Did I get any petrol?

Struth, are you going to keep wittering about this ad infinitum? Which bit of "your trip computer or fuel tank level data is not 100% accurate and could never be expected to be" do you still not get?

No not 100% accurate, but in this case it would need to show 7.55 litres for just 5.01 litre in the tank if that; this would be *very* inaccurate.

Well go & buy a metal container & find out. Apart from anything else you don't know what level of fuel in the tank is it's nominal capacity vs its real capacity.

Until and unless you demonstrate the accuracy (or otherwise) of either the petrol pump or your fuel gauge... yes.

I am basing my belief on what I feel are reasonable assumptions. You appear to be basing yours on the assertion that your fuel guauge is connected to a computer with a mileage read-out, and is thus unfallable.

But remember that the mileage read-out that I have is essentially a gallon read-out. As I have explained, any error in the mpg is not important since it cancels out. Say mpg = X. The read-out is thus mileage = gallon*X. From which I deduce gallon = mileage/X, since X is a known number.

ONlyh if you've got a consistent X. But that still doesn't prove the gua= ge =

is accurate.

But if "gallon" in the very first instance is a pile of poo? GIGO.

It is not important what X is since it is also shown on the read-out.

Example1: X = 30, Gauge = 10. Read-out mileage = 300, read-out mpg = 30, deduction = 300/30 = 10 gallons.

Example2: X = 25, Gauge = 10. Read-out mileage = 250, read-out mpg = 25, deduction = 250/25 = 10 gallons.

Yes, Sir. On the other hand, it is technically easier to make an accurate volume measurement that an accurate flow measurement. The pump measures flow and deduce the volume received. The car tank measures the volume directly.

It's technically easier to make a volume measurement only when that volume measurement is based on length measurements of an accurately defined geometrical shape. Fuel tanks are not accurately defined shapes or even consistently sized shapes. They are either mild steel fabrications or plastic mouldings with a comparatively large tolerance on size and shape compared to a fully machined component.

Perhaps, unless there's an unknown safety factor incorporated into the mileage calculation.

So what? It is still relying on an accurate measurement of the volume of fuel remaining in the tank -- which will NOT be as accurate as a carefully calibrated petrol pump!

Two methods can be applied. Either using a fluid to calibrate the odd shape, or simply use the finite element representation of the tank to compute the volume/level relationship. Not a big deal for a large manufacturer with a comprehensive CAD/CAM. I accept that there are tolerances, but flow measurements are a great deal harder to get correct.

Do you honestly believe that any manufacturer would go to the trouble of doing this? If if so, for what reason?

Chris

As said, they already have the finite element representation of the tank or any large component affecting the packaging of the car.

That answers the question "Are they able to...", not "...would they...".

Chris

That still doesn't tell you anything about the accuracy of the guage =

though.

It measures the depth of fuel in the tank, not the volume. Using an old = =

float attached to an old sliding resistor

That may depend on the car. I remember that on an early WV Beetle, all you got was a twig to dip into the tank. The usual system was just a handle to engage the reserve tank if you ran out :-)

But why would they bother -- when the result would be pointlessly inaccurate as soon as you were on anything other than a perfectly flat, straight road at a constant speed?