Paper - IAC-12 - C4.6.2 - 284628 Delffi
Paper - IAC-12 - C4.6.2 - 284628 Delffi
Paper - IAC-12 - C4.6.2 - 284628 Delffi
IAC-12-C4.6.2
Angelo Cervone
Space Systems Engineering – Delft University of Technology, The Netherlands, [email protected]
The performance of several micro-propulsion systems is compared when applied to the specific needs of the
DelFFi project, a formation flying technology demonstration expected to take place in the framework of the QB50
mission. The evaluated options include the T3µPS cold gas micro-propulsion system developed by TNO, Delft
University of Technology and the University of Twente and a silicon-based MEMS resistojet design proposed by TU
Delft. The analysis clearly shows that a resistojet technology, if combined with the cool gas generator units used in
the T3µPS system, leads to significant advantages in terms of the total propulsion system mass and is still complaint
to the stringent volume requirements of micro- and nano-satellites. These advantages are particularly evident for low
values of the total velocity change required by the mission, as in the case of the DelFFi project.
IAC-12-C4.6.2 Page 1 of 8
63rd International Astronautical Congress, Naples, Italy. Copyright ©2012 by the International Astronautical Federation. All rights reserved.
In this paper, the performance of several micro- this scenario is about 1 km, and the velocity change for
propulsion systems (including the T3µPS cool gas and maintaining the baseline distance between the satellites,
the MEMS resistojet concept developed at TU Delft) in the low altitude orbit foreseen for QB50, is expected
will be compared when applied to the specific needs of to be about 0.1 m/s per day. Considering a 30-days
the DelFFi project, a formation flying technology formation flying mission and accounting for maneuvers,
demonstration expected to take place in the framework contingencies and margins, a total velocity change of
of the QB50 program, an international mission 6.3 m/s per satellite is expected to be required.
incorporating a network of 50 CubeSats in Low-Earth However, several different alternative formation flying
orbits for lower thermosphere and re-entry research. In configurations could be considered depending on the
particular, it will be shown that an advanced T3 stakeholders needs and requests, leading to consequent
propulsion system including a MEMS micro-resistojet is variations of the velocity change requirements. The
capable to meet the requirements relative to drag control candidate micro-propulsion system for the in-orbit
for relative formation flying and controlled re-entry of demonstration payload is presently represented by the
the twin DelFFi satellites, while still adhering to the T3µPS cold gas system, which will be described more in
mission mass constraints. detail in the next Section.
II.I. QB50 Main Requirements
II. THE DELFFI PROJECT
The present mass budget allocated to the in-orbit
The QB50 mission is intended to establish an
demonstration payload is 420 g per each QB50 satellite,
international network of 50 nano-satellites with several
out of a total satellite mass of 3600 g. In particular, a
mission objectives, including multi-point measurements
mass of 330 g is available for the propulsion system,
in the lower thermosphere and re-entry research tasks.
with a maximum available volume of 3x10x10 cm3. On
The Delft University of Technology plans to contribute
the other side, the power budget allocated to the in-orbit
to QB50 by means of two satellites, as space segment
demonstration payload is 290 mW per satellite: 100 mW
for the so-called DelFFi project10.
are required by the radio-frequency navigation sensor,
The main objective of the DelFFi project is an
while an average of 190 mW are allocated to the
autonomous formation flying demonstration between
propulsion system. Taking into account the general
the two satellites, using innovative concepts for their
project requirements, the estimated torque disturbances
navigation, guidance and control. Two identical triple-
and the possible misalignment and assembly errors, the
unit CubeSats are expected to be used to this purpose,
required thrust per satellite can be estimated to be 5 mN
including an in-orbit demonstration payload constituted
as a maximum. The presence of a deployable array leads
by a micro-propulsion system (for controlling the
to a relatively low value of the maximum allowed
relative motion of the two satellites) and a radio-
acceleration, equal to 2 m/s2. No specific requirements
frequency navigation sensor (for relative ranging and
related to the safety aspects are presently given:
inter-satellite communications).
therefore, there is not yet a maximum allowed pressure
and there are no indications on the propellants which
cannot be used (if any).
IAC-12-C4.6.2 Page 2 of 8
63rd International Astronautical Congress, Naples, Italy. Copyright ©2012 by the International Astronautical Federation. All rights reserved.
