Download Failure Diagnosis for Demagnetization in Interior Permanent Magnet

1
Failure Diagnosis for Demagnetization
in Interior Permanent Magnet
Synchronous Motors
1.
2.
3.
4.
5.
6.
7.
8.
Takeo Ishikawa Gunma University, Japan
Background and purpose
Rotor configuration of PMSM with permanent magnet defect
Electromotive force of the demagnetized motor
Simulation for demagnetized motor
Stator voltage and current under the constant V/f control
Stator voltage and current under the vector control
High frequency impedance
Conclusions
2
1. Background and Purpose
(1/3)
Background
• Permanent magnets of PM synchronous motors can be
demagnetized by temperature, large stator currents, large shortcircuit currents produced by the inverter or stator faults, and the
aging of magnet itself.
• Because the PM material is often magnetized after it is inserted
into the rotor assembly, there is a possibility that the PM is not in
a state of complete magnetization.
Since a high degree of reliability is necessary to PMSMs, the
detection of a precursor to the demagnetization of PMs is very
important.
3
1. Background and Purpose
(2/3)
Previous research
• Rajagopalan et al. created a magnet defect by chipping off a part
of the magnet and then measured the stator current.
in Proc. Inst. Elect. Eng. Conf. Publ., vol. 2, pp. 668-673, 2004.
• Urresty, Romeral, Ortega, et al. analyzed the stator current of a PM
motor using several methods.
in Proc. IEEE 33rd Ann. Conf. Ind. Electron. Soc., pp. 1149-1154, 2007.
IEEE Trans. Energy Convers., vol. 25, no. 2, pp. 312-318, 2010.
The degree of PM demagnetization in these past studies was
relatively large.
4
1. Background and Purpose
(3/3)
Purpose
PM volume is decreased by 2.5%, 5.0%, and 7.5%, so as to mimic the
effect of demagnetization.
• To clarify physical variables to be available for diagnosis for very
slight demagnetization in IPMSMs
• To simulate under a constant V/f control strategy and under a
vector control strategy
• To verify simulation results experimentally
5
2. Rotor configuration of PMSM
(1/2)
(a) Rotor configuration.
(b) Photograph of rotor.
Fig. 1. Rotor of the experimental PMSM.
Motor's rating is 1.5 kW, 3000 min-1, 4.8 N.m, 5.6 A, four-pole machine.
The PM volume of one of four poles is altered so as to mimic the effect of
demagnetization.
When performing experiments with several motors, slight differences between
motors can affect the relative motor performance especially if the amount of
demagnetization is small. In order to avoid this problem, we perform experiments
on the same stator and the same rotor using the different PMs.
6
2. Rotor configuration of PMSM
PM
(2/2)
Non-magnetic
material
Radial demagnetization
Radial demagnetization
Non-magnetic
material
PM
PM
(a) Rotor configuration. (b) permanent magnet
Axial demagnetization
Fig.2 Rotor and PM.
Radially demagnetized magnet, whose one pole is composed of four PMs, and the
thickness of the four PMs that compose one pole is reduced by 10%, 20%, and 30%.
Total reduced volume of PM is 2.5%, 5% and 7.5% (R-2.5%, R-5.0%, R-7.5%)
Axially demagnetized magnet, where the axial length of two of the four PMs per
pole is reduced by 20%, 40%, and 60%. Total reduced volume of PM is 2.5%, 5% and
7.5% (Z-2.5%, Z-5.0%, Z-7.5%)
7
3. Electromotive force
(a) Radial demagnetization
(1/4)
(a) Radial demagnetization
(b) Axial demagnetization
(b) Axial demagnetization
Fig.4.Measured EMF waveform
Fig.3.Calculated EMF waveform
In 3-D FEM, the analysis model is 1/2 of the model in the axial direction because of
unsymmetry. Node: 132,813, element: 225,758.
8
3. Electromotive force
(2/4)
Table1. Dimension of PMs and measured flux linkage
PM
V [mm3]
V/Vhealthy
Φ [Wb]
Φ/Φhealthy
Healthy
16070
1.0
0.2157
1.0
R -2.5%
15704
0.977
0.2108
0.978
R -5.0%
15339
0.955
0.2055
0.954
R -7.5%
14974
0.932
0.2002
0.930
Z -2.5%
15661
0.975
0.2104
0.976
Z -5.0%
15252
0.950
0.2048
0.950
Z -7.5%
14843
0.924
0.1994
0.925
The reduction of the flux linkage of the motor with demagnetized PMs is equal to
the reduction of PMs.
