Electr Eng

DOI 10.1007/s00202-015-0354-1

ORIGINAL PAPER

Lumped parameter method of permanent magnet synchronous

generator for wind energy
Yasemin Oner1 · İbrahim Şenol1

Received: 16 October 2014 / Accepted: 22 November 2015

© Springer-Verlag Berlin Heidelberg 2015

Abstract The breakthroughs in permanent magnet syn- 1 Introduction

chronous machines through the latest technologies, espe-
cially about machine design, raise the importance of those Technical developments of electrical machine design
machines. Moreover, developments in materials technology increase related to magnetic material developments. Dur-
implement the development of cost effective and profitable ing the last decade, different types of magnetic materials
products on electric machines. Permanent magnet synchro- have been used, including pure iron and its alloys, such as
nous machines are divided into two parts according to the Fe–Ni, Fe–Ni–P, Fe–Nd–B and Fe–Si, and soft and hard fer-
situation of slot/pole combination, which is integer or frac- rites, such as Ni–Zn, Mn–Zn and Ba ferrites [1]. Considering
tional. Permanent magnet synchronous machines are subject the technical literature, it can be seen that for conventional
of interest due to low torque ripple and low cogging torque. machines, laminated steel is used as rotor and stator core
In this paper, the 2D lumped parameter method (LPM) of material. In recent years, with the development of powder
permanent magnet synchronous machine was composed and technology, the use of soft magnetic composite material is
stated, and lumped parameter was verified by finite element increasing. Although this material has low permeability, its
method. When SMC materials are used in electric machines, effect on permanent magnet synchronous machines’ perfor-
flux flows in 3D. Because of this and according to literatures, mance is lower than asynchronous and reluctance motors’
3D method is used to analyze the SMC-used machines. 3D performances. Field effect of permanent magnet machines
finite element method takes a quite long time. Thanks to are provided by magnets, therefore, SMC materials are ideal
2D LPM analysis, which was used in our work; the analysis for permanent magnet synchronous machines.
period become remarkably shorter. In addition, torque and Flux of SMC materials is 3-dimensional (3D). Consider-
back EMF of permanent magnet synchronous machine were ing the technical literature, magnetic equivalent circuit for
calculated through the developed LPM method. Prototype is machines with SMC material is 3D. 3D magnetic equivalent
tested. It is seen by the results that the produced prototype circuit increases the analysis time and process complexity.
and the developed algorithm are matching properly. Magnetic circuit models have been used to study a
wide variety of machine types and configurations includ-
Keywords Lumped parameter method · Permanent ing switched reluctance, induction, surface PM variable
magnet machine · Wind energy · Soft magnetic composite reluctance, brushless dc (BLDC), and flux switching PM
machines. The majority of the reported LPM is used for pre-
diction of basic machine performance characteristics such as
torque production and terminal voltage–current characteris-
tics [2].
B Yasemin Oner LPM was originally proposed and developed in [3–6].
[email protected] A synchronous machine model was presented in [7–10]. In
1 [11,12], mesh-based LPM was discussed. In [13–16], flux
Department of Electrical Engineering, Electric and
Electronics Faculty, Yildiz Technical University, Davutpasa leakage was modeled by different means, which is crucial in
Campus, Istanbul, Turkey the analysis of motors. An accurate yet simple method for

123
 Electr Eng

predicting the flux density distribution and iron losses in lin-

ear PMSM was presented in [17]. However, 2D LPM was
not used for PM machine with SMC material in literature. In
addition, SMC material has been used for axial machine for
PM machine. In this paper, SMC material is used for radial
flux PM machine.
In this study, back EMF and torque of machine that
designed with SMC material in stator is analyzed and com-
pared with prototype results. Using the 2D LPM, analyzing
time is shortened. With the LPM method, PMSM torque
and induced voltage are calculated. It can be seen from the
obtained results that the prototype and the algorithm are com-
Fig. 1 Used SMC material’s untreated state
patible.

