Chapter 27

SINGLE-PHASE IM DESIGN

27.1 INTRODUCTION

By design we mean dimensioning. That is, the finding of a suitable

geometry and manufacturing data and performance indexes for given
specifications. Design, then, means first dimensioning (or synthesis), sizing,
then performance assessment (analysis). Finally if the specifications are not met
the process is repeated according to an adopted strategy until satisfactory
performance is obtained.
On top of this, optimization is performed according to one or more
objectives functions, as detailed in Chapter 18 in relation to three phase
induction machines. Typical specifications (with a case study) are
Rated power Pn = 186.5 W (1/4 HP)
Rated voltage Vsn = 115 V
Rated frequency f1n = 60 Hz
Rated power factor cosn = 0.98 lagging service continuous or
shortduty
Breakdown p. u. torque 1.3-2.5
Starting p. u. torque 0.5-3.5
Starting p. u. current 5-6.5
Capacitor p. u. maximum voltage 0.6-1.6
The breakdown torque p.u. may go as high as 4.0 for the dual capacitor
configuration and special-service motors.
Also starting torques above 1.5 p. u. are obtained with a starting capacitor.
The split-phase IM is also capable of high starting and breakdown torques
in p. u. as during starting both windings are active, at the expense of rather high
resistance, both in the rotor and in the auxiliary stator winding.
For two (three) speed operation the 2 (3) speed levels in % of ideal
synchronous speed have to be specified. They are to be obtained with tapped
windings. In such a case it has to be verified that for each speed there is some
torque reserve up to the breaking torque of that tapping.
Also for multispeed motors the locked rotor torque on low speed has to be
less than the load torque at the desired low speed. For a fan load, at 50% as the
low speed, the torque is 25 % and thus the locked rotor torque has to be less
than 25%.
We start with the sizing of the magnetic circuit, move on to the selection of
stator windings, continue with rotor slotting and cage sizing. The starting and
(or) permanent capacitors are defined. Further on the parameter expressions are
given and steady state performance is calculated. When optimization design is
performed, objective (penalty) functions are calculated and constraints are

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

verified. If their demands are not met the whole process is repeated, according
to a deterministic or stochastic optimization mathematical method, until
sufficient convergence is reached.

27.2 SIZING THE STATOR MAGNETIC CIRCUIT

As already discussed in Chapter 14, when dealing with design principles of

three phase induction machines, there are basically two design initiation
constants, based on past experience
The machine utilisation factor Cu in W/m3 D02L
The rotor tangential stress ft in N/cm2 or N/m2
D0 is the outer stator diameter and L-the stator stack length.
As design optimization methods advance and better materials are produced,
Cu and ft tend to improve slowly. Also, low service duty allows for improved in
Cu and ft.
However, in general, better efficiency requires larger Cu and lower ft.
Figure 27.1 [1] presents standard data on Cu

C u = D 02 L (27.1)

for the three phase small power IMs.

Figure 27.1 Machine utilization factor Cu=D02L (cubic inches) for fractional/horsepower three-phase
IMs.

Concerning the rated tangential stress there is not yet a history of its use but
it is known that it increases with the stator interior (bare) diameter Di, with
values of around ft = 0.20 N/cm2 for Di = 30 mm to ft = 1 N/cm2 for Di = 70 mm
or so , and more.

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

In general, any company could calculate ft (Di) for the 2, 4, 6, 8 single phase
IMs fabricated so far and then produce its own ft database.
The rather small range of ft variation may be exploited best in our era of
computers for optimization design.
The ratio between stator interior (bore) diameter Di and the external
diameter Do depends on the number of poles, on Do and on the magnetic (flux
densities) and electric (current density) loadings.
Figure 27.2. presents standard data [1] from three sources T. C. Lloyd, P.
M. Trickey and Reference 2. In Reference 2, the ratio Di/Do is obtained for
maximum airgap flux density in the airgap per given stator magnetisation m.m.f.
in three phase IMs (Di/Do = 0.58 for 2p1 = 2, 0.65 for 2p1 = 4, 0.69 for 2p1 = 6,
0.72 for 2p1 = 8).
The Di/Do values of Reference 2 are slightly larger than those of P. M.
Trickey, as they are obtained from a contemporary optimization design method
for 3 phase IMs.
In our case study, from Figure 27.1, for 186.5 W (1/4 HP), 2p1 = 4 poles,
we choose

C u = D o2 L = 3.5615 10 3 m 3

with L/Do = 0.380, Do = 0.137 m, L = 0.053 m.