EMF and flux linkage of the motor with radial demagnetization is approximately
the same as that of axial demagnetization.
9
3. Electromotive force
(3/4)
Fig. 5. The magnetic flux density at the center of the air-gap
and axial location (z = 0).
The flux density is reduced from 0.516 T to 0.481 T for the radial demagnetization
and from 0.516 T to 0.479 T for the axial demagnetization at θ = 315°.
There is a very slight difference between the radial demagnetization and the axial
demagnetization.
10
3. Electromotive force
(4/4)
Fig. 6. The magnetic flux density at the center of the air-gap
and θ = 315°.
The magnetic flux density is almost flat in the z-direction.
Even for axial demagnetization, the magnetic flux density is 0.479 T at z = 0, and
0.477 T at z = 42 mm.
11
4. Simulation for demagnetized motor
(1/4)
Fig. 7. MATLAB model for demagnetized motor under constant V/f control
An S-Function makes the reference of three stator voltages proportional to the
inverter frequency f. It is converted to a duty ratio for the transistors, and is input
to a block PWM inv. The PWM voltage produced by PWM inv. is input to a block
PM which is the model of the IPMSM provided in SymPowerSystem.
The demagnetized situation can be introduced by changing the flux linkage in the
PM block. Ld and Lq are assumed to be the same as those of a healthy motor.
12
4. Simulation for demagnetized motor
(2/4)
Δ Vdc [V]
0
-2
-4
-6
-8
0
0.5
1
1.5
2
2.5
I dc [A]
3
Fig. 9. DC voltage drop across the inverter.
In Fig.8, the PWM voltage is generated by
comparing the voltage reference and a triangular
carrier waveform. The stator voltage drop is taken
into account by considering an input voltage drop
and an inverter voltage drop.
In Fig.9, this voltage drop can be represented by a
linear equation, and is represented by a block Vdc
drop.
Fig. 8. Subsystem for PWM inv.
and subsystem for va, vb and vc
13
4. Simulation for demagnetized motor
(3/4)
Shifted down by v _ Tr  i  r _ Tr Shifted down by v _ D  i  r _ D
Vdc/2
Vdc/2
Vdc/2
Vdc/2
Case 1: s=1, i>0
Shifted up by v _ D  i  r _ D
Case 2: s=0, i>0
Shifted up by v _ Tr  i  r _ Tr
Vdc/2
Vdc/2
Vdc/2
Vdc/2
Case 3: s=1, i<0
Case 4: s=0, i<0
Fig. 10. Analysis of the output voltage of a
phase inverter leg.
Fig. 8. Subsystem Voltage drop
We assume that the v-i characteristics of the IGBT and diode are represented by
constant forward voltage drops, v_Tr and, v_D respectively and resistance, r_Tr and
r_D respectively. s denotes a switching function for the transistor.
14
4. Simulation for demagnetized motor
(4/4)
Fig. 11. MATLAB model for demagnetized motor under vector control
Two control loops are used; one is an inner loop to regulate the stator currents by
converting to the d and q axes with the rotor position, and the other is an outer
loop to control the motor speed. Stator voltage drop is also taken into account by
the subsystem for idc.
In the experiment, a broken line part was replaced by DSP blocks.
15
5. V and I under constant V/f control
Hysteresis brake Torque meter
(1/3)
IPMSM with encoder
Fig. 12. Experimental setup
In the experiment, a broken line part was replaced by DSP blocks.
Three signals to the inverter were connected to the DS1102PWM block of a
DS1102 DSP board.
The stator currents were detected through the DS1102ADC block, and the motor
speed and rotor position were detected through the DS1102ENC_POS_C1 block.
16
5. V and I under constant V/f control
110
4
Voltage [V]
Current [A]
5
3
2
Healthy
R-2.5%
R-5.0%
R-7.5%
1
0
(1/3)
0.5
1
1.5
Load torque [N.m]
Healthy
R-2.5%
R-5.0%
R-7.5%
105
100
95
90
0
0.5
1
1.5
Load torque [N.m]
1500min-1
Fig. 13. Simulated results under constant V/f control,
(a) left: stator current, (b) right: stator voltage
There is a difference in stator currents. The stator current of the motor
demagnetized by 7.5 % is the largest value.
There is no difference in stator voltages.
This is because the stator current is not controlled under constant V/f control.
