Electrical Machine Design

Chapter.1 PRINCIPLES OF ELECTRICAL MACHINE DESIGN

Introduction
The magnetic flux in all electrical machines (generators, motors and transformers) plays
an
important role in converting or transferring the energy. Field or magnetizing winding of rotating
machines produces the flux while armature winding supplies either electrical power or mechanical
power. In case of transformers primary wing supplies the power demand of the secondary.

The basic design of an electrical machine involves the dimensioning of the magnetic circuit,
electrical circuit, insulation system etc., and is carried out by applying analytical equations.

A designer is generally confronted with a number of problems for which there may not be one
solution, but many solutions. A design should ensure that the products perform in accordance
with the requirements at higher efficiency, lower weight of material for the desired output, lower
temperature rise and lower cost. Also they are to be reliable and durable.

A practical designer must effect the design so that the stock (standard frames, punching etc.,) is
adaptable to the requirements of the specification. The designer must also affect some sort of
compromise between the ideal design and a design which comply with manufacturing conditions.
A electrical designer must be familiar with the,
a. National and international standards
Indian Standard (IS), Bureau of Indian Standard (BIS), India
British Standard (BS), England
International Electrotechnical Commission (IEC)
NEMA (The National Electrical Manufacturers Association).

b. Specifications (that deals with machine ratings, performance requirements etc., of the
consumer)
c. Cost of material and labour
d. Manufacturing constraints etc.

A designer can refer to Design Data Handbook (Electrical Machine Design Data Book, authored
by A Shanmugasundaram and others , New Age International Publishers, Reprint 2007, or any
other such handbooks) which is a source of design procedure, properties of materials, ranges of
design parameters etc., and manufacturer’s brochure.

As the design involves a number of assumptions and constraints, final design values can be
obtained only by iterative methods. Computer plays a vital role in arriving at the final values. By
Finite Element Method (FEM), the effect of a single parameter on the dynamical performance of
the machine can be studied. Furthermore, some tests, which are not even feasible in laboratory
setup, can be virtually performed by Finite Element Method.

The design problems, that have been considered to solve in the latter chapters, are of different
nature from the design worked out in detail in respect of any machine. However, these test

1
problems provide adequate elementary skills in design, which is an indication that a student has a
fair knowledge to deal with the entire design.

Factors for consideration in electrical machine design

The basic components of all electromagnetic apparatus are the field and armature windings
supported by dielectric or insulation, cooling system and mechanical parts. Therefore, the
factors for consideration in the design are,
1. Magnetic circuit or the flux path: Should establish required amount of flux using
minimum mmf. The core losses should be less.
2. Electric circuit or windings: Should ensure required emf is induced with no
complexity in winding arrangement. The copper losses should be less.
3. Insulation: Should ensure trouble free separation of machine parts operating at
different potential and confine the current in the prescribed paths.
4. Cooling system or ventilation: Should ensure that the machine operates at the
specified temperature.
5. Machine parts: Should be robust.

The art of successful design lies not only in resolving the conflict for space between iron,
copper, insulation and coolant but also in optimization of cost of manufacturing, and operating
and maintenance charges.
The factors, apart from the above, that requires consideration are
a. Limitation in design ( saturation, current density, insulation, temperature rise etc.,)
b. Customer’s needs
c. National and international standards
d. Convenience in production line and transportation
e. Maintenance and repairs
f. Environmental conditions etc.

Limitations in design
The materials used for the machine and others such as cooling etc., imposes a limitation
in
design. The limitations stem from saturation of iron, current density in conductors,
temperature, insulation, mechanical properties, efficiency, power factor etc.

a. Saturation: Higher flux density reduces the volume of iron but drives the iron to
operate beyond knee of the magnetization curve or in the region of saturation.
Saturation of iron poses a limitation on account of increased core loss and
excessive
excitation required to establish a desired value of flux. It also introduces harmonics.
b. Current density: Higher current density reduces the volume of copper but
increases
the losses and temperature.
c. Temperature: poses a limitation on account of possible damage to insulation and other
materials.
d. Insulation (which is both mechanically and electrically weak): poses a limitation on
account of breakdown by excessive voltage gradient, mechanical forces or heat.
2
 e. Mechanical strength of the materials poses a limitation particularly in case of large and
high speed machines.
f. High efficiency and high power factor poses a limitation on account of higher capital
cost. (A low value of efficiency and power factor on the other hand results in a high
maintenance cost).
g. Mechanical Commutation in dc motors or generators leads to poor commutation.
Apart from the above factors Consumer, manufacturer or standard specifications may pose a
limitation.

Materials for Electrical Machines

The main material characteristics of relevance to electrical machines are those associated with
conductors for electric circuit, the insulation system necessary to isolate the circuits, and with
the specialized steels and permanent magnets used for the magnetic circuit.

Conducting materials
Commonly used conducting materials are copper and aluminum. Some of the desirable
properties a good conductor should possess are listed below.
1. Low value of resistivity or high conductivity
2. Low value of temperature coefficient of resistance
3. High tensile strength
4. High melting point
5. High resistance to corrosion
6. Allow brazing, soldering or welding so that the joints are reliable
7. Highly malleable and ductile
8. Durable and cheap by cost
Some of the properties of copper and aluminum are shown in the table-2.
Table-2
Sl.
Particulars Copper Aluminum
No 2
1 Resistivity at 200C 0.0172 ohm / m/ mm 0.0269 ohm / m/ mm2
2 Conductivity at 20 C
0
58.14 x 106S/m 37.2 x 106S/m
3
3 Density at 200C 8933kg/m 2689.9m3
4 0.393 % per 0C 0.4 % per 0C
Temperature coefficient Explanation: If the temperature increases by 1oC, the
(0-100oC) resistance increases by 0.4% in case of aluminum
5 Coefficient of linear 16.8x10-6 per oC 23.5 x10-6 per oC
expansion (0-100oC)
6 Tensile strength 25 to 40 kg / mm2 10 to 18 kg / mm2
7 Mechanical property highly malleable and not highly malleable and
ductile ductile
8 Melting point 10830C 6600C
0 0
9 Thermal conductivity 599 W/m C 238 W/m C
(0-100oC)
10 Jointing can be easily soldered cannot be soldered easily

3
For the same resistance and length, cross-sectional area of aluminum is 61% larger than that of
the copper conductor and almost 50% lighter than copper.
Though the aluminum reduces the cost of small capacity transformers, it increases the size and
cost of large capacity transformers. Aluminum is being much used now a days only because
copper is expensive and not easily available. Aluminum is almost 50% cheaper than Copper
and not much superior to copper.

Magnetic materials

The magnetic properties of a magnetic material depend on the orientation of the crystals of the
material and decide the size of the machine or equipment for a given rating, excitation
required, efficiency of operation etc.
.
The some of the properties that a good magnetic material should possess are listed below.

1. Low reluctance or should be highly permeable or should have a high value of relative
permeability r.
2. High saturation induction (to minimize weight and volume of iron parts)
3. High electrical resistivity so that the eddy emf and the hence eddy current loss is less
4. Narrow hysteresis loop or low Coercivity so that hysteresis loss is less and efficiency of
operation is high
5. A high curie point. (Above Curie point or temperature the material loses the magnetic
property or becomes paramagnetic, that is effectively non-magnetic)
6. Should have a high value of energy product (expressed in joules / m3).

Magnetic materials can broadly be classified as Diamagnetic, Paramagnetic, Ferromagnetic,

Antiferromagnetic and Ferrimagnetic materials. Only ferromagnetic materials have properties
that are well suitable for electrical machines. Ferromagnetic properties are confined almost
entirely to iron, nickel and cobalt and their alloys. The only exceptions are some alloys of
manganese and some of the rare earth elements.

The relative permeability r of ferromagnetic material is far greater than 1.0. When
ferromagnetic materials are subjected to the magnetic field, the dipoles align themselves in the
direction of the applied field and get strongly magnetized.
Further the Ferromagnetic materials can be classified as Hard or Permanent Magnetic materials
and Soft Magnetic materials.

a) Hard or permanent magnetic materials have large size hysteresis loop (obviously
hysteresis loss is more) and gradually rising magnetization curve.
Ex: carbon steel, tungsten steal, cobalt steel, alnico, hard ferrite etc.
b) Soft magnetic materials have small size hysteresis loop and a steep magnetization
curve.
Ex: i) cast iron, cast steel, rolled steel, forged steel etc., (in the solid form).
-Generally used for yokes poles of dc machines, rotors of turbo alternator etc., where
steady or dc flux is involved.

4
 ii) Silicon steel (Iron + 0.3 to 4.5% silicon) in the laminated form. Addition of silicon in
proper percentage eliminates ageing & reduce core loss. Low silicon content steel or
dynamo grade steel is used in rotating electrical machines and are operated at high flux
density. High content silicon steel (4 to 5% silicon) or transformer grade steel (or high
resistance steel) is used in transformers. Further sheet steel may be hot or cold rolled. Cold
rolled grain oriented steel (CRGOS) is costlier and superior to hot rolled. CRGO steel is
generally used in transformers.
c) Special purpose Alloys:
Nickel iron alloys have high permeability and addition of molybdenum or chromium
leads to improved magnetic material. Nickel with iron in different proportion leads to
(i) High nickel permalloy (iron +molybdenum +copper or chromium), used in
current transformers, magnetic amplifiers etc.,
(ii) Low nickel Permalloy (iron +silicon +chromium or manganese), used in
transformers, induction coils, chokes etc.
(iii) Perminvor (iron +nickel +cobalt)
(iv) Pemendur (iron +cobalt +vanadium), used for microphones, oscilloscopes, etc.
(v) Mumetal (Copper + iron)

d) Amorphous alloys (often called metallic glasses):

Amorphous alloys are produced by rapid solidification of the alloy at cooling rates of about a
million degrees centigrade per second. The alloys solidify with a glass-like atomic structure
which is non-crystalline frozen liquid. The rapid cooling is achieved by causing the molten
alloy to flow through an orifice onto a rapidly rotating water cooled drum. This can produce
sheets as thin as 10µm and a metre or more wide.
These alloys can be classified as iron rich based group and cobalt based group.
Maximum Saturation Coercivity Curie Resistivity
Material permeability magnetization A/m temperature Ωm x 108
µ x 10-3 in tesla oC

3% Si grain oriented 90 2.0 6-7 745 48

2.5% Si grain non -oriented 8 2.0 40 745 44

<0.5% Si grain non oriented 8 2.1 50-100 770 12

Low carbon iron 3-10 2.1 50-120 770 12
78% Ni and iron 250-400 0.8 1.0 350 40

50% Ni and iron 100 1.5-1.6 10 530 60

Iron based Amorphous 35-600 1.3-1.8 1.0-1.6 310-415 120-140

5
Insulating materials

To avoid any electrical activity between parts at different potentials, insulation is used. An
ideal insulating material should possess the following properties.

1) Should have high dielectric strength.

2) Should with stand high temperature.
3) Should have good thermal conductivity
4) Should not undergo thermal oxidation
5) Should not deteriorate due to higher temperature and repeated heat cycle
6) Should have high value of resistivity ( like 1018 cm)
7) Should not consume any power or should have a low dielectric loss angle
8) Should withstand stresses due to centrifugal forces ( as in rotating machines), electro
dynamic or mechanical forces ( as in transformers)
9) Should withstand vibration, abrasion, bending
10) Should not absorb moisture
11) Should be flexible and cheap
12) Liquid insulators should not evaporate or volatilize

Insulating materials can be classified as Solid, Liquid and Gas, and vacuum. The term insulting
material is sometimes used in a broader sense to designate also insulating liquids, gas and
vacuum.