CanX spacecraft program of the University of Toronto, shapes. The “thrusters” themselves consist of a
and it is especially designed for use in formation flying piezoelectric valve and an integrated nozzle.
missions11. The propellant can be stored in the tank at its vapour
The system (Fig. 2) is completely made of pressure, equal to about 15 psia (1.03 bar), but higher
commercially-off-the-shelf components. The propellant storage pressures are also possible; Vacco declares as
is Sulphur Hexafluoride (SF6), working in cold gas “operational” tank storage pressure a value of 40 psia
mode (but the possibility of heating the propellant is (2.75 bar), at which a thrust of 55 mN can be obtained.
considered in the design of the system); it is stored in a The estimated vacuum specific impulse of the system is
supply tank and flows through a thruster valve and a 60 s. The total system mass is 509 g, out of which 53 g
regulator valve. A vent valve and a pressure sensor are are propellant. The corresponding reported tank volume
also present. A frame holds together the entire system, is 95 cm3.
which typically consists of 4 thrusters. The design
storage pressure of the propellant inside the tank is 34.5 III.III. T3µPS Cold Gas System
bar; the overall wet mass of the system (including The T3µPS (Fig. 3) consists of a PCB and a plenum
propellant) is 500 g. The required DC power for the chamber in which the cool gas generators, the thruster
entire system is 4 W. The flight demonstrated average and the thrust valve are mounted. In 2010, a
specific impulse is 46.7 s, while the maximum measured qualification model of the system has been successfully
in-flight thrust was 35 mN and the minimum impulse bit tested in vacuum at TU Delft14. In this model, a flat
ranged from 0.07 to 0.15 mNs depending on the body Titanium buffer vessel is used as tank as well as
propellant pressure12. The tank volume, estimated from main structure of the system; the tank internal volume is
the sketch shown in Fig. 2, is 28.3 cm3 (cylindrical tank, 28 cm3. Each cool gas generator is loaded with an
3 cm diameter x 4 cm length). amount of usable propellant equal to 0.125 g, producing
0.1 normal liters of gaseous N2. A total of 8 cool gas
generators are present in the model, for a total system
mass of 140 g. The maximum expected operating
pressure of the system is 6 bar, and the operational
thrust level is about 6 mN in the model tested at TU-
Delft (but can be increased up to a theoretical maximum
of 150 mN, by changing the size of the micro machined
nozzle). The reported vacuum specific impulse of the
system is 68 s.
A further development of the cool gas generator
units, presently ongoing at TNO, has allowed for scaling
them up to a mass of 16.3 g, of which 4.21 g are usable
propellant. This has led to a significant improvement in
the performance of the system, and the analysis
presented in the present paper will therefore consider
these updated values.
IAC-12-C4.6.2 Page 3 of 8
63rd International Astronautical Congress, Naples, Italy. Copyright ©2012 by the International Astronautical Federation. All rights reserved.
Fig. 4: A fully packaged MEMS resistojet device (from NANOPS Vacco T3µPS MEMS
Tittu Varghese et al.8). Resistojet
MP [g] for:
∆v = 6.3 m/s 54.09 42.16 37.22 24.38
NANOPS Vacco T3µPS MEMS ∆v = 10 m/s 85.50 66.71 58.92 38.62
Resistojet ∆v = 40 m/s 331.1 260.2 230.5 152.3
Propellant SF6 Iso- N2 N2
butane Table 2: Required propellant mass (with 10% margin)
Isp [s] 46.7 60 68 104 for each DelFFi satellite, for different values of the
pT [bar] 34.5 2.75 N/A 5 total required velocity change, when one of the
pMAX [bar] 34.5 2.75 6 5 candidate micro-propulsion systems is used.
VT [cm3] 28.3 95 28 N/A
T [mN] 35 55 6 0.38 In order to scale the total propulsion system mass
MTOT [g] 500 509 140 N/A starting from the baseline values provided in the
Table 1: Summary of the main characteristics of the previous Section (see Table 1), it is necessary to
candidate micro-propulsion systems. consider not only a different amount of propellant
needed, but also a different size (and mass) of the
IAC-12-C4.6.2 Page 4 of 8
63rd International Astronautical Congress, Naples, Italy. Copyright ©2012 by the International Astronautical Federation. All rights reserved.
propellant storage tank. As a first approximation, the the design of the CGG units, has a lower total mass with
tank can be scaled by using the following simple respect to the other ones for sufficiently low values of
relationship: the required velocity change (∆v < 20 m/s). For the
design ∆v of 6.3 m/s, a total of 9 CGG units are
VT pT
Km = [1] estimated to be necessary. The estimated mass of the
MT Vacco propulsion system is slightly lower than the one
of the Canadian NANOPS due to its higher specific
where Km depends only on the tank geometry, the impulse, despite of the lower propellant density and the
tank material and the safety factor used, but remains higher mass-volume ratio of the tank. The MEMS
constant for all combinations of values of the tank resistojet, in turn, seems to be a better option with
volume VT, the tank pressure pT and the tank mass MT. respect to the NANOPS and the Vacco propulsion
Furthermore, the following additional assumptions system only for values of ∆v lower than 10 m/s.
have been made for the analysis:
• For the NANOPS, a cylindrical stainless steel tank is
assumed, with a burst safety factor of 3.5. The
corresponding value of Km is 8000 m2/s2. The
propellant density is taken equal to 1880 kg/m3.
