Fuel balancing

Game Talk The Bunker
21

TL;DR: car-brand-car, drove one full overmap top to bottom, used half a tank in 2 hours. In line with what Coolthulhu suggested in reply#1. My thoughts are below the last break line.

[hr]

Release: Built from git on Arch, commit ebe6fa5 (“Drop incorrect fuel definitions from blazemod”).

Worldgen: CITY_SIZE 1, CITY_SPACING 8, otherwise defaults.

Vehicle:

[ul][li]type: cheat-spawned car,[/li]
[li]general stats: like new, mass 1130 kg, O/S/T speed 156/179/179 km/h,[/li]
[li]K coefs: aero 62%, friction 93%, mass 96%, offroad 54%,[/li]
[li]fuel system: 60-liter gasoline tank, 2-liter gasoline engine, car alternator.[/li][/ul]

Test conditions:

[ul][li]cheat-revealed map,[/li]
[li]teleported to top-right corner,[/li]
[li]drove on road all the time to bottom-right corner,[/li]
[li]at 160 km/h speed (slightly above optimal).[/li][/ul]

Results:

[ul][li]general stats: dented, mass 1095 kg, O/S/T speed: 156/182/182 km/h;[/li]
[li]K coefs: unchanged except offroad 55%.[/li]
[li]distance travelled: ~ one “full overmap height”, or ~ 180 “overmap tiles”, or ~ 4350 “player-sized tiles” (80% of the way straight, the rest diagonal);[/li]
[li]fuel used: 34.5 liters;[/li]
[li]travel time: approx. 2 hours in-game.[/li][/ul]

Taking the few small cities and “diagonal” roads into account, I’ll assume distance travelled was about 200 ovemap tiles (4800 tiles). That’s 172.5 ml of gasoline per overmap tile, or 7.1875 ml per tile.

[hr]

I kind of expect a full tank to bring me a long, long way. Gasoline feels “less efficient than it used to be”.

I couldn’t make a trip around the overmap without stopping to siphone fuel at every car. Meaning “supply runs” in any given direction would be possible a few times tops. After that, there’d simply be no more fuel.

Default worldgen settings would alleviate that somewhat (I only used SIZE 1 / DIST 8 to get a long road without towns).

EDIT: ran a quick test on 0.C (stable) - crossing the overmap used up about 5% of fuel in car-brand-car.

22
Taking the few small cities and "diagonal" roads into account, I'll assume distance travelled was about 200 ovemap tiles (4800 tiles). That's 172.5 ml of gasoline per overmap tile, or 7.1875 ml per tile.
what is the length in meters of one tail? 1 overmap tail is 10 tail - i guess
23

This process that I’m seeing here feels like shooting blind, fumbling with values until we get ones that seem/feel right. We have to take a more result-oriented approach. What do we want from these cars? Take several stock vehicles and ask yourselves what their performance should be with regard to fuel consumption and maximum distance under “moderate payload”. Consider these vehicles:

Car
Humvee
Military Cargo Truck
Main Battle Tank
Electric Car

Finally think about an “average deathmobile”, the presumed endgame vehicle that most players likely build. To me that is a fully modified Military Cargo Truck or a Fire Truck. Its mass is about 12,000 kg, with all my possessions in it. How far should I be able to go with that when it has one full 60L tank of gasoline and two V12 engines? How much fuel do you guys keep in your vehicles anyway? I’ve never had more than three 60L fuel tanks.

We possibly should be dealing with curves here, not linears. Either that or we ease the extra engine restrictions.

To me there’s a bunch of unknowns. I don’t know the power differences between various engines. How much stronger is V12 compared to V8? How many V6s is equal to a V12? How much more fuel does the V12 gulp compared to V8? These are critical values that should be shown in the game. We’re not seeing the reported fuel consumption for any mass.

Does vehicle mass affect fuel consumption? If it doesn’t, then should it?

