See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/261320497

A software for the evaluation of winding factor harmonic distribution in high

efficiency electrical motors and generators

Conference Paper · March 2013

DOI: 10.1109/EVER.2013.6521571

CITATIONS READS

20 1,648

3 authors, including:

Antonino Oscar Di Tommaso Rosario Miceli

Università degli Studi di Palermo Università degli Studi di Palermo
173 PUBLICATIONS   1,898 CITATIONS    381 PUBLICATIONS   5,020 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Innovative Lead-Acid Battery View project

diagnosis of failures in squirrel-cage induction motors View project

All content following this page was uploaded by Antonino Oscar Di Tommaso on 29 March 2018.

The user has requested enhancement of the downloaded file.

EVER 2013 1

A Software for the Evaluation of Winding Factor

Harmonic Distribution in High Efficiency Electrical
Motors and Generators
A. O. Di Tommaso, F. Genduso, R. Miceli, Member, IEEE

Abstract—In this paper a software developed for the determi- Nc number of coil sides per phase;
nation of winding connections, the calculation of winding factors νx maximum harmonic order;
and the evaluation of their harmonic distribution of three and bs is the slot opening expressed as a fraction of the slot
two phase windings is presented by using as input data simply
the number of slots, pole pairs and phases. The software is pitch;
implemented as a function in MATLAB environment. By means Zsk number of skewing sections;
of some examples, covering the most relevant winding types, σ0 differential leakage factor.
the capabilities of this software will be shown. Particularly,
integer and fractional single and double layer winding can be I. I NTRODUCTION
treated. Some examples are presented and briefly discussed.
ONG-TERM operations of electrical motors and gen-
The connections, the winding factor harmonic distribution, the
electro-motive force (e.m.f) THD calculations results are shown
together with an example on how is possible to determine the
L erators are expected to have high energetic efficiency
as one of the most important goals to reach [1], [2], [3], [4].
e.m.f. induced in a winding from the air gap flux distribution In this paper a software strictly developed for the purpose of
and to estimate the differential leakage inductance.
winding factor harmonic distribution evaluation is presented.
Index Terms—winding, winding factor, winding connections, As known winding factors, and consequently the harmonic
harmonics. content of the electrical quantities involving such machines,
are determined by the type of winding employed in their con-
L IST OF S YMBOLS struction. Particularly, the harmonic distortion of voltages, cur-
N number of slots; rents, flux densities, etc., in electrical machines -which depend
p number of pole pairs; on the type of winding- affect in a significant manner both the
r star pitch; dynamic and steady state performances and their efficiency,
t greatest common divider (GCD) between N and p by introducing, for example, asynchronous, synchronous and
t=GCD(N,t); torsional torques, axial and radial forces producing vibrations
ν harmonic order; (and noise), and, last but not least, producing an increase in
m number of phases; copper and iron losses. It is, therefore, important to have a
q number of slots per pole and per phase; clear overview on the winding factor harmonic distribution
kdν distribution factor at the ν-th harmonic order; characterizing an electrical machines indeed just in the design
kwν group factor at the ν-th harmonic order; phase, i.e. when the right type of winding is to be chosen
kslν slot opening factor at the ν-th harmonic order; [5], [6]. In this sense the proposed software is conceived as
kskν skewing factor at the ν-th harmonic order; an aid to simplify and accelerate the design procedures in
kcν winding connection factor at the ν-th harmonic order; electrical machine windings. The software is also capable to
Es e.m.f. induced in a group of slot conductors; represent numerically the winding connections, in the case of
Ci,j phase component of Es ; single and double layer structures, and to perform some useful
Si,j quadrature component of Es ; calculation about phase leakage inductances. Some examples
will be presented in this paper to show and to explain the
Antonino Oscar Di Tommaso is with the Department of Energy, Information capabilities of the software, dealing both with integer and
engineering and Mathematical models (DEIM) - University of Palermo,
Palermo, Italy, e-mail: [email protected].
fractional winding types.
Fabio Genduso is with the Department of Energy, Information engineering In section II the mathematical expressions of the winding
and Mathematical models (DEIM) - University of Palermo, Palermo, Italy, factors are presented and the software procedure of the wind-
e-mail: [email protected].
Rosario Miceli is with the Department of Energy, Information engineering
ing connection construction is briefly outlined. In section III
and Mathematical models (DEIM) - University of Palermo, Palermo, Italy, some examples are presented and discussed.
e-mail: [email protected].
This work was financially supported by MIUR - Ministero dell’Istruzione II. W INDING FACTORS AND P ROCEDURE FOR W INDING
dell’Università e della Ricerca (Italian Ministry of Education, University and
Research) and by SDESLab (Sustainable Development and Energy Saving C ONNECTIONS
Laboratory) of the University of Palermo. In this section the mathematical aids used to determine the
winding connections, the method of winding factors calcula-
EVER 2013 2