2 Soft magnetic material

The main structure of SMC is coherence of iron filings of high

purity. These dust particles get in coherence with organical
material covering. The covered iron filings are compressed
into solid material through die. This kind of materials is nat-
urally powder in form, therefore, generally they are magnetic
isotropic, and this provides the simplicity in design [18,19].
Soft magnetic materials include iron filings covered with iso-
lated film coating. These iron filings are dissociated with
electrical isolation; therefore, this kind of materials has high Fig. 2 Comparison of electrical steel and SMC material’s B–H curves
electrical resistance. They are structured by the compression
of iron filings. The resistivity of SMC depends on material’s
ferromagnetic and mechanical characteristics, size of iron fil- method is faster than finite elements method, its accuracy is
ings, density, insulating material, compression process and lower than the finite elements method. However, it is one of
thermal behavior transformation. SMC material’s magnetic the fastest methods for beginning design and optimization.
and mechanical characteristics not only depend on iron fil- LPM models always treat the permanent magnet (PM) as
ings but also on the proportion of injected oil, resin, and cold one common source and all branches of LPM are coupled
or hot compression process. If the SMC material is not sin- together to become a large LPM network. If flux leakage is
tered, endurance of material becomes lower than laminated also considered, the complexity of these models is increased
SMC or sintered SMC. SMC material’s characteristics can and the analysis process would become extremely compli-
be made suitable for special applications [20]. SMC materi- cated as well [21].
als are isotropic, therefore, they have low conductivity and The main parameters of magnetic equivalent circuit
higher hysteresis loss than conventional steel. SMC mater- method are machine type, geometry, and material charac-
ial’s electrical, thermal and magnetic characteristics depend teristics. In principle, modeling is done with electric circuit
on material’s production process. Soft magnetic material’s analogy. Each part of machine is modeled with reluctance.
B–H curve is shown in Fig. 2. As seen from the figure, satu- Flux way is important. Flux way of permanent magnet syn-
ration point of SMC material is low thus while designing chronous machine with radial flux is radial from rotor to
machine with SMC material, magnetic flux density must stator. Reluctance frequency, although slows down the analy-
be lower than 1.2 T for areas that include SMC material. sis of the machine, gives more accurate results. Principal
Used SMC material’s brand is Hoganas and untreated state magnetic equivalent circuit is shown in Fig. 3. Principal mag-
is shown in Fig. 1. netic circuit model ignores leakage flux, but flux not only
moves from rotor to stator but also between slots, between
magnets and between slot and magnet, and thus the leakage
3 SMC permanent magnet generator LPM analysis flux occurs. Leakage flux between slots is shown in Fig. 4. To
model the permanence network in Fig. 5, Kirchhoff’s current
Magnetic equivalent circuit method is one of the oldest law is applied at each node to create a system of nonlineare-
methods used for studying magnetic analysis. Although this quations.

123
Electr Eng

Fig. 5 Non-linear magnetic equivalent circuit

Fig. 3 Principal magnetic equivalent circuit

(Ui − Ui+1 )G so + (Ui − Ui−1 )G so + (Ui − U1 )G i,1
+ · · · + (Ui − Ul )G i,l = φi (1)
     
U j − U j+1 G r o + U j − U j−1 G r o + U j − U1 G i,1
 
+ · · · + U j − Uh G i,h = φ j (2)
AG u = φ (3)
 
 AG 1 AG 2 
 
AG =   (4)
 AG 3 AG 4 

Fig. 4 Leakage flux between slots

 
 l 
 2G so + G i, j −G so −G so 0 
 
 i=1 
  l 
 −G so 2G so + −G so 
 G i−1, j 0 
 
AG 1 = i=1
 (5)
  l 
 −G so 0 2G so + G i+1, j 0 
 
 i=1 
 l 
 0 −G so 0 2G so + G i−2, j 

i=1

 
In this system, first, permanent magnet flux represents input,  G i, j −G i, j−1 −G i, j+1 −G i, j−2 

while nodal magnetic potentials represent outputs. Second, G −G i−1, j−1 −G i−1, j+1 −G i−1, j−2 
AG 2 =  i−1, j 
 G i+1, j −G i+1, j−1 −G i+1, j+1 −G i+1, j−2 
winding source flux represents input, while nodal magnetic
potentials represent output. Therefore, in this way, armature  G i−2, j −G i−2, j−1 −G i−2, j+1 −G i−2, j−2 
reaction is discussed. As mentioned in Fig: (6)

123
 Electr Eng
 
 G i, j −G i−1, j −G i+1, j −G i−2, j  • Hysteresis and Eddy currents are neglected.