Solid: Used with field, armature, transformer windings etc. The examples are:

1) Fibrous or inorganic animal or plant origin, natural or synthetic

paper, wood, card board, cotton, jute, silk etc.,
rayon, nylon, terelane, asbestos, fiber glass etc.,
2) Plastic or resins. Natural resins-lac, amber, shellac etc.,
Synthetic resins-phenol formaldehyde, melamine, polyesters, epoxy, silicon resins,
bakelite, Teflon, PVC etc
3) Rubber : natural rubber, synthetic rubber-butadiene, silicone rubber, hypalon, etc.,
4) Mineral : mica, marble, slate, talc chloride etc.,
5) Ceramic : porcelain, steatite, alumina etc.,
6) Glass : soda lime glass, silica glass, lead glass, borosilicate glass
7) Non-resinous : mineral waxes, asphalt, bitumen, chlorinated naphthalene, enamel etc.,

Liquid: Used in transformers, circuit breakers, reactors, rheostats, cables, capacitors etc., & for
impregnation. The examples are:
1) Mineral oil (petroleum by product)
2) Synthetic oil askarels, pyranols etc.,
3) Varnish, French polish, lacquer epoxy resin etc.,

6
Gaseous: The examples are:
1) Air used in switches, air condensers, transmission and distribution lines etc.,
2) Nitrogen use in capacitors, HV gas pressure cables etc.,
3) Hydrogen though not used as a dielectric, generally used as a coolant
4) Inert gases neon, argon, mercury and sodium vapors generally used for neon sign lamps.
5) Halogens like fluorine, used under high pressure in cables

No insulating material in practice satisfies all the desirable properties. Therefore a material
which satisfies most of the desirable properties must be selected.

Classification of insulating materials based on thermal consideration

The insulation system (also called insulation class) for wires used in generators, motors
transformers and other wire-wound electrical components is divided into different classes
according the temperature that they can safely withstand.
As per Indian Standard ( Thermal evaluation and classification of Electrical
Insulation,IS.No.1271,1985,first revision) and other international standard insulation is
classified by letter grades A,E,B,F,H (previous Y,A,E,B,F,H,C).

Insulation class Maximum

operating Typical materials
Previous Present temperature
in ℃
Cotton, silk, paper, wood, cellulose, fiber etc., without
Y 90
impregnation or oil immersed
The material of class Y impregnated with natural resins,
A A 105 cellulose esters, insulating oils etc., and also laminated wood,
varnished paper etc.
Synthetic resin enamels of vinyl acetate or nylon tapes,
E E 120
cotton and paper laminates with formaldehyde bonding etc.,
B B 130 Mica, glass fiber, asbestos etc., with suitable bonding
substances, built up mica, glass fiber and asbestos laminates.
The materials of Class B with more thermal resistance
F F 155
bonding materials
Glass fiber and asbestos materials and built up mica with
H H 180
appropriate silicone resins
Mica, ceramics, glass, quartz and asbestos with binders or
C C >180 resins of super thermal stability.

The maximum operating temperature is the temperature the insulation can reach during
operation and is the sum of standardized ambient temperature i.e. 40 degree centigrade,
permissible temperature rise and allowance tolerance for hot spot in winding. For example, the
maximum temperature of class B insulation is (ambient temperature 40 + allowable
o
temperature rise 80 + hot spot tolerance 10) = 130 C.

Insulation is the weakest element against heat and is a critical factor in deciding the life of
electrical equipment. The maximum operating temperatures prescribed for different class of

7
insulation are for a healthy lifetime of 20,000 hours. The height temperature permitted for the
machine parts is usually about 2000C at the maximum. Exceeding the maximum operating
temperature will affect the life of the insulation. As a rule of thumb, the lifetime of the winding
insulation will be reduced by half for every 10 ºC rise in temperature. The present day trend is
to design the machine using class F insulation for class B temperature rise.

8
 Chapter.3 Design of Commutator and Brushes

The Commutator is an assembly of Commutator segments or bars tapered in section. The

segments made of hard drawn copper are insulated from each other by mica or micanite, the
usual thickness of which is about 0.8 mm. The number of commutator segments is equal to the
number of active armature coils.

The diameter of the commutator will generally be about (60 to 80)% of the armature diameter.
Lesser values are used for high capacity machines and higher values for low capacity
machines.

Higher values of commutator peripheral velocity are to be avoided as it leads to lesser

di
commutation time dt, increased reactance voltage dt and sparking commutation.
RV = L
The commutator peripheral velocity vc = DC N / 60 should not as for as possible be
more
than about 15 m/s. (Peripheral velocity of 30 m/s is also being used in practice but should be
avoided whenever possible.)

The commutator segment pitch C = (outside width of one segment + mica insulation between
segments) = DC / Number of segments should not be less than 4 mm. (This minimum
segment pitch is due to 3.2 mm of copper + 0.8 mm of mica insulation between segments.) The
outer surface width of commutator segment lies between 4 and 20 mm in practice.

The axial length of the commutator depends on the space required

1
 1) by the brushes with brush boxes
2) for the staggering of brushes
3) for the margin between the end of commutator and brush and
4) for the margin between the brush and riser and width of riser.

If there are nb brushes / brush arm or spindle or holder, placed one beside the other
on the
commutator surface, then the length of the commutator LC = (width of the brush wb + brush
box thickness 0.5 cm) number of brushes / spindle + end clearance 2 to 4 cm + clearance for
risers 2 to 4 cm + clearance for staggering of brushes 2 to 4 cm.

If the length of the commutator (as calculated from the above expression) leads to small
dissipating surface DC LC, then the commutator length must be increased so that the
temperature rise of the commutator does not exceed a permissible value say 550C.

The temperature rise of the commutator can be calculated by using the following empirical
formula.
2
o 120 watt loss / cm of dissipating surface D L
θC= C C
1 0.1 v C

The different losses that are responsible for the temperature rise of the commutator are (a)
brush contact loss and (b) brush frictional loss.

Brush contact loss = voltage drop / brush set Ia

The voltage drop / brush set depend on the brush material – Carbon, graphite, electro graphite
or metalized graphite. The voltage drop / brush set can be taken as 2.0 V for carbon brushes.

Brush frictional loss (due to all the brush arms)

= frictional torque in Nm angular velocity

2
 2 N
= frictional force in Newton x distance in
60
metre
2
= 9.81 PbAball D C N = 9.81 PbAballvC
2 60

where = coefficient of friction and depends on the brush material. Lies between 0.22 and 0.27
for carbon brushes
Pb = Brush pressure in kg / m2 and lies between 1000 and 1500
2
Aball = Area of the brushes of all the brush arms in m
= Ab number of brush arms
= Ab number of poles in case of lap winding
= Ab 2 or P in case of wave winding
Ab = Cross-sectional area of the brush / brush arm

Brush Details

Since the brushes of each brush arm collets the current from two parallel paths, current
collected by each brush arm is 2 I and the cross-sectional area of the brush or brush arm or
a

A
2 a
holder or spindle Ab = I cm2. The current density b depends on the brush material and
A b

can be assumed between 5.5 and 6.5 A / cm2 for carbon.

In order to ensure a continuous supply of power and cost of replacement of damaged or worn
out brushes is cheaper, a number of subdivided brushes are used instead of one single brush.
Thus if

i) tb is the thickness of the brush

ii) wb is the width of the brush and
iii) nbis the number of sub divided brushes

thenAb = tbwbnb

As the number of adjacent coils of the same or different slots that are simultaneously
under
going commutation increases, the brush width and time of commutation also increases at the
same rate and therefore the reactance voltage (the basic cause of sparking commutation)
becomes independent of brush width.

With only one coil under going commutation and width of the brush equal to one segment
width, the reactance voltage and hence the sparking increases as the slot width decreases.
Hence the brush width is made to cover more than one segment. If the brush is too wide, then
those coils which are away from the commutating pole zone or coils not coming under the
influence of inter pole flux and under going commutation leads to sparking commutation.
3
Hence brush width greater than the commutating zone width is not advisable under any
circumstances. Since the commutating pole zone lies between (9 and 15)% of the pole pitch,
15% of the commutator circumference can be considered as the maximum width of the brush.

It has been found that the brush width should not be more than 5 segments in machines less
than 50 kW and 4 segments in machines more than 50 kW.

The number of brushes / spindle can be found out by assuming a standard brush width or a
maximum current / sub divided brush.

Standard brush width can be 1.6, 2.2 or 3.2 cm

Current/subdivided brush should not be more than 70A

A
Thus with the brush width assumed, nb= b . With the current / sub divided brush assumed
t w
b b
2 I and wb = A b
nb = a

A x 70 t n
b b

Note :

A) Staggering of Brushes :
Because of the current flowing from commutator segments to the brush, copper is
eaten
away leading to formation of ridges between the subdivided brushes of the same brush
arm. Since it is not possible to avoid eating away copper by the arc, eating away of copper
must be made to take place over the entire axial length of the commutator to ensure
uniform commutator surface. This is achieved by displacing all the positive brushes in one
direction and all the negative brushes in the other direction or by staggering of brushes in
pairs as shown below.

4
 B) Brush materials and their properties

Peripheral Current Voltage drop Coefficient of

Material velocity m/s density in per brush set friction
A/cm2 in volts
Normal carbon 5 to 15 5.5 to 6.5 2.0 0.22 to 0.27
Soft graphite 10 to 25 9.0 to 9.5 1.6 0.12
Metalized graphite 5 to 15 15 to 16 0.24 to 0.35 0.16
(copper carbon mixture)
Electro graphite 5 to 15 8.5 to 9.0 1.7 to 1.8 0.22
(Graphitized by heating)

C) Step by step design procedure of commutator and brushes

1) Diameter of the commutator DC = (0.6 to 0.8) D and must be such that the
peripheral
velocity of the commutator vC = DC N / 60 is not more than 15 m/s as far as possible.

2) The commutator segment pitch C = DC / Number of segments should not be less than 4
mm from the mechanical strength point of view.

3) The number of commutator segments is equal to number of active armature coils.

4) Length of the commutator LC = (width of the brush + brush box thickness 0.5 cm)
number of brushes / spindle nb + end clearance 2 to 4 cm + clearance for risers 2 to 4 cm +
clearance for staggering of brushes 2 to 4 cm.
2I a

5) Cross-sectional area of the brush / spindle or arm or holder Ab = A cm2. The current
b

2
density in the brushes b lies between 5.5 and 6.5 A / cm for carbon brushes.

6) Maximum thickness of the brush tb max = 4 Cfor machines greater than 50 kW

= 5 C for machines less than 50 kW

7) With standard brush width Wb assumed, the number of brushes / spindle nb

=
Ab
t W
b b
8) Total commutator losses = Brush contact loss + Brush frictional loss
= voltage drop / brush set Ia + 9.81 PbAballvC
where
voltage drop / set = 2.0 V for carbon brushes
= coefficient of friction and lies between 0.22 to 0.27 for carbon brushes
Pb = Brush pressure and lies between 1000 and 1500 kg / m2
5
9) Temperature rise of the commutator
0
C = Cooling coefficient x watt loss / dissipating surface
120
1 0.1 watt loss / cm2 of dissipating surface DC LC
=
v C

10) Temperature rise should be less than about 550C.

*******

6
 DESIGN OF COMMUTATOR AND BRUSHES

Example.1

A 500kW, 500V, 375 rpm, 8 pole dc generator has an armature diameter of 110 cm
and the
number of armature conductor is 896. Calculate the diameter of the commutator, length of the
commutator, number of brushes per spindle, commutator losses and temperature rise of the
commutator. Assume single turn coils.

Diameter of the commutator DC = (0.6 to 0.8) D = 0.7 x 110 = 77cm

Length of the commutator LC = (width of the brush Wb + brush box thickness 0.5 cm) number
of brushes / spindle nb + end clearance 2 to 4 cm + clearance for risers 2 to 4 cm + clearance
for staggering of brushes 2 to 4 cm.

Armature current Ia = kW x 10 = 500 x 10 = 1000A

3 3

V 500

Note : An armature current of 1000 A obviously calls for a lap winding.

