Minimize PRECISE

PRECISE (chemical-µPRopulsion for an Efficient and accurate Control of Satellites for Space Exploration)

PRECISE is a European program, a dedicated approach to design and develop a MEMS-based monopropellant µCPS (micro Chemical Propulsion System), an enabling technology and in particular, a necessity for the realization of future distributed space architectures, consisting of a network of formation-flying entities in support of a common goal. 1) 2)

The µCPS development roadmap was already elaborated by ESA in the European Space Technology Harmonization on Chemical Propulsion and Micro-Propulsion in 2008, with main emphasis on the harmonization of European activities under the technical coordination of an industrial partner. On this basis, the setup of a competence network to advance the development of µCPS in Europe was started. Compared with the current technology of a conventional single monolithic spacecraft, the realization of distributed systems of several smaller spacecraft entails obvious benefits like increasing reliability, flexibility, low cost solutions and a reduced development time.

The objective of the PRECISE project is the development of micropropulsion thrusters for high precision attitude control and precise approach maneuvers. PRECISE combines European capabilities and know-how from universities, research organizations, experienced European companies and a Russian company for the research and development of a µCPS (micro Chemical Propulsion System) for the future market demands. PRECISE provides a stepping stone along the ESA-µCPS roadmap. Basic research will be conducted aiming at improving crucial MEMS technologies required for µCPS. 3)

The EC (European Commission) is funding the PRECISE program within the FP7-Space (Seventh Framework Program). The PRECISE consortium consists of the following project partners: DLR (German Aerospace Center), AST (EADS Astrium Space Transportation), Germany, CNRS (Centre National de la Recherche Scientifique), France, NSP (NanoSpace AB), Sweden, SSC (Surrey Space Center), University of Surrey, UK, LCO (University of Poitiers), France, UTW (University of Twente), The Netherlands, NPO Mashinostroyenia, Moscow region, Russia. DLR and AST are the coordinators of the program.

The availability of µCPS forms the basis for defining new mission concepts such as formation flying, advanced robotic missions and rendezvous maneuvers. These concepts require propulsion systems for precise attitude and orbit control maneuverability. Basic research will be conducted aiming at improving crucial MEMS technologies required for the propulsion system. Research and development will also focus on the efficiency and reliability of critical system components. System analysis tools will be enhanced to complement the development stages. Finally, the µCPS will be tested in a simulated space vacuum environment. These experiments will deliver data for the validation of the numerical models. In addition, application-oriented aspects will be addressed by two endusers who are planning a formation flying mission for which the propulsion system is crucial. 4) 5)

Overview of the model mission:

The University of Surrey and NPO Mashinostroyenia complete the concept of PRECISE with their collaborative formation flight mission. For the realization of this mission, the availability of a µCPS is crucial. The mission consists of two small satellites, where one spacecraft carries a solar sail, referred to as the SSSC (Solar Sail Spacecraft), and the other, named InspectorSat, is orbiting the SSSC for observation and inspection, as illustrated in Figure 1. 6)


Figure 1: Schematic view of the solar sail spacecraft and the InspectorSat (image credit: PRECISE consortium)

The µCPS will be used for all stages of the mission, this includes attitude maneuvers including detumbling after launch vehicle separation, support for rendezvous and orbital maintenance maneuvers; thus,demonstrating long thruster burns in addition to the shorter bursts required for attitude control. Finally, analysis of orbit maintenance and drift compensation maneuvers will allow the SRP (Solar Radiation Pressure) on the solar sail spacecraft to be determined.

Studies have shown, that a total ΔV budget of ~27m/s for the whole mission results in a propellant requirement, which is feasible for the small spacecraft, nominally of up to 50 kg in mass.The thrust of µCPS is ~10 mN with a minimum specific impulse of 180 s. An analyses indicated, that the use of the µCPS thruster will provide an accurate, precise AOCS actuator with performance characteristics which are ideal for the on-orbit examination of co-orbital targets.

Historically, such manoeuvres have been performed by various macroscopic thruster technologies, including electrostatics and electromagnetics, arcjets, resistojets, monopropellants, liquid and solid chemical propellants and cold gas jets. The relative performance of these technologies is shown in Figure 2, illustrating how monopropellant thrusters fill a performance gap between cold gas and liquid/solid chemical thrusters. A monopropellant thruster, therefore, provides the mechanical simplicity of a cold gas system while allowing the utilisation of the propellant chemical energy through a catalytic breakdown mechanism.