G −G i−1, j−1 −G i+1, j−1 −G i−2, j−1  • Current is regarded as sinusoidal.
AG 3 =  i, j−1 
 G i, j+1 −G i−1, j+1 −G i+1, j+1 −G i−2, j+1  • SMC material is used for stator core.
 G i, j−2 −G i−1, j−2 −G i+1, j−2 −G i−2, j−2 
(7)

 
 l 
 2G r o + G i, j + G m −G r o −G r o 0 
 
 i=1 
 l 
 −G r o 2G r o + G i, j + G m −G r o 
 0 
 
AG 4 = i=1
 (8)
 l 
 −G r o 0 2G r o + G i, j + G m 0 
 
 i=1 
 l 
 0 −G 0 2G + G + G 
 ro ro i, j m
i=1

 T 4 Back EMF analysis

U = Ui Ui−1 Ui+1 Ui−2 U j U j−1 U j+1 U j−2 (9)
φ = [0 0 0 0 Fm G m Fm G m Fm G m Fm G m ]T (10) The cross-section of the machine with nine slots and ten poles
φ = [N i s N i s N i s N i s 0 0 0 0] T
(11) is shown in Fig. 7. In this analytical model, flux direction is
from magnet pole to armature. The generated voltage and
Two-dimensional cross-section of machine is shown in inducted current are in the same direction with flux. Rotor
Fig. 6. The stator has windings concentrated around each position depends on magnet poles. The center of the slot
stator pole. The number of winding turns of each phase is between A-phase winding and B-phase winding builds up
216. The rare-earth magnet “Nd–Fe–B” is buried in the rotor reference axis [22].
iron. Residual magnetic flux density Br and coercive force Ideal magnetic flux density is shown in Fig. 8.
Hc of the permanent magnet are about 1.3 T and 1000 kA/m, The distribution functions of the magnetic flux density
respectively. The following assumptions have been made for depending on rotor angle:
the analysis:
⎧ ⎫
• The used materials have homogeneous properties. ⎨ Bg max for 0 < θ < θ1 and θ4 < θ < θ5 ⎬
Bg (θ ) = −Bgmax for θ2 < θ < θ3
• Ns = Nm ⎩ ⎭
0
• First, no load is applied.
• The stator teeth and permanent magnets are rigid; no (12)
deformation due to radial and tangential force is expe-
rienced by these components.

Fig. 7 Cross-section of the machine whose slot/pole ratio is 9/10

Fig. 6 2D cross-section of the permanent magnet synchronous

machine Fig. 8 Change of magnetic flux density depending on rotor angle

123
Electr Eng

Fig. 11 Inducted voltage waveform of A phase obtained by FEA model

Fig. 9 Changes of air gap and magnetic flux density that were obtained
and magnetic equivalent circuit method
with FEA model and magnetic equivalent circuit method

Winding distribution factor:

1 
sin 2 σ n

kd,n = (16)
z sin nσ
2z
π Nm
γs = (17)
Ns
ε = π − γs (18)
Fig. 10 Magnetic flux density from FEA
Pitch factor:
In Eq. 12,
ken = cos(nε) (19)
θm p θm p θm p kw,n = kd,n ∗ ke,n (20)
θ1 = , θ2 = θ p p − , θ3 = θ p p + ,
2 2 2 Ns √
θm p En = kwn ωm L stk rr o Bgn 2 (21)
θ4 = 2θ p p − , θ5 = 2θs p N ph
2 ea = E 1 sin (ωt) + E 3 sin (3ωt) . . . (22)
The air gap and flux density distribution of the nine slots ten
poles machine is shown in Fig. 9. Inducted voltage waveform of the nine slots ten poles struc-
Figure 10 shows the flux contour profile and FEA distrib- tured machine is shown in Fig. 11
ution of the air gap in the permanent magnet machine.
Change in the magnetic flux density depending on rotor
rotation is shown in Fig. 9. Flux function obtained from flux 5 Torque analysis
density,
Cogging torque and electromagnetic torque of a PM machine

−−−→ −−−→ are calculated according to Maxwell’s stress tensor method
ψ (θ ) = Bg (θ )dS (13)
as,

According to Faraday law, change of flux depending on time L
T = r Bn Bt dl (23)
gives the induced voltage. Thus, μ0
S
dψ (θ )
e (θ ) = − (14) where S is air gap surface constituted by the layers between
dθ
stator and rotor, g is air gap length. According to Virtual work
Important parameters of inducted voltage are magnetic flux method, electromagnetic torque is equal to the derivative of
density and winding factor. Winding factor calculated with the magnetic co-energy Wm’ with respect to rotor angle at
below statements; constant current [19].