Cross-sectional area of the brush per spindle or brush arm or holder Ab =

2 a 2
since the current density lies between 5.5 and 6.5 A/cm for carbon brushes,
I
A b

let it be 6 A/mm2
2
Ab = 2 x 1000 = 41.66 cm
8 x 6
maximum thickness of the brush = 4 C

Commutator segment pitch C = DC / Number of segments or coils

Number of coils = Z / 2 x number of turns per coil = 896 / 2 x 1 = 448

Therefore C x 77 = 0.54 cm
448
=
Maximum thickness of the brush = 4 x 0.54 = 2.16 cm

Let the thickness of the brush tb= 2.0 cm

If a brush width of 2.2 cm (a standard value) is assumed then Wb = 2.2 cm

Therefore, number of brushes / spindle

Ab 41.66
nb= = = 9.46 and is not possible
t W 2 x 2.2
b b

Let the number of brushes / spindle be = 10

7
Therefore LC = (2.2 + 0.5) 10 + 2 + 2 + 2 = 33 cm

Brush contact loss = voltage drop / brush set x Ia = 2 x 1000 = 2000 W

Brush frictional loss = 9.81 PbAball vC

Let the coefficient of friction = 0.25 as it lies between 0.22 to 0.27 for carbon brushes.

Let the brush pressure Pb= 1215 kg/m2 as it lies between 1000 to 1500 kg/m2

Aball = Area of the brushes of all the brush arms

= tbwbnb x number of brush arms or number of poles as the number of brush
arms number of poles for a lap winding
= 2 x 2.2 x 10 x 8 x 10-4 = 0.0352 m2

Brush frictional loss = 9.81 x 0.25 x 1215 x 0.352 x 15.1 = 1583.8 W

Therefore commutator losses (total) = 2000 + 1583.8 = 3583.8 W

Temperature rise in degree centigrade

2
= 120 x watt loss / cm of the L
commutator dissipating surface DCC

1 0.1 v C

= 120 x 3583.8 / x 77 x 33
= 21.46
1 0.1 x 15.1
Example.2

A 20 Hp, 4 pole, 250V, 1000 rpm wave wound D.C. machine has the following
design
data. Diameter of the armature = 25 cm, number of slots = 41, number of coil sides / slot =
4, turns / coil = 2. Calculate the number of segments, outside width of one segment and
mica, brush thickness, length of the commutator and brush contact loss.

Number of segments = number of active coils for a wave winding

i) Number of coil sides (total) = 41 x 4 = 164
Number of coils = 164 / 2 = 82 as each coil will have 2 coil sides OR

ii) Since the coil is of 2 turns, each coil side will have 2 conductors and therefore the
number of conductors per slot = 4 x 2 = 8.
Total number of conductors = 41 x 8 = 328
Number of coils = total number of 328 82 OR
conductors 2 x
2 x number of 2
turns / coil

iii) Number of coils = number of slots x number of coil sides / layer

= 41 x 2 = 82
8
 C 1
For a wave winding YC = must be an integer. With the number of coils calculated, YC
p
82 1
= 2 is a fraction. Therefore a wave winding is not possible. However a wave winding can
be made possible by considering one of the coils as dummy. Therefore number of active coils =
81 and number of commutator segments = 81.

Outside width of one segment and mica = Commutator segment pitch

=
D C

number of
segments
= with the assumption that
x 0.7 x 25
81
DC = 0.7 D
= 0.68 cm

Maximum thickness of the brush = 5 times the commutator segment pitch

= 5 x 0.68 = 3.4 cm

Let the thickness of the brush tb = 2.5 cm

Hp x 20 x
Armature current Ia = 66.3 A
746 / 250 746 /
V 0.9
2 =
Cross-sectional area of the brush / spindle Ab = a
2 x 66.3 = 11.65 cm2
I
A b 2 x 6
Let the standard brush width Wb = 1.6 cm
Ab 11.65
Number of brushes / spindle nb = = = 2.76 and is not possible
t W 2.5
x 1.6
b b
= 3 (say)
Length of the commutator LC = (1.6 + 0.5) 3 + 2 + 2 + 2 = 12.3 cm
Brush contact loss = Voltage drop / brush set x Ia = 2.0 x 66.3 = 132.6W

Example.3

A 600 kW, 6 pole lap connected D.C. generator with commutating poles running at 1200 rpm
develops 230V on open circuit and 250V on full load. Find the diameter of the commutator,
average volt / conductor, the number of commutator segments, length of commutator and brush
contact loss. Take Armature diameter = 56 cm, number of armature conductors = 300, number
of slots = 75, brush contact drop = 2.3 V, number of carbon brushes = 8 each 3.2 cm x 2.5 cm.
The voltage between commutator segments should not exceed 15V.

[ Note :
1. The D.C. generator is a cumulative compound one, with 230V on open circuit and 250V on
full load. Therefore while calculating the load current, 250V is to be considered.
9
2. The number of commutator segments or coils and hence the number of turns / coil must be
so selected that the voltage per segment is not greater than 15V.
3. For a given voltage between segments, the volt / conductor goes on reducing as the number
of turns / coil goes on increasing. Thus the volt / conductor is maximum when the turns / coil is
minimum or turns / coil is one.
4. Volt / conductor = (voltage between segments) / (conductors /coil) or (2 x number of turns
per coil)
5. There are 8 brushes / spindle of width 3.2 cm or 2.5 cm. ]

********

10
 Department of Electrical and Electronics Engineering

Notes on Lessons

Class : VI semester B.E. Sub. EE2355 Design of Electrical Machines

UNIT -1

INTRODUCTION

Flux (Φ) = (MMF/ Reluctance) , Wb ; Reluctance(S) = (l/aµ) , A/Wb ;

Permeance = ( 1/S) , Wb/A ; where l = length of the flux path, m ;

A = area of cross section for the flux path, m ; µ = permeability = µo µr ;

2

µo = absolute permeability = 4π x 10 H/m and µr = relative permeability.

-7

H = Ampere turns / m = MMF/ l = S Φ/ l = S Ba /l = (l/aµ) Ba/l = B/µ .

or B = µH

Series magnetic circuit : S = S1 + S2+…

Parallel magnetic circuit: Per= Per1+ Per2+...

Leakage Coeff. = total flux/useful flux; total flux = useful flux + leakage flux

Expressions for reluctance: Sg = lg/ µol ys

1. Various configurations of slotting

i) Smooth iron surface on both sides of the air gap : y s = ys

l

ii) slotted armature : ys = ys - Ws = Wt (no fringing)

l

iii) slotted armature : ys = ys – Kcs Ws ; Ws = slot width , Wt = tooth width

l

Kcs = Carter’s Coefficient for slots depends on the ratio of slot opening /airgap length or
the empirical relation is 1/ {1+ (5lg/Ws )}

If radial ventilating ducts are provided: Ll = L – Kcd nd Wd

where Kcd = Carter’s Coefficient for ducts,
n No.of ducts, W = width of each duct
d= d

1
 If ducts are provided on the stator and on the rotor, then K should be based cd

on half the air gap.

Considering the effect of both slotting and ducts, K ( gap contraction factor) = K K
g gs gd

where Kgs ={ ys / ys } and Kgd ={L /Ll }

l

If slots are provided on both sides of the airgap, K = Kgss Kgsr (ss and sr denoting stator
gs

and rotor slots respectively).

MMF for airgap = H K l ={ B/ µ }K l = 8,00,000 B K l .

g g o g g g g

Effect of Saliency : K f =Field form factor = Bave / Bg

= ψ = pole arc/pole pitch ;

pole pitch = πD/P ; B = Φ /( πDL/P)

ave

_______________________________

2. MMF calculation for teeth

The calculation of MMF for producing flux in the Teeth of the machine is difficult
because :
i) the teeth are tapered when parallel sided slots are used and this results in variation in
the flux density over the depth of the tooth.
ii) the slots provide another parallel path for the flux flow, the teeth are normally worked
in saturation and hence µr becomes low.

Following methods are usually employed for the calculation of MMF required for the
tapered teeth:-

i) Graphical method :AT = Mean ord. x l

t t

Mean ord. is the mean ord. of “at” variation with tooth depth.

ii) Simpson’s rule :at mean = (at +4 at + at )/6 A/m

1 2 3

iii) Bt1/3 method : ATt = at1/3 x lt ,

where at1/3 = MMF for corresponding to B at 1/3 height from the narrow end.
rd

3. Real and Apparent flux densities

Breal = B - 4π x 10 -7 at (K –1) ;where Ks = Atotal / Airon = Ly / L W

app s s i t

Specific Permeance : λ = Permeance per unit length or depth of the field.

Parallel sided slot :

2
 λ=µ [(h /3w ) + (h /w )+ {2h /w +w )} + (h /w ) ]
o 1 s 2 s 3 s o 4 o

4. Leakage reactance of transformers

Concentric winding
 Total leakage reactance of the transformer referred to the primary =
Xp = 2πf µoTp (Lmt/Lc) {a+ (bp + bs)/3}
2

The per unit reactance can also be calculated as εx = Ip Xp /Vp ; where Ip , and Vp are
rated phase current and voltage respectively.

Sandwich winding:

Xp = πf µo(Tp /n)(Lmt/w) {a+ (bp + bs)/6}

2

5. Temperature rise calculations

Q = Power loss(heat produced ), J/s or W
G = weight of the active material of the Machine, kg
h = specific heat, J/kg-◦C
S = cooling surface area, m2
λ = specific heat dissipation, W/ m2-◦C
c = 1/ λ = cooling coefficient, m -◦C / W
2

θ = final steady temperature rise, ◦C

m

The temperature of the machine rises when it is supplying load. As the temperature
rises, the heat is dissipated partly by conduction, partly by radiation and in most cases
largely by air cooling. The temperature rise curve is exponential in nature. Assuming the
theory of heating of homogeneous bodies ,

Heat developed = heat stored + heat dissipated

Q dt = Gh dθ + Sλθ dt ; solving → θ = θm ( 1- e –t/Th ) + θi e –t/Th

If the machine starts from cold, θi = 0 →θ = θm ( 1- e –t/Th )

Th=heating time constant (time taken by the machine to attain 0.632 times θm) = Gh/Sλ

T = cooling time constant (time taken by the machine to fall to 0.368 times θi)
c

Cooling curve is an exponentially falling curve.

6. Rating of machines

IS: 4722-1968: specification for Rotating Electrical machinery:

1. continuous duty
2. short time duty (T‹‹ T )h

3. intermittent periodic duty

4. intermittent periodic duty with starting
5. intermittent periodic duty with starting and braking
6. continuous duty with intermittent periodic loading
7. continuous duty with starting and braking
8. continuous duty with periodic speed changes

7. Determination of motor rating

From the point of calculation of motor rating, the various duty cycles listed earlier, can
be broadly classified as
i) Continuous duty ii) Fluctuating duty and iii) Short time and intermittent duty.

Continuously duty motors work

These motors work with the same load through out the duty cycle.

Fluctuating loads
The motor is switched on for a period T and kept off for a period T .
1 2

To calculate the power rating of motors to be used with fluctuating loads,

the commonly used methods are :

i) Method of average losses ii) Equivalent current method

iii) Equivalent torque method.

Short time duty

The motor carries a load much higher than the rated continuous load for a short time.
The time for which the motor may be allowed to carry the short time higher load is t = h

Th log {ph/ (ph – 1) } l

where ph = θm / θm ; θm = final steady temperature that would be attained if the machine
l

is allowed to run indefinitely at its short time rating.

8. Cooling of rotating electrical machines

In most cases, the cooling electrical machines is carried out by air flow and this cooling
is called ventilation. In high speed machines such as turbo alternators, hydrogen is used
for cooling.

Advantages of hydrogen cooling:

Compared with air, hydrogen has the following properties:-

4
i) (1/14) density thereby the windage losses and noise reduced ii) 14 times specific
th

heat and 1.5 times heat transfer leading to improved cooling iii) 7 times thermal
conductivity resulting in reduced temperature gradient iv) reduced corona effect v)
will not support combustion so long as the hyd /air mixture exceeds 3/1.