Table I
tion and the determination of their harmonic distribution are C ONNECTION SCHEME OF A FRACTIONAL SLOT WINDING WITH N=24,
presented. P =5, R =5

As known from the theory of electrical machines [7], [8], a -c b -a c -b

[9], [10] each conductor or group of conductors in a slot is 1 6 11 16 21 2 7 12 17 22 3 8 13 18 23 4 9 14 19 24 5 10 15 20

characterized by a particular phasor of the induced e.m.f.. All

Table II
these phasors can be represented in a polar diagram called P ER - PHASE CONNECTION SCHEME OF A FRACTIONAL SLOT WINDING
slot star or e.m.f. star. If we define t as the greatest common WITH N=24, P =5, R =5
divider between the slot number N and the pole pair number
a b c
p, t slot e.m.f. phasors will have the same phase and a star of
+ 1 6 11 16 17 22 3 8 9 14 19 24
- 13 18 23 4 5 10 15 20 21 2 7 12
N
S= (1)
t
phasors with different phases will be obtained. The phase In the case of a group of S coils, all with a diametrical
angle between e.m.f. of two adjacent phasors within the slot coil pitch, the distribution factor (group factor) of the ν-th
star relative to the fundamental wave is, therefore, harmonic is [11]
t sin νS πp π



α0 = 360° (2) N
sin ν 2m
N kdν = = (6)
S sin ν πp
 
π
N q sin ν 2mq
while the phase angle between between e.m.f. of adjacent
slots is where m is the number of phases of the winding and q the
number of slots per pole and per phase
p p
α= 360° = α0 . (3)
N t N
q= . (7)
For all other harmonic orders it is sufficient to multiply both 2pm
α or α0 by the harmonic order coefficient ν. If the coil pitch is shortened then the resulting winding
With the help of (3) it is possible to determine a cor- factor kwν is obtained by multiplying kdν by the shortening
respondence between the slot and star phasors numbering. pitch factor defined as
Therefore, if the slots are numbered in the order Z=1, 2, 3, ...
the phasors of the star will be numbered by jumping over p/t-1  π 
2pη π

star phasors, meaning that adjacent slot e.m.f. have a distance kpv = sin ν · sin ν f or ν odd (8)
2 N 2
of p/t steps in the star.
An important quantity for the determination of a e.m.f. star  π  
is the star pitch defined as 2pη π
kpv = −cos ν · sin ν f or ν even (9)
2 N 2
g Nt + 1
r= p = integer (4) kwν = kpν · kdν . (10)
t

with g=0, 1, 2,... where η is the coil pitch expressed in terms of slot numbers.
By applying (4) it is possible to evaluate the slot e.m.f. star Normally the shortening pitch factor can be calculated apply-
and from this, with the help of a procedure similar to that used ing (8) when q is an integer, but, when dealing with fractional
to realize the Tingley’s scheme [8], the winding connection slot windings, in which the spatial distribution of the magnetic
order. The last is expressed as a matrix of numbers with each flux density has no half period symmetry, also the even order
row corresponding to the phase a, b and c. Each element of a harmonics must be considered [7], [9], [10]. This aspect can
row is referred to a slot number and the number sequence of often yield to tedious calculations.
each row represents how the coil sides (located within each In case of complex winding configurations, as for example
slot) are to be connected in series to form the phase winding with fractional slot windings where the e.m.f.s do not show
section symmetries (see Fig. 1 a) and b)), the winding factors will
  be made up by more complex formulas [9], [10], [7]. In
a1 a2 a3 ... aN/m order to gain generality and to simplify the algorithm, the
Connh,i =  b1 b2 b3 ... bN/m  . (5) winding factor definition is considered as the ratio between
c1 c2 c3 ... cN/m the geometrical and the algebraic sum of the e.m.f. induced
in a winding section:
Tables I, II and III summarize the procedure of building
the connection matrix Connh,i in the case of a single layer
fractional winding. Table III
C ONNECTION MATRIX FOR N=24, P =5, R =5
Once the winding connections are defined it is possible to
determine all the winding factors as it will be explained below. a 1 6 11 16 -13 -18 -23 -4
Traditionally, in literature [7], [8], [9], [10] many formulas b 17 22 3 8 -5 -10 -15 -20
have been proposed for the calculation of these factor. c 9 14 19 24 -21 -2 -7 -12
EVER 2013 3

slot pitch

Figure 1. Difference between a) symmetrical and b) unsymmetrical phase

e.m.f. systems.