z=
Ns
(15) ∂W
GCD(Ns , Nm .N ph ) T = (24)
∂θ

123
 Electr Eng

Fig. 15 Fractional slotted

permanent magnet synchronous
machine’s stator with SMC
material drawings

Fig. 12 Torque that obtained by FEA model and magnetic equivalent

circuit method Fig. 15 and molding of stator, gouging and opened states
of the slots is shown in Fig. 16. CNC opened the slots. Motor
body is made of aluminum and shaft is made of stainless steel
(Figs. 17, 18).
Winding structure is shown in detail in Fig. 19. Wind-
ings of the fractional slotted permanent magnet synchronous
machine with SMC material are concentrated type and each
of them is wound surrounding the tooth (Figs. 20, 21).

7 Test arrangement of permanent magnet

Fig. 13 Torque–speed generator analysis

The block diagram and test arrangement of the system to

measure the back EMF are shown in Figs. 22 and 23. Test
arrangement includes actuating stepper motor and driver,
torque sensor to do torque data acquisition, NI usb analog dig-
ital data converter and fractional slotted permanent magnet
synchronous machine with SMC material to measure back
EMF values.

8 Comparison of FEA, LPM and test result

Fig. 14 Power–speed

In this part, to do back EMF tests, stepper motor rotates at 500

According to Eq. (24); rpm speed and instantaneous values are measured at wind-
ings. These values are compared with the values that were
∂W
∂ W  ∂t ∂t P ea i a + eb i b + ec i c acquired from FEA analysis and lumped parameter method
T = = ∂θ
= = (25) (LPM). Comparison of back EMF values and torque values
∂t ∂θ ∂t
ω ω
is shown in Figs. 24 and 25.
is obtained. Comparison of the peak values of the back EMF is shown
According to the torque results given in Fig. 12, the dif- in Fig. 25 and Table 1
ference between the results of LPM and FEA is less than 1 %
torque–speed and power–speed curve is shown in Figs. 13
and 14. 9 Conclusion

Permanent magnet synchronous generators and conventional

6 Production of permanent magnet synchronous synchronous generators have some differences such as, PM
generator do not require the excitation windings, and to have magnet on
rotor. Permanent magnet synchronous generators do not have
Drawings of fractional slotted permanent magnet synchro- excitation windings because they do not need an additional
nous machine’s stator with SMC material are shown in source that generates excitation current. This characteristic

123
Electr Eng

Fig. 16 Production process of

fractional slotted permanent
magnet synchronous machine’s
stator with SMC material. a
Molding of stator. b Opening of
slots. c Slots

Fig. 17 Shaft of the fractional Fig. 19 Winding layout

slotted permanent magnet fractional slotted permanent
synchronous machine with SMC magnet synchronous machine
material with SMC material

Fig. 20 Fitting the windings

into the stator of fractional
slotted permanent magnet
synchronous machine with SMC
material

Fig. 18 Stator and rotor of the fractional slotted permanent magnet

synchronous machine with SMC material
machine stator; 2D equivalent circuit method is used for
machine analysis, 3D finite elements method (FEM) is used
decreases machine losses and machine size thus permanent for comparison.
magnet synchronous generators have high efficiency and According to analysis results, it can be seen that the dif-
high power density which makes them advantageous for low ference between 2D LPM values and test values is less than
power applications. 5 %. SMC material is low priced and does not require mold-
Prototype machine’s type is surface magnet synchronous ing cost, thus SMC is advantageous when compared to steel.
machine; its slot number for each phase and pole is frac- Because SMC material is easily shapeable; its prime cost is
tional. Soft magnetic composite material (SMC) is used for low.

123
 Electr Eng

Fig. 21 Prototype of fractional

slotted permanent magnet
synchronous machine with SMC
material

Fig. 25 Comparison of torque values with test results

Table 1 Comparison of the

peak values of the back EMF LMP 11.7 V
FEA 11.14 V
Test 11.8 V

To make the output EMF similar to sinusoidal form, ten

Fig. 22 Block diagram of the test arrangement poles and nine slots configuration is preferred. This configu-
ration is preferred according to the biggest one of the greatest
common diviser of pole and slot must be chosen rule in lit-
erature. According to analysis results, the inducted voltage
and torque difference between Matlab software and FEM
software is 1 %

