ICCAS2005 June 2-5, KINTEX, Gyeonggi-Do, Korea

Performance Analysis of Magnetic Torquer for Spacecraft Control

Seon-Ho Lee*, Hyun-Ho Seo**, and Seung-Wu Rhee***
* Department of Satellite Control, Korea Aerospace Research Institute, Taejon, Korea
(Tel : +82-42-860-2035; E-mail: shlee71@kari.re.kr)
** Dept of Satellite Systems & Applications, University of Science and Technology, Taejon, Korea
(Tel : +82-42-860-2557; E-mail: seo2h@ust.ac.kr)
*** Department of Satellite Control, Korea Aerospace Research Institute, Taejon, Korea
(Tel : +82-42-860-2447; E-mail: srhee@kari.re.kr)

Abstract: This paper presents the overall design, manufacture, and test result of the high capacity (more than 150 Am^2)
magnetic torquer for the use of satellite control. To provide an electrical current to the magnetic torquer, the driving electronics is
also constructed. The integration and test of the magnetic torquer and its driving electronics are performed via the magnetic field
measurement according to the distance and the magnetic torque measurement with torque-meter. To compare the performance
obtained from the test results we also perform computer simulation with three-dimensional model.

Keywords: Magnetic Torquer, Magnetic Field, Magnetic Dipole Moment, Driving Electronics

1. INTRODUCTION The magnetic torquer is an electromagnet which is

composed of ferromagnetic cylindrical core wound with the
The magnetic torquer (or torque rod) is a magnetic dipole excitation coil as shown in figure 1. The excitation coil current
moment generating device which produces torque with an (I) induces the magnetic field (B) and the volume integral of
interaction of geomagnetic field in space. The magnetic the magnetic field generates the magnetic dipole moment (M).
torquer is uses in spacecrafts for attitude control instead of Thus, as shown in figure 2 the dipole moment produces the
reaction wheels and thrusters, momentum dumping of reaction torque (T) which rotates the spacecraft in space by interacting
wheels, and compensation of magnetic residual inside the
with the geomagnetic field (BE) such as T = M × BE .
spacecraft in order to achieve a stable camera pointing
capability during the mission [1-2]. Although the magnetic
torquer produces smaller torque level than the reaction wheel T

and the thruster, it is mounted on the most satellites as a basic M BE BE

actuator owing to its simple structure and high reliability. In T = M x BE M
this paper, we present the details of design, manufacture, and MTQ ON

initial test result of the high capacity magnetic torquer for the Figure 2 Interaction of MTQ with geomagnetic field (BE)
use of small satellites. Currently, three magnetic torquers
whose momentum is expected to be more than 150 Am^2, In most cases, the current coil of the magnetic torquer is
have been manufactured as a laboratory model and they are wound dually for primary and redundant usage to cover the
under various test. To provide a current to the magnetic single failure situation. In electrical point of view, the
torquer, the driving electronics is constructed. Furthermore, magnetic torquer is modeled as a serial connection of resistor,
the computer simulation and experimental results are provided inductor, and residual capacitor between the coils. Moreover,
to compare with the requirements. This paper consist of the the magnetic torquer is mathematically modeled with a single
design and manufacture of the magnetic torquer and the current loop approximately. In figure 3, the measurement
driving electronics in Section 2, the computer simulation position angle from the axial axis at center of the loop is
analysis in Section 3, the experimental results in Section 4, denoted by θ, the distance of measurement position from the
and the Conclusions in Section 5. center by R, and the total measured magnetic field at given
position by B. Here, we note that BR and Bθ are the radial and
2. DESIGN & MANUFACURE tangential components such as
G G
2.1 Magnetic Torquer (MTQ) B( R, θ ) = BR R + Bθ θ
µ0 M G G (1)
= 3
(2 cos θ R + sin θθ )
Ferromagnetic core
4π R
Moreover, the dipole moment is computed as
Connector
1 2π R 3
Excitation
coils
M =
µ0 ∫
V
BdV = B ( R, 0)
µ0
(2)

Mounting The above equation is an approximated results by assuming

Block that R is sufficiently larger than the current loop radius where
650m
m0 (=4pµ10-7 H/m) is the vacuum permeability. Particularly,
for the cases of θ=0 and p/2, we obtain

Figure 1 Structure of magnetic torquer

 ICCAS2005 June 2-5, KINTEX, Gyeonggi-Do, Korea
µ0 M G
B ( R, 0) = 3
R
2π R
MTQ-X
Pri/Red Selection

