Sie sind auf Seite 1von 6

Australasian Universities Power Engineering Conference (AUPEC 2004)

26-29 September 2004, Brisbane, Australia

PERFORMANCE ANALYSIS OF A
PERMANENT MAGNET LINEAR SYNCHRONOUS MOTOR

G. Gentile*, A. Ometto** and N. Rotondale**

* Department of Electrical Engineering


University of Naples “Federico II”, Italy

** Department of Electrical Engineering


University of L'Aquila, Italy

Abstract

A modified one-dimensional mathematical model based on the space-vectors theory is used to analyse
the performance of permanent magnet linear synchronous motors. The end-effects are taken into
account by assuming that the length of the iron core is infinite and the length of the primary winding is
finite. A resulting differential equation set is obtained which allows both steady state and transient
operations to be analysed. The harmonic analysis of the currents and the average value of the thrust are
carried out. The mathematical model is used to analyse the performance of four low power motors.

1. INTRODUCTION computed. The model is used to compute the average


value Fx of the thrust of four low power motors,
It is well know that end-effects influence the
emphasizing the end-effects.
performance of linear machines. In linear induction
motors with sinusoidal supply voltage, the magnetic 2. MATHEMATICAL MODEL
asymmetry produces a backward magnetic field that
2.1 Assumptions and PM modelling
causes unbalanced, but still sinusoidal, primary currents
and a reduction of the thrust. In permanent magnet The mathematical model of the PMLSM has been
linear synchronous motors (PMLSM) there are also time developed by considering a secondary sheet too (Fig. 1)
harmonic components in primary currents induced by and it is based on the following assumptions:
the permanent magnets. Many papers on LSMs can be • iron reluctance is negligible;
found in technical literature; they mainly focus magnetic • the length of the iron core is infinite, so that the
field and current distributions applying mathematical length of the magnetic lines can be assumed
models with different degrees of complexity [1,2,3]. constant. Primary m.m.f. is acting on a finite portion
of the iron core;
Aim of this paper is the investigation of the motor • the primary winding is described by an impulsive
performance, thrust characteristic, by taking into current density distributed over the air-gap
account the end-effects. A modified one-dimensional separating surface;
mathematical model based on the space-vectors theory • end-effects are taken into account by assuming the
is used. By this approach, it is possible to study easily period of the Fourier expansion 2τ m greater than the
the interactions between the different space and time
harmonic fields by starting from the geometric structure motor length: τ m = npτ p (n even).
of the motor. The end-effects are taken into account by By referring to the space-vector theory, it is possible to
assuming that the length of the iron core is infinite and separate space and time variables. The generic
the length of the primary winding is finite. To represent electromagnetic quantity Y(x,t) (B, A or J) can be
this winding, each conductor of every single slot is expressed as:
separately considered. A resulting differential equation ⎧∞ π ⎫
− jν x
⎪ ⎪
set is obtained which allows both steady state and
transient operations to be analysed. If steady-state ∑
A( x, t ) = Re⎨ Aν (t ) ⋅ e τ m ⎬
⎪⎩ν =1 ⎪⎭
(1)
operations are considered, an infinite system of complex
algebraic equations, referred to each frequency, can be where Aν (t ) is the space vector of order ν in the
deduced and the currents harmonic spectrum can be armature frame, x (Fig. 1).
2τ m
y 2 pτ p
x
z
−τ m y′′ τm
δ N S N S N S N S d
x′′
z′′
−τ m τm
x
bp

N S hp b p = βτ p

τp
Fig.1 – Scheme of a PMLSM and reference frames.

