External sensor for fluid in pipe?

Something that has come up from a couple of guys I chat with for Picaxe applications. Is there a sensor that can be attached to the outside of a line/pipe/tube that will detect the presence or absence of fluid in the tube? Flow rate is a nice bonus, but not required. Relative density might also be nice (one would like to detect various gasses).

An example. Say you have a gas (petrol) tank on a vehicle that has no fuel gage. The tank feeds to a reserve, which, in turn feeds to the engine. Once the primary tank is empty, you would like a warning that you are now on reserve. You do not want to put anything foreign in the fuel line, nor do you want to pierce it, creating a failure point for leaks.

The fuel line could be metal or rubber, I am guessing.

One would think there would be ultrasonic or RF (radar-like), inductance, capacitance, proximity, or other sensors that could accomplish this, but after some "quality" google time, I am at a loss. One of the guys, an engineering type had suggested that the flowing fuel will create a static buildup, and by testing the static voltage in a section, you could tell if fuel was flowing. Stops flowing, you are out.

A strap-on solution would be ideal.

Just dumbfounded here. Any illumination, hints, etc. appreciated.

Wreno

Eesh...I hope theres no static buildup in your fuel lines! They should hopefully be metal as far as ca be, and well bled where they're not. Ok...anywhoo...

I'm sure there may be...but I cant think of it. Well, that's not entirely true. Form a big enough magnetic field around the fuel line and you can force the atoms to align in such a way that you can control the absorption and release of RF energy to a measurable degree.

All you need are some copper-beryllium oxide (I think that's what it is)wire, some liquid helium, and a generator in trailer behind the car to power it all <img src="wink.gif" width=15 height=15 align=middle>.


Ok..in all seriousness...for most atomically inert fuels there is no property that would really allow this to take place. Your best bet is to tap into the pipe/line and use sensors. Dangerous? Yeah...probably.

For gasses, it may be possible to pick up sound as the gas passes through the pipe. I guess it all depends on the setup, though. Without being able to get into the pipe, it makes it REALLY difficult. If you could get in there, there are plenty of ways of gaging not only the flow on and off, but also speed and density, etc.

--Andy P

Edit because &quot;inert fuels&quot; sounded ridiculous.

Edited by - andypro on 24/03/2007 15:45:09

I remember in my old physics college days that we did a bit on ultrasonic Doppler measurements of blood flow. (Details have left my cerebellum). Not quite as hazardous as fuel unless its your own personal blood!
I'd search on ultrasonic fluid flow, but I bet the cost of a ready-made unit will cause an increased flow of blood!

Be interesting if you could make a home-made one, but I suspect it would only measure flow rather than presence/absence.
If you try it, I'd like to see your results.

Something that might be possible on a non-metallic fuel line (most engines have a flexible section somewhere because of vibration):
Place two coils of wire around the non-metallic portion of the line, spaced a couple of inches apart. Drive one with an pulse train. Connect the other coil to a rectifier and filter (possibly with an amplifier) and check the voltage produced with/without fluid in the line. Adjust the pulse frequency and number of turns and the spacing between the coils to get the desired operation. If you have an oscilloscope, you can do most of the testing in a matter of minutes - set up the PICAXE to generate a sweep frequency output (100Hz to 100KHz) and watch the scope.

No guarantees that this will work, but it's possibly a one chip solution.

John

Take a look at Quantam Research Groups capacitive sensor chips. For example:

http://www.qprox.com/downloads/datasheets/qt114_104.pdf

They have lots of good supporting technical data, and are available from Digikey.

I've used some of their chips for galvanicly isolated touch sensors in displays, and fluid level detection and control circuits. If you have nonconductive section of fuel line it should be easy to make a removable &quot;line empty&quot; detecter.

If the containers were not metal - external sensors are used on recreational vehicle water and waste tanks to detect fluid levels.

i can't find any details, but i have definately come across sensors for volumetric flow rate in pipes that do not come into physical contact with the fluid. a friend was trying to explain to me how it works, he was doing some hocus pocus academic research. suffice to say there is a sensor out there.

