Too bad. After increasing precision I found that our magic engine doesn't use any fuel as long as you don't use the throttle. When you drive at idle rpm there's no consumption.

Quote from Keling :

Too bad. After increasing precision I found that our magic engine doesn't use any fuel as long as you don't use the throttle.

It does, just very slowly.

Try econometrics from MadCatX. You will see that the calculations are funny in some ways. GTR engines are consuming at idle but road not (or not possible to see on 2 decimal). Also if you shut down the engine you will get still the same consumption depends of your throttle and revs...

Quote from amp88 :

It does, just very slowly.

It doesn't.

Prof added.

Quote from Keling :

I doesn't.

Prof added.

Was the engine running all the time or did it automatically cut out?

Quote from amp88 :

Was the engine running all the time or did it automatically cut out?

Running. I kept it running manually.

Here's an other test. The car was running in a circle for more than 2 min.

Sorry for the typo, BTW.

we cant see revs in pic, i just tested and amount of fuel is constantly falling down

OK, found problem, when you use 13% and more fuel, your car wont consume fuel in idle. this is probably due to float precision limit

so guys always use more than 13% fuel :P

(XFG case)



edit: my engine had restriction, thats why its at 13% no idle consumption
on normal engine fuel consumption starts at 8,86%

Thought that. Its coded with some basic representative formula and in the extremes its probably very far from real.
So I should have added another rule that you should not use motor in idle revs as actuator. Or something like this.

I was also seeing fuel usage at idle with 1% in the tank. A rounding error caused by single precision values seems like a reasonable explanation of the behaviour. So, the LFS engine does use fuel at idle, but the change is so small as to not be able to be represented in single precision. Seems like a decent bug report. Anyone want to claim it?

Not sure if it could be that because a normal car takes 0.9l/h idling. Dirivng at 5th gear 20 km/h idle should be like 5 litres/100km... Its zero... The fuel consumption should be reworked as I said before it even doesnt take any difference if you apply the throttle with shut down engine still rotating above idle rev.

Using a 32bit float seems fine to me. LFS uses fixed range from 1.0f - 0.0f where 1.0f means full tank, so I doubt there are any rounding problems. The formula to calculate consumption however seems to be optimized to work at racing engine speeds which causes the odd behaviour when the engine is at idling or even lower RPM.

My XRG does use fuel when idle

P.S. Can I shut down the engine when needed to save some fuel?

Tips for economy setup:

Tyre pressure to max
Camber adjust to min (because you want minimal surface touching ground) (lower rolling resistance)
Toe to 0
hard suspensions and some optimal height

Think everything is correct

Also what transmision would be best (diff, lsd,...) ?

Quote from Sobis :

My XRG does use fuel when idle

it uses because you have only 1% fuel in tank, try something like 50%

Not a lap, oh well at least I took part.

Quote from MadCatX :

Using a 32bit float seems fine to me. LFS uses fixed range from 1.0f - 0.0f where 1.0f means full tank, so I doubt there are any rounding problems.

How else do you explain why fuel is only 'used' at idle below a certain percentage in the tank? The only logical explanation is that closer to zero the proportion of numbers which can be represented by single precision floating point increases (which is true for IEEE 754 single precision).

Quote from Sobis :

My XRG does use fuel when idle

P.S. Can I shut down the engine when needed to save some fuel?

In RL, if you use the starter to start the engine, you lose energy stored in the battery, which will be refilled with the output of the engine (cost fuel). In LFS this doesn't happen. Kinda unfair...

Restarting without the starter should be fine.

In LFS you have nearly zero consumption while having engine in idle. If you are starting you get some little peak so its better to run it without restart.

Dont know how many times you can restart a 1.8 liter 4 cylinder engine on a battery in RL.

Hey guys. Why are you actually so bothered about using fuel while idling? I mean.. this is a racing simulator.. You gotta be so bored.

Quote from Kristi :

Hey guys. Why are you actually so bothered about using fuel while idling? I mean.. this is a racing simulator.. You gotta be so bored.

It's more about deficiencies in the modelling of certain aspects of LFS (and what knock-on consequences those deficiencies may have) than it is about just how much fuel you use while idling.

Quote from Kristi :

Hey guys. Why are you actually so bothered about using fuel while idling? I mean.. this is a racing simulator.. You gotta be so bored.

Because the problem is not isolated to idling, the consumption is wrong throughout the whole low rev range which allows people to cheat their way to the pits with only drops of fuel left. Some endurance racers might be particularly interested in this getting fixed.

Quote from amp88 :

How else do you explain why fuel is only 'used' at idle below a certain percentage in the tank? The only logical explanation is that closer to zero the proportion of numbers which can be represented by single precision floating point increases (which is true for IEEE 754 single precision).

Even though this is true, it still should not cause any problems as the difference of 1.10^-7 corresponds to 4.5 uL of fuel. Moreover, when the fuel levels actually begin to fall, the consumption is about 0.09 l/hour which is about 10 times less than what an ordinary 4 cylinder engine should consume at idle. (These figures are for XFG). I conclude that the precision would be OK if the consumption was calculated correctly.