The position of the inductor frame (x′′) in the armature The induced e.m.f ek is the sum of the induced voltage
frame is x and a permanent magnet north pole begin in in each conductor:
x′′ = 0 (Fig. 1). The relative magnetic permeability of the
d ⎡ ∞ ∨ ⎤
magnet µ r, rec is assumed to be equal 1 and the ek (t ) = Re ⎢ z k qk
dt ⎢⎣ ν =1
∑ξ k ,ν Φν (t )⎥
⎥⎦
(7)
permanent magnets are modelled by means of a constant
current Ie located at the field poles edges [1]:
where Φν (t ) represents the space vector of the yoke
Br
Ie = hp (2) flux given by:
µ0 ⋅ µ r , rec
Φν (t ) =
where Br is the remanent magnetic flux density.
Such a current gives a linear current density that is ⎧ ⎡ 3 π
jν x ( t ) ⎤ ⎫
H ⎪1 ⎢ z k qk ξ&k ,ν ik (t ) + 2npξ&eν I e ⋅ e τ m ⎥ + dJν (t )⎪⎬
considered impulsive: ⎨
ν 2 ⎪τ m

⎢ k =1 ⎥ ⎪
π ⎩ ⎣ ⎦ ⎭
⎧∞ − jν x ′′ ⎫
⎪ ⎪

Ae ( x′′) = Re⎨ Ae′′ν ⋅ e τ m ⎬
⎪⎩ν =1 ⎪⎭
(3)
2
(8)
⎛τ ⎞
with H = L⎜ m ⎟ µ 0 λ0 .
2np & ⎝π ⎠
Ae′′ν = ξ eν I e (4)
τm
Jν (t ) denotes the space-vector of the inductor sheet
2 np −1 ⎧ π
R⎡
π ⎤⎫
⎪ jν jν β
⎪ current density and λ0 is the inverse of the air-gap.
∑ ⎨⎪(− 1)
1 R +ν
ξ&eν = ⋅ e np ⎢1 − e np ⎥ ⎬ (5)
2np ⎢ ⎥⎪
R =0
⎩ ⎣ ⎦⎭ For the inductor sheet the voltage equation is given by:
where Ae′′ν represents its space vector in the inductor d Φν (t ) L π
+ Jν (t ) − j vν Φν (t ) = 0 (9)
frame and ξ&eν is the complex winding factor of the dt σ τm
inductor winding equivalent to the permanent magnet
where v is the motor speed.
poles.
Finally, the overall electromagnetic thrust is computed
2.2 Resulting equations
by considering the effect of each space harmonic:
The armature voltage equation is: ∞
⎧⎡ ∨ ∨ ⎤ ⎫
⎛ d⎞
vk (t ) = ⎜ rk + lσk ⎟ik (t ) + ek (t ) k = 1,2,3 (6)
Fx (t ) = π ∑ Im⎨⎩⎢⎣ A ν (t ) + d ⋅ J ν (t )⎥⎦ Φν (t )⎬⎭
ν =1
e, (10)
⎝ dt ⎠
2.3 Steady-state operations angle ρ, are shown in Fig.2a and Fig. 2b respectively.
The thrust-angle has been defined as the angle between
With sinusoidal supply voltage at angular frequency
the supply voltage and the fundamental time harmonic
ω = 2πf, at constant synchronous speed vs = 2fτp the
of the e.m.f induced by the permanent magnets. The
inductor position is:
same quantities are reported in Fig. 3, 4 and 5 for motor
x (t ) = x0 + vs t [m] M2, M3 and M4 respectively. The dashed lines have
been obtained neglecting the end-effects and the dotted
or sector of the curves refer to operating conditions with
π π currents values greater than the rated one.
γ (t ) = x (t ) = x0 + ωt = γ 0 + ωt electrical radians.
τp τp
From Fig. 2a and 4a, it comes out that the end-effects
The armature voltages induced by the permanent cause a significant derating of the motors in case of
magnets have angular frequencies ωm = m⋅ω/(n⋅p) (for machines without inductor sheet (-21.7% for M1 and
m = n⋅p, ωm = ω): -21.6% for M3, if rated currents are considered). On the
contrary, the derating is limited when motors with the
⎡ j ωt ⎤
m
inductor sheet are considered (Fig. 3a and 5a), proving
d ⎢ & np ⎥
e0 k , m (t ) = Re M k , m I e ⋅ e (11) that the inductor sheet reduces the end-effects. That is
dt ⎢ ⎥
⎣ ⎦ also proved by the comparison between Fig. 2b and 3b
and Fig. 4b and 5b from which it is evident that the
with input current unbalancing is reduced by the inductor
m sheet. It must be also underlined that the inductor sheet
∨ j γ0
2 H &
M& k ,m = z k qk ξ k ,m ξ em ⋅ e np (11) gives a further positive thrust Fx,s (Fig. 3a and 5a).
τp m 2