Making the fuel line &quot;sacred - not to be broken&quot; makes things tricky as it implies one is stuck with the rubber/metal that the line happens to be made of, which could be of variable thickness.

There are lots of sensors that have something spinning and the spinning metal object is detected from the outside eg with a hall effect device. A quick check on Radio Spares searching for 'flow sensor fuel' comes up with a few, so there will be lots out there.

There are other flow sensors out there using magnetically coupled paddles, spinning things, the cooling effect of flow on a small resistor etc. I use several flow sensors to turn off pumps if there is no flow for 5 seconds. They work fine and they don't leak at 20 atmospheres.

Is your question a generic question about flow or a specific one about fuel lines?

The specific application posed to me was a fuel line sensor for a type of antique aircraft. FAA rules apparently prohibit introducing foreign objects into the fuel line (I can see why), thus no turbine or paddle flow meters.

However, it is my understanding that a portion of the line is rubber or plastic, and thus the chip mentioned for an external sensor might well be just the ticket. I will pass it on to him.

Thanks for all the help. I figured it could be done, just the details were a level or two above my current pay grade. Doppler ultrasonic flow meter (were I to find one as apparently cheap and easy) would also work, though knowing the flow rate is not required. Just that there is flow.

Again, thanks.

Wreno

Edited by - Wrenow on 25/03/2007 13:59:46

I have measured fluid flow in plastic pipes with ultrasonics. From memory it was a commercial system squirting U Sonics into the flow and looking for reflected waves frequency shifted by the flow and the pipe was about 50mm dia.

google throws up a number of possibilities.

Edited by - rickharris on 25/03/2007 07:46:25

You may have some success with a capacitive sensor. I think someone mentioned the method in passing earlier in the thread, but here is a link with some details:
http://home.chattanooga.net/~cdp/level/level.htm

It is a method I intend to use to measure varying levels in a container, but this would involve it probe touching the liquid (in my case water). Being capacitive, I see no reason why the capactive plates could not be successfully utilised on the outside of some rubber hose.

If you want to do it on the cheap, use a clear tube for the fuel line and a couple drops of dark food coloring in the fuel.

Then just use a LDR or photodiode to see if the tube is light or dark. You'll probably need an LED to shine trough the tube.

If you are lucky, it might even work with a standard IR photo-interruptor.

Edited by - Larry on 26/03/2007 01:02:54

Edited by - Larry on 26/03/2007 01:03:40

Larry,

No need for food coloring (plus you do not want to intentionally contaminate your avgas). A &quot;clear&quot; tube would probably be relatively easy to cobble with photodetectors and a light source, but I am not sure that &quot;clear&quot; tubing is allowed in the fuel line specs. I am, however, reasonably certain some rubber-like tubing is, as it was mentioned by the guy submitting the puzzler to me.

The capacitance sensor is intriguing, and relatively simple and straightforward if it works in this application. In fact, something like it was gnawing around the corners of my mind. That and some of the little &quot;radar&quot; RF proximity detectors I have heard exist but have never seen...

Wreno

another possibility if you have a flexible section of fuel line is to discern its effective mass by incorporating it in a mechanically vibrating structure and measuring the change in vibration frequency with and without fuel. How discernable this is would depend on the difference of the mass of the line with and without the fuel in it.... There are mass flow meters based on Coriolus forces with a vibrating tube... probably too complex for your use...

My local hobby shop has clear and translucent silicone tubing rated for fuel, so that's not a problem. (I buy it and use it to make air muscles. It works nicely for that use as well).
You may find that just the fuel is enough to change the readings on a photosensor. As to the food coloring, fuel is tinted with organic dies all the time without fear of gumming up the works. in the USA, diferent types of fuel and different Octane ratings are tinted different colors. A couple drops of food coloring won't hurt. Local fliers sometimes tint the different fuel mixtures to easily distinguish them.

