1. This section is for Ideas and Suggestions related to official game content. Don't request licensed vehicles here please. Read the rules here
    Dismiss Notice

In-Depth Powertrain Customization

Discussion in 'Ideas and Suggestions' started by A12_Roadrunner, Sep 9, 2019.

Tags:
  1. A12_Roadrunner

    A12_Roadrunner
    Expand Collapse

    Joined:
    Aug 11, 2013
    Messages:
    19
    I think it would be cool to see more development being made in the complexity of the engine and transmission systems. I feel that with the game trying to achieve maximum realism, it should lean more towards how the individual components and systems react, rather than "Engine + Level 1 Upgrade = More Power & Cool Backfires". I'm not saying the developers should implement a full air flow and combustion simulator into the game, but more on par with other driving simulators that are currently on the market.

    I have attached a screenshot showing an example of how I have my stock Grand Marshal engine set up, and then a screenshot of the available parameters you can modify with an aftermarket engine and transmission controller.

    Another item on my 'wishlist' for the game would be transmission thermals and driveline efficiency. This may be difficult to develop on a case-by-case basis, but I'd be willing to help out your programming team and provide examples for things like spin-loss, torque converter efficiency, inertia data, etc.

    Thanks for making such a great game!

    Engine_Customization.jpg Vehicle_Tuning.jpg
     
    • Like Like x 2
    • Staff Pick Staff Pick x 1
    • Agree Agree x 1
  2. Kueso

    Kueso
    Expand Collapse

    Joined:
    Oct 14, 2016
    Messages:
    3,335
    "BeamNG- but we performed the fusion dance with Automation"
    and this would be hell for mod makers, trying to update all of their engines to this new proposed version.
    However, in the long run, this will most likely happen. From what I can tell.
    plus all of the separate models- oh my i cant even imagine
     
    • Agree Agree x 3
  3. Dr. Death

    Dr. Death
    Expand Collapse

    Joined:
    May 12, 2016
    Messages:
    1,652
    Spark timing i think would be 100% useless to add unless engine knock gets added. Since that's the main cause of spark advance/delay. Unless on older cars.Then it could be to slightly adjust the peak torque. The AFR i think would be the EASIEST to add to the game. Leaner overheats more but consumes less and gives less power, 12.5/12.7 is best power for everything, richer consumes more overheats the least but it reduces power. Unless you are with forced induction, in which case the more boost the more richer the mix gives more power.
    --- Post updated ---
    If it was up to me i would consider the possibilities of having camshaft as an option that just picks the best peak torque. Its weird to have a 300kmh+ achiever 700 hp SBR or so that has a peak torque in the middle of the range and be unable to change it to higher RPM-
     
    • Agree Agree x 2
  4. Kueso

    Kueso
    Expand Collapse

    Joined:
    Oct 14, 2016
    Messages:
    3,335
    well knock is in the game, its just very hard to achieve before the cylinders lock up from overheating.
    i agree about the camshaft additions, it would be very cool to see more customization in that regard
     
  5. Diamondback

    Diamondback
    Expand Collapse
    Moderator
    BeamNG Team

    Joined:
    Apr 8, 2014
    Messages:
    1,447
    Hey,

    thanks for your detailed and well thought out post! :)

    We'd absolutely love to have the engine and transmission a bit more customizable. However, there's a bunch of issues with that atm:
    1. Our combustion engine model does not support parameters like spark timing, different cam/valve properties or varying metals for the hardware. This means that we can't have these parts properly influence the engine behavior, all we can do atm is change the torque in some basic ways and adjust something like the efficiency. Additionally, having multiple parts affect the same parameter is very tricky and not easily done.

    2. Even IF we were to try to implement more in-depth parts based on top of the current engine model, it would be absolute hell for QA to test the different combinations of parts of each engine and each vehicle. We'd instantly scale up the work by at least a factor of 10, including a similarly scaled up chance for bugs.
    Now, if we had a different (mathematical) engine model, things might be different, but so far, we couldn't find something that fits our need. We can't really use models based on actual air particle flow calculations because that is way too complex, takes too much CPU time and is overall just way too detailed (you'd somehow have to deal with geometry of parts inside the engine then etc).
    Other models are (too) simple (like the one we use atm), but they are cheap to compute and easy to feed with data, so at least they are usable.

    Finding the right middle ground between "too simple" and "too complex" has not worked out yet. (Please note: such models are not just readily available, that stuff is topic of ongoing research on universities etc)

    So if you by any chance know of a good paper about simulating ICEs (and their accesories like turbos and superchargers) that is complex enough to support your ideas but not too complex to calculate dozens of models at 2000hz, feel free to share :)
     
    • Informative Informative x 6
    • Like Like x 3
  6. A12_Roadrunner

    A12_Roadrunner
    Expand Collapse

    Joined:
    Aug 11, 2013
    Messages:
    19
    Thanks for your replies everyone

    I agree that ignition timing would be unnecessary, I just made that as a proof of concept for myself. The wide-open-throttle (WOT) AFR tuning I have currently in my program scales the power curve, fuel economy, and visible exhaust emissions relative to the baseline at 14.7:1. Peak power is achieved at 12.5:1 with a crude absolute value function decreasing the scale factor as it moves away towards leaner (max 15.0:1) or richer (min 10.0:1). I'm terrible at explaining things like this without visuals though, hahaha.

