.

Water cooling of the stator: In the stator winding bars are inserted, consisting of

transposed conductor elements (according to Roebel's principle)

hollo\v conductors ior

Fig.

1.4.4.

Devices for

water-

cooled system:

a)Section

through a water-cooled Roebcl

bar;

solid conductor

elcment;

insiilation of

conductor ele-

ment;

cooled conductor cle-

ment

(hollow conductor);

half

insulation;

filling;

semi-

conductive

potential

coating;

nlain

insulation;

bcnd un-

derlay.

Transfur

unit

(watcr

head)

for feed and discharge of

thc

cooling

water

to and from

the rotor winding. c) Rotor coil

With

square, hollow copper con-

ductor. (Drawing courtesy

Brown, Bovcri

Cornpanic

Ltd.

Badcn, Switzerland.)

thc cooling watcl- (Fig. 11.4.Ja). The cooling watr:r is supplied to

the

st:\tor wind'

and rernovcd by a closed circuit system of pipus or

ring main spjlcrn arranged

the wlncli~igb

thc stator frame. The insulation bclwccn thc stator windings

cooling systcm

provided by Tcfloll hoses.

The iron losses of the stator laminatiol~ are likewise relnovcd by coolinp water.

Water cooling of the rotor: The cooling water is kd to the rotor winding and

remo,,o;i

via

transfer unit on the end of the shaft (Fig. 11.4.4 b). concentric stainless steel

in the shaft bore carry the treated cooling water via flexiblc conncctioas to the manifoldr

nloui1ted on the rotor. The water then flows via insulating boses into the hollow conduo

tors of the pole windings (Fig. 11.4.4~). The iiquid cooling sysl-m consists of two circui(r

The first circuit supplies treated, i.e., de-ionized, oxygen-free water Lo the stator and

rot&

winding system. The other circuit supplies untreated fresh water to the stator end

plalo

and the back of the stator iron.

The conductivity of the treated water, the relative llu~nidity inside the stator frame

the circulation of the liquid are monitored continuously.

Fig. 11.4.5 a shows the sectional drawing of the

fi~lly

water-cooled alternator at

~~~~t~b,

Ncrway, built by

BBC,

to prove the reliability of this design. Fig. 11.4.5b shows

perspective sectional drawing of the alternator ltaipfi, Rio Parank Brazil, at present one

of the largest water-cooled alternators in the world, built by

BBC

and Siemens.

But

here,

only the stator is water cooled.

A.7.

Rotor construction

The output, runaway speed, and the required fly-wheel effect largely determine

the

construction of rotors. Assembly and transport limitations are often additional criteria

which affect rotor constructio11. The shaft dirnensions are determined by the critical

flesural and torsional speed, which must be in adequate distance from the interfering

excitation frequencies. Fig.

11.4.6 shows different bearing arrangements.

Rotor body: Rotors for small number of poles and sniall bore

diameters

are constructed

as solid forgings (Fig.

11.4.7.1). The shaft and central body form one unit.

rotors with

a larger number of poles and larger bore

dianleters the heavy central section is replaced

cast or forged hollow body with shaft ends flanged on to both sides (Fig.

11.4.7.2).

Rotors for large capacities and medium number of polcs are preferably constructed

disk rotors (Fig. 11.4.7.3). The central body is formed of several rolled steel plates centred

in relation to each other by spigots or mating rings and compressed by prestressed

clalnping bolts. The shaft ends are also flanged to the rotor body by means of clamping

bolts.

The central section can be replaced by

cast steel spider with several shrunk-on forged

steel rings (Fig.

11.4.7.4). Rotors for larger diameters and medium to low speeds are

constructed

with laminated rims. They have a continuous shaft with a shrunk-on spider,

or shaft ends flanged to the latter (Fig.

11.3.7.5). The spider consists of a cast, forged or

welded hub with attached arms, which can

onscrewed or are welded to the hub

depending on transport limitations. The rim consists of segments of steel laminations

each several

millinleters thick. l'he punched segments overlap each other and are corn-

pressed by clamping bolts

form a ring resistant to bending. With the usual overlapping

of one pole pitch. at the joint between the individual segments, the critical cross-section

for the strength analysis deviates less from the full ring cross-section, the greater

the

number

pole pitches in

segment.