If you go the route of sensing the mass/period of vibration of the tube as premelec suggests, accerometer sensors are used in that fashion. Both the ADXL202 sensors and Memsic sensors have datasheets that show how. But you may find that the motor vibration is what you will sense, not the tube vibration. It's a lot higher magnitude.

Is keeping the the fuel line as it really sacrosanct ? I presume that if there is a pipe then there are already two points of potential failure where the pipe joins the tanks. Inserting a sensor into the pipe shouldn't cause any greater problem; I'd have thought it was more a case of doing it properly than being inherently prone to failure. No different to adding an in-line fuel filter really.

The biggest difference is the level of quality/durability required for aircraft use.

Remember that there are required rebuild periods (X thousand hours) for aircraft engines and that maintenance must be done by a *certified* mechanic (engine/airframe). The rules may not be as strict for historical craft, but my guess is that they are still more stringent than &quot;shade tree mechanic&quot; ;-)

John

Wreno,

I breadboarded a QT114 capacitive sensor IC to detect fluid (water) in a length of 1/8&quot; ID plastic tubing. The foil rings around the tubing are about 1&quot; long, with about 1/4&quot; between them. The circuit draws 20 uA with fluid in the tube, and 1.4 mA (due to the lighted LED) when empty. R2 adjusts the detection sensitivity. I posted the schematic is at:
http://i7.tinypic.com/2rxgtgp.jpg
<A href='http://i7.tinypic.com/2rxgtgp.jpg' Target=_Blank>External Web Link</a>

Ken

simple solution .
replace a section of the tubing
with a copper sections and a plastic section in the middle literally wire each copper section to an adc or the picaxe adc pins and use readadc10

avgas does conduct ,given not very well but good enough to detect if there is any in the tube
i just used a similar setup to replace the fuel float in my mini's fuel tank

KMoffett,
Interested in your capacitive sensor.
The application I am looking for is to sense water flowing, as against being stationery. Did you trial this? and does if differentiate between moving &amp; still water?
Thanks
Gonzo

Papaof2 is dead on. The FAA rules are very stringent on what can be added and by whom, and the restrictions on &quot;vintage&quot; aircraft are what we are dealing with. Bottom line, you can probably add something to the outside of the fuel line, but certainly cannot add anything within the fuel flow. The guy who asked set the parameters - he is an A&amp;E (Airframe and Engine) mechanic for the specific type of plane (and one of the foremeost authorities on the Ercoupe).

I like the idea of checking the resistance, as I understand there is a nonmetallic portion connecting two metallic portions already. If it works, easy schmeasy. Though they may share a common ground. I will have to ask him some more.

Thanks again, on both my behalf and his, for all the thoughts.

Wreno

Edited by - Wrenow on 08/04/2007 15:41:18

Summertown,

I don't think the capacitive sensor could distinguish between flowing or stationary fluids. It works on variations in the dielectric constant of fluid vs air. What's you appplication? Pipe/tubing size, flow rates?

Ken

Summertown,

You tweeked my curiosity about flow vs no-flow sensing with my capacitive fluid detector.

See: http://www.builditsolar.com/References/Measurements/DIYFlowMeter.htm
<A href='http://www.builditsolar.com/References/Measurements/DIYFlowMeter.htm' Target=_Blank>External Web Link</a>

If the capacitive detector were attached to one leg of the U-tube manamometer in this flow meter, you could detect flow vs no-flow...or even a specific flow rate by moving the electrodes up or down the manometer. This would not involve a Picaxe, except for outboard control functions.

Also, the thought of using self-heated thermistors in a half-bridge, one in ambient and one in the fluid stream, and fed to a Picaxe ADC10 could detect flow.

I have 2 applications, and the capacative system will be fine for one.

The other involves 50 mm pipes, water that varies from dirty to clean at about 100 litres/min

That one calls for some of Dr_acula's solutions.
Thanks guys