Another oversight is the lack of easy comparison when making vehicle modifications. The only way to find out how the performance changes is by experimentation and relying on your memory, if you can’t be arsed to take before&after screenshots. Why can’t we upgrade an engine, and see the upgrade results before we make the upgrade? Players are deprived of this information and for a good reason, but now that same reason sort of bites us in the ass, and we have to trial-and-error this thing. I’m just saying it doesn’t have to be like that.

(As for a proper testing environment, I suggest coming up with a mod that perhaps reduces the vehicle part installation times to 0. For quicker study and experimentation.)

24

[quote=“BeerBeer, post:23, topic:12965”]Car
Humvee
Military Cargo Truck
Main Battle Tank
Electric Car
…[/quote]

Should a motorcycle be added to this list? None of those vehicles seem like they would properly simulate fuel economy like a motorcycle would. Though honestly, I don’t know. Mechanics and vehicles are beyond my expertise, coding is beyond my expertise, and the calculus and physics that might be necessary are also beyond my expertise. I have always had a motorcycle in game because I wanted to model and approximate the real world advantages of having a motorcycle over a regular car; they are heaps more fuel efficient. With the previous engine mechanics, I could install a V-twin for fuel consumption 1 and I would choose that over any other sort of engine (If I understood the previous system correctly).

25
This process that I'm seeing here feels like shooting blind, fumbling with values until we get ones that seem/feel right. We have to take a more result-oriented approach. What do we want from these cars?

It’s not that simple, because in-game autonomy of a given vehicle is not ONLY controlled by something as specific as fuel efficiency. There are other factors at play like:

  • Are drivers allowed to ALWAYS go at optimum speed? (ie Can a player reach maximum travel distance with a given vehicle?) No.

  • Are vehicle templates allowed to mount the most efficient engine/gearbox for their mass? No.

IMO, until the above 2 topics are “fixed”… Any attempt to “balance” fuel efficiency looking at how specific in-game templates should mimic IRL behaviour is spending time in something you will have to revise later.

What I would do is paying attention to the process itself on how players (and their worlds) are affected by this ongoing effort:

  1. Keep fuel efficiency really high so overall fuel consumption in ongoing players/worlds is very low (ie prevent CURRENT players from becoming dry while the whole process is been worked on).

  2. Ensure players can drive at optimum speed by offering alternative Control paradigms.

  3. Collect current templates and divide them into discrete “weight classes”. DO NOT include Mobile Bases/Deathmobiles as a superheavy class (Reassoning at the End).

  4. Define average cruise speed for each “weight class”. Create one and ONLY one engine+gearbox for each weight class so vehicles of such class travel at that cruise speed at optimum fuel efficiency.

  5. Establish percentage of “fuel mass” an average autonomy template should come with (IRL most Civilian Landbased vehicles “spend” around 5% total mass in fuel capacity). Establish average autonomy HOURS (IRL civilian landbased vehicles are expected to have 48h of autonomy on a “FULL” tank).

  6. Use the above figures to assign to each engine+gearbox PER CLASS the “optimum fuel” consumption rate at cruise speed. You now have a figure of how much weight of “optimum fuel” your engines should consume per hour of optimum use.

  7. Create engine power vs fuel consumption “inneficiency” curves for our “ONE per Weight class” set of engines/gearboxes when player drives at a different speed than optimal.

Well, now you have set a of average optimum performance figures you can balance… So it’s time to “disrupt it” for the sake of variety. This should also be done in steps:

  1. Fuel Diversification: Diesel = Optimum Fuel. Gasoline = More weight per hour but better top acceleration (I don’t know if the Vehicle simmulation accounts for this… But IRL, engines offering the same power DOESN’T mean that can change their RPMs at the same rate… Precissely that’s the main operative difference between Diesel, Gasoline and other “more energetic” fuels). Alternatively you can convert weight to volume and use “weight” as balance so Gasoline is lighter but less efficient per Hour and Diesel heavier but more efficient per hour (Meaning “light” vehicles should enjoy better overall efficiency by reducing the overall mass by employing Gasoline… While Heavy vehicles, that can afford heavier fuel payloads, should look at Diesel as the most optimal solution. This has the advantage that if, in the future, aerial vehicles are included that have a much higher difference between dry and loaded masses, a “lighter” fuel would be mandatory). Meanwhile, Batteries, should exhibit the worse efficiency per weight because of their, potential, self-generation (in comparisson with the limited ammounts of the other 2 common fuels). OFC, more exotic sources can be balanced in the same way by been “better” or “worse” than the 3 “standard” fuels.