Figure 2: Comparison of thruster technology performance - exhaust velocities as a function of typical vehicle accelerations (image credit: George P. Sutton, Oscar Biblarz) 7)



The µCPS concept overview:

The term chemical micropropulsion is used for propulsion systems with thrust levels in the order of microNewton (µN) up to several milliNewton (mN) and generating thrust by means of chemical energy of the propellant. The primary objective is the development of a µCPS necessary for highly accurate attitude control maneuvers of satellites as they were not feasible until today with propulsion systems of this size. The tasks are performed under stringent quality assurance aspects. Further aspects need to be considered for a consistent development, these are (Ref. 4):

- Definition of requirements and specifications, comprising S/C demands

- Research of propulsion aspects like catalysis

- Development of crucial components

- Test facility infrastructure and diagnostics

- Numerical development, simulation and comparison

- Manufacturing, assembly, integration and testing of the µCPS

A revolutionary feature of µCPS using MEMS technologies is the very compact, low mass and modular architecture. The micro thruster has a mass of only a few gram, it is etched on a silicon wafer as illustrated in Figure 3.


Figure 3: Illustration of a µthruster prototype in comparison to a one cent coin (image credit: image credit: PRECISE consortium)

µValve component development:

The thorough development of single components is an important pillar in the development of efficient and reliable µCPS. The µCPS component development within PRECISE has two main objectives:

• the first is to develop the individual components like the µValve, the µHeater and an actuator for propellant injection

• the second is to develop the manufacturing methods and interfaces to facilitate a high level of integration for the complete µCPS.

The Swedish consortium partner NanoSpace is responsible for the µValve and µHeater component development within PRECISE. Various micro fluid control devices such as proportional flow control valves, thermal flow controllers, isolation- and pressure relief valves, filters, heaters and control orifices have already been developed as single components.

The selected valve concept is based on the PCM (Phase Change Material) actuation principle. An enclosed cavity is filled with PCM which increases in volume when it goes from solid to liquid phase. The phase change is achieved by resistive heating and the melting point can be chosen to fit the thermal requirements of the system. The PCM actuation mechanism is evidently a technique that can provide both large forces (>1 N) and displacements (>100 µm). The flow is modulated proportionally by adjusting the applied power to heaters in a PCM cavity, which actuates on the valve seat due to the expansion. The valves are fabricated using four fusion bonded silicon chips and each valve stack has four integrated PCM-actuated normally closed valves. The valve interfaces are designed for bonding towards the nozzle package and electrically to the pod assembly connections.

The valve seat package consists of two fusion bonded square chips with the purpose to interface the nozzle package and also act as a normally closed valve seat. The valve seat chips also include holes and channels to lead gas from the valves to the nozzles. The inlet and outlets on the valve seat package are shown in Figure 4.


Figure 4: The valve seat package, 22 mm in side, seen from the nozzle side (image credit: PRECISE consortium)

The actuator package is a two-wafer stack with cavities for the PCM and internal heaters as essential integral parts. The cavities are formed at the interface between the two actuator chips, furthest down in the flow control valve stack. Paraffin was finally chosen as the phase change material. The filling channels and the internal heaters are found in the bottom chip. The paraffin heaters are mechanically fixed and electrically connected to the thin film conductors on the chip surface by soldering. When the paraffin melts through resistive heating of the PCM-heaters, it expands and forces the rigid center diaphragm of the valve seat package to deflect. The manufactured prototype chip is shown in Figure 5.


Figure 5: µValve chip seen from the backside to the left and top side to the right (image credit: PRECISE consortium)

µHeater component development:

The implementation of a µHeater to heat the µChamber is investigated to improve the performance during the start phases of a micro thruster at low operational temperatures. The microheaters could also be categorized in different ways, and one way is whether the heater is applied on the external surface of the MEMS component or inside the actual flow path or between component layers. The latter is referred to as internal heaters. External heaters might reduce the complexity of the interfaces and hence, might reduce the technical risks involved.

