How to improve fuel efficiency?

And the car's accuracy also varies according to tyres. No 2 brands are identical, though close. And there will be a variation with tyre wear, and in extreme cases, tyre pressures.

I use ODOMETER reading - as that's the best I've got. Plus $$ spent and fuel cost for my spreadsheet. I do check TRIP A as a double check, and it's always been right.

 

I really need to see some actual test results on tire wear. It is already clear to me, from several paths, that tread diameter does not match up with rolling circumference according to the standard circle equations.

This is also blatantly clear in the 'revolutions per mile' item on tire spec sheets. As I pointed out in another thread, even after shaving off the entire tread down to 0/32nds, the remaining tire diameter is still larger than a circle matching the specified revs-per-mile.

Remember that the contact patch on the ground is not round. It is flat. So it should be no surprise that the circle equations don't work right.

Odometers can be calibrated against GPS units, or against surveyed roads markers, to get better accuracy. Unfortunately not all road mileposts are accurately placed, and road re-construction and realignment frequently 'bust' the original route survey.

I guess I should try out multiple odometer calibrations over the entire wear life of a tire set, to see just how it changes.

 

These figures are hypothetical. I'm not going to sit here and argue about it. So here is an example.

Okay... You have 0 miles. You go 300 miles. Get gas and the pump shuts off at 10 gallons. 30 mpg. Next fill up, you go 300 miles. Get gas and pump shuts off at 9.5 gallons. 31.6 mpg. 5% difference... Hmmm.

 

No need to argue.
The facts are the facts.

The possible discrepancy in your example gradually disappears as you fill the tank more and more times........partly because the way the math of averages works and also because one pump that shuts off early will be made up for on the next fill when more goes in with a properly working one.

 

Repeat for 10,000 miles, figure the average for the whole distance. That particular error item shrivels to 0.3%.

 

BUT.....if you assume that the pump used for the next fill makes up that missing half gallon.......then the "error" might disappear completely.

 

We all have Moms and so "old Lady" sounds like "Drive like an young person on a cell phone in the left lane with flasher on with a 20 vehicle tailback"

 

I won't assume that a somewhat random process will exactly compensate on the very next fillup. Rather, I'll assume that there will always be some random scatter. But in this process, the total accumulated scatter should be constrained within a reasonable fixed range, not growing ever larger as the number of samples (fillups) increases.

Previous fillup errors should be approximately compensated on the next fillup, but usually not exactly.

So if the pump cuts off somewhere in the +/- half a gallon range, this may look like a large 5% error after just 10 gallons. But after 300 gallons, it will look like no more than a 0.3% error.

 

You forgot to account for rising temperatures causing the fuel volume to increase. It is not worthwhile to get accuracy down to the molecular level.

 

Forget? That is a separate error term. And because the fuel station's gasoline is stored in underground tanks where the temperature doesn't change much, it isn't a large error term.

 

You seemed to be trying to get to such a precision level I thought you forgot how temperatures cause a given quantity of fuel to vary in volume and density.

 

I was merely applying some math to Gas Mizer's hypotheticals in post #43, to show how averaging over consecutive fillups improves the accuracy of that one particular error term.