(3) In-X Out-X

µM G
π
-10 ~ -15 ~
P R
12 CS +15V P

B ( R, ) = 0 3 θ
+10V Power
DAC 12 bits Amp.
Numerical
R
2 4π R
command
Zener P
Protection
R
According to the above equations, the dipole moment can be Relay

calculated from the measured magnetic field distribution and Magnetic

Driving
the generated torque is estimated. Electronics Torquer

Figure 5 Circuit of driving electronics

Bθ BR
Inside the driving electronics, the electrical circuit board is
composed of 12-bit digital-to-analog converter, power
B ( R, θ )
R amplifier, and the two relay making it easy to select the
primary and redundant coils of the magnetic torquer as shown
MTQ θ in figure 5. Owing to the driving electronics, bang-bang and/or
x proportional control action according to the specific control
M algorithm can be implemented. Outside the box, three LED
I eqv
lamps are installed to indicate the on/off status of the primary
and redundant coil excitation in the front panel and three
Figure 3 Magnetic field of single current loop input/output connectors (X,Y,Z) are installed in the back
panel.
The requirement specification of the designed magnetic
torquer with high dipole moment capacity is described as
follows:
- Coil resistance: 78 Ohm
- Coil inductance: 13 H
- Generated magnetic dipole moment: > 150 Am^2
- Max applied voltage: +/- 15V
Figure 6 Photo of driving electronics
- Coil turn number: 6546
- Coil length: 33.8 3. Computer Simulation
- Core: Ni-Alloy with high permeability In this paper, we execute the computer simulation for the
- Core dimension (D x L): 22mm x 650mm performance analysis of the magnetic torquer. As shown in
- Total mass: 5 kg figure 7, the three-dimensional model of the magnetic torquer
- Max power: < 3 W is performed with the following steps: (i) define structure of
magnetic torquer, (ii) set boundary conditions, (iii) define
current sources, (iv) solve the magnetic field density, (v)
As shown in figure 2, the manufactured magnetic torquer is compute the magnetic dipole moment by post-processing.
equipped with two mounting blocks for spacecraft installation Figure 8-9 show the estimated magnetic field BX(X=0,Y,Z) at
and a D-type connector for the driving current interface with Y-Z plane and BX(X,Y=0,Z) at X-Z plane. The simulation
on-board computer. results reveal that the magnetic field are induced dominantly
inside the core and the maximum field magnitude along the
axial direction is about 1.1Tesla (=11,000 Gauss). We note
that the geomagnetic field normally amounts to -0.6~+0.6
Gauss on ground.
Figure 4 Photo of magnetic torquer

2.2 Driving Electronics (TDE)

The driving electronics is designed to provide the coil Y
current to the magnetic torquer according to the control X
command. It reads the 12-bit digital command from the Z
on-board computer of the satellite and outputs a corresponding
analog voltage in the range of -15V~+15V to the magnetic
torquer.

Figure 7 Three-dimensional model of magnetic torquer

 ICCAS2005 June 2-5, KINTEX, Gyeonggi-Do, Korea
Precision
MTQ Magnetometer

DSP
Board

Figure 10 Photo of magnetic field measurement test setup

The test setup consists of magnetic torquer, driving electronics,

DSP board, control PC, and precision magnetometer. The test
is done by measuring the magnetic field magnitude along the
axial direction of the magnetic torquer. Recall that
µ M a µ M (5)
Bx = 0 3 = , (a  0 )
2π | x | | x |3 2π
where a is a scale factor to be determined. By gathering the
measured magnetic field data and its corresponding distance
Figure 8 Magnetic field Bx at X-Z plane (when Y=0) and applying to the above equation, we can plot figure 11. By
1st order curve fitting of the measurement data, we obtain
a = 3.7 × 10−5
(6)
2π
M = a× = 186 Am2
µ0

The obtained dipole moment from the test results in (6) is

quite similar with the estimated one from the computer
simulation in (4).

Figure 9 Magnetic field Bx at Y-Z plane (when X=0) Slope

Using the simulation data, we calculate the magnetic dipole

moment by volume integration of the magnetic field
distribution as follows.
1
M =
µ0 ∫ BdV
V
Figure 11 Calculation of slope (a) with measurement data

2π (4)
∫∫ B(r, z)rdrdz
= 4.2 Magnetic Torque Measurement
µ0
= 180 Am2
According to the above calculation, we can check the
estimated dipole moment satisfies the minimum requirement
of 150Am^2.
4. Experiment
4.1 Magnetic Field Measurement

One of the well known methods to check the dipole

moment is to measure the magnetic field magnitude depending
on the measurement distance from the excited magnetic
torquer. This method is very simple but it needs magnetically Figure 12 Photo of magnetic torque measurement test setup
calm environment by minimizing the external magnetic
disturbance source and noise coming from the magnetic We make an experiment on the magnetic torquer with a
materials. torque-meter. To compute its dipole moment, the geomagnetic
field should be measured through the precision magnetometer,
which has a different coordinate from the magnetic torquer.
The magnetic torquer and the torque-meter are instrumented
 ICCAS2005 June 2-5, KINTEX, Gyeonggi-Do, Korea
on the table as shown at figure 12. The following formula
represents the computing procedure of dipole moment and the
coordinate transformations.
G G G
T = M × BE
TX  cos θ − sin θ 0  M X 
T  = sin θ cosθ 0   M Y  ×
 Y 
TZ   0 0 1   M Z  (7)
 cos 90 s in 90D 0  cos180D
D
0 − sin180 
D