17
5. V and I under constant V/f control
110
4
Voltage [V]
Current [A]
5
3
2
0.5
95
90
1
1.5
Load torque [N.m]
4
3
2
Healthy
R-2.5%
R-5.0%
R-7.5%
1
0
100
0.5
1
1.5
Load torque [N.m]
0
0.5
110
Voltage [V]
Current [A]
5
Healthy
R-2.5%
R-5.0%
R-7.5%
105
Healthy
R-2.5%
R-5.0%
R-7.5%
1
0
(2/3)
1
1.5
Load torque [N.m]
Fig.13. Simulated
Healthy
R-2.5%
R-5.0%
R-7.5%
105
100
95
90
0
0.5
1
1.5
Load torque [N.m]
Fig.14. Measured
Experiments were carried out several times.
The measured results of axially demagnetized motor were the same as above.
The measured results are approximately the same as those of the simulation.
18
5. V and I under constant V/f control
(3/3)
Table 2 Failure diagnosis under V/f control
Stator current
Stator voltage
Load
0%
10%
20%
30%
R-2.5%
[%]
19.6
9.13
-0.602
-2.83
R-5.0%
[%]
40.4
30.8
2.08
-3.45
R-7.5%
[%]
61.5
57.3
5.18
-2.64
Load
0%
10%
20%
30%
R-2.5%
[%]
0.00
0.401
0.500
0.273
R-5.0%
[%]
-0.517
0.711
0.963
0.168
R-7.5%
[%]
-0.445
0.604
0.718
-0.169
: It is possible to distinguish the existence of demagnetization
When the torque is 0, the increase of stator current is proportional to the amount
of demagnetization. Therefore, the stator current under constant V/f control at
no-load can be used for the failure diagnosis of demagnetization.
There is no difference between radial demagnetization and axial demagnetization.
19
5. V and I under vector control
12
i d , i q [A]
8
Stator voltage [V]
130
: Healthy
: R-2.5 %
: R-5.0 %
: R-7.5 %
10
iq
6
4
: Healthy
: R-2.5 %
: R-5.0 %
: R-7.5 %
120
110
100
90
80
2
0
(1/6)
70
id
0.5
1
1.5
2
2.5
3
3.5
Load torque [N･m]
4
60
0
0.5
1
1.5
2
2.5
3
3.5
Load torque [N･m]
4
Fig. 15. Simulated results under vector control,
(a) left: current, (b) right: voltage
The stator current iq is approximately proportional to the load torque.
The difference due to demagnetization are very small when the load torque is
small, and becomes large when the load torque becomes large.
There is a difference due to demagnetization in stator voltages.
When the load torque is small, the stator voltage of the motor demagnetized by
7.5% has the lowest value and that of the healthy motor has the highest value.
In contrast, when the load torque is large, the stator voltage of the motor
demagnetized by 7.5% has the highest value.
20
5. V and I under vector control
12
i d , i q [A]
8
Stator voltage [V]
130
: Healthy
: R-2.5 %
: R-5.0 %
: R-7.5 %
10
iq
6
4
0.5
1
1.5
2
2.5
3
3.5
Load torque [N･m]
90
: Healthy
: R-2.5 %
: R-5.0 %
: R-7.5 %
iq
4
2
1
1.5
2
2.5
3
3.5
Load torque [N･m]
0.5
1
1.5
4
2
2.5
3
3.5
Load torque [N･m]
4
Fig.15. Simulated
120
110
100
90
80
70
id
0.5
0
130
6
0
100
60
4
Stator voltage [V]
i d , i q [A]
8
110
70
id
12
10
: Healthy
: R-2.5 %
: R-5.0 %
: R-7.5 %
120
80
2
0
(2/6)
60
0
0.5
1
1.5
: Healthy
: R-2.5 %
: R-5.0 %
: R-7.5 %
2
2.5
3
3.5
4
Load torque [N･m]
Measured results are approximately the same as simulated ones.
The simulation results have been verified by experiments.
Fig.16. Measured
21
5. V and I under vector control
(3/6)
Table 3 Failure diagnosis under Vector control
Stator current
Stator voltage
Load
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
R-2.5% [%]
0.00
2.21
2.16
2.53
2.40
2.07
2.55
2.42
2.65
2.40
2.49
R-5.0% [%]
0.00
4.63
4.88
4.92
4.93
4.65
4.81
5.25
5.30
5.09
5.15
R-7.5% [%]
0.00
7.35
7.80
7.72
7.78
7.22
7.51
7.83
8.05
7.98
7.91
Load
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
R-2.5% [%]
-2.25
-2.41
-1.68
-1.12
-0.715
-0.505
0.389
0.379
1.13
0.938
1.16
R-5.0% [%]
-4.85
-4.61
-3.53
-2.48
-1.46
-0.872
0.397
1.22
1.91
2.21
2.65
R-7.5% [%]
-7.15
-6.62
-5.04
-3.82
-1.92
-0.988
0.603
1.60
2.96
3.59
4.34
It is possible to diagnose the amount of demagnetization
It is possible to distinguish the existence of demagnetization
Stator current is approximately proportional to the amount of demagnetization. Therefore,
the stator current can be used for the failure diagnosis of demagnetization.