References

1. Shokrollahi H, Janghorban K (2007) Soft magnetic composite

materials (SMCs). J Mater Process Technol 189(1–3):1–12. ISSN
0924-0136
2. Bianchi N, Bolognani S, Grezzani G (2006) Design considera-
tions for fractional-slot winding configurations of synchronous
machines. IEEE Trans Ind Appl 42(4):997–1006
3. Slemon GR (1953) Equivalent circuits for transformers and
machines including non-linear effects. Proc IEEE 100:129–143
4. Fiennes J (1973) New approach to general theory of electrical
Fig. 23 Test arrangement of the fractional slotted permanent magnet machines using magnetic equivalent circuits. Proc IEEE 120:94–
synchronous machine with SMC material 104
5. Ostovic VA (1986) method for evaluation of transient and steady
state performance in saturated squirrel cage induction machines.
IEEE Trans Energy Convers EC–1:190–197
6. Laithwaite ER (1967) Magnetic equivalent circuits for electrical.
Proc IEEE 114:1805–1809
7. Carpenter CJ (1968) Magnetic equivalent circuits. Proc IEEE
115:1503–1511
8. Slemon GR (1990) An equivalent circuit approach to analysis of
synchronous machines with saliency and saturation. IEEE Trans
Energy Convers 5:538–545
9. Shima K, Ide K, Takahashi M, Oka K (2003) Steady-state magnetic
circuit analysis of salient-pole synchronous machines considering
cross-magnetization. IEEE Trans Energy Convers 18:213–218
10. Qin R, Rahman MA (2003) Magnetic equivalent circuit of PM
hysteresis synchronous motor. IEEE Trans Magn 39:2998–3000
11. Delale A, Albert L, Gerbaud L, Wurtz F (2004) Automatic gen-
eration of sizing models for the optimization of electromagnetic
Fig. 24 Comparison of back EMF values with test results devices using reluctance networks. IEEE Trans Magn 40:830–833

123
Electr Eng

12. du Peloux B, Gerbaud L, Wurtz F, Leconte V, Dorschner F (2006) 18. Zhu J, ve Guo Y (2004) Study with magnetic property measurement
Automatic generation of sizing static models based on reluctance of soft magnetic composite material and its application in electrical
networks for the optimization of electromagnetic devices. IEEE machines. IEEE industrials application conference, 3–7 Oct. 2004
Trans Magn 42:715–718 19. Maeda T, Toyoda H, Igarashi N, Hirose K, Mimura K, Nishioka T,
13. Hwang CC, Cho YH (2001) Effects of flux leakage on magnetic Ve IA (2005) Development of super low iron-loss P/M soft mag-
fields of interior permanent magnet synchronous motors. IEEE netic material. SEI Tech Rev 60:3–9
Trans Magn 37:3021–3025 20. Hultman LO, ve Jack AG (2003) Soft magnetic composites—
14. Tsai WB, Chang TY (2001) Analysis of flux leakage in a brush- materials and applications. IEEE international electric machines
less permanent-magnet motor with embedded magnets. IEEE Trans and drives conference, IEMDC’03, pp 516–522, Madison, USA,
Magn 35:543–547 1–4 June 2003
15. Qu R, Lipo TA (2004) Analysis and modeling of air-gap and zigzag 21. Qian J, Chen X, Chen H, Zeng L, Li X (2013) Magnetic field analy-
flux leakagees in a surface-mounted permanent magnet machine. sis of Lorentz motors using a novel segmented magnetic equivalent
IEEE Trans Ind Appl 40:121–127 circuit method. Sensors 13:1664–1678
16. Zheng L, Wu TX, Sundaram KB, Vaidya J, Zhao L, Acharya D, 22. VuXuan H, Lahaye D, Ani SO, Polinder H, Ferreira JA (2011)
Ham CH, Kapat J, Chow L (2005) Analysis and test of a high-speed Effect of design parameters on electromagnetic torque of PM
axial flux permanent magnet synchronous motor. In: Proceeding of machines with concentrated windings using nonlinear dynamic
international electric machines and drives conference (IEMDC), FEM. IEEE international electric machines & drives conference
San Antonio, TX, USA, pp 119–124, 15–18 May 2005 (IEMDC’11), pp 383–388, Niagara Falls, ON, 15–18 May 2011
17. Vaez-Zadeh S, Isfahani AH (2007) Enhanced modeling of linear
permanent-magnet synchronous motors. IEEE Trans Magn 43:33–
39

123