Figure 2. To clear the meaning of skewing sections. Here the skewing is

Pνx PN q made by 4 sections (Zsk = 4).
2 + S2
Ci,j
ν=1 j=1 i,j
kwν = (11)
Nc · Es
with ν = 1, 2, 3, ..., νx , which is the general formula
where for winding factor calculation implemented in the MATLAB
Connh,j program here presented.
Ci,j = k · Es · cos(i · 2π · ) (12) Equation (18) can also be used to compute the so called
N
differential leakage factor as [9], [12], [8], [7]
Connh,j
Si,j = k · Es · sin(i · 2π · ) (13) p−1  2 νx  2
N X kwν X kwν
σ0 = + . (19)
and ν=1
νkwp ν=p+1
νkwp
 The developed MATLAB software mainly consist of three
+1 if Connh,j > 0
k= . (14) subroutines: the first one accepts as inputs the N, p and m
−1 if Connh,j < 0
values and gives as output the connection matrix Connν,j ; the
In (12), (13) and (14) h can be set equal a or b or c second one accepts as inputs a single rows of the connection
indifferently. It is to be noted that (11) is valid for the matrix and p, and gives as an output the σ0 differential leakage
determination of the resultant winding factor also in the case factor; the third one accept as inputs a row of the connection
of coil pitch shortening; as a matter of fact, information about matrix, the number of pole pairs and the radial component
pitch shortening is already contained in the matrix Connh,j . of the flux density along the air gap, giving as outputs the
Only if one will take also the effects of slot openings and of induced e.m.f. in the winding and the THD factor of the same
pole or slot skewing into account, further correction factors e.m.f.. The next section illustrates some examples, referred
must be introduced, i.e., respectively: to 4 particular cases of stator windings, whose results were
π

obtained with the developed software. The first three examples
sin ν · bs · N involve the first and the second subroutines, therefore the
kslν = 2 · (15)
ν · bs · 2π
N results are independent from the machine type, while the
and fourth example is related to the case of an particular IPM
machine.
π


sin ν · N
kskν =   (16)
Zsk · sin ν · Zskπ N III. E XAMPLES
In this section some examples used to test the program
where Zsk is the number of skewing sections in the case are presented and discussed. Particularly, first the connection
of a discrete skewing (see Fig. 2). Furthermore, taking the matrix will be shown and then the winding factor spectral
winding phase connections into account (concatenated e.m.f.) distribution will be analyzed.
the following factor is introduced:
 π · p Example 1. Three phase double layer winding with N=27
kcν = sin ν · . (17) slots, p=3, q=1+1/2.
m By introducing these data as input for the MATLAB func-
In this case (11) becomes tion the following 3x6 connection matrix is obtained.
 
Pνx  PN q 2  1 −5 2 −6 −6 1
kcν · kslν · kskν · 2
Ci,j + Si,j
ν=1 j=1  4 −8 5 −9 −9 4 
kwν = (18) 7 −2 8 −3 −3 7
Nc · Es
EVER 2013 4

1.8
odd
The matrix has a reduced dimension because this is a even
base winding type, i.e. the complete winding is obtained by 1.6

repeating these connections p=3 times.

1.4

1
odd 1.2
even
0.9

Amplitude
1
0.8
0.8
0.7

0.6 0.6
Amplitude

0.5 0.4

0.4 0.2

0.3
0
0 5 10 15 20 25 30 35 40 45 50
0.2 ν ′=p⋅ν

0.1 Figure 5. Resultant winding factor spectrum for N=27 and p=3 ( kres =
kcν · kskν · kslν · kwν ).
0
0 5 10 15 20 25 30 35 40 45 50
ν ′=p⋅ν

Figure 3. Winding factor spectrum for N=27 and p=3.

In Fig. 3, 4 and 5 the spectral distributions of the winding

factors are represented in the order: kwν , kcν · kwν · kslν and
kcν · kslν · kskν · kwν , with bs equal to half the slot width and
Z=4 skewing sections.
1.8
odd
even Figure 6. Three phase double layer winding of the IPMSM for N=27 and
1.6 p=3.