In operation, the fans mounted on the rotor circulate hydrogen through the ventilating
ducts and internally mounted gas coolers. The required gas pressure is maintained by a
regulator. The precaution to be observed is the stator frame must be gas tight and
explosion proof and oil film gas seals at the rotor shaft ends are essential.

Induced and Forced ventilation: In induced ventilation, the fan produces decreased air
pressure inside of the machine, causing air to be sucked into
the machine under the external atmospheric pressure ; and in the forced ventilation, the
air is forced into the fan by the fans mounted internally or externally.

The ventilation can also be classified as i) Radial, ii) axial and iii) combined radial and
axial.

9. Quantity of the cooling medium employed

Volume of air required = Va = (Q/cp θ) 10 x (760/H) x {( θi + 273)/273} V, m /s

3 3

cp = 995 and V = 0.775 m3 assumed; then

V = 0.78 (Q/ θ) x (760/H) x {( θ + 273)/273} , m3

a i

The capacity of the fan required = Pfan = P Va /ηfan, W

Similar calculations can be made for the volume of hydrogen or air or oil used for
cooling the machine.

5
 UNIT II
DC MACHINES

D = stator bore or armature diameter, m

L = stator core length, m
p = number of poles
Z = Total numbers of armature conductors
Iz = current in each conductor(Ia/A) , A
E = induced EMF, V
P = machine rating (power output),kW
Pa = power developed by the armature, kW
Q = kVA rating of the machine
Φ = flux per pole, Wb
τ = pole pitch (π D/p), m

Total Electric loading = pΦ ; Total Magnetic loading = I zZ

Specific Electric loading = ac = IzZ

(amp.conductor/m) πD

Specific magnetic loading = Bave = pΦ

(Wb/m ) 2
π DL

1. Output equation

Pa = E Ia x 10 -3 = (p/A)(ΦZN/60) Iz A 10 -3

= {π2 Bave ac x10-3}D2 L n = C D2 L n

o where C = output coefficient
o

___________________________________

Generator: Pa = (P/η) – (FW & Iron losses) ; Motor : Pa = P + (FW & Iron losses)

For large machines : FW & Iron losses are neglected i.e.,

P = P/ η (Generator)
a

= P (Motor)

For small machines : FW & Iron losses can be taken as 1/3 rd of the total losses. So,

P = (P/η) – (1/3) P (1-η)/ η = P(2+ η)/(3 η) …generator

a

= P + (1/3) P (1-η)/ η = P(1+2η)/(3 η) …motor

2. Choice of specific magnetic loading (Bave)

6
 i) B max in the iron part of the magnetic circuit:
B max ≤ maximum allowable density
i.e., Bt = BaveYs /wt (non-salient pole machines); Ys = slot-pitch and wt = tooth width.

&Bt = (Bave /ψ)Ys /wt (salient pole machines)

where ψ = pole arc /pole-pitch ratio.
If Ys =2 wt and ψ = 0.667 , then
Bt = 3 Bave : For example,if Bt is to be limited to 2.2 Wb/m , 2

B should excced 2.2/3 = 0.73 Wb/m2.

ave

ii) Magnetizing current:

large B → high magnetizing current → large core loss
ave

3. Choice of specific electric loading (ac)

i)Temperature rise(θ)
θ depends on Q (losses),which in turn depends on ac.
Allowable θ depends on insulating material used.

ii) Cooling coefficient (C)

θ is also proportional to the cooling coefficient;
a machine with a better ventilation has a lower C and then higher ac can be used.

iii)Operating voltage (V)

In high voltage machines, the slot space factor, S is less and so only smaller ac can
f

be used. It also depends on the shape of the conductors, circular or rectangular in

cross section.

iv) Current Density (δ)

choice of δ depends on cooling; higher C → higher value in the choice of ac.

4. Constraints in the design of DC machines

i) Peripheral speed,v ≤ 45 m/s

ii) Frequency of flux reversal, f ≤ 50 Hz
iii) Current per brush arm ≤ 400 A
iv) Armature MMF per pole ≤ 7000 A

The MMF required for the airgap = 50% of the armature MMF and gap
contraction factor = 1.1.

The current per brush arm (Ib) = 2Ia/p , A

For square poles: L = ψ π D/p

In the design process, choose p based on f& Ib and then calculate D and L

7
5. Armature Design
Considerations in choice of number of armature slots :

i) mechanical difficulties ii) cooling of armature

iii) pulsation of flux iv) cost v)commutation (Slots/pole ≥ 9 )
vi) slot pitch (ys =25 to 35 mm) vii) slot loading (Iz Zs ≤ 1500 Amp-cond)
viii) suitability of the winding – doublelayer Lap or wave )

Slot dimensions:

i) the slot area should accommodate the armature conductors and the required
insulation depending on the operating voltage ii) Bt1/3 ≤ 2.1 Wb/m2 iii) deep slots
cause eddy current losses iv) slot opening should be narrow to reduce the flux
pulsation and hence to reduce eddy current losses.

Armature voltage drop = Ia ra ;

ra = (Z/2) ρLmt /(a az) ;

2

where Lmt = length of mean turn, m = 2L + 2.3τ +5d ; a = no.of parallel paths and a =
s z

area of each conductor, m2.

6. Design of the Field system

Area of each pole (Ap) = Flux in the pole body / Flux density = cl φ / Bp
where cl = leakage coefficient

Width of the pole = A /L ;

p i where L = net iron length = 0.9 L
i

Height of the pole(hf) chosen based on the MMF to be provided by the pole at full-load.

( ATf at full-load) /(ATArm. at full-load) =1.0 to 1.25

(to overcome armature reaction)

7. Tentative design of Field winding

Copper loss in the field winding = If Rf 2

= (δfaf)2 {ρLmt Tf}/af = δf ρLmt Tf af

2

= δf ρLmt Sf hfdf ; ………………….(1)

2

where df = depth of the pole winding. Tf = no.of turns in each field coil.
Permissible loss = S qf = 2Lmt hf qf ..(2)
_______
Equating (1) and (2) ; δf 4 √ qf / Sf df
=10

8
 MMF per metre of the field winding = AT / h = I T /h = δ (a T )/ h f f f f f f f f f

= δf Sf hfdf/hf = δf Sf df = Sqrt {2 qf Sf df /ρ}

8. Design of commutator and brushes

i) The number of commutator segments is equal to the number of coils.

ii)The commutator diameter D = 60 – 70 % of the armature diameter (D)

c

iii)Peripheral speed of commutator = π DcN/60 ≤ 20 m/s

iv) Pitch of the commutator segment = β = π D /C ≥ 4.0 mm c c

(3.2 mm for the conducting portion and 0.8 mm for mica separator)

v) Distance between brush spindles = π D /p (25 – 30 cm) c

vi) Length of the commutator = Lc= nb (wb + cb) + c1+ c2

where nb = number of brushes per brush arm, wb = width of the brush, cb = clearance
between brushes , c1 = clearance for staggering (10–30 mm), c2 = clearance for end
ply (10-25 mm).

vii) Current carried by each brush spindle

= I = 2I /p = A δ ; where A = n w t
b a b b b b b b

= total brush contact area ; δb = current density in the brushes (≈0.1 A/mm ). 2

nb is selected such that each brush does not carry more than about 70 A .
viii) The thickness of brush = (2-3) βc

ix) Com. surface area = S = π D L m2

c c c

x) Commutator surface losses:

1.brush contact loss = W = V Ia ; V = brush contact drop ≈ 1.0 V/brush
bc b b

2.brush friction loss =Wbf = μ ρbpAbVc

μ = coefficient of friction (0.1 - 0.3) ; ρ = brush pressure (10 – 15 kN/ m2)
b

3. Cooling coefficient = c =k/(1+0.1Vc) ; k = 0.015 – 0.025

4. Temperature rise θ = Q c /S
c c c

9. Losses and efficiency in DC machines

1. I2R losses : copper loss in i) armature ii) Field iii) Inter pole winding

9
2. Rotational losses :i) Friction &windage

ii)Iron loss a) Hysterises =Kh Bm f b) Eddy current loss =Ke Bm f t

1.6 2 2 2

For the calculation of copper losses , the total length and area cross section of each
of the windings should be first calculated.
 10

UNIT III

TRANSFORMERS

1. Single-Phase Transformers

Voltage per turn = E/T = 4.44fΦm = Et

The window in a single-phase transformer contains one primary and one secondary
winding. The total copper area in the window:

A = T a + T a = ( T I + T I )/δ
c p p s s p p s s since a = I / δ and a = I / δ
p p s s

= 2AT/ δ ; since TpIp = TsIs neglecting magnetizing current

A = total window area ; K = window space factor = A / A

w w c w

=K A
Ac w w = 2AT/ δ → AT = Kw Aw δ/2

Rating in kVA = Q = Vp Ip x 10 -3 I
= Ep p x 10 -3 = Et ( Tp Ip )x 10 -3

= 4.44 fΦ ( K A δ/2 ) 10 -3
m w w ; where Φ = B Ai
m m

= 2.22 fΦm (Kw Aw δ ) 10 -3

2. Three-phase transformers

Each window contains two primary and two secondary windings

Ac =2 (Tpap + Tsas ) = 4 AT / δ → AT = Kw Aw δ/4

Rating in kVA = Q = 3Vp Ip x 10 -3 = 3Ep Ip x 10 -3 =3 Et ( Tp Ip )x 10 -3

= 3 x 4.44 fΦ ( K A δ/4 ) 10 -3
m w w where Φ = B Ai m m

= 3.33 fΦm (Kw Aw δ ) 10 -3

Using the output equation it can also be shown that

_____ __________
E t = K √ kVA where K =√ 4.44 f r 10 ; r = Φm / AT 3

r is a constant for transformer of a given type ,service and method of connection, since
Φm determines the core section and AT fixes the total copper area.

3. Ratio of Iron loss to copper loss

 11

Copper loss/m = (ρl/a) (current) /(la) = (ρl/a) (aδ) /la

3 2 2

= ρδ2 W/m3 = ρδ2/density W/Kg

Taking ρ =0.021 x 10 and density = 8.9 x 10 Kg/m

-6 3 3

Copper loss/Kg = specific copper loss = p c

= 2.36 x 10 12 δ2 W/Kg . Then total copper loss =

Wc = pcGc where Gc weight of copper , Kg

In addition to the above , we must add stray losses which may be 5 to 25 % of copper
loss.

The total iron loss / Kg = specific iron loss (pi) can be found from the iron loss curves.
Then the total iron loss = Wi = pi Gi
Where Gi = weight of iron . Ratio of Iron loss / copper loss = pi Gi / pcGc

4. Design of the core

The core section of the core type transformer may be rectangular , square or stepped.
Shell type transformers use cores with rectangular cross section. For the rectangular
core the ratio of depth to width 1.4 to 2.0.

Square or stepped coreWhen circular coils are required for high voltage transformers ,
square and stepped cores are used . Circular coils are preferred because of their superior
mechanical characteristics.

As the size of the transformer increases, it becomes wasteful to use rectangular cores.
Square cores are used and the surrounding circle, representing the inner surface of the
tubular form carrying the windings, is called the circumscribing circle. Even now a lot of
useful space is wasted and the length of the mean turn increases causing higher I2R
losses.

With larger transformers, cruciform cores, with better utilization of the space, are used. It
should be born in mind that two different types of laminations are used in cruciform
cores. With still larger transformers, further step sizes are introduced to utilize the core
even more effectively. However, larger step sizes → larger number of lamination sizes →
higher labor cost.

Assuming a stacking factor for iron = 0.9 , Net core area / area of circumscribing circle
= 0.637 for square core and = 0.710 for stepped core with a= 0.851d , b= 0.526 d ;
where d = diameter of the circumscribing circle.