    I don't know about some of the more complex models, but I got a few graphs and histograms for things like accessory drive loss, internal friction, and thermal efficiency of an automatic transmission (based on TCC state, gear ratio, etc.). I'll attach a few images that are basically ripped from some vehicle calibrations I have in HP Tuners. I'll check at work tomorrow for other stuff, hopefully I can find some info that isn't GM Confidential or above. It could just be a bunch of obscure studies from the '90s, GM Powertrain Division was a lot more interesting back then.

    All the torque readings in the attached images are in N*m.

    Ford 4.6L Engine Coolant Torque Loss
    FordModularCoolantTempFrictionAdder.jpg

    Chevrolet 5.7L L31 Internal Friction Loss
    GM350InternalFriction.jpg

    Chevrolet 5.7L L31 Oil Temp Friction Multiplier
    GM350OilTempFrictionMultiplier.jpg

    Chevrolet 5.7L L31 Accessory Drive Loss
    GM350AccessoryDriveLoss.jpg

    Chevrolet 5.7L L31 AC Compressor Loss
    GM350AirConLoss.jpg
     
    • Like Like x 3
    • Informative Informative x 2
    • Staff Pick Staff Pick x 1
  7. SebastianJDM

    SebastianJDM
    Expand Collapse

    Joined:
    Apr 9, 2017
    Messages:
    950
    the coolant/oil temp losses should be in the game for sure. warming up your vehicle before driving does nothing in beam afaik. accessory drive is fairly useless, since all the accessories are on the engines permanently, and if you install a supercharger it does include losses. the friction loss based on manifold pressure sounds overly complex, but idk. it's an insignificant detail if you ask me. just something that has to be added to all engines, including modded ones, lots and lots of work for an unnoticeable and minuscule difference.
     
  8. atv_123

    atv_123
    Expand Collapse

    Joined:
    Aug 5, 2012
    Messages:
    1,251
    Well... I don’t know if calculating them on the fly is fully necessary... might be a bit overkill... however, as for calculating engines... I might be able to help with that.

    https://drive.google.com/file/d/1wJw9iVlgfVFJ11cIQhGLIKAHDYO-mQNr/view?usp=drivesdk

    This is something I have been working on for years now as engines are kinda my bread and butter. I don’t know when the last time I uploaded to my drive was, but the more time that passes, the better it gets.
     
    • Like Like x 2
    • Informative Informative x 2
    • Staff Pick Staff Pick x 1
  9. A12_Roadrunner

    A12_Roadrunner
    Expand Collapse

    Joined:
    Aug 11, 2013
    Messages:
    19
    I hope you don't mind if I just add that to my collection of awesome spreadsheets. If you like designing engines, you should look into IHRA Drag Racing 2 on the PS2. I have it on my emulator and it's basically an engine building simulator that lets you drag race.

    Anyways, I found some useful graphs and tables from studies we did in the early 2000s that could go a long way into implementing transmission thermals.


    Torque Converter RPM Differential Efficiency
    TorqueConverterSummary1.jpg TorqueConverterCompare.jpg


    Spin Loss From Increased Pump Load (more throttle means requires more clutch holding pressure)
    SpinLossComparisonsForAFwdTransmission.jpg


    Temperature Increase from Excessive Shifting
    BakerGradeTest_NoShiftStabilization.jpg BakerGradeTest_ShiftingStabilized.jpg


    Gear Mesh Efficiency
    HeatGenerationThroughGearEfficiency.jpg ASampleOfGearEfficiencies.jpg


    Situational Heat Generation (aka Why You Shouldn't Tow Trailers in Overdrive)
    HeatGenerationComparison.jpg


    How Heat Leaves the Gearbox
    LowSpeedConditionExample.jpg HighSpeedConditionExample.jpg
     
    #9 A12_Roadrunner, Sep 11, 2019
    Last edited: Sep 11, 2019
    • Like Like x 3
    • Informative Informative x 2
    • Staff Pick Staff Pick x 1
    • Agree Agree x 1
  10. Dr. Death

    Dr. Death
    Expand Collapse

    Joined:
    May 12, 2016
    Messages:
    1,652
    in theory if the peak torque @X rpm vs peak torque was co-relative it would be easy. As in, depending on the base displacement and friction/heat efficiency and the AFR (incluiding boost into displacement or having it be separate) you could have a slider that adjusted power depending based on where you want peak torque to be. I dont think that as you change the peak torque RPM the peak torque amount remains the same, but if that's still relative you could have some parameters from where you can adjust it. Adding to that would be, of course, the boost, in which something that i dont like what the game does is that more pressure, or at least a bigger turbo, should increase the peak torque RPM.
     
  11. Diamondback

    Diamondback
    Expand Collapse
    Moderator
    BeamNG Team

    Joined:
    Apr 8, 2014
    Messages:
    1,447
    Nice, keep these coming, they don't really help with the issue I posted about above, but having real world data for these things is always good and often not easy to find :)
     
    • Agree Agree x 1
  12. Dr. Death

    Dr. Death
    Expand Collapse

    Joined:
    May 12, 2016
    Messages:
    1,652
    i wish we could get real world data of the same engine with different peak torque RPM showing if the torque really changes or just the HP, however outside automation i dont think that many people even did that test.
     
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice