How to calculate fuel consumption

[Rhodusite]

It's super easy! For everyone who does't know. You take the delta of your fuel consumption, and divide it by your RPS, the smaller your final number, the better! if you want an example of this in game let me know.

[Ra-Ra-Rasputin]

...but your fuel consumption is already the delta of your fuel tank. Dividing it by RPS only gives you a division of your fuel consumption based on the RPS of the engine.

If you want relatively accurate data on fuel consumption, you multiply the delta by 60 to get a consumption rate by second, or 360 to get it by the minute.

If you want actually accurate and readable figures with high response rate and no excessive feathering, you average the delta of your fuel tank across 10 ticks and multiply that by 6 for 1s rate.

If you want fuel consumption based on distance travelled, you need to look at your m/s figure and time. You can look at the consumption rate during 100m of travel and extrapolate your range from there.

If you want the most accurate figures, you take the value of the fuel tank into memory, and compare that value to the fuel tank 1 minute later. It's not responsive, but it's easy and accurate.

[Jorg Hammond]

That's convenient, I was recently wondering how one would go about this.

Averaging a value over time (not only fuel delta but also speed) would imply some logic shenanigans with memory and timer modules I assume ? I've toyed around this a bit before collapsing out of confusion.

I had a main 60s timer to start and stop the whole process, was reading the amount of fuel at start and finish, and was trying to average speed at regular intervals with a sub clock triggering each x ticks to try a get a reasonably accurate figure for the final calculations, but totally failed at that last part.

I guess I could figure it out of my own if I spent enough time on it, but I could certainly use some pointers on how to implement the thing.

[Ra-Ra-Rasputin]

Yes, to average a value, simply do a looping add on a memory, then divide that memory on the amount of ticks that have passed.
You can also do this math by seconds, but it's significantly slower as a result. The lowest clocktime is 0.1/s while ticks are 1/60s.

[Rhodusite]

Dividing the delta of your fuel consumption by your RPS negates the increase in consuption by increase in explosions per second, leaving ONLY the efficiency of the engine at any given RPS. I will show example in next post.

[Rhodusite]

Logic:
https://imgur.com/a/C4pB4Db

[Ra-Ra-Rasputin]

Yeah, that's what we call a magic number. It means it's just whipped together to get a desired value or function out. The actual fuel consumption is just the delta value.

[DoNotPanic]

it's wrong to use RPS because you also should take into account throttle (how much fuel burns each cycle).
I used calculated from fuel deltas after each own tick (10 seconds)
https://steamcommunity.com/sharedfiles/filedetails/?id=2042470052

[agrimes]

Here's my implementation, I just add the deltas into a memory register and then dekay it at 1/60th of a second... It would be interesting to check this method against your methods...

https://steamcommunity.com/sharedfiles/filedetails/?id=2042688612

[Rhodusite]

I never though about that, taking into account the throttle, cause it's not just more RPS, its also more fuel being injected into each cylinder!

[Rhodusite]

I'm trying to figure out the most efficient RPS for each engine, the general rule of thumb ive found is that they all seam to be most efficient at 7.25 RPS

[Ra-Ra-Rasputin]

Not for their power or torque output. According to the graph i currently have, 8.32 RPS seems to be a relatively exact sweet spot for the small diesel, but it is absolutely different for medium, which i have not completely mapped out yet.

If i go by lerping the values, it should be around 10.2

[Rhodusite]

I mapped it for the large, i think i'm remembering incorrectly and it is actually 8.25 that the sweetspot is at.

[agrimes]

I'd been using 7.5 rps based on generator tests. It could be different for direct drive applications, if so this could be a very interesting insight!

[Johannes]

If I'm not completely wrong, you only meed to connect the fluid value to a PID, with P=0,I=0,D=1 to get the flow. This is due the derivative value of the tank gives the speed (fuelflow) . Another PID with same setting will give acceleration og fuelflow.

[Rhodusite]

:O huh?