Sound Fuel Gauge?

Interesting, I would have thought that maybe the environment would be too noisey for an acoustic transducer, but maybe its possible.....

I thought most aircraft fuel tanks were monitored by capacitive sensors? These can be made in various shapes which together with baffle plates to prevent fuel sloshing around would be suitable??

John.

I too have often pondered this problem of measuring the fuel remaining in a flat shallow horizontal tank. I think in aircraft wing tanks they measure levels at multiple points and a microcomputer works out the quantity of fuel in spite of the changing pitch of the craft. But I am not fully convinced that this would give a very accurate value. May be the best idea is to transfer fuel into and out of a tank via a very good positive displacement type of flow meter so that an accurate account is possible.

Would it be better to accurately meter the fuel in and out? You can get some very good flow gauges, and given the problems inherent in measuring the volume of an active fluid in an irregular shaped tank I would imagine it would be a more reliable method. After all, this is how we pay for our natural gas and gasoline.

it is a very good Idea. But, as pointed out by other it would probably be better to use a flow meter so that you could more accurately guage what goes out. If you know what went in then its just a matter of math. I do wonder though where else could this be applied? I seems sound enough of prinicple.

I think the main reason for not using flow meters would be if there is a leak in the fuel tank. Flow meters are used for consumption. Remaining available fuel is a concern if there are possibilities of fuel tank damage. I do think this idea has merit.

Hahahaha very good point 'Guest'........

I think its good to know how much has gone in.. and how much has gone out but as the previous poster said it doesn't really tell you how much is left.....

Does it?? John.

Also if you are using fuel injection, some of the fuel is returned to the tank. Of course you could measure the return fuel too. This would not take into account tank leaks and malfunctions that may blow fuel out a vent tube. Most pilots use the flow meters for mixture adjustments.

I hadn't thought of leaks!

However, I think it is highly desirable NOT to have a leak on an aircraft, (remember Concord?) and that an exact figure for fuel left in the tanks would be of secondary importance.

Actually, there should be little effect on fuel level due to banking. Unless the plane is flown in an uncoordinated matter, the fuel will remain level in the tank, just as a passenger does not lean from side-to-side relative to the plane. Pitch and acceleration can affect local levels, though.

In practice, one does not trust fuel level gauges in aircraft. This is true both in very simple general aviation aircraft (where one monitors fuel usage based on engine rpm and manifold pressure, fuel flow meters usually being entirely absent), and in most jets, where fuel flow is measured accurately and "totalized". Where a totalizer is present, a pilot will always trust it before trusting the level gauge. I don't know that you could change that perception.

Given the variable nature of the noise generated by the aircraft itself, calibration for different power levels, gear and flap positions, etc. would be time consuming. But it is an interesting concept, and certainly worth experimenting with, on a small aircraft to start. I'd think it might be quite difficult to improve upon ultrasonic level sensors, though.

Thanks Ken, you made quite a few valid points. Since I am building the aircraft using composite materials, I am aware of the many different problems encountered while sorting out sound patterns and sources. These planes use baffles in the tanks, they also have other fuel system options such as auxiliary tanks that can be pumped in to or out of, and fuel injection return lines keeping levels in flux. Usually the need to know fuel levels are at the beginning and end of the flight. The stance of these airplanes changes when parking (nose down) so the vertical guage is changed.

So the normal guages aren't going to be as accurate, they could probably be altered. I just figured that acoustics in an enclosed area is a good place to start for finding levels in liquids where the container moves around.

I may have misunderstood your intent. I thought you were proposing this as an alternative to the status quo in commercial aircraft, where fuel level monitoring is something of a non-issue, in that there are systems in place (and backups for those systems) that work well. It sounds as if you are actually the one doing the building of this composite aircraft, in which case you will have considerable freedom to try new systems, without having to go through all the paper work for certifications, STCs and so forth. As long as you can convince your local inspector that what you are doing is not clearly unsafe, then you should be able to use whatever type of level monitoring you desire. My personal favorite is the cork mounted in the tank of a J3 Piper Cub. The "gauge" was a bent piece of wire, outside the cockpit, on the other side of the windscreen from the pilot and directly in front him/her.