• For the Vacco micro-propulsion system, the tank is
again assumed to be made of stainless steel but, due
to its non-cylindrical shape, the value of Km is now
4000 m2/s2 for the same safety factor of 3.5. The
propellant density is taken equal to 556 kg/m3.
• For the T3µPS, the only variable mass is assumed to
be the mass of the cool gas generators (CGG); it is
assumed to install the CGG units outside the
propulsion system plenum, the size and mass of
which remains therefore the same whatever the
needed amount of propellant (and, therefore, of
CGG) is. Each CGG unit is loaded with 4.21 g of
usable propellant and has a total mass of 16.3 g and Fig. 6: Required propellant mass for the four candidate
a volume of approximately 16.7 cm3. propulsion systems as a function of the required
velocity change (cyan=NANOPS; blue=Vacco;
• For the MEMS-Resistojet, a cylindrical stainless
green= T3µPS; red=MEMS resistojet).
steel tank is assumed to be used, similarly to the case
of the NANOPS. The corresponding value of Km is
therefore, also in this case, 8000 m2/s2. The
propellant is assumed to be stored in gaseous state at M TOT = A ⋅ ∆v + B
ambient temperature, leading to a density equal to
5.613 kg/m3. In order to achieve a thrust level
comparable to that of the T3µPS, it is assumed to use
16 thruster units in parallel. The mass of some
components (like the valves and pipes for feeding
the propellant to the thruster units) is not precisely
known yet, and is therefore not taken into account in
this analysis; however, this additional mass can be
estimated to be not higher than 100 g.
The results of the analysis are shown in the
following Figures. More in detail, Fig. 6 shows the
required propellant mass (with margin) for the candidate
propulsion systems as a function of the required velocity
change, while the estimated total propulsion system Fig. 7: Estimated total mass for the four candidate
mass is shown in Fig. 7. propulsion systems as a function of the required
Analysis of Fig. 7 clearly shows that the T3µPS velocity change (cyan=NANOPS; blue=Vacco;
system, thanks to the significant latest improvements in green= T3µPS; red=MEMS resistojet).
IAC-12-C4.6.2 Page 5 of 8
63rd International Astronautical Congress, Naples, Italy. Copyright ©2012 by the International Astronautical Federation. All rights reserved.
The curves of Fig. 7 can be approximated as straight In order to overcome this problem and to take
lines, with the fitting coefficients A and B given in advantage of the higher specific impulse obtained by
Table 3. means of the propellant heating, one first option is
represented by increasing the pressure at which the
NANOPS Vacco T3µPS MEMS gaseous propellant is stored in the tank. This is in fact a
Resistojet viable option, because preliminary tests have shown that
Α [g*s/m] 10.1053 7.2692 22.2493 46.0461 the maximum tank pressure can be increased up to at
Β [g] 437.418 451.225 131.33 8.3218 least 20 bar for both the T3µPS and the MEMS
Table 3: Fitting coefficients for the total propulsion resistojet; in addition, as already recalled in the previous
system mass curves of Fig. 7. Sections, no definite requirements are given so far for
the maximum allowed pressure in the QB50 satellites.
Another significant parameter is the total thrust time: To further investigate this option, Fig. 9 shows the
for the design value of the velocity change (6.3 m/s) it estimated tank volume for the MEMS resistojet, as a
can be estimated to be equal to 708 s for the NANOPS, function of the tank pressure, for the design value of the
451 s for the Vacco propulsion system, 4138 s for the velocity change (6.3 m/s). The Figure shows that
T3µPS and 4091 s for the MEMS resistojet. Note, significant improvements are indeed produced by
however, that this last value becomes as high as 65456 s increasing the pressure, but the tank volume is still in
if only one resistojet unit is considered: in general, the the order of magnitude of 103 cc for pressures up to 20
higher thrust time for the T3µPS and the MEMS bar, and it therefore remains not compliant to the QB50
resistojet is due to their significantly lower thrust level. requirements.
The relatively poor performance of the MEMS-
resistojet system shown by Fig. 7 is mainly a
consequence of the assumption of storing the propellant
(N2) in gaseous state at a low pressure and the
consequent very low propellant density. For the same
reason, the estimated tank volume is extremely high if
compared to the other two candidate propulsion systems
where a tank is present, as shown by Fig. 8. The tank
volume required by the MEMS resistojet is
approximately 2 orders of magnitude higher than the
others and, in any case, not compliant to the QB50
requirements. This particular option is therefore not
applicable to the DelFFi project case.
IAC-12-C4.6.2 Page 6 of 8
63rd International Astronautical Congress, Naples, Italy. Copyright ©2012 by the International Astronautical Federation. All rights reserved.
IAC-12-C4.6.2 Page 7 of 8
63rd International Astronautical Congress, Naples, Italy. Copyright ©2012 by the International Astronautical Federation. All rights reserved.
IAC-12-C4.6.2 Page 8 of 8