The outer stator punching diameter Do might not be free to choose, as the
frames for single phase IMs come into standardized sizes [3].
For 4 poles we choose from fig. 27.2 a kind of average value of the three
sets of data Di/Do = 0.60. Consequently the stator bore diameter Di = 0.6
0.137 = 82.8 10-3 m.
The airgap g = 0.3 mm and thus the rotor external diameter Dor = Di-2g =
(82.8-20.3)10-3 = 82.2 10-3 m.
The number of slots of stator Ns is chosen as for three phase IMs (the rules
for most adequate combinations Ns and Nr established for three phase IMs hold
in general also for single phase IMs see chapter 10). Let us consider Ns = 36 and
Nr = 30.
The theoretical peak airgap flux density Bg = 0.6-0.75 T.Let us consider Bg
= 0.705 T. Due to saturation it will be somewhat lower (flattened).
Consequently, the flux per pole in the main winding m is
2
m K dis Bg L (27.3)

The pole pitch,
Di
= = 82.2 10 3 64.57 10 3 m (27.4)
2p1 4

With Kdis = 0.9

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

Figure 27.2 Interior/outer diameter ratio Di/Do versus Do.

2
m = 0.9 0.705 64.57 10 3 0.053 1.382 10 3 Wb

Once we choose the design stator back iron flux density Bcs = 1.3-1.7 T, the
back iron height hcs may be computed from

m 1.382 10 3
h cs = = = 8.69 10 3 9 10 3 m (27.4)
2 Bcs L 2 1.5 0.053

The stator slot geometry is shows on Figure 27.3.

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

hcs=9
W2s =8

hss=16
h ts =18
b ts =2.5
W1s =5
hos=1

hws=1
Wos =1.8

Di = 82.8
D0=137

Figure 27.3 Stator slot geometry in mm

The number of slots per pole is Ns/2p1 = 36/(22) = 9. So the tooth width bts
is
m 2p1
b ts = (27.5)
N s B ts L

With the tooth flux density Bts (0.8-1.0) Bcs

1.382 10 3 4
b ts = 2.55 10 3 m (27.6)
36 1.3 0.053
This value is close to the lowest limit in terms of punching capabilities.
Let us consider hos = 1 10-3 m, wos = 6g = 6 0.3 10-3 = 1.8 10-3 m.
Now the lower and upper slot width w1s and w2s are
(D i + 2(h os + h ws )) (82.8 + 2(1 + 1))
W1s = b ts = 2.55 10 3
Ns 36 (27.7)
5.00 10 3 m

(D o 2h cs ) (137 2 9 )
W2s = b ts = 2.5 10 3 8.00 10 3 m (27.8)
Ns 36
The useful slot height is

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

Do D 137 82.8 3
h ts = h cs h ws h os i = 92 10
2 2 2 2 (27.9)
16.0 10 3 m
So the active stator slot area As is

As =
(W1s + W2s ) h =
(5.00 + 8.00) 16 10 6 = 104 10 6 m 2 (27.9)
ss
2 2
For slots which host both windings or in split phase IMs some slots may be
larger than others.

27.3 SIZING THE ROTOR MAGNETIC CIRCUIT

The rotor slots for single phase IMs are either round or trapezoidal or in
between (Figure 27.4).
Wor ~1 Wor ~1 Wor ~1
h or=1
hor hor
r r1
h 2=10 W1r =6
h2

r2

W2r =3.5

Figure 27.4 Typical rotor slot geometries

With 30 rotor slots, the rotor slot pitch sr is

D r 82.2 103 4
sr = = = 8.6 10 3 (27.10)
Nr 30

With a rotor tooth width btr = 2.6 10-3 m the tooth flux density Btr is
2p1 1.388 103 4
Btr = = = 1.343T (27.11)
N r b tr L 30 2.6 10 3 0.053

The rotor slots useful area is in many cases (3560) % of that of the stator
slot
As Ns 36
A r = 0.38 = 0.38 104 10 6 = 47.42 10 6 m 2 (27.12)
Nr 30

The trapezoidal rotor slot (Figure 27.4), with

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

h or = Wor = 1 10 3 m; h or1 = 1 10 3 m
(27.13)
has W1r sr b tr = (8.6 2.6 ) 10 3 = 6 10 3 m

Adopting a slot height h2 = 10 10-3 m we may calculate the rotor slot

bottom width W2r
2A r 2 47.40 103
W2 r = W1s = 6 10 3 = 3.48 10 3 m (27.14)
h2 10 10 3

A geometrical verification is now required.