I like your idea. I wonder if it might be better to send out a single frequency though, and to use the technology employed in noise-canceling headsets (like Bose) to reject all the other frequencies. (Or maybe that's overkill -- maybe simply filtering out all but the frequency of interest would be simpler?) It seems that correlating the sound signatures to specific operating conditions might require a lot of processing power. In other words, a certain fuel level might have a great many sound signatures, depending upon whether the engine is running or not, throttle settings, aircraft configurations, etc. I'd guess that looking at a single frequency might reduce the ambiguities somewhat. The harmonics of the base frequency would change with empty space in the tank, I'd think, so you'd analyze at just a few harmonic frequencies. (I'm not an acoustics engineer, so I'm guessing here.)

Are you building your own design, someone else's design, or a kit? Many years ago, I owned a BD5 kit. I was the third owner, I think, and had planned to build new wings and use a reliable engine (by the time I bought it there were several alternatives available). But like owners one and two, I could never find the time and funds: when I had the time, I didn't have the funds; when I had the funds, I didn't have the time.

BTW, I'd be pretty sure you are aware of this, but you will want to be sure that your transducer is intrinsically safe even if remotely connected through tubing, so that if there is a failure such as a short circuit, things don't go boom.

As far as commercial/certified aircraft is concerned, status quo is fine and good enough for "government work" and for the process involved, would be better left alone. Change in certified systems, parts and just about anything else carries such a high price tag, that it is no wonder we are flying in aircraft little changed (at least in GA) in 50 years! I am attempting a CozyIV, a Rutan derivitave plans built pusher. There are several of these being built and there are several systems recommended but are approximations at best.

The reason I felt pink noise would be a good signal injection and detection is because:

1. An audio spectrum analyzer (freeware available on the internet) can be used easily and readily converted to indicate a defined range.

2. Pink noise generator can be easily applied.

3. Chamber effect of sound waves in an enclosed area add gain to the sound injected and reduces other noises effect on the chamber.

4. The sound deadening effect of the fiberglass composite structure in the strakes area also minimizes the opportunity for interference from audible frequencies.

5. As was already pointed out, the fuel doesn't really slide to one side or another, especially with the baffles, but with a fuel injection return into one tank or another, and/or pumping in and out of different tanks, a current situ indication would be nice at-a-glance.

I spent my early years of student flying in a J-3 Cub out of Bob Shank Field in Indianapolis $12 an hour dual/wet (Shank field has been gone for 30 years!) that "coat hangar" wire signalled the end of lessons and time to back to the field!

What happened to the BD5?

All good points. I suppose, also, that if you pulsed the sound you could continuously recalibrate for changing background noise.

I'm envious -- the Cozy is a pretty ambitious project. The first flight will be a little tense, no doubt, but what great satisfaction!

The BD5 was sold by my first wife as part of a divorce settlement. There was someone in her office who was interested, and she pocketed the money -- which at the time was fine with me, figuring we'd balance things out in the end. As it happens, her lawyer was a sleeze bag, and did everything possible to make the divorce contentious. So just a few weeks after the plane was sold, I recieved a summons to appear in court to explained why I had "secreted" the money from the plane sale!??! That was about 20 years ago... time wounds all heals.

Wow, $12/hr! When I was learning to fly (1979), I flew a Tomahawk, and it was $38/hr wet. One of my most memorable flights was in a J3 from Pittsburgh PA to Melbourne FL. 2 entire days, dawn to dusk.

Assuming that the holes in the tank baffles are big enough to allow fuel to flow at a rate slightly more than the engine(s) need it (or even more maybe...) and that the baffling compartments are kept as small as possible - then how about 4 ultrasonic Tansmitter / Receivers in a diamond pattern, mounted in the tank "Ceiling" and as far apart as possible. Each must be able to measure the depth (or height - depending upon your perception) of the fuel at that point.

Fill the tank carefully with 1 gallon (could be clean water at this moment!) say at a time (or in pounds for aircraft I believe!) and note the outputs of all four sensors in a table, not just in a "flat" flying condition, but also when banking (simulated of course!) left and right, diving and climbing, till the tank is full. I would say at least 5 values for each gallon of fuel - Flying Flat, 15° left, 15° right, 30° up and 30° down..... I am sure that a small (Basic Stamp?) computer, programmed to read these sensors and having the table you have made would then be able to calculate the present contents to within a very small margin of error! At least near enough for government work! The only problem I see is if you want to fly upside down for extended periods!!!!