  2. Engine Diversification: 2 extra variants of engine per class should be available (Better simmulated if gearboxes were a component themselves). High RPM versions, designed to provide better acceleration at the cost of a reduced O/S/T speeds. And Low RPM versions, designed to provide worse accelerations but higher O/S/T speeds. I repeat that, ideally, gearboxes should be made a component themselves so even more variety between engine power and speeds could be offered without having to flood .json files with miriads of engine definitions.

  3. Template Diversification: Armed with the above diversification tools. You can now return to existing templates and put there all the new elements. The small differences between total vehicle weight should take care of “disturbing” the average figures providing variety. Then, as a last finishing touch, MOST Military vehicles are usually designed to contain extra fuel tanks for all kinds of “Oh! Shit!” situations and better autonomy than their Civilian counterparts.

The final step, OFC, is puting your Average Fuel figures back into the game so the whole system works as intented and your players can then jump into their “less fuel efficient” worlds.

And what about Mobile Bases/Deathmobiles?..

…IMHO, they are the peak of player progression and as such, they shouldn’t be balanced at all… EACH WORLD has far more parameters affecting them than what a “base fuel efficiency” simulation should have to deal with:

  • Luck based part accesibility.

  • Average fuel sources density, including distribution of fuel types and the locations they appear (Basically the size and density of cities as Main Factor).

  • Player mechanics skill (And supporting skills) progression.

What we lack ATM is an open-ended propulsion mechanism, because multiple engines can’t be linked together. Once we have this back, EACH player will get their own Deathmobile version based on how “well” a given world can support it. There is no need to “disturb” the common set of Templates/Rules for something that is very unique and SHOULD be heavily influenced by the resources each player can get.

Sorry for the length.

26

Wholly cows Did you guys up fuel consumption or what. This hurts players outside of cities, since fuel is so plentiful inside them that its easy to grab a full tank.

It hurt motorcycles and light vehicles too.

27

I’ve lost track of whats going on with the development at the moment and find myself questioning how matters are being handled.

I’ve always had good faith in the devs. Unsure what to think.

28

Here’s my idea of verifying that fuel “works”:

[ul][li]Spawn some artificial map made entirely of roads[/li]
[li]Spawn vehicle with full tank[/li]
[li]Drive it until the tank empties[/li]
[li]Note how much it traveled[/li][/ul]

As I noted before: I want the regular car-brand car to get a 1 overmap’s side length worth of distance. More than 1 but less than 2 is fine.

29

[quote=“Wally-kun, post:27, topic:12965”]I’ve lost track of whats going on with the development at the moment and find myself questioning how matters are being handled.

I’ve always had good faith in the devs. Unsure what to think.[/quote]
Same.
Also, is fuel balancing fixed in the latest experimentals?

30

Enabling the “megavehicles” mod makes fuel consumption more manageable, but still a bit quick.

31

For diesel and gasoline engines, if you believe that what Coolthulhu has stated to be what he wants as consumption under optimal conditions to be correct, then yes.

If you are using an electric engine, or you disagree with that measure, then no, but it’s easily modable, just 1 json line per fuel type.

Electric engines do have a problem without modding, since “batteries” don’t have an “energy” field defined, the item_factory assigns them a value of 1 Kilojoule per unit. While Diesel has 30 and gasoline has 20. Once you factor in engine efficiency, for vehicles driving at optimum speed, a car with a base gasoline engine will consume 10 times less fuel units than an electric one, and a base diesel one will consume a whooping 22.5 times less fuel units. Even a v12 gasoline engine will consume 8 7.2 times less fuel units.

