I was in the process of making a program that displays Outsim and Outgauge Data.

And what i found was that when the car has no throttle or is idle, the car consumes no fuel

Shouldnt the fuel be more RPM's Based not Throttle Based???

They do, just very slowly. As a test put a hugely powerful engine in a BF1 (using tweak) and just sit there - the fuel does go down

My 2.0 litre 'race' engine uses 0.039 litres per minute at idle when warm. That's

2.34 litres per hour (that's around 1.6kg)
56 litres per 24 hours (that's around 40kg)
1 litre ever 25 minutes 39 seconds (that's around 0.7kg)
One 25 litre tankful in 10 hours, 41 minutes and 2 seconds (that's around 17.5kg)

In other words, fuel load won't change very quickly.

Well, it also reverts to the idle setting as soon as you come off the accelerator at full revs, so maybe thats something that could be added, like the way it goes up but in reverse

At 7000rpm my injector pulsewidth is considerably smaller than my idle pulsewidth, although I don't turn off the fuel altogether. As such my fuel usage is (according to the datalogger and maths channels) 0.157 litres per minute (more revs, so more injector pulsewidths per second).

Most modern road cars completely cut the fuel a second or two after the throttle is lifted until the revs fall to ~1500rpm to catch the stall (also felt by a small drivetrain shunt if your in gear and simply coasting at the time).

So again, not much fuel is used with a closed throttle regardless of RPM in most cases. Even anti-lag systems won't use much. F1 hot-exhaust-blowing will use more, but still considerably less than full throttle for a given RPM.

You're mixing up two different things here. As Tristan said, cars with EFI use none to very small amount of fuel when the throttle is fully closed. LFS models this quite correctly.

Fuel consumption at idle is, however, ridiculously low. It's most apparent when you enable autoclutch and crawl at 10 km/h on 5th gear. It can get a low as few milliliters/100 km. I suppose this comes down to LFS fuel consumption model being optimized for racing engine speeds and throttle positions.

http://www.lfsforum.net/showthread.php?p=1668355#post1668355

solution: using double for fuel in tank ?

edit: even better would be to use unsigned int. For example that value of 1 000 000 equal to 1 L of fuel. That will provide much higher resolution than float as well as constant interval.