The primary selection criteria for µHeaters are actually not the melting point of the resistor material. It is again the compatibility with the rest of the fabrication steps. A very common MEMS manufacturing step is oxidation or annealing at T >1000°C in equipment where metal is not allowed. Hence the resistors need to be applied subsequently to any oxidation/annealing step. For this reason an external micro heater is preferred as a baseline and starting point. Ti/Cu is selected as a baseline metal stack, where Ti will act as adhesion and diffusion layer and Cu as conducting layer.

The design of the thin film heaters is schematically shown Figure 6. They are shaped by evaporation or by sputtering through a shadow mask. The shadow masks are made from 500 µm thick silicon wafers. The manufactured prototype chip is shown in Figure 7.


Figure 6: CAD drawing showing the shadow mask design of the thin film heaters evaporated or sputtered onto a silicon substrate (image credit: PRECISE consortium)


Figure 7: Photo of µHeater prototype chips (image credit: PRECISE consortium)


MEMS-based Coriolis mass flow sensors:

To ensure the reliable characterization of micropropulsion systems generating the required ultra low thrust levels and impulse bits, it is necessary to accurately measure and control the propellant mass flow. Therefore, a fluid channel technology is needed that is compatible with the used propellant hydrazine and that allows the integration of flow sensors in these channels. A suitable candidate is the so-called surface channel technology, which was developed at the University of Twente and results in channels with an almost circular cross-section and a channel wall of silicon nitride. The nearly circular cross-section results in low pressure drops. Furthermore, the thin silicon nitride wall (1 µm) allows integration of sensor and actuator structures very close to the fluid but without actual contact to the fluid. Figure 8 shows a cross-section of a fabricated channel.


Figure 8: SEM photo of a fabricated surface channel with a diameter of ~20 µm and a silicon nitride wall thickness of 1 µm (image credit: PRECISE consortium)

For the measurement of the propellant mass flow, a most interesting solution is the so-called Coriolis flow sensor. An important advantage of Coriolis sensors is their sensitivity to the true mass flow, independent of the flow profile, pressure, temperature and the properties of the fluid (density, viscosity, etc.). Figure 9 shows a SEM (Scanning Electron Microscopy) photo of a Coriolis sensor prototype, based on a rectangular tube shape. The tube is actuated in a torsion mode. The Coriolis force induces a “flapping mode” vibration with an amplitude proportional to the true mass flow.


Figure 9: SEM photo of the micro Coriolis flow sensor fabricated at University of Twente. Right: Measurement results for three different fluids: water, ethanol and white gas (image credit: PRECISE consortium)

For PRECISE, the derived required flow range of 6 mg/s is about a factor 30 higher than presented in the designs above. Therefore, for the redesign, the proven basic sensor element will be used combined with an on-chip bypass to reduce the sensitivity. Reduction factors ranging from 20 to 100 are designed, so that the least sensitive devices will be suitable for use with four thrusters together. Figure10 shows the mask design of a typical device. In this case, the calorimetric flow sensors are integrated on the same chip.

For a calorimetric flow sensor, the required flow range of 6 mg/s is extremely large. A design was made that could be integrated on the same chip as the Coriolis sensors. In this way the calorimetric sensor can provide a reliable signal in the lower part of the measurement range. A measurement up to the full range of 6 mg/s will not be possible using the relatively standard implementation with a Wheatstone bridge. Therefore, instead of a Wheatstone bridge, all heating and sensing resistors can be connected separately, to allow electronic feedback with constant temperature operation. First wafers with both test structures and functional devices were completed. Preliminary test runs revealed ,that the bypass tubes function as expected. Filling the tubes with water was easier than expected, and the estimated reduction in sensitivity is very close to the designed value.


Figure 10: Mask design for the combined micro Coriolis and calorimetric flow sensor to be used in PRECISE (image credit: PRECISE consortium)


MEMS-based plume measurement sensor:

For the plume analysis, which will be conducted during the final hot firings, it is not possible to use a standard type flow sensor, since the flow cannot be confined in a tube. Therefore, a Pitot tube sensor is designed and realized for this purpose. The same basic fabrication technology will be used, although an adaptation is necessary to allow for an open tube end. Various readout mechanisms (e.g. capacitive, thermal) are considered to detect small pressure variations. A design of a Pitot tube sensor has been made, having a tube diameter of 40 µm based on DSMC-CFD calculations conducted by DLR. These calculations indicate a pressure range up to about 400 mbar. Such pressures can be measured using a deflecting membrane with capacitive readout. Figure 11 shows a CAD drawing of the sensor design. The tube is fabricated using the same techniques as the ones used for the Coriolis type flow sensor.