1.4

1.2
In Fig. 7, 8 and 9 the spectral distributions of the winding
factors are represented in the order: kwν , kwν · kslν and kslν ·
Amplitude

1 kskν ·kwν , with bs equal to half the slot width and Z=4 skewing
sections.
0.8

0.6 1
fractional
odd
0.9 fractional
0.4
even
0.8
0.2
0.7
0
0 5 10 15 20 25 30 35 40 45 50
ν ′=p⋅ν 0.6
Amplitude

0.5
Figure 4. Resultant winding factor spectrum for N=27 and p=3 (kres =
kcν · kslν · kwν ).
0.4

The winding connection scheme is represented in Fig. 6. 0.3

0.2
Example 2. Three phase double layer winding with N=27
0.1
slots, p=2, q=2+1/4.
By introducing these data as input for the MATLAB func- 0
0 20 40 60 80 100
tion the connection matrix is obtained. For simplicity only the ν ′=p⋅ν

first row of this matrix is shown, which is used to determine

Figure 7. Winding factor spectrum for N=27 and p=2.
the winding factor spectral distribution by means of (11) or
(18).
Example 3. Three phase double layer winding with N=27
1 −7 15 −21 2 −8 16 −22 3 ... slots, p=4, q=1+1/8.
[ ]
−9 −8 12 −22 1 −9 14 −23 2 By introducing these data as input for the MATLAB func-
EVER 2013 5

1.8 1.6
fractional fractional
odd fractional
1.6 fractional fractional
1.4
even odd
fractional
1.4 fractional
1.2
fractional
even
1.2
1
Amplitude

Amplitude
1
0.8
0.8

0.6
0.6

0.4
0.4

0.2 0.2

0 0
0 20 40 60 80 100 0 50 100 150 200
ν ′=p⋅ν ν ′=p⋅ν

Figure 8. Resultant winding factor spectrum for N=27 and p=2 (kres = Figure 10. Resultant winding factor spectrum for N=27 and p=4 ( kres =
kcν · kslν · kwν ). kcν · kskν · kslν · kwν ).

1.8
fractional
odd Table IV
1.6 fractional R ATED VALUES OF THE IPMSM.
even

1.4 Item [unit] Specification

Rated speed [rpm] 4000
1.2 Rated current [A] 3.6
Rated phase voltage [V] 77
Rated torque [Nm] 1.8
Amplitude

1
Number of poles 6
0.8 Number of conductors per slot 20

0.6
Table V
0.4 S TRUCTURE PARAMETERS OF IPMSM. A LL MEASUREMENTS ARE
EXPRESSED IN MM .
0.2
Item Specification
0 External stator diameter 81
0 20 40 60 80 100
ν ′=p⋅ν Internal stator diameter 49.6
External rotor diameter 48
Figure 9. Resultant winding factor spectrum for N=27 and p=2 ( kres = Internal rotor diameter 18.46
kcν · kskν · kslν · kwν ). Axial rotor length 59
PM width 13.45
PM thickness 3
Air-gap thickness 0.8
tion the connection matrix is obtained. For simplicity only the Slot depth 9.2
first row of this matrix is shown, which is used to determine Slot mean width 4
Stator screw hole diameter 2
the winding factor spectral distribution by means of (11) or Rotor screw hole diameter 5
(18). Shaft flux barrier thickness 1

1 −4 8 −11 15 −18 22 −25 2 ... Fig. 12 represents the spatial distribution of the radial
[ ]
−5 −18 21 −25 1 −5 8 −12 15 component of flux density along the air-gap, which is produced
For brevity in this example only the spectral distributions by the interaction between the permanent magnet field and the
of the resultant winding factor is represented in Fig. 10 (i.e. armature reaction field produced at full load. Despite the heavy
kslν · kskν · kwν , with bs equal to half the slot width and Z=4 distortion produced by the slot harmonics, by the armature
skewing sections). reaction and the saturation of iron, the induced e.m.f. shape
maintains quite sinusoidal, due to the effect of winding factors,
Example 4. Regarding Example 1 here an other example is as shown in Fig. 13. The quality of e.m.f. is proved by THD%
reported in order to show how the air-gap e.m.f. of an electrical = 0.85, while, applying (19), the differential leakage factor is
machine can be estimated by knowing machine parameters and σ0 = 0.0036.
its spatial distribution of the air-gap flux density. The machine
is a IPMSM whose main parameters are listed in Table IV and IV. C ONCLUSIONS
Table V while a cross section of the machine is depicted in In this paper a software written in MATLAB language was
Fig. 11. synthetically presented and tested by means of 4 examples.
EVER 2013 6

250

200

150

100

50

e.m.f. [V]
0

−50

−100

−150

−200

−250
0 100 200 300 400 500 600 700 800 900 1000
Electrical angle [degrees]

Figure 13. IPMSM induced e.m.f. at an angular speed of 4000 rpm.

