15 replies to this topic

[Pit Babe]

In her ever-optimistic fashion, PB wonders if there might be some information out there on the Sauber fuel tank problem. Apparently, it is difficult to tell how much gas is in the tank. Why? That seems elementary to me, but the fact that I think so is enough reason to doubt the assessment.

Still, I seek some explanation(s) for what happened at Hungary other than evident incompetence on the part of the pit crew. Are there any?

------------------
PB ;-)
Isobel's Girls F1 Page!

If you can't laugh at yourself, others are certain to oblige.

[SlowDrivr]

I don't know what happened with Sauber. However, I read about the fuel metering problem in one of the UK magazines, Race Tech or Racecar Engineering. Even though it seems like it should be simple, it turns out not to be.

A float type system like what's in street cars wouldn't work, since a race team needs to know how much fuel is left down to the fraction of a gallon.

Since the engine computer (ECU) controls the injectors, they program it to keep track of how much fuel it has squirted into the engine. This is reset at pitstops. Problem is, that even this isn't as accurate as they would like. Somehow there's enough variation in something that there are errors in how much fuel they think is left.

Also, since this method only calculates how much fuel has been used, and doesn't measure it, if there's an error in how much fuel is put in in a pitstop, they can run out (for instance, meant to put in 30 gallons, but only 29 went in, so if the computer says 28 have been used they think they could make it one more lap, but they can't...).

Maybe the rules prevent this, but I would try putting a video camera in the tank which would allow them to broadcast it back to the pits so they could see how much fuel is left....

[PDA]

The tanks are foam filled, so that wouldn't work.

they usually have a good measure of how much fuel was put in, and they should know how much is being used. Looks like something went wrong on this occassion, but I doubt it will ever be reported.

[Christiaan]

Why then is Sauber the only team that makes this blunder?

I was thinking- how about a damped oeizoelectic material at the bottom of the fuel tank. The derived pressure reading from such a sensor would very accurately relate to the height of the column of petrol above it. The problem of course is that the car swings all over the place ,shaking the petrol; but on a straight you could quiet easily compute the true height even with the petrol jiggling.

[Pit Babe]

What does "oeizoelectic" mean?

[Ursus]

Hmmm, piezoelectric maybe.

[Pit Babe]

Okay, then what does "piezoelectric" mean?

[SlowDrivr]

Piezoelectricity is when a substance gets an electrical charge when it is compressed. It also works the other way- if a voltage is placed across it, the substance contracts (or expands). Crystal oscillators and the needles (stylus) in record players use piezo substances (quartz).

The problem with trying to use it for a fuel meter (as a scale) is that it only responds to dynamic loads. There would be an output as the fuel is put into the tank, but "x" gallons of fuel sitting on a piezo sensor would give no output.

[Christiaan]

yes, peizoelectric

I hadn't thought of that dynamic load bit. So if the sensor is put on a narrow section then it might be able to measure fuel consuption?

[Yelnats]

The fuel volumn can be indirectly measured by measuring the air volume in the tank. This can be done by pumping a measured volume of air into the tank until a predetermined pressure change is reached. This has the advantage of being independant G forces and actually measures fuel available, not just the amount used as present techniques do.

The old riding mechanicas were aware of this principal when they pumped air into the tanks to maintain fuel pressure. They soon realised that an empty tank took a lot of pumping to get the pressure up and got a rough idea of the remaining fuel by the feel of the pumping pressure.

[PDA]

The electronics in the engine management system measure and record exactly how much fuel has been used. In the pits, they know exactly how much fuel has been loaded, so there is no need to try and measure level in the tank.

[Jecko]

PDA is right,

You don't need any fancy detectors or sensors telling you how much fuel is in the tank. You just need to know three things. How much fuel you put in, the rate at which the fuel is consummed, and a pocket calculator.

Cheers

Jecko

[Christiaan]

My experience with such sensors is that they are to within 5-10% accuracy for relatively expensive sensors. I do not know how what they use in F1, but 5-10% is the reason why Jean and HHF run out of fuel.

[MPH]

You also need to know how much fuel you can get out of the tank, not just how much you put in.

[PDA]

MPH - very good point. I believe that most test each car to running out of fuel at least once so that any fuel pickup problem can be identified. Obviously not during race practice, as then they cannot afford to lose any laps.

[Yelnats]

PDA, Per your comment about "there is no need to know how much fuel is in a tank" (above)

Tell that to John Alesi! Any system that detects how much fuel is in a tank will be welcomed by a team manager. Metering is not 100% accurate in fueling or racing. The smallest errors become critical when strategy requires a run be extended to the limit. I wouldn't like to drive a car that relied on the meter principle, why should a F1 driver have to?

[This message has been edited by Yelnats (edited 10-08-1999).]