For m ≠ n⋅p, the armature winding is short-circuited and


there are currents at ωm as well as the currents at the Table I – Motors parameters
supply angular frequency ω. Motor M1 M2 M3 M4
p 3
3 NUMERICAL RESULTS
Q 36
The proposed model has been used to analyse the
influence of the inductor sheet and of the PM poles zk 54
width on the motor performance. To this aim, four delta- Pr [W] 900
connected, single layer winding motors (M1, M2, M3 Vr [V] 65 72 68 74
and M4) have been designed (Tab. I). They have the
same rated values of power, speed and current. Ir [A] 12
Moreover, all the motors have essentially the same fr [Hz] 30
armature; only the armature width is different in order to vr -1
[m⋅s ] 3
obtain the same power with the same direct and
quadrature axis currents (Iq,1 = Ir and Id,1 = 0) [4]. τp 51
Br [T] 1.2
Two values of the PM pole width have been considered: bp [mm] 36 31
36 mm for motors M1 and M2 and 31 mm for motors
M3 and M4. As regards the inductor sheet, motors M2 hp [mm] 1.5
and M4 have a 2 mm thickness inductor sheet whereas δ [mm] 2.1 4.1 2.1 4.1
motors M1 and M3 do not have any inductor sheet at all. d [mm] − 2 − 2
The performance of the four motors has been also
computed neglecting the end-effects, ideal machines, in lσ,k [mH] 6.3 11.5 6.9 12.4
order to understand their influence. L [mm] 33 65.5 36.5 71.5
rk [Ω] 1.52 1.99 1.57 2.08
With reference to motor M1, the thrust and the rms
values of the fundamental time harmonic of the input σ -1
[S⋅m ] 30.6⋅10 6

currents, i.e. the line currents, as function of the thrust-


500 14
Fx [N] Pa (31.5, 8.64) I b,1
Pb (31.5, 12.04) Pb I1

input current [A]


400 12 Pc (31.5, 9.47)
P
(32.5, 296.3)
P (32.5, 12.0) I c,1
300
10 Pc
200
(31.5, 232.1) I a ,1
8
Pa
100

6
0

-100 4
-20 0 20 40 60 80 100 0 5 10 15 20 25 30 35 40
ρ° ρ°
a) Thrust vs. thrust-angle. b) rms values of the input currents vs. thrust-angle.
Fig. 2 – Motor M1. Dashed curves are obtained neglecting the end-effects.

400 40 13
Fx Fx ,s Pa (33.5, 12.03)
Pa
[N] input current [A] Pb (33.5, 11.39) P
(33.5, 299.4) [N] 12
300 30 Pc (33.5, 11.62)
(35.5, 297.4) Pc
P (35.5, 12.0) I a ,1
Pb
Fx 11
200 20 I c ,1
I1
Fx , s 10
I b,1
100 10
(33.5, 6.7) 9

0 0
8

-100 7
-20 0 20 40 60 80 100 0 5 10 15 20 25 30 35 40
ρ° ρ°
a) Thrust vs. thrust-angle. b) rms values of the input currents vs. thrust-angle.
Fig. 3 – Motor M2. Dashed curves are obtained neglecting the end-effects.

500 14
Pa (33.3, 8.71) I b,1
Fx [N]
input current [A]

Pb (33.3, 12.07)
400 Pb I1
12 Pc (33.3, 9.62)
P (34.3, 12.0) P
(34.3, 299.3) I c ,1
300
10 Pc
(33.3, 234.8)
200 I a ,1
Pa
8
100

6
0

-100 4
-20 0 20 40 60 80 0 5 10 15 20 25 30 35 40
ρ ° 100 δ°
a) Thrust vs. thrust-angle. b) rms values of the input currents vs. thrust-angle.
Fig. 4 – Motor M3. Dashed curves are obtained neglecting the end-effects.
400 40 13
Fx Fx , s Pa (35.3, 11.95)
Pa P
[N] Pb (35.3, 11.29)

input current [A]