5. Choice Flux density and Current density( Bm and δ)

12
Bm determines the core area.
Higher B → smaller area → smaller L → saving in the cost of iron and copper. But
m mt

higher Bm increases the iron loss and temp rise. For Distribution transformer Bm = 1.1 to
1.35 Wb/m . For Power transformer Bm = 1.25 to 1.45 Wb/m .
2 2

The area of conductors for the primary and secondary windings determined after
choosing a suitable value for δ which depends on the method of cooling.

6. Types of Windings

i) Cylindrical winding with circular conductors

ii) Crossover winding with circular or rectangular conductors
iii) Continuous disc type winding with rectangular conductors
iv) Helical winding

7. Design of insulation

i) Electrical insulation: depends on the operating voltage

ii) Eddy current loss in the conductors and tank walls
iii) Mechanical considerations: high mechanical forces during fault
iv) Thermal considerations: depends on cooling

Major insulation : between windings and core (grounded)

Minor insulation ; between turns and layers
Materials : cotton thread, cotton tape, leatheriod paper, millinax paper etc.

8. Window dimensions

a = width of the largest samping ; d = dia of the circumscribing circle

D = distance between centres of adjacent limbs
Ww , Hw = width and height of the window ( length of the window)
H = height of the yoke
y

For core type: D = d + Ww ; W = D+a ; H = Hw + 2Hy

For core type, the yoke section may be either rectangular or stepped. In rectangular yoke
sections, depth of the yoke = depth of the core; when stepped cores are used
the core depth = width of the largest stamping; area rectangular yokes = Ay
= Dy Hy = a Hy

For three phase transformers : W = 2d + a

Window space factor, Kw = Ac/Aw = 10 /(30+kV)

Area of the window = A = H W ;
w w w

Aw = 2apTp/ Kw (single-phase) = 2apTp/ Kw ( three-phase)

13
For shell type : D = b ; Hy = a ; W = 2W +4a ; H = H + 2a
y w w

9. No-load current calculations

The phasor sum of the magnetizing current (Im) and the loss component of current (I ) ; l

Im is calculated using the MMF/m required for the core and yoke and their respective
length of flux path. Il is determined using the iron loss curve of the material used for the
core and yoke and the flux density employed and their weight.

10. Temperature rise of transformers

Losses dissipated in transformers in the core and windings get converted into thermal
energy and cause heating of the corresponding transformer parts.The heat dissipation
occurs as follows: i) from the internal heated parts to the outer surface in contact with oil
by conduction ii) from oil to the tank walls by convection and iii) from the walls of the
tank to the atmosphere by radiation and convection.

11. Transformer oil as a cooling medium

The specific heat dissipation due to convection of oil

= λconv = 40.3 (θ /H) ¼ W/m2- oC ; = temp difference of the surface relative to the oil and
H = height of the dissipating surface.

Experimentally found that a plain tank surface dissipates 6.0 W/m2-oC by radiation and
6.5 W/m2-oc by convection (.for a temp rise of 40 oC above an ambient temp of 20 oc).
Thus a total of 12.5 W/m - C is taken.
2 o

The temp rise θ = total loss/ (λ St) = (Pi + Pc) / (λ St)

Where S = heat dissipating surface area of the tank.
t

For small transformers , plain walled tank is enough to dissipate the losses. As the rating
of the transformer increases, the volume increases as he cube of the linier dimensions but
the heat dissipating surface area increases only as square of the linier dimensions. So
above certain rating, plain tank becomes inadequate to dissipate losses and the area is
increased by providing tubes. For larger ratings forced air cooling is used.

If tubing is provided, the oil circulation is improved due to the head of the oil, and this
causes an additional dissipation by convection of about 35 % .

Let x S be the area of the cooling tubes. Then

t

Loss dissipated by the tank surface 12.5 St W/0C

Loss dissipated by the tubes (1.35 x 6.5) x St W/0 C = 8.8 x St W/0C
Total loss dissipated by the tank and oil tubes

14
= (12.5 St + 8.8 x St ) W/ C
0

Hence θ = ( Pi + Pc)/ (12.5 St + 8.8 x St ) →Total tube area x St

= (1/8.8) [ {(Pi + Pc)/ θ} – 12.5 St ]
The number of tubes = nt = Total tube area /(π dt lt )

The arrangement of the tubes on tank side walls should be made uniformly with a
spacing of usually 75 mm. Examples of calculation of nt and the arrangement of the tubes
should be studied.
UNIT IV

INDUCTION MOTORS
1. Output Equation

Q(kVA) = m EphIph 10-3

= m (4.44f Φ T K ) {(ac πD)/( m 2T )}

m ph w ph

= 4.44 (pNs/120) (Bave π DL/p) Kw (ac πD/ 2) x 10 -3

= 1.11 π2 Bave ac K x 10-3D2L ns

w

=C D
= 11 Bave ac Kw x 10-3 D2L ns o 2 L ns

Co = output coefficient , Q is calculated as ( hp x 0.746 )/(η cosΦ)

2. Choice of Bave
i) Low Bave → large size machine for a given hp
ii) high Bave → large magnetizing current → low power factor
iii) high B → high iron loss
ave

iv) high Bave → high Φm→ less Tph→ low leakage reactance
→ larger diameter for the circle diagram→ larger over load capacity
For 50 Hz motors B : 0.3 to 0.6 W /m2
ave b

3. Choice ac (ampere conductor /m)

Low ac → large size machine for a given h p

High ac → higher copper loss and temp rise

High ac → large T → large leakage reactance

ph

→ lower diameter for the circle diagram→ lower over load capacity

For 50 Hz motors ac : 10,000 to 45,000 amp.cond/m

The value ac chosen depends on the ventilation and cooling

It should be remembered that the Power factor (PF) and efficiency(η) of the motor at full
load increases with the rating of the machine. Again η and Pf are higher for high speed
motors compared to low speed motors.

4. Separation of D and L and air gap length

i) L/τ can be assumed

ii) for best power factor τ = Sqrt( 0.18 L) or D= 0.135p√ L
iii) D can be chosen based on peripheral speed = πDN/60

The stator winding can be connected in star or delta. The motors meant for starting with
star- delta starter should be designed with delta connected stator winding.

16
Stator slot –pitch = y = π D/S where S =number of stator slots
ss s s

Stator outer diameter Do = D + 2 dss + 2dcs

Where dss = stator slot depth and dcs = stator core depth

Air gap length = 0.2 + 2 sqrt(DL) where Dand L are in metre.

After designing the main dimensions, the following calculations can be made:

i) The flux per pole, Φm =

B
ave π DL/p

ii) Turns in series per phase, Tph = V/ (4.44f ΦmKw )

iii) Number of slots per pole per phase can be suitably assumed

iv) Slot pitch should not exceed about 24 mm

v) Number of conductors per slot should be rounded off

vi) T recalculated
ph

vii) Iph = [Q x 10 /(√3 400)] for star

3

and [Q x 103 /(√3 400)] /√3 for delta connection

viii) Assuming suitable current density (δs) , conductor area required is

calculated (as = Iph / δs )

ix) Total copper area in the slot = (conductors /slot) as

x) Total area of the slot = copper area /slot space factor

Slot space factor is to account for the space for insulation

xi) A suitable tooth flux density is assumed and tooth width is calculated

xii) slot width = slot pitch- tooth width (at different diameters)

xiii) The core flux(Φm/2) , core area and core depth are calculated
assuming a core flux density. Core depth = core area/ Li

xiv) The rotor bar and end rings are designed (sq.cage rotor)

xv) The rotor winding is designed similar to stator winding (wound rotor)

5. Losses in the Induction motor

17
i) stator copper loss ii) rotor copper loss iii) iron loss in the stator teeth and
core iv) friction and windage loss (1- 1.5 % of output)

The rotor resistance in stator terms can be obtained as

rotor copper loss/ I2’ ; where I2’ = 0.85 I1

6. No-load current

Iron loss component = Il =(Iron loss/phase) /Vph

Magnetizing component = Im =(0.427 p ATtotal )/( K T ) ws ph

ATtotal = sum of the ampere turns for airgap ,stator tooth and core and rotor tooth and
core.
No-load current = In = { Il + Im }1/2
2 2

The leakage reactance calculations are made to find x , x and x .

1 2 01
UNIT V

SYNCHRONOUS MACHINES

1. Construction

Stationary armature, rotating field type of construction is preferred.

High speed alternators have non-salient pole rotor (Turbo alternators) and they have
either 2-pole or 4-pole. Slow speed alternators have salient pole rotor (water
wheel alternators) and they have more than 4 poles.

2. Output Equation

Q = m EphIph 10-3
=C D
= 11 Bave ac Kw x 10-3 D2L ns o 2L ns

Where C = output coefficient

o

(see Unit -4 for derivation)

D and L are separated using L/τ ratio or maximum specified peripheral

speed of the rotor.

Damper winding is used for starting and damping rotor oscillations that occur during
sudden load changes.

3. Choice of specific magnetic loading (Bave)

i) High B → high flux density in the teeth and core →

ave

high iron loss → higher temperature rise.

ii) high Bave → low Tph → low leakage reactance (Xl )

→ high short circuit current

iii) In high voltage machines slot width required is more to accommodate

thicker insulation →smaller tooth width → small allowable B ave

iv) stability : Pmax =VE/Xs . Since high Bave gives low Tph and hence low Xl
P increases and improves stability.
max

v) Parallel operation : Ps = (VE sinδ)/Xs ; where δ is the torque angle. So

low Xs gives higher value for the synchronizing power leading stable parallel
operation of synchronous generators.

Guide lines : Non-salient pole alternator : 0.54 – 0.65 Wb/m 2

Salient – pole alternator : 0.52 – 0.65 Wb/m2

4. Choice of specific Electric loading (ac)

i) Copper loss and temperature rise: High value of ac → higher copper loss
leading high temperature rise. So choice of depends on the cooling method used.
ii) Operating voltage : High voltage machines require large insulation and so the slot
space available for conductors is reduced. So a lower value for ac has to be chosen.

iii) Synchronous reactance (X ) : High value of ac results in high value of

s

Xs , and this leads to a) poor voltage regulation b) low steady state stability limit.

iv) Stray load losses increase with increase in ac.

Guide lines : Non-salient pole alternators : 50, 000 – 75,000 A/m

Salient pole alternators : 20,000 – 40,000 A/m

5. Short Circuit Ratio (SCR)

SCR = Field current required to produce rated voltage on opencircuit

Field current required to produce rated current on short circuit

= 1/ direct axis synchronous reactance = 1/Xd

Thus SCR is the reciprocal of X , if X is defined in p.u.value for rated voltage and
d d

rated current. But Xd for a given load is affected by saturation conditions that then
exists, while SCR is specific and univalued for a given machine.

Non-salient pole alternators : 1- 1.5 ; Salient pole alternators : 0.5 – 0.7

Effect of SCR on machine performance

i) Voltage regulation : A low SCR → high Xd → large voltage drop

→ poor voltage regulation..

ii) Parallel operation : A low SCR → high Xd → low synchronizing

power → parallel operation becomes difficult.

iii) Short circuit current : A low SCR → high Xd →low short circuit
current. But short circuit current can be limited by other means not
necessarily by keeping a low value of SCR.

iv) self excitation : Alternators feeding long transmission lines should

not be designed with small SCR as this would lead to large terminal
voltage on open circuit due to large capacitance currents.

Summarizing ,high value of SCR leads to i) high stability limit ii) low voltage
regulation iii) high short circuit current iv)large air gap

The present trend is to design machines with low value of SCR, this is due to the
recent development in fast acting control and excitation systems.
 6. Length of airgap

The length of air gap very much influences the performance of a synchronous
machine. A large airgap offers a large reluctance to the path of the flux produced by
the armature MMF and thus reduces the effct of armature reaction. Thus a machine
with large airgap has a small Xd and so has
i)small regulation ii) high stability limit iii) high synchronizing power which makes
the machine less sensitive to load variations iv) better cooling at the gap surface v)
low magnetic noise and smaller unbalanced
magnetic pull.