Figure 11: CAD model of the Pitot tube sensor (image credit: PRECISE consortium)

The chip (excluding tube) measures 4 mm x 2.6 mm and includes 4-point contacts for the on-tube thermistors, a reference tube for the reference pressure sensor and the pressure sensors itself. For the high pressure range of 400 mbar a pressure sensor design was made, making use of the microchannels to develop a membrane, which can deflect due to the applied pressure. The capacitance is measured between the comb-finger electrode structures, giving higher sensitivity, when measuring capacitance relative to the bulk Si wafer. The Si wafer in this design can be used as electrical guard. Down-scaling a pressure sensor results in a membrane, that is comparatively stiff, and the 400 mbar pressure sensor, based on microchannels, cannot be scaled easily to measure with 1% accuracy on a full scale of 1 Pa. Thus, in the final design, Pirani type pressure sensors were included to accommodate the lowest measurement range.

The Pitot tube chip design was first evaluated using FEM simulations. Processing in the MESA+ cleanroom started in August 2012, and the first fabricated devices were available in January 2013. Figure 12 shows a photo of a completed chip.


Figure 12: Photo of the Pitot tube sensor with a 5 cent coin as size reference (image credit: PRECISE consortium)


µCatalyst development:

The state-of-the-art techniques of catalyst manufacturing, as well as currently used configurations, are not suitable to micropropulsion applications. Classical systems typically use 3D components in contrast to the 2D chip design of the proposed µCPS unit. When moving to a MEMS-based hydrazine µCPS, the size of the decomposition chamber and feeding part has to be decreased by at least one order of magnitude. Hydrazine decomposition initiation is a key point for the operation of the µCatalyst. Various possibilities have to be taken into consideration:

- heat up of liquid hydrazine

- direct vaporisation of gaseous hydrazine in the decomposition chamber

- heat up and catalytic/thermal decomposition of gaseous hydrazine.

During the hydrazine decomposition process, hydrazine is decomposed into ammonia and nitrogen, followed by a further decomposition of ammonia into hydrogen and nitrogen. Thus, it can basically be distinguished between a catalytic decomposition, a thermal decomposition, or a mixture of the processes. Thermal decomposition of liquid hydrazine alone occurs at much higher temperatures than a catalytic decomposition. This requires more energy input compared to a catalytic decomposition, since the liquid hydrazine must be pre-heated.

To minimize the energy, catalytic decomposition appears to be a better choice. However, from a technological point of view, this process implies several coatings and thus a challenging manufacturing process. On the other hand, high initial temperatures can be avoided, from which the overall system might benefit and the efficiency of the catalyst is increased.

To maximize the reacting catalytic surface area, the µCatalyst bed unit might be designed with either micro pillars or microchannels. The residual stresses and thus the stability of the final µCatalyst is highly dependent on the materials used for the coatings and on their deposition method. The coatings have to be deposited in different layers, e.g. catalyst and catalyst supporting layer, protective and adhesion layer. The PRECISE consortium partner, IC2MP (Poitiers Institute of Chemistry, Materials and Natural Resources) of CNRS, is investigating and testing various concepts, to identify the best procedure, material and deposition method to produce the µCatalyst by ensuring both mechanical and thermal stabilities.

Two different µCatalyst supports have been used. Flat silicon support, with or without a silicon oxide (SiO2) layer, and a metallic alloy with microchannel patterning. An alumina washcoat has been developed on these supports by a sol-gel process, prior to iridium active phase deposition.

Two different alumina preparations have been established from boehmite (or pseudoboehmite) synthesis, followed by thermal treatments to induce the boehmite to alumina phase transition. The setup used to synthesize the boehmite is presented in Figure 12. Two methods have been performed to deposit the pseudoboehmite on the supports: a) by a droplets deposition, and b) by immersing the sample for a specific time period into the solution. The first method has the advantage to control precisely the pseudoboehmite quantity deposited onto the support. The second one leads to a better homogeneity of the boehmite washcoat onto the different supports.


