Accelerometers of various types and sizes have become essential to modern life, from sensors that inflate automobile airbags to devices that measure vibration and shock in machinery to inertial navigation for spaceflight, missiles, and aircraft, among many others. Not surprisingly, demand is rising for inexpensive, high-precision instruments that can be embedded in ever-smaller dimensions.
That is why NIST researchers have developed and are testing an optomechanical accelerometer only one-quarter of a millimeter thick designed to deliver measurements directly traceable to the SI with uncertainties better than 1 part per 1000 – as good as any laboratory acceleration device in the world.
Accelerometers typically function by measuring the motion of a “proof mass” with respect to a fixed local reference. For example, the mass might move to compress a piezoelectric material, generating a current, or it might stretch a sheet of insulator so that its electrical resistance increases.
The NIST instrument consists of a tiny Fabry-Perot optical cavity made of silicon. At one end of the cavity is a fixed hemispherical mirror about 300 micrometers (µm) deep, 500 µm wide, and with a surface smoothness that varies by no more than 1 nanometer. At the other end, about 250 µm away, is a mirror-coated mass that is suspended in the cavity by flexible supports: silicon nitride “beams” that are 20 µm wide, 1.5 µm thick, and 40 to 100 µm long. The length of the beams, together with the dimensions of the proof mass, determine the fundamental resonant frequency of the structure.
A tunable diode laser beam is sent into the cavity and is reflected between the two mirrored surfaces. A single optical fiber carries both the incoming and reflected light. When the device is at rest or moving at a constant velocity, the cavity length remains constant. When the device is accelerated, however, the suspended mass is displaced either toward or away from the fixed mirror, and the distance between the two surfaces changes. That change in cavity length changes the resonant frequency, which is detected by the laser beam, and the magnitude of the change is a measure of the acceleration.
Because the device relies on fundamental principles of optics, it is intrinsically self-calibrating. And because the components are the same size as those routinely produced in microelectronics or MEMS fabrication, the eventual production cost of a complete unit about 1 cm3 should be low. But before then, the NIST scientists will have to overcome a number of obstacles.
One is the demanding time scale involved. As the cavity dimensions change, the tunable laser will have no more than about 100 microseconds to scan through a range of frequencies before the on-board feedback devices locks it onto the new resonant frequency. Finding an inexpensive laser with those capabilities is another challenge. So is making a robust optical fiber connection to a device that vibrates at 1000 cycles per second – and eventually perhaps 10 times faster.
Finally, sudden heating caused by the motion of the mass might affect the measurements because change in temperature alters the cavity dimensions. Interestingly, the scientists think that by incorporating another NIST-on-a-Chip device – a photonic thermometer – they may be able to correct for heat effects in real time.