86 The changing accelerometer As the name implies, quantum gravimeters are, in essence, extremely sensitive accelerometers based on quantum technology. These can output phenomenally accurate readings at a low rate; fusing that with a traditional accelerometer allows highbandwidth, high-accuracy acceleration measurements at a rate not yet achieved in any commercially available technology. For reference, today’s high-end commercially available IMUs can take three to five days to drift by 1 nautical mile, whereas one powered by current prototypes of a quantum gravimeter hybridised with a more standard accelerometer takes around six months to experience such drift. Access to such systems is currently available albeit with significant restrictions, but some suggest future colonisations of the Moon and Mars will depend on such quantum inertial sensing technology (particularly if autonomous manufacturing facilities pumping out uncrewed robots and satellites are to pave the way for human settlements, as has been suggested to us in past issues). For such technology to make it to space, however, making IMUs lighter and cheaper comes first, and a critical part of doing that is designing accelerometers with tighter, miniaturised packaging relative to their performance and cost parameters. This is prompting increased usage of quartz in the production of accelerometers. MEMS made from quartz crystals tend to be bigger than those made from silicon, but do not require tuning for bias sensitivity to input range, as they exhibit the same bias parameters and hence similar performance, whether they have been designed for measurement input ranges of 10 g, 20 g, 30 g or more. As previously suggested, larger sensing instruments tend to provide higher performance, hence quartz MEMS can offer slightly better bias stability and other parameters over silicon devices of comparable unit cost. Such are the advantages of quartz MEMS over silicon that some silicon MEMS gyroscope manufacturers view quartz as a possible gyro material of the future and are producing prototypes to that end. Whether of silicon, quartz or another material, it is generally accepted that open-loop accelerometers – in which there is no clear feedback as to whether the output reading has been processed correctly relative to the input value from the sensing element – will be gradually phased out in favour of closed-loop devices. These can use feedback sources such as electrostatic force readings, or more effectively balance the structure of the sensing element around a nominal position to minimise the impact of nonlinearities in the sensor design, and thus achieve more consistently correct data on inertia changes. MEMS production Today’s MEMS devices are usually made via photolithography, in much the same way as most integrated circuits. The technique is often leveraged due to its ability to install extremely fine patterns (a few nanometers in size) with close control of the shape and size of objects created across wafers, as well as being able to do so at high speed and low cost. Such photolithography typically begins with a 100 µm-thick silicon wafer, with a photoresist (a light-sensitive mask) then placed on top of it, for either wet February/March | Uncrewed Systems Technology MEMS made from quartz crystals tend to be bigger than those made from silicon, but do not require tuning for bias sensitivity to input range MEMS devices are typically fabricated similarly to most integrated circuits, and calibration is increasingly automated for both MEMS and FOG IMUs (Image courtesy of Microstrain by Hottinger Bruel & Kjaer)
RkJQdWJsaXNoZXIy MjI2Mzk4