Propane Sensor Fuel Metering

[MrCrowley]

Hey guys, long time no see! Hope the New Year is treating you all well.

I've become slightly addicted to YouTuber AvE and his videos have spurred some creation inside of me (at least that's what I hope it is). I feel like getting stuck in with an Arduino project and was trying to think of ideas. It got me thinking... why has no one made a digital propane meter using an Arduino knock-off for a hybrid?

I had a look around at sensors such as the MQ-6, which detects propane gas (among others I believe), and I noticed that the ppm range is about 10 to 10,000 and this is common on similar sensors. If my guesswork is correct, 10k ppm equates to 1% so you'd need a sensor that reads up to about 100k ppm for a hybrid fuel meter using propane. Now my chemistry is atrocious so I'm not sure if it's as straight-forward as this but would 42,000 ppm of propane (assuming its only detecting propane) equate to 4.2% propane to 95.8% air?

I foresee at least two issues: pressure and temperature. I have no idea whether these sensors can work in a pressurised chamber (even if its only a few PSI for the initial metering). I'd also be worried that temperature fluctuations may change the readings as I hear these sensors often need time to warm up before they start working optimally.

Does anyone know if these sensors typically refer to mass fraction, mole fraction or volume fraction in terms of ppm? I guess that's kind of important too

If this can be done relatively inexpensively I'd be keen to give it a shot, though I figure someone would've surely done it by now if it were straight-forward.

[jimmy101]

Does anyone know if these sensors typically refer to mass fraction, mole fraction or volume fraction in terms of ppm? I guess that's kind of important too

If this can be done relatively inexpensively I'd be keen to give it a shot, though I figure someone would've surely done it by now if it were straight-forward.

Mole fraction and volume fraction are the same for gases.

I would think that the problem will be the pressure and that the sensors actually respond to both the propane mole fraction and the propane's pressure (hence the overall pressure and the propane pressure are linked via the propane mole fraction).

The other problem is, if I am following your plan correctly, that you want to monitor fuel in the pressurized combustion chamber which means you have to have effective mixing. The system would probably need to inject some fuel, wait a short period of time, measure fuel, inject more, wait, ...

If you wanted a computerized fueling system I would go with either total chamber pressure (and not specifically measuring the fuel pressure) or controlling the pressure in a meter pipe followed with automated valves.

[MrCrowley]

Thanks for the reply. Had found a few useful links yesterday that further dampened my hopes, as you say the mixing would be particularly troublesome. Some have said that the values of the sensor can vary quite significantly when in a closed chamber depending on how the gas is mixed and thus its proximity to the sensor.

The one upside is that it look like you may be able to use the MQ sensors but that they output analogue values so you would need to calibrate it yourself and then make assumptions about the correlation of those values to the actual amount of gas in the chamber (i.e. that its a linear relationship). That's probably a bit of a stretch considering the sensor is influenced both by humidity and temperature.

A fun project but probably not useful for spudding

[ToasT]

I had thought about making an electronic fuelling device, using a pressure transducer and a meter pipe. Could be further automated with solenoid valves, but how close do you want your flammable gasses to your electricity?

Further ideas I had;

The mix number can be controlled with a dial (rotary encoder) with some representative number shown on a 7-segment display or similar. Some rotary encoders have a push switch built in that could then trigger the fuelling sequence.

Further complicating this is to input details on the projectile, and a desired speed then magic up some formula and see how close you can get. That magic up can be some look up table that you manually create from previous measured shots.