Gauges that would measure liquid levels optically have been proposed for use in aircraft fuel tanks. These gauges would contain no moving parts (no floats) and no wiring inside the tanks. Their overall function could be characterized as that of permanently immersed, self-reading dipsticks.
The proposed gauges are intended to supplant the capacitance probes now used to measure liquid-fuel levels in such tanks. Capacitance probes are mounted at several locations inside a tank and are connected to external instrumentation via wiring. The probes and wiring are usually reliable, but fail occasionally. Because replacement of capacitance probes and/or wiring involves intrusion into the tank, the aircraft could be out of service for days.
In a gauge of the proposed type, the only part intruding into the tank would be a rodlike assembly, mounted from the outside of the tank, that would provide optical access to the liquid inside. The rodlike assembly would include a baffle plus a rod made of a suitable transparent material. The rod would be etched or scored at prescribed intervals along its length to provide optically reflective fiducial marks at known levels. Light would be coupled into the rod from a source at the outer end to illuminate the fiducial marks. A camera or other imaging device would be mounted adjacent to the source of light and would be aimed along the rod to observe the illuminated marks.
The rod material would be chosen so that its index of refraction would approximately match that of the liquid in the tank. As a result, the fiducial marks immersed in the liquid would appear dark to the imaging device, while those above the surface of the liquid would appear bright to the imaging device. The liquid level would thus be assumed to lie between the lowest bright mark and the dark mark just below it. The output of the imaging device would be processed to into an indication of the liquid level in increments of depth between fiducial marks.
A mass-produced gauge of this type would likely include a miniature imaging device containing an active-pixel sensor, plus input/output circuits, all integrated on a single chip. An application-specific integrated circuit (ASIC) for processing the image-sensor output could also be included. Clock and command signals and signal input voltage would be supplied to the chip from external instrumentation. The overall size of the unit on the outer end of the rod assembly (including the ASIC) would be of the order of 1 in.3 (=16 cm3).
In a typical case, it would be necessary to place gauges at several locations. Then the fuel-level readings from the several locations could be processed by an algorithm that would take account of the shape of the tank in determining the amount of fuel remaining. It should also be possible to implement some form of autocalibration in software. The level readings or the final calculated quantity of fuel could be integrated or averaged before being displayed in nearly real time (update every few seconds).
With respect to initial costs, the proposed gauges would be competitive with capacitive fuel gauges. However, recurring costs of the proposed gauges would be much lower because their rodlike assemblies could be replaced in minutes instead of days.
This work was done by Philip Moynihan, Paul Henry, Tien-Hsin Chao, William Lincoln, William King, and Lloyd Adams of Caltech for NASA's Jet Propulsion Laboratory.