1. Thread posting restricted: Only the BeamNG team members can post new threads in this sub-forum in order to prevent spam. Posting is available on unlocked threads.

Powertrain Feedback Collection

Discussion in 'Microblogs' started by Diamondback, Apr 20, 2021.

  1. default0.0player

    default0.0player
    Expand Collapse

    Joined:
    Nov 30, 2018
    Messages:
    1,441
    Found another problem about differential, when using the active locking differential, the default torque distribution is always 50:50. When using open, LSD and viscous differentials, the "diffTorqueSplit" is adjustable, but when using "activeLock", the "diffTorqueSplit" does not work.

    In the differential.lua both LSD and viscous have the device.diffTorqueSplit
    Code:
    device.outputTorque1 = inputTorque * device.diffTorqueSplitA - openTorque * outputAV1threshold
    device.outputTorque2 = inputTorque * device.diffTorqueSplitB - openTorque * outputAV2threshold
    However in the activeLock
    Code:
      device.outputTorque1 = inputTorque * 0.5 - lockTorque
      device.outputTorque2 = inputTorque * 0.5 + lockTorque
    The torque split is always 0.5, editing the 0.5 to device.diffTorqueSplitA/B enable users to make a planetary center diff with active locking.
     
  2. florencetravis

    florencetravis
    Expand Collapse

    Joined:
    Jun 19, 2021
    Messages:
    2
    The theme to raise is that you can have 2 motors and 2 transmissions all at once, yet just 1 of the transmissions is shiftable, which makes this choice totally pointless.

    This happens when your vehicle has 2 motors and each has its own transmission, when you shift just one of them is influenced, different stays in unbiased always, while the choke input influences the two motors. Also, even in the arcade gearbox conduct, the subsequent transmission never leaves impartial stuff.

    Rather it should move regularly in arcade conduct, and perhaps moving one of the transmissions physically ought to likewise influence the other one, or it very well may be feasible to tie separate contributions to it some way or another without making an entirely different custom transmission LUA like some modders have done previously.

    My specs are not important on the grounds that it happens paying little heed to the specs, however I can make reference to that I shift with a console. (Indeed I realize this post is idiotic and futile yet in any event it's better than 99% of different ones so far haha)
     
  3. Agent_Y

    Agent_Y
    Expand Collapse

    Joined:
    Jul 10, 2020
    Messages:
    6,520
    Did you just copy my post
     
    • Agree Agree x 3
  4. The Gas Station

    The Gas Station
    Expand Collapse

    Joined:
    Jul 14, 2016
    Messages:
    254
    Well, to be fair, plenty of people, including me, have discussed this in the past.

    Here's what the devs have answered me when I raised a very similar matter. However, it was about a year ago, things might have changed...
     
  5. Agent_Y

    Agent_Y
    Expand Collapse

    Joined:
    Jul 10, 2020
    Messages:
    6,520
    I mean he LITERALLY copy+pasted what I said and changed some words
     
  6. Turbo49>

    Turbo49>
    Expand Collapse

    Joined:
    Apr 1, 2021
    Messages:
    2,163
     
    • Like Like x 1
  7. florencetravis

    florencetravis
    Expand Collapse

    Joined:
    Jun 19, 2021
    Messages:
    2
    brother it was a mistake will be corrected pls
     
  8. Cento186

    Cento186
    Expand Collapse

    Joined:
    Feb 15, 2014
    Messages:
    37
    The topic/issue to bring up is...
    Clutches for manual transmissions seem too basic for a game like this and in my opinion the drivelines are all too strong.

    This happens when...
    You can take any car/truck whatever and upgrade the power to massive levels and the clutch and gearbox still has no problem with it at all. It appears that all clutches once engaged are just locked and will not slip or overheat. The only time they behave in a realistic way is when the clutch isn't fully engaged. I should not be able to tune a 100BHP car to 1000BHP on a stock clutch with zero slip in gear and on top of that not even snap the factory driveshafts or propshafts. (Without some serious effort and maybe drag tyres).

    Instead it should...
    have different strength/upgraded clutches that can be fitted in the configurator, along with heavy duty driveshafts, gearboxes and diffs etc. Also maybe even gearbox thermals, especially for automatics.

    My specs are...
    Ryzen 3600, RTX 2060, 16GB RAM, Dualshock 4
     
    • Agree Agree x 5
  9. Turbo49>

    Turbo49>
    Expand Collapse

    Joined:
    Apr 1, 2021
    Messages:
    2,163
    I've got the clutch to overheat many times, but it was by driving in a wall and accelerating while stuck or something like that.
     
