Aerospace and Electronic Systems Magazine June 2017 - 39

Macˇ iulis and Buzas
tion wheel, a magnetometer, high precision sun sensors, a nadir
sensor, and 3 magnetorquer coils.
This design solution attains required pointing accuracy in two
steps. At first, the satellite enters a so called Y-Thomson spin mode
[6] to ensure that the momentum wheel's axis aligns with the orbit normal, which represents a minimum energy solution. This is
achieved through simultaneous application of wheel bias and Bdot rate-damping controller. During the second step, the momentum wheel operates in reaction mode to slew the satellite around its
minor axis (ostensibly aligned with the orbit normal). Magnetorquers are used to trim momentum in the wheel [7].
The performance of such design was demonstrated on a 3U
CubeSat during the Canadian CanX-II mission in 2010. The satellite was able to achieve at least ±2 degree payload pointing accuracy and managed to detumble from the initial angular rate of 50
deg/s in 7.5 orbits, i.e., within half a day [8].
Although initially considered as a proper solution for LituanicaSAT-2 mission, this system has several disadvantages that later
led to other ADCS design choices. The main drawbacks are the
complexity and high cost of such system. It requires a very accurate and space qualified momentum wheel while the pointing
accuracy is highly dependent on the precision of reference sensors - both sun and nadir sensors are required if the system is to be
accurate during eclipse and sunlight periods of the orbit. The cost
of the system was estimated to be around 20k EUR at that time
(excluding software development costs).
During later stages of the project it was decided to use a semipassive aerodynamic stabilization approach for ADCS design.
This principle takes advantage of the restoring torque created by
atmospheric particles colliding with the satellite to align its center of pressure with velocity vector direction. The feasibility of
such ADCS design for stabilization of CubeSats in low Earth orbit
(LEO) has been investigated in the works of Rawashdeh [9]-[12]
and Armstrong [13]. The main advantage is the ability to control
the attitude passively without the need of active actuators and sensors [11]. However, damping of oscillations is required, as aerodynamic torque by itself cannot dissipate rotational kinetic energy
of the satellite at very low densities. This is usually done by using
either soft magnets or magnetic torquers.
In 2010 the U.S. Naval Research Laboratory launched two 3U
CubeSats which successfully demonstrated the capability to attain
stable velocity vector pointing mode in 300 km altitude by employing deployable aerodynamic fins or the so-called space dart
configuration [14].
The primary motivation factor that led to the choice of this
ADCS design approach for LituanicaSAT-2 is the simplicity and
low cost of the system. Other advantages are increased area for
solar power generation and reliability as the performance is not
dependent on other subsystems or software. The view of the final
aerodynamically stable satellite shape with deployable thins (used
as solar panels) is shown in Figure 7.

ADCS Subsystem Design Description
ADCS is a complex subsystem that consists of ADCS controller,
software, memory storage, actuators, attitude, and inertial sensors.
The ADCS system block diagram can be seen in Figure 8. The
JUNE 2017

Figure 7.

LituanicaSAT-2 CAD model (shown in flight configuration).

Figure 8.

ADCS system block diagram.

ADCS system utilizes several sensors including the MPU6050
gyroscope & accelerometer, L3GD20H gyroscope, four digital
fine sun sensors, and 2 HMC5883L magnetometers-one internal
and the other external, mounted on the back of the satellite structure. Collected measurements are fused into attitude estimate by
unscented Kalman filter (UKF). The system then uses B-dot or
3-axes control algorithms to control external magnetorquers via
microcontroller's PWM outputs.
For X and Y axes actuators magnetorquer rods composed of
soft ferrite ferromagnetic cores and 4000 turns of 0.14 mm diameter copper wire are used while for Z axis a magnetorquer coil with
the same wire but 270 turns was selected and occupies a complete
spacecraft cross section for easier integration and maximum efficiency (see Figure 9). Polarity of magnetorquer can be reversed
using H-bridge drivers. Each magnetorquer is rated to 40 mA current and can provide up to 0.3 Am2 magnetic dipole strength for the
rods and 0.5 Am2 for the coil.
64 MB of nonvolatile NOR flash memory are used for storing
flight data, parameters, backups, and other ADCS data. 256 KB of

IEEE A&E SYSTEMS MAGAZINE

Table of Contents for the Digital Edition of Aerospace and Electronic Systems Magazine June 2017