But as the airgap length increases, a large value of Field MMF is required resulting in
increased cost of the machine.

7. Number of stator slots

Factors to be considered in the selection of number of slots :

1. Balanced 3-phase winding to be obtained
2. With large number of slots i)→large number of coils → increased
labor cost ii) cooling is improved iii) tooth ripples are less iv) Flux
density in the iron increases due to decreased tooth width.

Guide lines : Slot pitch (ys )≤ 25 mm for low voltage machines;

≤ 40 mm for machines upto 6 kV ; ≤ 60 mm for machines upto 15 kV.

8. Methods of Eliminating Harmonics

By using i)distributed windings ii) fractional coil pitch

iii) fractional slot windings iv) skewing v) large airgap

Further calculations needed after determining D and L :

i) Flux per pole = Φ = Bave (π DL/p )

ii) Tph is calculated from the EMF equation taking Eph= Vph
iii) Iph = (Qx 103 ) / √ 3 V
L

iv) Armature MMF/pole = Ata = 2.7 Iph TphKw /p

v) Effective area per pole = 0.6 – 0.65 times actual area

9. Field Design (Salient poles)

Data needed for the design of the Field winding :

i) Flux density in the pole core ii)Winding depth (d )iii) Leakage factor (pole flux/gap
f

flux) iv) Field winding space factor (Sf) v) Power dissipation (qf)in W/m2 v) The ratio

21
of field MMF to armature MMF vi) Allow about 30 mm for insulation , flanges and
height of the pole shoe.
MMF per unit height of the winding = 104 Sqrt (Sf df qf )

Computer Aided Design of Electrical Machines

The process of design any electrical may be broadly divided into three major aspects: i)
Electrical design ii) Mechanical design iii) Thermal design. Even though, these
problems can be solved separately, there are many inter- related features.

The advantages of computer aided design are : i) The computer can handle large
volume of data to make a number of trial designs. And speed and accuracy of
calculations are very high. iii) It can be programmed to satisfying take logical decisions
iv) An optimized design with least cost and the required performance can be easily
obtained.

Generally any design method can be

i) analysis method ii) Synthesis method iii) Hybrid method

In the analysis method of design , a preliminary design is made by the designer

regarding the machine dimensions, materials and other constructional features and
these are given as input data to the computer and the performance quantities are
calculated. The designer examines the performance and accordingly alters the input
data and then feed them to the computer again. The computer calculates the new
performance with the revised data. This process is repeated till the required
performance is achieved.

In the synthesis method, the required performance values are also given to the
computer as input. The computer through an iterative process alters the dimensions till
the required performance is obtained.

In the hybrid method, by some human intervention, a combination of analysis and

synthesis methods are adopted.

The method of design optimization using computers :

i) Choice of independent variables

ii) Variable transformation
iii) Forming the constraint functions for the performance

22
iv) Forming the objective function (OBJ)
v) Applying the minimization technique till the OBJ becomes with in the chosen
tolerance.

Example of Design of optimization of Induction Motors

The independent variables which has a significant effect on the performance are

stator core diameter, stator core length , stator core depth, stator slot depth, stator slot
width, rotor slot depth, rotor slot width, end ring depth, end ring width, . airgap length
and airgap flux density.

The other variables in the design are either taken as constants dased on the voltage and
power rating of the machine or they are in some way related to the above 11 variables.

During the course of optimization when the variables undergo incrementing or

decrementing, they should also be constrained to be with in practical ranges. This is
obtained by variable transformation. For example for airgap Xact = X tran + Lg min ;
where they respectively denote actual and transformed values and Lg min = minimum
airgap required.

Performance Specifications:

1. Starting torque 2. maximum torque 3. Full- load power factor 4. full -load efficiency
5. full load slip 6.tooth and core flux densities 7. starting current 8. temperature rise 9.
cost of the machine.

Objective function

The objective function is formed by comparing the specified and calculated values of
the performance quantities at each iteration. Objective function minimization can be
carried out either using conventional methods such as Powel’s algorithm or
Rosenbrock method or the recent techniques such as Genetic algorithm.

It should be noted that the independent variables or the performance specifications vary
with the type of machine and its application.

23
 Chapter.2 DESIGN OF DC MACHINES

Details to be specified while ordering a DC machine or consumer’s specification

1. Output : kW (for generators), kW or Hp (for motors)

2. Voltage : V volt
3. Speed : N rpm
4. Rating : Continuous or Short time
5. Temperature rise: 0C for an ambient temperature of 400C
6. Cooling : Natural or forced cooling
7. Type: Generator or motor, separately excited or self-excited-shunt, series, or
compound, if compound type of connection – long or short shunt, type of
compounding – cumulative or differential, degree of compounding – over, under or
level. With or without inter poles, with or without compensating windings,with or
without equalizer rings in case of lap winding.
8. Voltage regulation ( in case of generators) : Range and method
9. Speed control ( in case of motors ) : range and method of control
10. Efficiency: must be as for as possible high (As the efficiency increases, cost of the
machine also increases).
11. Type of enclosure: based on the field of application – totally enclosed, screen protected,
drip proof, flame proof, etc.,
12. Size of the machine etc.,

Size of the DC machine

The size of the DC machine depends on the main or leading dimensions of the machine viz.,
diameter of the armature D and armature core length L. As the output increases, the main
dimensions of the machine D and L also increases.

1
 Figure.1.Armature of a dc machine Figure.2. Yoke and pole arrangement of a
dc machine

OUTPUT EQUATION

Note: Output equation relates the output and main dimensions of the machine. Actually it
relates the power developed in the armature and main dimensions.

Derivation:
Nomenclature: E : emf induced or back emf
Ia : armature current
: Average value of flux / pole
Z : Total number of armature conductors
N : Speed in rpm
P : Number of poles
A : number of armature paths or circuits
D : Diameter of the armature
L : Length of the armature core

Power developed in the armature in kW = E Ia x 10-3

ZNP I 10 -3

a
60 A
-3
IZ N x 10
a
60 ....... (1)
(P ) A

The term P represents the total flux and is called the magnetic loading. Magnetic loading/unit
area of the armature surface is called the specific magnetic loading or average value of the flux
density in the air gap Bav. That is,

2
 P 2
B av Wb/m or tesle denoted by T
DL
Therefore P ................ (2)
B DL
av

The term (Ia Z/A) represents the total ampere-conductors on the armature and is called
the
electric loading. Electric loading/unit length of armature periphery is called the specific electric
loading q. That is,

I Z
q a
ampere - conductors / m
A
D
Therefore I Z/A q D ............ (3)
a
Substitution of equations 2 and 3 in 1, leads to
-3
kW B DL q N 10
D
av
60
-4 2
1.64 10 B q D L N
av
2
C D LN
0
where C is called the output coefficeint of the DC machine and is
0
-4
equal to 1.64 x 10 B q.
av
2 kW 3
Therefore D L -4 m
1.64 10 B q N
av

The above equation is called the output equation. The D2L product represents the size

of the
machine or volume of iron used. In order that the maximum output is obtained/kg of iron used,
D2L product must be as less as possible. For this, the values of q and Bav must be high.

Effect of higher value of q

Note: Since armature current Ia and number of parallel paths A are constants and armature
diameter D must be as less as possible or D must be a fixed minimum value, the number of
increases.
armature conductors increases as q I a
Z/
A D

a. As q increases, number of conductors increases, resistance increases, I2R loss increases

and therefore the temperature of the machine increases. Temperature is a limiting factor
of any equipment or machine.
b. As q increases, number of conductors increases,conductors/slot increases, quantity of
insulation in the slot increases, heat dissipation reduces, temperature increases, losses
increases and efficiency of the machine reduces.
c. As q increases, number of conductors increases, armature ampere-turns per pole
 AT / pole = (I Z / 2 A P) a
a increases, flux produced by the armature increases, and
therefore the effect of armature reaction increases. In order to overcome the effect
of

3
 armature reaction, field mmf has to be increased. This calls for additional copper
and
increases the cost and size of the machine.
d. As q increases, number of conductors and turns increases, reactance voltage
proportional to (turns)2 increases. This leads to sparking commutation.

Effect of higher value of Bav

a. AsBav increases, core loss increases, efficiency reduces.

b. AsBav increases, degree of saturation increases, mmf required for the magnetic
circuit increases. This calls for additional copper and increases the cost of the
machine.

It is clear that there is no advantage gained by selecting higher values of q and B av. If the
2
values selected are less, then D L will be large or the size of the machine will unnecessarily be
high. Hence optimum value of q and Bav must be selected.

In general q lies between 15000 and 50000 ampere-conductors/m.

Lesser values are used in low capacity, low speed and high voltage machines.
In general Bav lies between 0.45 and 0.75 T.

SEPARATION OF D2L PRODUCT

Knowing the values of kW and N and assuming the values of q and Bav, a value for
2 kW can be calculated.
DL = -4
1.64 10 B q N
av
3
Let it be 0.1 m .

Since the above expression has two unknowns namely D and L, another expression relating D
and L must be known to find out the values of D and L.

Usually a value for the ratio armature core length L to pole pitch is assumed to separate D2L
product. The pole pitch refers to the circumferential distance corresponding one pole at
diameter D. In practice L / lies between 0.55 and 1.1.

Therefore L = (0.55 to 1.1)

= (0.55 to 1.1) D/P

If L/ = 1.0 and P = 4, then L = 1.0 D / P = 1.0 D/4 = 0.785D.

Therefore D2 0.785 D = 0.1 or D = 0.5m. Thus L = 0.785 0.5 = 0.395 m.

2
Note: The D L product can also be separated by assuming a value for the peripheral velocity
of the armature.
4
LIMITATIONS OFD AND L

As the diameter of the armature increases, the peripheral velocity of the armature
v = π DN / 60 m / s , centrifugal force and its effects increases. Therefore the machine must be
mechanically made robust to withstand the effect of centrifugal force. This increases the cost of
the machine. In general for normal construction, peripheral velocity should not be greater than
30 m/s as for as possible.

To prevent arcing between commutator segments, voltage between the commutator segments
should not be greater than about 20V on open circuit. If a single turn coil is used
then the
voltage/conductor e = Bav L v should not be more than 10V.
(Bav – Average value of flux density in the air gap in tesla, L – Length of the
conductor or
gross armature core length in metre and v – Peripheral velocity of armature in m/s).

e
Therefore, armature core length L = B v should not be greater than 10 / (0.75 30) = 0.44m
av
for normal design.

Ventilating ducts and net iron length of armature core

To keep down the temperature rise of the machine parts, the core is provided with both radial
and axial ventilating ducts. Radical ducts are created by providing vent or duct spacers in
between core packets of width 5-7 cm. Width of the radial duct lies between 0.8 and 1.0 cm in
practice.

It is clear from the figure that the net core length = (L – nvbv)
wherenv – number of ventilating ducts and bv – width of the ventilating duct.

Since the core is laminated to reduce the eddy current loss and flux takes the least reluctance
path, flux confines to the iron only. Therefore the length over which the flux passes is not L or
(L -nvbv) but less than that. Generally the thickness of insulation on the laminated core will be

5
around 10% of the gross core length. Therefore the net iron length will be 0.9 times
the net
core length or net iron length,
Li = Ki (L - nvbv)
where Ki is called the iron or stacking factor and is approximately 0.9 in practice.

Estimation of number of ventilating ducts

Determine the number of ducts if the length of armature core is 35 cm.

Since the width of the core packet lies between 5 and 7 cm, let the width of the core packet be
6 cm. Therefore number of core packets is equal to 35 /6 or 6 say. Six core packets lead to 5
ducts. If the width of the duct is assumed to be 1.0 cm, then the net core length will be (35 – 1
5) = 30 cm and each core packet after revision will be 30 / 6 = 5 cm in width.