(35.3, 295.8) [N] 12
300 30 Pc (35.3, 11.52)
(38.3, 296.5)
P (38.3, 12.0) I a ,1 Pc
Fx 11 Pb
200 20 I c ,1
I1
Fx , s 10
100 10 I b,1
(35.3, 7.1) 9

0 0
8

-100 7
-20 0 20 40 60 80 100 0 5 10 15 20 25 30 35 40
ρ° ρ°
a) Thrust vs. thrust-angle. b) rms values of the input currents vs. thrust-angle.
Fig. 5 – Motor M5. Dashed curves are obtained neglecting the end-effects.

As regards the analysis of the input current harmonics, 4. CONCLUSIONS


no sub-harmonics (m < n⋅p) have been found. The
A mathematical model has been proposed to analyse the
significative current harmonics in the range 1 < m < 42
performance of PMLSMs. The model takes into account
occurs for m = n⋅p=6 (fundamental wave), m = 5 n⋅p=30 both the end-effects and the inductor sheet presence.
(the 5th harmonic) and m = 7 n⋅p=42 (the 7th harmonic). The results highlight that, without inductor sheet, the
The rms values of these current harmonics for the end-effects decrease the electromagnetic thrust and give
operating point at the highest fundamental input current rise to an input current unbalancing. On the contrary, the
Imax,1 = max(Ia,1, Ib,1, Ic,1) = Ir are reported in Tab. II. The inductor sheet presence limits these negative effects.
inductor sheet causes an increasing of the 5th and 7th Finally, the input current harmonics have been evaluated
harmonic but, as previously seen, reduces the and discussed too.
unbalancing of the fundamental currents. Moreover, it
can be deduced that the variation of the width of the 5. LIST OF SYMBOLS
permanent magnet poles influences in different way the
5th and 7th harmonic. ∨
a complex conjugate of a;
bp width of the permanent magnet poles;
d thickness of the inductor sheet;
Table II – rms values of the line current harmonics
fr rated frequency;
Motor Phase 30 Hz 150 Hz 210 Hz hp thickness of the permanent magnet poles;
a 8.64 A 0.11 A 0.08 A ik current of phase k;
M1 b 12.04 A 0.10 A 0.08 A j imaginary unit;
c 9.47 A 0.11 A 0.08 A p number of pole pairs;
a 12.03 A 0.24 A 0.19 A qk number of slots of phase k;
M2 b 11.39 A 0.25 A 0.19 A vr rated speed;
c 11.62 A 0.25 A 0.19 A zk number of slot conductors of phase k;
a 8.71 A 0.16 A 0.03 A Re(a) real part of a;
M3 b 12.07 A 0.15 A 0.03 A Ix,1 rms value of the fundamental time harmonic of
c 9.62 A 0.16 A 0.03 A the phase x (a,b,c) input current;
a 11.95 A 0.36 A 0.07 A Ir rated current;
M4 b 11.29 A 0.37 A 0.07 A L width of the armature core;
c 11.52 A 0.37 A 0.07 A Pr rated power;
Q number of armature slots;
Vr line to line rated voltage; [2] Mosebach, H., “Direct two-dimensional
δ overall air-gap; analytical thrust calculation of PM excited
linear synchronous machine”. Proc. LDIA’98,
µ0 magnetic permeability of the free space;
Tokyo, Japan.
ν space harmonic order;
σ inductor sheet conductivity; [3] Gieras, I.A., and Gieras, J.F., “Influence of
τp pole-pitch; permanent magnet materials on performance
ω angular frequency of supply voltages. characteristics of a linear synchronous motor”.
Proc. LDIA’01, Nagano, Japan.
6. REFERENCES
[4] Rotondale, N., and Villani, M., “Guideline for
[1] Gieras, J.F., and Piech Z.J., “Linear the PM linear synchronous motors design”.
synchronous motors”, CRC press, 1999. Proc. SPEEDAM’02, Ravello, Italy.