The math gets only worse when you count that storage batteries hold 40k units, while a car tank holds 60k. If a gasoline v12 car will run out of fuel in 2 hours, replacing engine & tank for electric versions will get you 10 11.1 minutes.

Even while redlining at it’s worst, that v12 is going to be 3 times more efficient-per-unit than the electric engine, and 4.5 times more efficient per tank. I ask what the hell.

Thankfully, this is easily solved with just giving batteries an energy value (like the megaveh mod does), tho it’ll take quite some fiddling to get a value that feels “right”, it’s currently somewhat high in that mod in comparison to other fuel types.

If you want, say, batteries to be half as good per “tank” as gasoline in optimum conditions, a value of 45 (base), or 135 (megaveh mod) would do the trick. Current megaveh value is 150.

EDIT: fixed values of v12, had missed it copied from v6 which copied from v4, instead of straight copying from base engine. Also added consumption comparison when redlining for said engine.

32

As of current gamestate, what is the advantage, if any, to using a smaller or large engine?

33

In most cases, you get more power the bigger the engine, alongside higher weight, and battery cost to start it. Power lets you move bigger loads, so just pick the smallest you can get away with to reduce vehicle weight. The exceptions are the following:

The largest diesel engine nets you lower speed/acceleration than the others, but it’s more efficient when operating outside optimum rpm and of course, provides a helluva lot of power.

The tiniest gasoline engine is weird, it’s inefficient as all hell, it barely has any power, but it gets you silly-high speed and acceleration when it can actually move the thing it’s powering. It also doesn’t require power to start. I guess you could carry one along in a foldable frame and use it to jumpstart other vehicles, but you might as well use pedals for that and save yourself the gas.

Once you hit V6’s in gasoline, efficiency starts to go down. Once you hit Wankel and V12s, you can get higher RPMs (and speed), the former gets you faster but is less efficient, the later gets you more power.

34

Okay. Ive made what I hope is the right move.

If I -we- assume a car can last for 72 hours one a full tank idling, it now does for me. 12% for 8 hours, so between 11.50 and 12.49. Within toleration for what I set out to do.

Because 8 percent gas in a tank consumed an hour is a littttle silly.

Anywho, I bumped up energy by a factor of ~3.3 250 for diesel and 175 for gas. I tweaked the idle of the engines a little too, but …

35

Cool

the thing is different cars can idle for different amounts of time depending on

Engine type
Fuel tank size etc etc

A v12 using a 1ltr tank will run out of fuel before a 1 cylinder engine using same size tank and amount of fuel etc.

36

Theres till variations due to engine size and such.

I said a car -refering to a car car. genric car.

I dont mind gas rotting slowly or increased fuel consumption on the road. I do mind an engine guzzling gas by simply turning over.

37

Fair enough

And yeah a car going through a ton of fuel when idling ain’t right.

38

I’ve got a PR in progress that handles the alternator and idling issues - fuel consumption becomes proportional to how much work the engine has to do to refill the battery.

39

I am a mechanic/mechanical engineer if you lot want me to go in depth on the workings of an engine etc let me know depends how deep ya want to go really.

40

[quote=“pisskop, post:34, topic:12965”]Okay. Ive made what I hope is the right move.

If I -we- assume a car can last for 72 hours one a full tank idling, it now does for me. 12% for 8 hours, so between 11.50 and 12.49. Within toleration for what I set out to do.

Because 8 percent gas in a tank consumed an hour is a littttle silly.

Anywho, I bumped up energy by a factor of ~3.3 250 for diesel and 175 for gas. I tweaked the idle of the engines a little too, but …[/quote]

i definitely agree that the fuel usage before the change you’re talking about was… well… insane and not in the good way, however i’ve heard there is a replacement to the bugatti veyron being made that depletes it’s entire fuel capacity in 7 minutes when traveling at top speed.
in other words, maybe jet engines when we have them fit with extreme fuel usage and other very high performance applications as well, but it sounded like fuel usage was for a bit heading towards being mostly a nuisance.