Stator voltage is approximately proportional to the amount of demagnetization when the
load torque is small. Therefore, the stator voltage can be also used for the failure diagnosis of
demagnetization.
:
:
22
5. V and I under vector control
(4/6)
Discussion: The voltage and torque equations are expressed by
 r La  id   0 

   
Ra  PLa  iq   r  
Te  piq  p ( Ld  Lq )id iq
When id is controlled to be 0,
vd   Lq iq
vq  R  PLq  iq   
T  p iq
(1)
(2)
(3)
(4)
(5)
12
: Healthy
: R-2.5 %
: R-5.0 %
: R-7.5 %
10
i d , i q [A]
 vd   Ra  PLa
 
 vq 
   r La
8
iq
6
4
2
0
id
0.5
1
1.5
2
2.5
3
3.5
Load torque [N･m]
4
Fig.15 (a). Simulated current
From (5), iq is inversely proportional to flux linkage, and the flux linkage is
proportional to the magnet volume, that is, demagnetization situation.
Therefore, the demagnetization of PM can be estimated by the stator current under
vector control with id = 0.
23
5. V and I under vector control
(5/6)
Discussion: The voltage and torque equations are expressed by
 r La  id   0 

   
Ra  PLa  iq   r  
Te  piq  p ( Ld  Lq )id iq
When id is controlled to be 0,
vd   Lq iq
vq  R  PLq  iq   
T  p iq
(3)
(4)
(5)
130
Stator voltage [V]
 vd   Ra  PLa
 
 vq 
   r La
: Healthy
: R-2.5 %
: R-5.0 %
: R-7.5 %
120
110
100
90
80
70
60
0
0.5
1
1.5
2
2.5
3
3.5
Load torque [N･m]
4
Fig.15 (b). Simulated voltage
From (3) and (4), vq is proportional to flux linkage and vd becomes 0, when the load
torque is small, that is, iq is small.
Therefore, the demagnetization of PM can be estimated by the stator voltage at noload under vector control with id = 0.
In contrast, when the load torque becomes large, that is, iq becomes large, vq is
affected by iq. As a result, the stator voltage of the demagnetized motor becomes
large.
24
5. V and I under vector control
T
p healthy faulty
 Lq  R
2
2
2
(6)
where  healthy and  faulty are the flux linkages of
healthy motor and demagnetized motor.
130
Stator voltage [V]
Discussion:
The torque at the intersection point is driven as
(6/6)
: Healthy
: R-2.5 %
: R-5.0 %
: R-7.5 %
120
110
100
90
80
70
60
0
0.5
1
1.5
2
2.5
3
3.5
Load torque [N･m]
4
Fig.15 (b). Simulated voltage
The intersection point changes when the motor parameter changes.
The diagnosis of demagnetization using stator voltage becomes difficult for the motor
with small flux linkage and large Lq.
25
6. High frequency impedance
(1/2)
The impedance was measured when a
high-frequency voltage is injected
between u and w phases. Here, the motor
is not rotating, that is, there is no
fundamental voltage and current.
The effect of demagnetization does not
appear at the impedance of 1000Hz.
Fig.17 Impedance at 1000 Hz
26
6. High frequency impedance
(2/2)
The effect of demagnetization appear2 at
the resistance of 250Hz.
Therefore, the resistance of low
frequency region is useful for the
diagnosis of demagnetization of PM.
Fig.18 Impedance at 250 Hz
27
Conclusions
We have investigated the diagnosis of the very slight PM
demagnetization.
1. The electromotive force is useful for the diagnosis of
demagnetized PM.
2. The resistance at relatively low frequency is useful.
3. Under constant V/f control, the stator current is useful.
4. Under vector control, the stator voltage is useful except in an
intermediate torque range, and the stator current is useful in a
high torque range. The intermediate torque has been expressed
by a simple equation.
5. The simulation results have been verified experimentally.
Thank you for kind attention