General rightsThis document is made available in accordance with publisher policies. Please cite only the published version using the reference above. Full terms of use are available:

No abstract

This paper describes the design, fabrication and characterization of a ceramic, heated cold-gas microthruster device made with silicon tools and high temperature co-fired ceramic processing. The device contains two opposing thrusters, each with an integrated calorimetric propellant flow sensor and a heater in the stagnation chamber of the nozzle. The exhaust from a thruster was photographed using schlieren imaging to study its behavior and search for leaks. The heater elements were tested under a cyclic thermal load and to the maximum power before failure. The nozzle heater was shown to improve the efficiency of the thruster by 6.9%, from a specific impulse of 66 to 71 s, as calculated from a decrease of the flow rate through the nozzle of 13%, from 44.9 to 39.2 sccm. The sensitivity of the integrated flow sensor was measured to 0.15 m sccm −1 in the region of 0-15 sccm and to 0.04 m sccm −1 above 20 sccm, with a zero-flow sensitivity of 0.27 m sccm −1 . The choice of yttria-stabilized zirconia as a material for the devices makes them robust and capable of surviving temperatures locally exceeding 1000 • C.

Thermal flow sensors have been designed, fabricated, and characterized. All bulk material in these devices is silicon so that they are integratable in silicon-based microsystems. To mitigate heat losses and to allow for use of corrosive gases, the heating and sensing thin film titanium/platinum elements, injecting and extracting heat, respectively, from the flow, are placed outside the channel on top of a membrane consisting of alternating layers of stress-balancing silicon dioxide and silicon nitride. For the fabrication, an unconventional bond surface protection method using sputter-deposited aluminum instead of thermal silicon dioxide is used in the process steps prior to silicon fusion bonding. A method for performing lift-off on top of the transparent membrane was also developed. The sensors, measuring 9.5 × 9.5 mm 2 , are characterized in calorimetric and time-of-flight modes with nitrogen flow rates between 0 sccm and 300 sccm. The maximum calorimetric sensor flow signal and sensitivity are 0.95 mV and 29 μV sccm −1 , respectively, with power consumption less than 40 mW. The time-of-flight mode is found to have a wider detectable flow range compared with calorimetric mode, and the time of flight measured indicates a response time of the sensor in the millisecond range. The design and operation of a sensor with high sensitivity and large flow range are discussed. A key element of this discussion is the configuration of the array of heaters and gauges along the channel to obtain different sensitivities and extend the operational range. This means that the sensor can be tailored to different flow ranges.

Monopropellant ceramic microthrusters with an integrated heater, catalytic bed and two temperature sensors, but of various designs, were manufactured by milling a fluidic channel and chamber, and a nozzle, and screen printing platinum patterns on green tapes of alumina that were stacked and laminated before sintering. In order to increase the surface area of the catalytic bed, the platinum paste was mixed with a sacrificial paste that disappeared during sintering, to leave behind a porous and rough layer. As an early development level in manufacturing robust and high-temperature tolerant microthrusters, the influence of design on the temperature gradients and dry temperature tolerance of the devices was studied. On average, the small reaction chambers showed a more than 1.5 times higher dry temperature tolerance (in centigrade) compared to devices with larger chambers, independent of the heater and device size. However, for a given temperature, big devices consumed on average 2.9 times more power than the small ones. It was also found that over the same area and under the same heating conditions, devices with small chambers were subjected to approximately 40% smaller temperature differences. A pressure test done on two small devices with small chambers revealed that pressures of at least 26.3 bar could be tolerated. Above this pressure, the interfaces failed but the devices were not damaged. To investigate the cooling effect of the micropropellant, the endurance of a full thruster was also studied under wet testing where it was fed with 31 wt.% hydrogen peroxide. The thruster demonstrated complete evaporation and/or full decomposition at a power above 3.7 W for a propellant flow of 50 µl min−1. At this power, the catalytic bed locally reached a temperature of 147 °C. The component was successfully heated to an operating temperature of 307 °C, where it cracked. Under these firing conditions, and assuming complete decomposition, calculations give a thrust and specific impulse of 0.96 mN and 106 s, respectively. In the case of evaporation, the corresponding values are calculated to be 0.84 mN and 92 s.