  
 − sin 90 cos90 0  0
D D
1 0  Figure 14 Command voltage cycling to magnetic torquer
 0 0 1  sin180 0 cos180D 
  Table 1 Measured magnetic torque from torque-meter
cos θ sin θ 0   Bx ' 
   Torque (mNm)
 − sin θ cosθ 0   By '  Voltage
 0 0 1  Bz '  (V)
No. Rotation Angle θ (deg)
0 45 135
where +15 1 2.1 1.3 -1.2
G
T : Measured torque from torque-meter +0 2 0.0 0.0 0.0
 G -15 3 -2.9 -2.1 2.0
 M : Dipole moment -0 4 -0.9 -0.7 0.7
G
 BE : Geomagnetic field +15 5 2.0 1.3 -1.2
 +0 6 0.0 0.0 0.0
θ : Rotation angle from the reference frame
 -15 7 -2.9 -2.1 1.9
 X - Y - Z : Fixed reference coordinate -0 8 -0.9 -0.7 0.6
 x '- y '- z ' : Magnetometer coordinate +15 9 2.1 1.4 -1.3
+0 10 0.0 0.0 0.0
-15 11 -2.9 -2.0 1.9
Assuming that MY @ MZ @ 0, the dipole moment is simply -0 12 -0.9 -0.6 0.6
obtained as
TZ = M X ( Bx ' cos 2 θ + By ' sin θ cos θ + Table 2 Measured field data from precision magnetometer
(8)
Bx ' sin 2 θ − By ' sin θ cos θ ) Rotation Geomagnetic Field (Bx’)
T (9) Angle
MX = Z (deg) Left Center Right Average
Bx '
0 -23765 -22277 -22243 -22762
The geomagnetic field is measured on three positions (left,
45 -15550 -16888 -17859 -16766
center, and right) by putting the magnetometer on the plate in
90 89.9 -1959 -2605 -1491
figure 13. The torque-meter measures the generated magnetic
torque about the Z-direction. The command voltage to the 135 16583 13419 14044 14682
magnetic torquer varies from +15V to -15V and its sequence
is plotted in figure 14. As sown in Table 1, to minimize the Figure 15 shows the calculated dipole moment by
measurement error of the test results, the magnetic torque and post-processing of the gathered data according the command
geomagnetic field data are gathered at different rotation angle voltage cycling. The result reveals that the maximum magnetic
of the plate such as θ =0o, 45o, 135o. torque is about 105 Am2 on average and it seems to have a
residual dipole moment of approximately 20Am2. This result
which is different from (4) and (6), results from the
performance limitation of the torque-meter. To investigate it
more clearly, it should be noted that the measurement accuracy
of the torque-meter is less about 1mNm. Thus, the
torque-meter can’t measure the generated torque ranging from
-2mNm to 2mNm exactly, and the torque level corresponding
to the residual dipole moment is too small for the torque-meter
to measure the magnetic torque. Moreover, at the moment now
it is too rush to conclude that 20Am2 is the residual dipole
moment and it comes from the inherent hysteresis
characteristics of the magnetic core. Thus, it is believe that the
magnetic torque measurement test results obtained by the
Figure 13 Coordinate definition limited torque sensing equipment gives not the quantitative but
the qualitative performance of the magnetic torquer. To exact
measurement of the generated torque level, more accurate
 ICCAS2005 June 2-5, KINTEX, Gyeonggi-Do, Korea
torque-meter should be engaged in the magnetically uniform
environment without any noise source.

Figure 15 Dipole moment versus command voltage cycling

5. CONCLUSIONS
This paper presents the development of the magnetic
torquer which is used in spacecrafts for the attitude control
and momentum dumping. The magnetic torquer and its driving
electronics is designed and tested via magnetic field
measurement according to the distance and magnetic torque
measurement with torque-meter. Furthermore, this paper
performs computer simulation with three-dimensional model
to compare it with the real test performance. Throughout the
analysis and test experience as introduced in this paper, we
achieve the fully understanding of the nature of the magnetic
torquer and set up its development process of the future flight
models for spacecraft control.

REFERENCES
[1] J.R. Wertz, "Spacecraft Attitude Determination and
Control", D. Reidel Publishing Company, Holland,
1978.
[2] M.J. Sidi, "Spacecraft Dynamics and Control - A
Practical Engineering Approach", Cambridge University
Press, 1997.
[3] D.K. Cheng, Field and Wave Electro-magnetics,
Addison-Wesley Pub., 1989.
[4] J. Lee, A. Ng, R. Jobanputra, "On Determining Dipole
Moments of a Magnetic Torquer Rod - Experiments and
Discussion", J. Aeronautique et Spatial De Canada, vol.
48, no. 1, pp. 61-67, 2002.
[5] B.V. Rauschenbakh, M.Y. Ovchinnikov, and S.
Mckenna-Lawlor, Essential Spaceflight Dynamics and
Magnetoshperics, Kluwer Academic Publishers
[6] P. Nakmahachalasint, K.D.T. Ngo, and L. Vu-Quoc,
“Effective Magnetic Parameters in the Presence of
Hysteresis”, IEEE Trans Aerospace & Electronic
Systems, Vol. 40, No. 3, pp 1100-1104, 2004.