ACKNOWLEDGMENT
Figure 11. Cross section of the IPMSM.
The Authors wish to thank all the collaborators and the staff
of SDESLab (Sustainable Development and Energy Savings
1
Laboratory)-University of Palermo- and Emergenzamatic S.r.l..
0.8

0.6 R EFERENCES
[1] C. Cavallaro, A. O. Di Tommaso, R. Miceli, A. Raciti, G. Ricco
0.4
Galluzzo, and M. Trapanese, Efficiency Improvement of Inverter-Fed
0.2
Permanent Magnet Synchronous Motors, pp. 490–497. Energy Efficiency
in Motor Driven Systems, Springer Verlag, 1 ed., 2003.
Br [T]

0 [2] C. Cavallaro, A. O. Di Tommaso, R. Miceli, A. Raciti, G. Ricco

Galluzzo, and M. Trapanese, “Analysis a DSP Implementation and
−0.2 Experimental Validation of a Loss Minimization Algorithm Applied to
Permanent Magnet Synchronous Motor Drives,” in The 29th Annual
−0.4 Conference of the IEEE Industrial Electronics Society, 2003. IECON
’03., pp. 312–317, IEEE, IEEE, 2003.
−0.6 [3] C. Cavallaro, A. O. Di Tommaso, R. Miceli, A. Raciti, G. Ricco
Galluzzo, and M. Trapanese, “Efficiency Enhancement of Permanent
−0.8
Magnet Synchronous Motor Drives by on-Line Loss Minimization
Approaches,” IEEE Transactions on Industrial Electronics, vol. Volume
−1
0 5 10 15 20 25 52, pp. 1153–1160, Aug. 2005.
Slot number
[4] A. O. D. Tommaso, R. Miceli, G. R. Galluzzo, and M. Trapanese,
“Efficiency Control for Permanent Magnet Synchronous Generators,”
Figure 12. IPMSM radial component of the air-gap flux density. in IEEE International Conference on Industrial Technology, ICIT 2006.,
pp. 2079–2084, IEEE, IEEE, December 2006.
[5] A. O. Di Tommaso, G. Ricco Galluzzo, and R. Miceli, “Improvement
of IPMSM Performance through a Mixed Radial-tangential Rotor Struc-
The software can be used as a winding design tool as well as ture,” in IEEE International Symposium on Industrial Electronics (ISIE),
(Bari, Italy), pp. 1327–1332, IEEE, 2010.
a winding analysis tool, being it capable to: [6] Y. Gritli, A. O. Di Tommaso, F. Filippetti, R. Miceli, C. Rossi, and
1) determine schematically the connection order of the coil A. Chatti, “Investigation of motor current signature and vibration anal-
ysis for diagnosing rotor broken bars in double cage induction motors,”
sides located in machine slots; in International Symposium on Power Electronics, Electrical Drives,
2) determine the spectral distribution of winding factors; Automation and Motion (SPEEDAM), (Sorrento, Italy), pp. 1360–1365,
3) to estimate the shape of in a winding induced e.m.f., 20-22 June 2012.
[7] R. Richter, Lehrbuch der Wicklungen elektrischer Maschinen. Wis-
from air-gap flux density; senschaftliche Bücherei, Verlag und Druck G. Braun, 1952.
4) to estimate the THD and the differential leakage factor [8] H. Sequenz, Die Wicklungen elektrischer Maschinen, vol. 1. Springer
of both integer and fractional slot windings. Verlag, 1950.
[9] G. Müller, K. Vogt, and B. Ponick, Berechnung elektrischer Maschinen.
For the determination of the winding connections only few Elektrische Maschinen, Wiley-VCH Verlag GmbH & Co KGaA, 6 ed.,
2008.
elements are required as: the knowledge of the numbers of [10] G. Müller and B. Ponick, Theorie elektrischer Maschinen. Elektrische
slots, of pole pairs, the number of phases and the type of Maschinen, Wiley-VCH Verlag GmbH & Co KGaA, 6 ed., 2009.
winding (single or double layer). On the other hand for the [11] R. Richter, Allgemeine Berechunungselemente - Gleichstrommaschinen,
vol. 1 of Elektrische Maschinen. Basel, Stuttgart: Birkhäuser, 3 ed.,
e.m.f. induced in a winding the knowledge of more data is 1967.
necessary, such as the angular speed, the number of conductors [12] R. Richter, Die Induktionsmaschinen, vol. 4 of Elektrische Maschinen.
in a slot, the spatial distribution of flux density, etc.. Basel, Stuttgart: Birkhäuser, 2 ed., 1954.

View publication stats