CHOICE OF NUMBER OF POLES

In order to decide what number of poles (more or less) is to be used, let the different factors
affecting the choice of number of poles be discussed based on the use of more number of poles.

1. Frequency
PN
As the number of poles increases, frequency of the induced emf f = increases,
core
120
loss in the armature increases and therefore efficiency of the machine decreases.

2. Weight of the iron used for the yoke

Since the flux carried by the yoke is approximately /2 and the total flux T = p is a
constant for a given machine, flux density in the yoke
/2 T
1
B y
.
cross sectional area of the yoke 2PA y PA y

Ay

It is clear that Ay is 1/P

as By is also almost constant for a given iron. Thus, as the number of poles increases, Ay
and hence the weight of iron used for the yoke reduces.
6
3. Weight of iron used for the armature core (from the core loss point of view)

Since the flux carried by the armature core is /2, eddy current loss in the armature core
B 2f2
c

/2 2 T
PN
] [ ]
[ A ][c
f 2PA c
120
1
2 is independent of the number of poles.
A c

On the other hand, since the hysteresis loss in the armature core is
1.6 T1.6 PN 1 , the armature core area Ac
B f c ( ) x 0.6 A1.6

2PA c 120 P c
1
P0.6 / 1.6 decreases as the number of poles increases for a given hysteresis loss. Thus
the weight of iron used for the armature core reduces as the number of poles increases.

4. Weight of overhang copper

For a given active length of the coil, overhang pole pitch D/P goes on reducing as the
number of poles increases. As the overhang length reduces, the weight of the inactive
copper used at the overhang also reduces.

5. Armature reaction
Since the flux produced by the armature = AT /
a and armature ampere turns ATa
pole a
IZ reluctance
/ pole = a is proportional to 1 / P, a reduces as the number of poles increases. This
2AP
in turn reduces the effect of armature reaction.

7
6. Overall diameter
When the number of poles is less, ATa / pole and hence the flux, produced by the armature is
more. This reduces the useful flux in the air gap. In order to maintain a constant value of air
gap flux, flux produced by the field or the field ampere-turns must be increased. This calls for
more field coil turns and size of the coil defined by the depth of the coil df and height of the
coil hf increases. In order that the temperature rise of the coil is not more, depth of the field coil
is generally restricted. Therefore height of the field coil increases as the size of the field coil or
the number of turns of the coil increases. As the pole height, is proportional to the field
coilheight, height of the pole and hence the overall diameter of the machine increases with the
increase in height of the field coil.

Obviously as the number of poles increases, height of the pole and hence the overall diameter
of the machine decreases.

7. Length of the commutator

Since each brush arm collects the current from every two parallel paths, current / brush arm
= 2 Ia / A and the cross sectional area of the brush / arm
Ab = 2Ia / A b = 2Ia / P b
1/P
reduces as the number of poles increases.
As Ab = tbwbnb and tb is generally held constant from the commutation point of view, wbnb
reduces as Ab reduces. Hence the length of the commutator
Lc = (wbnb + clearances) reduces as Ab reduces or the number of poles increases.

wb – width of the brush, tb – thickness of the brush, nb – number of brushes per spindle

8
8. Flash over
As the number of poles increases, voltage between the segments
voltage between positive and negative brushes
E= increases. Because of the increased
b
number of segments / pole
value of Eb and carbon dust collected in the space where the mica is undercut, chances of
arcing between commutator segments increases. The arc between the segments in turn may
bridge the positive and negative brushes leading to a dead short circuit of the armature or
flash over.

9. Labour charges
As the number of poles increases cost of labour increases as more number of poles are to be
assembled, more field coils are to be wound, placed on to the pole, insulate, interconnect
etc.

It is clear that, when the number of poles is more, weight of iron used for yoke and
armature core, weight of inactive copper, overall diameter, length of commutator and effect
of armature reaction reduces. On the other hand efficiency reduces chances of flash over
increases and cost of machine increases.

Since the advantages outnumber the disadvantages, more number of poles is preferable.

Thus, though more number of poles is preferable, it is not advisable from the cost point of
view. In general the number of poles should be so selected that good operating
characteristics are obtained with minimum weight of active material and minimum cost of
construction.

SELECTION OF NUMBER OF POLES (For a preliminary design)

As the armature current increases, cross sectional area of the conductor and hence the eddy
current loss in the conductor increases. In order to reduce the eddy current loss in the

9
 conductor, cross-sectional area of the conductor must be made less or the current /
path
must be restricted.
For a normal design, current / parallel path should not be more than about 200A. However,
often, under enhanced cooling conditions, a current / path of more than 200A is also being
used. By selecting a suitable number of paths for the machine, current / path can be
restricted and the number of poles for the machine can be decided.
While selecting the number of poles, the following conditions must also be considered as
far as possible.
a. Frequency of the armature induced emf f = PN/120 should as for as possible between
25 and 50 Hz.
b. Armature ampere turns / pole = IaZ / 2AP should not be greater than 10000.

Example: Select a suitable number of poles for a 1000kW, 500V DC machine.

Armature current (approximately) Ia = 1000 x 10 2000A

500
In order that the current / path is not more than about 200A, lap winding has to
be used
2000
having number of parallel paths = 10 . Since the number of parallel paths is equal
200
to number of poles in a simplex lap winding, 10 poles can be selected for the machine as a
preliminary value. (The number of poles selected can also be like 8 or 12).

Example: Select a suitable number of poles for a 100kW, 500V DC machine.

Armature current (approximately) Ia= 1000 x 10 200A .

500
For a current of 200A a lap or wave winding can be used. Since the minimum number of
paths and poles is two, 2 poles are sufficient for the machine. However to gain more
advantages of more number of poles, let the number of poles be 4.

ARMATURE WINDING

The armature winding can broadly be classified as concentrated and distributed winding.
In case of a concentrated winding, all the conductors / pole is housed in one slot. Since the
conductors / slot is more, quantity of insulation in the slot is more, heat dissipation is less,
temperature rise is more and the efficiency of operation will be less. Also emf induced in the
armature conductors will not be sinusoidal. Therefore
a. design calculations become complicated (because of the complicated expression of non-
sinusoidal wave).
b. Core loss increases (because of the fundamental and harmonic components of the
non-
sinusoidal wave) and efficiency reduces.
c. Communication interference may occur (because of the higher frequency components of
the non-sinusoidal wave).

Hence no concentrated winding is used in practice for a DC machine armature.

10
In a distributed winding (used to overcome the disadvantages of the concentrated
winding),
conductors / pole is distributed in more number of slots. The distributed winding can be
classified as single layer winding and double layer winding.

In a single layer winding, there will be only one coil side in the slot having any number of
conductors, odd or even integer depending on the number of turns of the coil. In a double layer
winding, there will be 2 or multiple of 2 coil sides in the slot arranged in two layers. Obviously
conductors / slot in a double layer winding must be an even integer.

Since for a given number of conductors, poles and slots, a single layer winding calls for less
number of coils of more number of turns, reactance voltage proportional to (turn)2 is high. This
decreases the quality of commutation or leads to sparking commutation. Hence a single layer
winding is not generally used in DC machines. However it is much used in alternators and
induction motors where there is no commutation involved.

Since a double layer winding calls for more number of coils of less number of turns/coil,
reactance voltage proportional to (turn)2 is less and the quality of commutation is good. Hence
double layer windings are much used in DC machines.
Unless otherwise specified all DC machines are assumed to be having a double layer winding.
A double layer winding can further be classified as simplex or multiplex and lap or wave
winding.

CHOICE BETWEEN A LAP AND WAVE WINDING

ZNP
Since emf induced or back emf E = , number of armature conductors Z
60A
60EA
= = KA for given value of E, N and P . K is a constant. Therefore Zlap = KP, Zwave =
NP
P
K2 and Zlap= Zwave . It is clear that the number of conductors, coils and commutator
2
segments in a lap winding is more than that of a wave winding for the same number of turns
/coil in both cases. Also since a lap winding is generally equipped with equalizer rings, cost of
a lap winding is more than that of a wave winding.

11
Since a lap winding provides more number of parallel paths, it is generally used for
low
voltage and high current machines (Ex: machines used for electroplating, battery charging etc).
A wave winding is generally used for high voltage and low current machines.

Note:
a. A lap winding can be used for any capacity machine. But a wave winding is to be used only
for low capacity machines. As the cost of the wave winding is less, almost all low capacity
machines employ wave winding.

b. While selecting the type of winding the following conditions must also be considered as for
as possible.
i) Commutator segment pitch c = ( x diameter of the commutator Dc) / number of
commutator segments should not be less than 4mm. Otherwise the segments becomes
mechanically week. (The value 4mm is applicable to medium capacity machines and must
only be taken as a guiding value and need not be strictly followed).

ii) Voltage between the commutator segments Ebon open circuit

emf induce should not be more than 20V. Otherwise arcing between
d x
P the
number of
segments
segments might result. Obviously voltage / conductor should not be greater than 10V
on
open circuit.
12
 Sketch showing the coil end connection to commutator
segments

NUMBER OF ARMATURE CONDUCTORS

ZNP
Since the emf induced or back emf E = , number of armature conductors
60A
60 EA B π DL
av

Z= NP where = P Wb.For a preliminary design emf induced can be taken as the

terminal voltage in case of a generator and back emf can be taken as the applied voltage in case
of a motor.
The number of armature conductors can also be calculated by considering the specific electric

loading. In that case Z = q A D . The number of conductors must be an even integer.

I a

CROSS-SECTIONAL AREA OF THE ARMATURE CONDUCTORS

Since the armature conductors are connected in series in any of the parallel paths of
dc
machine, maximum value of the current to be carried by the conductor is Ia / A and the cross
sectional area of the conductor a = Ia / A mm2. The current density lies between 4.5 and 7.0
A / mm2 in practice.
Note:
As the current density increases, cross sectional area decreases, resistance of the conductor
2
increases, I R loss increases and temperature of the machine increases. If the cooling facility is
not good, temperature rise will unnecessarily be high. Hence higher value of current density
should be used in machines where the peripheral velocity or speed of operation is high or
where cooling facility is good.

For the calculated area ‘a’, a round, square or rectangular section conductor can be used.
(Trapezoidal section conductor is also used but only in case of squirrel cage rotor of an
induction motor for increasing the starting torque).

13
CHOICE BETWEEN A ROUND AND RECTANGULAR CONDUCTOR

By considering the coil space factor, what type of conductor section is preferable can
be
decided.
Coil space factor Sf is defined as the ratio of the area of copper in the coil to the area of the
coil. That is,

Area of copper in the coil a T f

f

S = < 1.0
f
Overall area of the coil h d
f f
where
af – cross sectional area of copper or conductor
Tf – number of copper sections (or turns) in the coil
df – depth of the coil and hf– height of the coil.

If d is the diameter of the bare conductor and di is the diameter with insulation, then in case of
round conductor
2
S ( d / 4) x 0.785 (d / d 2
T )
f
f 2 i
d x T
i f
and is always less than 0.785. In practice S f lies between 0.3 and 0.65 in case of
round
conductors.

If d and t are the depth and width of the rectangular / square conductor without insulation and
di and ti are the depth and width with insulation, then in case of a rectangular /square conductor
dt x T dt
Sf = S= f
f
=d t and lies between 0.7 and 0.85.
i i
d t x T
i i f
[Note:
Since the rectangular conductors can be wound closely without wasting any space in between
conductors, Sf in case of a rectangular conductor is more than that of a round conductor of
same cross-sectional area.]