[D r (h or + h or1 + h 2 ) 2]
W2 r = b tr =
Nr
(27.15)
[82.2 2(1 + 1 + 10)] 10 3
= 2.6 10 3 = 3.49 10 3 m
30
As (27.14) and (27.15) produce the same value of slot bottom width, the
slot height h2 has been chosen correctly. Otherwise h2 should have been
changed until the two values of W2r converged.
To avoid such an iterative computation (27.14)(27.15) could be combined
into a second order equation with h2 as the unknown after the elimination of
W2r.
A round slot with a diameter dr = W2r = 6 10-3 m would have produced an

( 2
)
area A r = 6 10 3 = 28.26 10 6 m 2 .
4
This would have been too small a value unless copper bars are used instead
of aluminum bars.
The end ring area Aring is
1 1
A ring A r = 47.42 10 6 = 114 10 6 m 2 (27.16)
p1 2
2 sin 2 sin
Nr 30

With a radial height br = 15 10-3 m, the ring axial length

A ring
ar = = 7.6 10 3 m .
br

27.4 SIZING THE STATOR WINDINGS

By now the number of slots in the stator is known Ns = 36. As we do have a

permanent capacitor motor the auxiliary winding is always operational. It thus
seems natural to allocate both windings about same number of slots in fact
20/16. In general single layer windings with concentrated coils are used. (Figure
27.5)

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

m m m mm a a a a m mmm m a a a a mmm mm a a a a mm mm m a a a a

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

Figure 27.5 Stator winding m-main phase; a-auxiliary phase.

With identical coils per slot and phase; the winding factors for the two
phases (Chapter 4) are

sin q m s sin 5
2 36
K wm = = = 0.9698

q m sin s 5 sin
2 36

sin q a s sin 4
2 36
K wa = = = 0.9810 (27.17)
s
q a sin 4 sin
2 36
There are 5 (qm = 5) slots per pole per phase for the main winding and 4 (qa
= 4) for the auxiliary one.
In case the number of turns/phase in various slots is not the same, the
winding factor can be calculated as shown below.
As detailed in chapter 4, two types of sinusoidal windings may be built
(Figure 27.6)
a a
r=1
Z1 Z1 Z1 r = 1 Z1
Z2 Z2 Z2 Z2
Z3 Z3 Z3 Z3

a a tooth axis
slot axis
a.) b.)
Figure 27.6 Sinusoidal windings a.) with slot axis symmetry; b.) with tooth axis symmetry.

When the total number of slots per pole per phase is an odd number,
concentrated coils with slot axis symmetry seem adequate. In contrast, for even
number of slots/pole/phase, concentrated coils with tooth axis symmetry are
recommended.
Should we have used such windings for our case with qm = 5 and qa = 4, slot
axis symmetry would have applied to the main winding and tooth axis symmetry
to the auxiliary winding. A typical coil group is shown on Figure 27.7.

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

From Figure 27.7 the angle between the axis a-a (Figure 27.6-27.7) and the
kth slot, k (for the th harmonic), is
2
k = + (k 1) (27.18)
Ns Ns

a a
Z1
Z2
Z3
L
l ec pole
length
Z3

De Z2 3
Di
Z1 2
1 a

Figure 27.7 Typical overlapping winding

The effective number of conductors per half a pole Z ' is [4]

n
Z ' = Z k cos k
k =1 (27.19)
N = 4p1 Z '

N is the effective number of conductors per phase.

The winding factor Kw is simply
n

Z ' Zk cos k
K w = n

= k =1
n
(27.20)
Zk Zk
k =1 k =1

The length of the conductors per half a pole group of coil, lcn, is [3]
n
l cn Z k L + ( 2)L ec + D e k1 (27.21)
k =1 2p1
So the resistance per phase Rphase is
4p1l cn
R phase = Co (27.22)
A con

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

Acon is the conductor (magnet wire) cross-section area. Acon depends on the
total number of conductors per phase, slot area and design current density.
The problem is that, though the supply current at rated load may be
calculated as
Pn 186.5
Is = = = 2.36A , (27.23)
n Vs cos 0.7 115 0.98

with an assigned value for the rated efficiency, n, still the rated current for the
main and auxiliary currents Im , Ia are not known at this stage.
Im and Ia should be almost 900 phase shifted for rated load and
Is = I m + Ia (27.24)

Im Na
Now as the ratio a = = , for symmetry, we can assume that
Ia Nm

Is
Im (27.25)
2
1
1+
a
With a the turns ratio in the interval a = 1.0-2, in general
I m = I s (0.700 0.90) .
The number of turns of the main winding Nm is to be determined by
observing that the e.m.f. in the main winding E m = (0.96 0.98) Vs and thus
with (27.3)

E m = 2 m N m k wm f1n (27.26)

Finally
0.97 115
Nm = = 313turns
21.382 10 3 0.9698 60
In our case, the main winding has 10 (p1qm = 2 5 = 10) identical coils. So
the number of turns per slot nsm is
Nm 313
n sm = = = 31turns / coil (27.27)
q m p1 5 2

Assuming the turn ratio is a = 1.5, the number of turns per coil in the
auxiliary winding is
N m a k wm 310 1.5 0.9698
n sa = = 57 turns / coil (27.28)
q a p1 k wa 4 2 0.9810

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

As the slots are identical and their useful area is (from 27.9) As = 104 10-6
2
m , the diameters of the magnetic wire used in two windings are

4 A s K fill 4 104 0.4

dm = = 10 3 1.3 10 3 m
n sm 3.14 31
(27.29)
4 A s K fill 4 104 0.4
da = = 10 3 1.0 10 3 m
n sa 3.14 57

The filling factor was considered rather large Kfill = 0.4.

The predicted current density in the two windings would be
Im 2.36
j Com = = = 1.4847 10 6 A / m 2
d 2m 2 2
1 1 1.3 6
4 + 1.5 4 10
(27.30)
Im 1.9696
j Coa = = = 1.6726 10 6 A / m 2
a d a2 1.5 12 10 6
4 4
The forecasted current densities, for rated power, are rather small so good
efficiency (for the low power considered here) is expected.
An initial value for the permanent capacitor has to be chosen.
Let us assume complete symmetry for rated load, with
I 1.96
I an = mn = = 1.3A .
a 1.5
Further on
Van = Vmn a = 115 1.5 = 172V (27.31)

The capacitor voltage Vc is

Vc = Vs2 + Van2 = 115 1 + 1.52 = 207.3V (27.32)

Finally the permanent capacitor C is

Ia 1.3
C= = 16F (27.33)
1VC 2 60 207.3

We may choose a higher value of C than 16 F (say 25 F) to increase

somewhat the breakdown and the starting torque.

27.5 RESISTANCES AND LEAKAGE REACTANCES

The main and auxiliary winding resistances Rsm and Rsa (27.22) are

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

l cnm
R sm = Co 4
d 2m
4 (27.34)
l cna
R sa = Co 4 2
d a
4
The length of the coils/half a pole for main and auxiliary windings lcnm
(27.21) are

1 3 2 4
l cnm n cm 2 + L + 3( 2)L ec + D e =
2 2p1 36 36
1 3 2 4
= 31 2 + 0.053 + 3( 2) 0.025 + 0.14 = 9.805m
2 4 36 36

(27.35)

2 2 4
l cna n ca 2L + 3( 2)L ec + D e =
2p1 36 36
(27.36)
2
= 57 2 0.053 + 3( 2) 0.025 + 0.14 = 11.38m
4 18 9
Finally
2.1 10 8 4 2 9.805
R sm = = 1.2417

1.32 10 6
4
2.1 10 8 4 2 11.38
R sa = = 2.435
2
1 10 6
4
The main winding leakage reactance Xsm
X sm = X ss + X se + X sd + X skew (27.37)

where Xss-the stator slot leakage reactance

Xse-the stator end connection leakage reactance
Xsd-the stator differential leakage
Xskew-skewing reactance

W12 L
X se + X ss = 2 0 1 ( ss + se ) (27.38)
pq
with (chapter 6 and Figure 27.3)

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

2h ss 2h ws h 2 16 2 1 1
ss = + + os = + + = 1.53 (27.39)
3(W1s + W2s ) W1s + W2s Wos 3(5 + 8) 5 + 8 1.8

Also (chapter 6, equation (6.28))

qm
0.67
se = 2 (l 0.64 ) (27.40)
ec
L
It is qm/2 as the q coil/pole are divided in two sections whose end
connections go in opposite directions (Figure 27.5).
The end connection average length lec is
l cn 2
m
l ec = L (27.41)
q m n cm

With lcn from (27.21)

9.805 2
m
l ec = 0.053 = 0.0735m
5 31

5
0.67
se = 2 (0.0735 0.64 0.065) = 1.008
0.053
The differential leakage reactance Xsd (chapter 6, Equation (6.2) and (6.9)
and (Figure 6.2-Figure 6.5)) is
X sd
= ds d = 2.6 10 2 0.92 = 2.392 10 2 (27.42)
X mm

for qm = 5, Ns = 36, Nr = 30, one slot pitch skewing of rotor.

Xmm is the nonsaturated magnetization reactance of the main winding (Chapter
5, Equation (5.115))
4 0
X mm = 1 (Wm K wm1 )2 L =
2 p1 g K C
4 4 10 7
= 2 60 (10 31 0.9698)2 0.053 0.065 = 72.085
2 2 0.3 10 3 1.3
KC-Carter coefficient; g-the airgap.
skew
sin 2 sin 2
X skew
(
= 1 K skew
2
= 1) 2 = 1 18 = 9.108 10 3 (27.43)
X mm ( skew )2
2

18
Finally the main winding leakage reactance Xsm is

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

2 60 310 2 0.053 (1.53 + 1.008)

X sm = 2 4 10 7 +
25
+ 72.085 (23.92 + 9.108) 10 3 = 4.253

The auxiliary winding is placed in identical slots and thus the computation
process is similar. A simplified expression for Xsa in this case is
2
W 57 8
X sa X sm a = 4.253 = 9.204 (27.44)
Wm 31 10

The rotor leakage reactance, after its reduction to the main winding (chapter
6, equations (6.86)-(6.87)) is

(
X rm = X rmd + X 'bem = 1 L rmd + L'bem ) (27.45)

The rotor-skewing component has been attached to the stator and the zig-
zag component is lumped into differential one Xrmd.
From Equation (6.16) (Chapter 6)

X rmd = dr 0 X mm = 2.8 10 2 X mm (27.46)

With Nr/p1 = 30/2 = 15 and one slot pitch skewing, from Figure 6.4

dr 0 = 2.8 10 2

The equivalent bar-end ring leakage inductance L' ben is (Equation 6.86)

12 K 2wm Wm2
L'bem = L bem (27.47)
Nr

D ir
From (6.91)-(6.92) L ben = 0 l b b + 2 0 ei 2 (27.48)
Nr

With the rotor bar (slot) and end ring permeance coefficients, b and ei
(from (6.18) and (6.46) and Figure 27.5)
h or 2h or1 2h 2 1 1 2 2 10
b = + + = + + = 1.987 (27.49)
Wor (Wo1 + Wo 2 ) 3(W1r + W2 r ) 1 1 + 6 3(6 + 3.5)

2.3 D ir 4.7 D ir
ei = log (27.50)
p a r + 2 br
4 N r L sin 2 1
N
r

With D ir D r b = (82.8 15) 10 3 = 67.8 10 3 m , a r = 7.6 10 3 m and

b r = 15 10 3 m from (27.16)

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

2.3 67.8 10 3 4.7 67.8

ei = log = 0.579
2 15 + 7.6
4 30 0.053 sin 2
15
so from (27.48)

2 67.8 10 3
L ben = 1.256 10 6 0.053 1.987 + 0.579 = 0.1425 10 6 H
30
Now from (27.45) and (24.76)

12K 2wm Wm2

X rm = 2.8 10 2 X mm + 1L ben =
Nr
12 0.96982 310 2
= 2.8 10 2 72.085 + 2 60 0.1425 10 6 = (27.51)
30
= 3.961
The rotor cage resistance

12K 2wm Wm2

R rm = R be (27.52)
Nr

With Rbe (equation (6.63))

R ring
R be = R b + (27.53)
p1
2 sin 2
Nr

lb 0.065
R b = Co = 3 10 8 = 4.112 10 5
Ar 47.42 10 6
(27.54)
l ring 2 67.8 10 3
R ring = b = 3 10 8 = 3.735 10 6
A ring 1.14 10 4 30

Finally

6
12 0.9698 310 = 3.048
2 2
3.735 10
R rm = 4.112 10 +
5

2 30
2 sin 2
30
Note The rotor resistance and leakage reactance calculated above are not
affected by the skin effects.

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

27.6 THE MAGNETIZATION REACTANCE X mm

The magnetisation reactance Xmm is affected by magnetic saturation which

is dependent on the resultant magnetisation current Im.
Due to the symmetry of the magnetisation circuit it is sufficient to calculate
the functional Xm (Im) for the case with current in the main winding (Ia = 0, bare
rotor).
The computation of Xm(Im) may be performed analytically (see Chapter 6 or
Ref. 4) or by F.E.M. For the present case, to avoid lengthy calculations let us
consider Xmm constant for a moderate saturation level (1+Ks) = 1.5
1 72.085
X mm = (X mm )unsat = = 48.065 (27.55)
1 + Ks 1.5

27.7 THE STARTING TORQUE AND CURRENT

The theory behind the computation of starting torque and current is

presented in chapter 24, paragraph 24.5

(Is )S=1 = Isc

(
t es C Zsc a 2 sin sc ) 2
t es2
1 + + 2 (27.56)
a cos sc a

Tes K a sin sc
t es = =
2p1 2
I sc (R r+ )S=1
(
a 1 + K a2 + 2K a cos sc )
1

1 Vs
Ka = ; I sc =
C a 2 Zsc 2 Zsc

Zsc = (R sm + R rm )2 + (X sm + X rm )2 =
(27.57)
= (1.2417 + 3.048)2 + (4.235 + 3.96)2 = 9.26

cos sc = 0.463
sin sc = 0.887

1
Ka = = 5.095
2 60 25 10 6 1.52 9.26

115
I sc = = 6.2095A; (R r + )S=1 = R rm = 3.048
2 9.26

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

2 2
Tsc = 6.20952 3.048 = 2.495Nm
2 60
The relative value of starting torque tes is
5.095 0.887
t es = = 0.095
(
1.5 1 + 5.0952 + 2 5.095 0.463 )
Tes = t es Tsc = 0.095 2.495 = 0.237 Nm

The rated torque Ten, for an alleged rated slip Sn = 0.06, would be
Pn p1 186.5 2
Ten = = 1.053Nm (27.58)
1 (1 Sn ) 2 60 (1 0.06)

So the starting torque is rather small as the permanent capacitance Ca = 25

10-6 F is too small.
The starting current (27.56) is
2
1
0.095 5.095 0.887 0.095 2
(Is )S=1 = 6.2095 1 + + 5.636A
1.5 0.463 1.5

The starting current is not large (the presumed rated source current Isn =
2.36 A,(Equation 27.23)) but the starting torque is small.
The result is typical for the permanent (single) capacitor IM.

27.8 STEADY STATE PERFORMANCE AROUND RATED POWER

Though the core losses and the stray load losses have not been calculated,
the computation of torque and stator currents for various slips for S = 0.04-0.20
may be performed as developed in chapter 24, paragraph 24.3.
To shorten the presentation we will illustrate this point by calculating the
currents and torque for S = 0.06.
m m m
First the impedances Z + , Z and Z a ((24.11)-(24.14)) are calculated

R
X mm jX rm + rm
S
Z + = R sm + j X sm + j =
R rm
+ j(X mm + X rm )
S
(27.59)
3.048
j 48.056 j 3.96 +
0.06
= 1.2417 + j 4.253 + = 23.43 + j 29.583
3.048
+ j (48.056 + 3.96)
0.06

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

R rm
X mm j X rm +
Z = R sm + j X sm (2 S)
+ j
R rm
+ j (X mm + X rm )
(2 S)
(27.60)
3.048
48.056 j 3.96 +
(2 0.06)
1.2417 + j 4.253 + j 2.7 + j 7.912
3.048
+ j (48.056 + 3.96)
(2 0.06)

m 1 R sa 1X 1
Za = R sm + j 2sa X sm j =
2 a2 2 a 2 a 2
C
1 2.435 j
= 1.2417 + 0 =
2 1.52 2 1 .5 2
2 60 25 10 6
= 0.08 j 23.6 j 23.6
The current components Im+ and Im- ((24.16)-(24.17)) are

j m

1 Z + 2Z a
V a = 1.808 j 2.414
I m+ = s (27.62)
2 Z + Z + Z am (Z + + Z )

j m

1 + Z + + 2Z a
V a = 0.304 j 8.197 10 3
I m = s (27.63)
2 Z + Z + Z am (Z + + Z )

Now the torque components T+ and T- are computed from (24.18)-(24.19)

2p1 2
Te+ = I m+ [R e (Z + ) R sm ] (27.64)
1

2p1 2
Te = I m [R e (Z ) R sm ] (27.65)
1

Te = Te+ + Te

The source current Is is

I s = I m + I a = I m+ + I m + j
(I m+ I m ) = 3.721 j 1.422 (27.66)
a
The source power factor becomes

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

R e (Is )
cos 1 = = 0.934 (27.67)
Is

2 2
Te+ = 9.096 [23.43 1.2417] = 2.1425Nm
2 60
2 2
Te = 0.09248 [2.7 1.2417] = 0.967 10 3 Nm
2 60
Note The current is about 60 % higher than the presumed rated current
(2.36 A) while the torque is twice the rated torque Ten 1Nm.
It seems that when the slip is reduced gradually, perhaps around 4 % (S =
0.04) the current goes down and so does the torque, coming close to the rated
value, which corresponds, to the rated power Pn (27.58).
On the other hand, if the slip is gradually increased the breakdown torque
region is reached.
To complete the design the computation of core, stray load and mechanical
losses is required. Though the computation of losses is traditionally performed
as for the three phase IMs, the elliptic travelling field of single phase IM leads
to larger losses [5]. We will not follow this aspect here in further detail.
We can now consider the preliminary electromagnetic design finished.
Thermal model may than be developed as for the three-phase IM (Chapter 12).
Design trials may now start to meet all design specifications. The
complexity of the nonlinear model of the single phase IM makes the task of
finding easy ways to meet, say, the starting torque and current, breakdown
torque and providing for good efficiency, rather difficult.
This is where the design optimization techniques come into play.
However to cut short the computation time of optimization design, a good
preliminary design is useful and so are a few design guidelines based on
experience (see [1]).

27.9 GUIDELINES FOR A GOOD DESIGN

In general a good value for the turn ratio a, lies in the interval between 1.5
to 2.0 except for reversible motion when a = 1 (identical stator windings).
The starting and breakdown torques may be considered proportional to the
number of turns of main winding squared.
The maximum starting torque increases with the turn ratio a.
The flux densities in various parts of the magnetic circuit are inversely
proportional to the number of turns in the main winding, for given source
voltage.
The breakdown torque is almost inversely proportional to the sum Rsm + Rrm
+ Xsm + Xrm.
When changing gradually the number of turns in the main winding, the
rated slip varies with W2m.
The starting torque may be increased, up to a point, in proportion to rotor
resistance Rrm increase.

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

For a given motor, there is a large capacitor CST which could provide
maximum starting torque and another one CSA to provide, again at start,
maximum torque/current. A value between CST and CSA is recommended
for best starting performance. For running conditions a smaller capacitor Ca
is needed. In permanent capacitor IMs a value Ca closer to Ca (Ca > Ca) is
generally used.
The torque varies with the square root of stack length. If stack length
variation ratio K is accompanied by the number of turns variation by
1 K , the torque remains almost unchanged.

27.10 OPTIMIZATION DESIGN ISSUES

Optimization design implies

A machine model for analysis (as the one described in previous
paragraphs);
Single or multiple objective functions and constraints;
A vector of initial independent variables (from a preliminary design)
are required in nonevolutionary optimization methods.
A method of optimization (search of new variable vectors until the best
objective function value is obtained).
Typical single objective functions F are
F1-maximum efficiency without excessive material cost
F2-minimum material (iron, copper, aluminum, capacitor) cost for an
efficiency above a threshold value.
F3-maximum starting torque
F4-minimum global costs (materials plus loss capitalized costs for given
duty cycle over the entire life of the motor).
A combination of the above objectives functions could also be used in the
optimization process.
A typical variable vector X might contain
1. Outer rotor diameter Dr
2. Stator slot depth hst;
3. Stator yoke height hcs;
4. Stator tooth width bts;
5. Stack length L;
6. Airgap length g;
7. Airgap flux density Bg;
8. Rotor slot depth hrt;
9. Rotor tooth width btr;
10. Main winding wire size dm;
11. Auxiliary winding wire size da;
12. Capacitance Ca;
13. Effective turns ratio a.
Variables 10-12 vary in steps while the others are continuous.
Typical constraints are
Starting torque Tes

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

Starting current (Is)S=1

Breakdown torque Teb
Rated power factor cosn
Stator winding temperature rise Tm
Rated slip Sn
Slot fullness kfill
Capacitor voltage VC
Rotor cage maximum temperature Trmax
Chapter 18 of this book presented in brief quite a few optimization methods.
For more information, see Reference 6.
Table 27.1 Design constraints

Constraint Limit Standard F1 F2 F3

Power factor > 0.92 0.96 0.96 0.95 0.94
Main winding < 90 86.2 82.2 86.0 (+) 84.8
temperature [0C]
Max. torque > 1.27 1.29 1.27 (+) 1.28 (+) 1.49
[Nm]
Start torque 0.84 0.878 0.876 0.84 (+) 1.06
[Nm]
Start current [A] < 11.5 10.5 10.9 11.5 (+) 11.5 (+)
Slot fullness < 0.8 0.78 0.8 (+) 0.66 0.8 (+)

k fill
4

(
to d 2 )
Capacitor < 280 263 264 246 261
voltage [V]
Efficiency > 0.826 0.826 0.859 0.828 (+) 0.836

According to most of them the objective function F (x) is augmented with

the SUMT, [7]
m
1
P(X k , rk ) = F(X k ) + rk (27.68)
j=1 G j (X k )

where the penalty factor rk is gradually decreased as the optimization search

counter k increases.
There are many search engines which can change the initial variable vector
towards a global optimum for P (Xk, rk) (Chapter 18).
Hooke-Jeeves modified method is a good success example for rather
moderate computation time efforts.
Reference 8 presents such an optimization design attempt for a 2 pole, 150 W,
220 V, 50 Hz motor with constraints as shown in Table 27.1.
Efficiency (F1) and material cost (F2) evolution during the optimization
process is shown in Figure 27.8 a). [8] Stack length evolution during F1 and F2
optimization process in Figure 27.8 b) shows an increase before decreasing
towards the optimum value.

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

Also, it is interesting to note that criteria F1 (max. efficiency) and F2

(minimum material costs) lead to quite different wire sizes both in the main and
auxiliary windings (Figure 27.9 a, b).
A 3.4% efficiency increase has been obtained in this particular case. This
means about 3.4% less energy input for the same mechanical work.
It has to be noted that reducing the core losses by using better core material
and thermal treatments (and various methods of stray load loss reduction) could
lead to further increases in efficiency.
Though the power/unit of single phase IMs is not large, their number is.
Consequently increases in efficiency of 3% or more have a major impact
both on worlds energy consumption and on the environment (lower
temperature motors, less power from the power plants and thus less pollution).

a.)

b.)
Figure 27.8 Evolution of the optimization process a.) Efficiency and material costs F1 and F2.
b.) Stack length for F1 and F2.

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

a.)

b.)

Figure 27.9 Main (a) and auxiliary (b) winding wire sizes evolution during the optimization process

27.11 SUMMARY

The design of single phase IM tends to be more involved as the magnet

field is rather elliptical, in contrast to being circular for three phase IMs.
The machine utilization factor Cu and the tangential force density (stress) ft
tend to be higher and, respectively, smaller than for three phase IMs. They
also depend on the type of single phase IM split phase, dual-capacitor,
permanent capacitor, or split-phase capacitor.
For new specifications, the tangential force density ft (0.1-1.5)N/cm2 is
recommended to initiate the design process. For designs in given
standardized frames the machine utilisation factor Cu seems more practical
as an initiation constant.
The number of stator Ns and rotor slots Nr selection goes as for three phase
IMs in terms of parasitic torque reduction. The larger the number of slots,
the better in terms of performance.

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

The stator bore to stator outer diameter ratio Di/Do may be chosen as
variable in small intervals, increasing with the number of poles above 0.58
for 2p1 = 2, 0.65 for 2p1 = 4, 0.69 for 2p1 = 6 and 0.72 for 2p1 = 8. These
values correspond to laminations which provide maximum no-load airgap
flux density for given stator m.m.f. in three phase IMs [2].
By choosing the airgap, stator, rotor tooth and core flux densities, the sizing
of stator and rotor slotting becomes straightforward.
The main winding effective number of turns is then calculated by assuming
the e. m. f. Em/Vs 0.95-0.97.
The main and auxiliary windings can be made with identical coils or with
graded turns coils (sinusoidal windings). The effective number of turns (or
the winding factor) for sinusoidal windings is computed by a special
formula. In such a case the geometry of various slots may differ. More so
for the split-phase IMs where the auxiliary winding occupies only 33 % of
stator periphery and is active only during starting.
The rotor cage cross section total area may be chosen for start as 35% to
60% of the area of stator slots.
The turns ratio between auxiliary and main winding a = 1.5-2.0, in general.
It is equal to unity (a = 1) for reversible motors which have identical stator
windings.
The capacitance initial value is chosen for symmetry conditions at an
assigned value of rated slip and efficiency (for rated power).
Once the preliminary sizing is done the resistance and leakage reactance
may be calculated. Then, either by refined analytical methods or by FEM
(with bare rotor and zero auxiliary winding current), the magnetisation
curve m (Imm) or reactance Xmm(Imm), is obtained.
To estimate the starting and steady state running performance (constraints)
such as starting torque and current, rated slip, current, efficiency and power
factor and breakdown torque, revolving field or cross field models are used.
The above completes a preliminary electromagnetic design. A thermal
model is then used to estimate stator and rotor temperatures.
Based on such an analysis model an optimization design process may be
started. A good initial (preliminary) design is useful in most
nonevolutionary optimization methods [5].
The optimization design (chapter 18) is a constrained nonlinear
programming problem. The constraints can be lumped into an augmented
objective function by procedures such as SUMT [7]. Better procedures for
safe global optimization are currently proposed [9].
Increases in efficiencies of a few percent can be obtained by optimization
design. Given the immense number of single phase IMs, despite the small
power/unit, their total power is significant. Consequently, any improvement
of efficiency of more than 1-2% is relevant both in energy costs and
environmental effects.

2002 by CRC Press LLC

Author Ion Boldea, S.A.Nasar....

27.12 REFERENCES

1. C. G. Veinott, Small electric motors, Chapter 6, Handbook of Electric

Machines, Editor S. A. Nasar, McGraw-Hill Company, 1987.
2. G. Madescu, I. Boldea, T. J. E. Miller, The Optimal Lamination Approach
(OLA) to Induction Machine Design Global Optimization, Record of IEEE-
IAS-1996, volume 2, pp. 574-580.
3. E.S. Hamdi, Design of Small Electrical Machines, Wiley & Sons, 1994, pp.
140.
4. J. Stepina, Single Phase Induction Motors, Chapter 5, Springer Verlag,
New York, 1981 (in German).
5. C. B. Rasmoussen and T. J. E. Miller, Revolving Field Polygon Technique
for Performance Prediction of Single Phase Induction Motors, Record of
ICEM-2000, Helsinki.
6. H. Huang, E. F. Fuchs, Z. Zak, Optimization of Single Phase Induction
Motor Design, Part I + II, IEEE Trans. Vol. EC-3, 1988, pp. 349-366.
7. G. I. Hang, S. S. Shapiro, Statistical Models in Engineering, John Wiley &
Sons, 1967.
8. F. Parasiliti, M. Villani, Design Procedure for Single Phase Capacitor
Motors Performance Improvement, Record of ICEM-1994, Paris, France,
pp. 193-197.
9. X. Liu and W. Xu, Global Optimization of Electrical Machines with the
Filled Function Methods, EMPS, vol. 29, 2001.

2002 by CRC Press LLC