Figure 13: Photo of the alumina preparation setup (image credit: PRECISE consortium)

The iridium active phase has then been deposited on these supports by wet impregnation CD (Chemical Deposition) or magnetron sputtering PD (Physical Deposition). In the wet impregnation process, the active metal iridium-based precursor, namely hexachloroiridic acid, was dissolved in an aqueous solution. The amount of precursor salt to be used was determined relatively to the porous alumina quantity and then added to the µCatalyst support in a reactor. Calcination and reduction of the iridium-based precursor are performed with an appropriate heat treatment, to obtain pure iridium nanoparticles onto the support.

In the magnetron sputtering process, used to perform the physical deposition, iridium particles are pulled from an iridium target to the support by a ballistic effect. This technique allows an accurate control of the growing process of iridium nanoparticles (size and shape) onto the µCatalyst with high adhesion energy. The obtained µCatalysts present a good iridium adhesion onto the different supports (mechanical stability). Prototypes were characterized by AFM (Atomic Force Microscopy)and SEM (Scanning Force Mictoscopy) processes to determine the iridium nanoclusters dimensions and dispersion.

Finally, droplets of pure hydrazine have been injected on the different prototypes at room temperature and pressure to highlight the catalytic activity. The metallic alloy prototype, including an alumina washcoat and a 4 nm iridium layer obtained by magnetron sputtering, seems to be a promising candidate. The higher catalytic activity, based on temperature measurements close to the µCatalyst chip surface, has been found for the flat silicon support with 4 nm iridium layer (PD). However, after liquid hydrazine decomposition, a sintering of the iridium active phase is suspected, as observed by optical microscopy. This could lead to a loss of catalytic activity of the prototype. This kind of prototype should be tested with gaseous hydrazine. The preliminary decomposition tests are still ongoing.


General overview of the program:

• The PRECISE program started officially on Feb. 1, 2012 for a duration of 24 months.

• The project is conducting system tests during the 2nd half of 2013.

• In Feb. 2014, a prototype µCPS device is expected capable of demonstrating the basic operational functions of the system.

• After approval from the funding agencies, the FM (Flight module) of µCPS must be built and tested.

• This may be followed with a formation flight mission. The projects InspectorSat of the University of Surrey, and SSSC of NPO Mashinostroyenia, are in the definition phase as of 2013.


2) “PRECISE — Chemical-µPropulsion for an Efficient and Accurate Control of Satellites for Space Exploration,” URL:

3) “Chemical-µPropulsion for an Efficient and Accurate Control of Satellites for Space Exploration,” CORDIS(Community Research and Development Information Service), Feb. 1, 2012, URL:

4) Markus Gauer, Dimitri Telitschkin, Ulrich Gotzig, Klaus Hannemann, Yann Batonneau, Pelle Rangsten, Mikhail Ivanov, Christopher Brunskill, Remco J. Wiegerink, “PRECISE - preliminary results of the MEMS-based µCPS,” 49th AIAA/ASME/SAE/ASEE Joint PropulsionConference and Exhibit, San Jose, CA, USA, July 14-17, 2013

5) M. Gauer, D. Telitschkin, U. Gotzig, Y. Batonneau, H. Johansson, M. Ivanov, P. Palmer, R. Wiegerink, “PRECISE - development of a MEMSbased monopropellant micro Chemical Propulsion System,” 48th AIAA/ASME/SAE/SAE/ASEE Joint Propulsion Conference & Exhibit, 29. July - 01. Aug. 2012, Atlanta, Georgia, USA

6) Christopher Brunskill, Phil Palmer, “The InspectorSat mission, operations and testbed for relative motion simulation,” Proceedings of the 5th International Conference on Spacecraft Formation Flying Missions and Technologies (SFFMT), Munich, Germany, May 29-31, 2013, URL of paper:, URL of presentation:

7) George P. Sutton, Oscar Biblarz, “Rocket Propulsion Elements,” Wiley & Sons, 7th edition, 2011, p. 41 URL:

The information compiled and edited in this article was provided by Herbert J. Kramer from his documentation of: ”Observation of the Earth and Its Environment: Survey of Missions and Sensors” (Springer Verlag) as well as many other sources after the publication of the 4th edition in 2002. - Comments and corrections to this article are always welcome for further updates.