If af = 100mm2 andTf = 48, then area of the coil,

14
 a T 100 x 48 2
, in case of round conductors with the
f f
hd= = = 8000 mm
f f
S 0.6
f
assumption that Sf = 0.6.
100 2
x 48 6000 mm , in case of a rectangular conductors withthe
0.8
assumption thatSf = 0.8

It is clear that rectangular conductors call for less space as compared to round
conductor of
same cross sectional area. Though a rectangular conductor is preferable to round conductor, it
is costlier than the round conductors. (Quantity of insulation required in case of a rectangular
conductor is more than that of the round conductor. Hence the cost is more). Hence rectangular
conductors are used when the conductor cross sectional area is more than about 10 mm2.

DETAILS OF INSULATION

Desirable properties of a good insulator:

1. Should have a high dielectric strength

2. Should withstand a high temperature
3. Should have a high resistivity
4. Should not absorb moisture
5. Should have a low dielectric hysteresis loss
6. Should withstand vibration, abrasion, bending
7. Should be flexible and cheap.

No insulating material in practice satisfies all the desirable properties. Therefore materials
which satisfy most of the desirable properties must be selected. Based on the temperature
withstanding ability, the insulating materials have been classified as indicated below according
to ISI (now BIS) – 1271, 1985, first revision.

Class of insulation – Y A E B H C
F
Maximum temperature 0C – 90 0C 105 0C 120 0C 135 0C 155 0C 180 0C >180

Class Y : Cotton, silk, paper, wood, cellulose, fiber etc., without impregnation or oil
immersed.

Class A : The material of class Y impregnated with natural resins, cellulose esters,
insulating oils etc., and also laminated wood, varnished paper etc.

Class E : Synthetic resin enamels of vinyl acetate or nylon tapes, cotton and paper
laminates with formaldehyde bonding etc.,

Class B : Mica, glass fiber, asbestos etc., with suitable bonding substances, built up
mica, glass fiber and asbestos laminates.
15
Class F : The materials of Class B with more thermal resistance bonding materials.

Class H : Glass fiber and asbestos materials and built up mica with appropriate silicone
resins.

Class C : Mica, ceramics, glass, quartz and asbestos with binders or resins of super
thermal stability.

Insulation on conductors:
In practice, the most important factor influencing the structure of the insulation is the
rated
voltage of the machine. Based on the rated voltage, electrical machines can be divided into
high voltage and low voltage machines. A low voltage machine is a machine the rated voltage
of which is below 1000V.

Insulation on the conductors serves as an inter turn insulation and may be enamel, cotton
(single cotton covered SCC or double cotton covered DCC), rayon, silk, glass fiber, asbestos
etc.

Addition of insulation to
a. Round wires or conductors – (0.015 to 0.075) mm for enamel covering
- (0.2 to 0.35) mm for dcc
b. Strip or rectangular conductors – (0.06 to 0.15) mm for enamel covering
- (0.45 to 0.5) mm for dcc

For commercial voltages (up to 1000V) enameled wires are generally used.
For further details chapter.3, Electrical Machine Design Data Book, A.Shanmugasundaram can
be referred.

TYPES OF SLOTS

The slots can be open, semi-closed or closed type. Slots may be parallel sided or tapered
(varying width). The tooth may also be of parallel sided or tapered. Whenever a double layer
winding is used, slots of parallel sided (or of constant slot width) are to be used.

Plain un-notched slots or slots without wedge portion are used only for low capacity machines.
In this case the coils are restricted from coming out of the slot by the use of steel binding wires.
This type of slots is used only in machines where the peripheral velocity is not more than about
25 m/s. For machines of higher peripheral velocity ear notched slots or slots with lip
and
wedge portions are used.

16
Since wide open slots increases the reluctance of the air gap and causes the flux to pulsate in
comparison with the closed or semi closed slots, open slots are not preferable. But open slots
offers the advantage of placing the coils into the slots and removing them from the slots
easilyfor repair. Also coils can be fully insulated or tapped before being placed into the slots.

With semi closed slots, the coils are to be placed into the slots, turn by turn and therefore it is
time consuming and expensive.

With closed slots, preformed coil placement and removal is very difficult and hence no closed
slots are used in practice for DC machine armature, stator of alternator or induction motor.
However closed slots are used in case of squirrel cage rotor of an induction motor.

In general open type of slots is used in case of high capacity machines and semi closed slots in
case of medium and low capacity machines.

17
NUMBER OF ARMATURE SLOTS

In order to decide what number of slots (more or less) is to be used, the following merits and
demerits are considered.
1. As the number of slots increases, cost of punching the slot increases, number of coils
increases and hence the cost of the machine increases.
2. As the number of slots increases, slot pitch
s = (slot width bs + tooth width bt)

= D/ number of slots S

decreases and hence the tooth width reduces. This makes the tooth mechanically weak,
increases the flux density in the tooth and the core loss in the tooth. Therefore
efficiency of the machine decreases.

If the slots are less in number, then the cost of punching & number of coils decreases, slot pitch
increases, tooth becomes mechanically strong and efficiency increases, quantity of insulation in
the slot increases, heat dissipation reduces, temperature increases and hence the efficiency
decreases.
It is clear that not much advantage is gained by the use of either too a less or more number of
slots.

As a preliminary value, the number of slots can be selected by considering the slot pitch. The
slot pitch can assumed to be between (2.5 and 3.5) cm. (This range is applicable to
only to
medium capacity machines and it can be more or less for other capacity machines).

The selection of the number of slots must also be based on the type of winding used, quality of
commutation, flux pulsation etc.

When the number of slot per pole is a whole number, the number slots embraced by each pole
will be the same for all positions of armature. However, the number teeth per pole will not be
same.

This causes a variation in reluctance of the air gap and the flux in the air gap will
pulsate.
Pulsations of the flux in the air gap produce iron losses in the pole shoe and give
rise to
18
 magnetic noises. On the other hand, when the slots per pole is equal to a whole number plus
half the reluctance of the flux path per pole pair remains constant for all positions of the
armature, and there will be no pulsations or oscillations of the flux in the air gap.

To avoid pulsations and oscillations of the flux in the air gap, the number of slots per pole
should be a whole number plus half. When this is not possible or advisable for other reasons,
the number of slots per pole arc should an integer.

Pole shoe

1 2 3 4 5

Number of teeth/pole shoe = 5 and flux passes through 5 teeth.

The reluctance of the air gap is inversely proportional to the area corresponding to 5 teeth.

Pole shoe

1 2 3 4 5 6

Number of teeth/pole shoe = 5 and flux passes through 6 teeth when the armature is moved half
tooth pitch to the right.The reluctance of the air gap is inversely proportional to the area
corresponding to 6 teeth. The reluctance in this case is less and the flux is more compared to
the former case. Therefore, the flux pulsates i.e. Varies in magnitude.

Pole shoe

1 2 3 4 5 6

Number of teeth/pole shoe = (5+ 0.5) and flux passes through 6 teeth.
The reluctance of the air gap is inversely proportional to the area corresponding to 6
teeth.

19
 Pole Pole shoe

1 2 3 4 5 6

Number of teeth/pole shoe = (5+0.5) and flux passes through 6 teeth when the armature
is moved half tooth pitch to the right.The reluctance of the air gap is inversely
proportional 6 teeth as before. The reluctance and the flux in both the cases remains the
same in all positions of the armature. However, the reluctance and the flux under the
tips of the pole are not the same for all the positions of armature. Therefore when the
armature rotates the flux under the pole oscillates between the pole tips. This produces
ripple in the voltage induced in the conductors moving under poles.

The flux pulsation under inter pole causes the sparking. A small tooth pitch helps to reduce the
effect of armature slots upon the inter poles.

To obtain good commutation, the flux density in the air gap must decrease gradually from
maximum value under the center of the pole to zero on the center line between two poles, and the
flux densities near the neutral point must be low. A field form that drops off rapidly
from
maximum value to zero not only leads to commutation difficulties but may also give rise
to
noises in machines with slotted armatures. In order to achieve good commutation the pole shoe is
designed to cover only certain percentage of the pole pitch.
The circumferential distance covered by the pole shoe on the armature surface is called the pole
arc. The ratio of the pole arc to pole pitch is called per unit embrace or enclosure. That is, per
Pole arc
unit enclosure Pole 1.0. In practice ψ lies between 0.6 and 0.7.
ψ=
pitch
S
In general, the slots between pole tips of two adjacent pole tips i.e. (1- should be at
ψ) P
least 3, or
S
(1-ψ) ≥ 3
P
If Ψ=0.66,the number of slots per pole,

S≥ 3 ≥ 8.82 or say 9
P (1-0.66)

Rule to select the number of slots

1. From satisfactory commutation point of view, minimum number of slots / pole can be
taken as 9.
2. To avoid flux pulsation, slots / pole should be an (integer + ½).
20
 3. Slot pitch s = D/S can assumed to lie between 2.5 and 3.5 cm.
4. For a lap winding, number of slots may be multiple of number of poles P or number of
pair of poles p (p=P/2) so that equalizer rings can used.
5. For a wave winding, number of slots should not be a multiple of number of pair of
poles p and should lead to a particular number of coils that makes the commutator or
average pitch Yc = [(C plex) / p], an integer.

[Note: With a particular number of coils, if Yc does not work out to be an integer, then a wave
winding can not be realized. However, it be realized by considering one of the coils as dummy.
In that case the number of commutator segments is equal to the number of active coils or (total
number of coils – 1). A dummy coil is coil which is electrically inactive. It is made inactive by
not connecting its coil ends to the commutator segments. Against its absence, its presence
ensures that all the slots are completely filled and armature is mechanically balanced.

Number of coils C = number of slots S x number of coil sides / layer

= number of coil sides / 2
Total number of armature conductors Z
=
number of conductors / coil
Z
=
2 x conductors / coil side
Z
=
2 x number of turns / coil
In general the number of commutator segments is equal to the number of active coils. ]

Size of the armature slots

Width and depth of the slot depends on the number of conductors per slot, cross sectional area
of the conductor, rated voltage etc.
[Note:Theinsulation within the slot can be roughly divided into two main categories: ground
wall or slot liner and that associated with conductor insulation and their arrangement in the
slot.
Slot liner: Protects the winding insulation against any mechanical damage due to roughness of
the slot walls and insulate the armature. It also galvanically separates the coil from the iron
core of the machine.Materials used generally are Leatheroid,Melinex® polyester films, horn
fiber etc.
Separator: Separates and insulates the two layers. Leatheroid, horn fiber, micanite etc., are
used.

21
Conductor insulation: Separates the wires and turns of a coil. Usually, standards are
not as
high for conductor insulation as they are for slot liner insulation, and therefore the former is
usually notably thinner than the latter. Consequently a conductor or turn to turn insulation is
usually, and in small machines in particular, a varnishing on the wire.]

Slot width bs = (Width or diameter of the conductor + insulation on it) number of conductors
along the width of the slot + (insulation over the coil or group of coils + slot liner + clearance)

Slot depth ht = (Depth or diameter of the conductor + insulation on it) number of conductors
along the depth of the slot + (insulation over the coil or group of coils + separator + slot liner +
clearance) + wedge 3 to 5 mm + lip 1 or 2 mm.

22
For a preliminary design, widthwise sum of insulation over the coil or group of coils, slot liner
and clearance can be taken as 1.5mm and depth wise sum of insulation over the coil or group
of coils, separator, slot liner and clearance 4mm.

For a given slot pitch, as the slot width and depth increases, the tooth width becomes less and it
will be mechanically weak. In general bs should not be greater than 60% of slot pitch. The slot
width and depth must be so selected that the flux density in the tooth at one third height from
the root of the tooth Bt1/3 is not greater than 2.1T. That is

Bt 1/3 b L x should not be more than 2.1T

S/P
t1/3i
b = width of the armature tooth at 1/3 height from the root of the tooth
t1/3
= – bs
s1/3
(D - 4/3 h )
t -b s

= S

23
Armature core depth dc:
/2
Flux density in the armature core lies between (1.0 and 1.5)T. With an assumed
dL
c i

Bc =

value of Bc , dc = /2 can be calculated.

BL
c i

Therefore the internal diameter of the armature = (D – 2ht - 2dc ).

**********

24