    • Agree Agree x 4
  10. Cento186

    Cento186
    Expand Collapse

    Joined:
    Feb 15, 2014
    Messages:
    37
    Yes this is my point exactly, it will overheat if youre at a standstill and the wheels can't spin or if the auto-clutch in manual mode is slipping it but it will never slip or overheat when it's engaged properly - For example when the vehicle is already moving.
     
    • Agree Agree x 4
  11. SKB

    SKB
    Expand Collapse

    Joined:
    Apr 22, 2017
    Messages:
    1,962
    The clutch slipping due to lots of power can be seen in some ways, like when you are wheelspinning and then your on the grippy tarmac and have instant grip, that makes the clutch slip and heat up quite a bit, but yeah, needing to have stronger clutches and reinforced gearboxes would be a good addition
     
  12. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    6,784
    Reson for that is that Clutch is automatically upgraded everytime you replace parts, so that it matches to your config.

    lockTorque The torque limit after which the clutch will slip. This value is calculated by finding the torque required to stop the engine from redline in 1 second with full power. Not necessary to change unless you want very high peak forces from the clutch.
    From:https://wiki.beamng.com/FrictionClutch

    So simulation is perfectly capable of doing what you would like, but it would need to be a mod as vanilla content has to deal with upkeep and management, so they have decided to keep it automatic setting.
     
    • Agree Agree x 4
    • Like Like x 2
  13. Josh Grindheim

    Josh Grindheim
    Expand Collapse

    Joined:
    Feb 19, 2016
    Messages:
    57
    The FCV seems to have an issue where the ESC/TC systems still use the brakes to stabilize the car - Even after they have been removed entirely. This only seems to happen to sets of wheels that are connected to an open differential, but it will do it with both the standard or race systems installed. Since the ESC and TC systems shouldn't exist at all when removed (and the behavior doesn't happen on other differential types under the same circumstances), I would expect them not to attempt to stabilize the vehicle.
    My specs are: Ryzen 9 3950X, RTX 2080 Ti and 32 GB of RAM. I have the game installed on a 970 Evo Plus and I use a controller (Astro C40TR) as my primary input device. I had tested this as of this morning in version 0.23.1.
     
    • Like Like x 1
  14. default0.0player

    default0.0player
    Expand Collapse

    Joined:
    Nov 30, 2018
    Messages:
    1,441
    This is because the ESC/TC system is slightly different than their IRL counterpart.
    This appears as early as 0.19 IIRC, and is not a bug, as explained above.
     
    #114 default0.0player, Jul 5, 2021
    Last edited: Jul 5, 2021
  15. Josh Grindheim

    Josh Grindheim
    Expand Collapse

    Joined:
    Feb 19, 2016
    Messages:
    57
    Turbos. More specifically, I'd like to see the turbo logic (and customization system) revamped in the future. I only have experience in daily driving one turbocharged car, so I may be incorrect with this, but: To me, the turbos in BeamNG spool too easily. The turbines seem to have nearly no friction or mass to them, allowing you to (in some cases) hit up to a full pound of boost higher than what the turbo is set to - Just from simply revving the car in neutral. I also noticed that with some cars (K-Series kct6 with S1T for example), while driving with very low throttle input (less than 30%), the turbos become close to (and sometimes even do) hitting boost as the engine revs higher, despite not having enough throttle input to do so at the lower RPMs. On top of that, every turbo in the game sounds very quiet - Especially the aftermarket ones with high-flow intakes.
    Lastly, (I'd make a separate reply or this, but it's strongly related and would get bundled with this one anyway) nearly all of the turbo models only appear to be massive, stage 3 turbos. It would be really nice if stage 1, 2 and 3 turbos all had different models to better match what people would expect with each turbo (stage 1 is factory size, stage 2 is slightly larger and has shorty headers, etc.). In addition to that last bit, being able to choose air filter colours and/or intercooler sizes would be a nice feature to have in the game, instead of the somewhat simple options that are available now.
    Specs: My specs are: Ryzen 9 3950X, RTX 2080 Ti and 32 GB of RAM. I have the game installed on a 970 Evo Plus and I use a controller (Astro C40TR) as my primary input device. I had checked to verify this as of this morning in version 0.23.4.
     
    • Like Like x 6
    • Agree Agree x 1
  16. Agent_Y

    Agent_Y
    Expand Collapse

    Joined:
    Jul 10, 2020
    Messages:
    6,520
    I appreciate the change that separated the friction from the torque loss coef, but it should have been metioned somewhere as the new friction is supposed to be made 100 times smaller than the value before the 0.23 update. Took me up until now to figure out what caused my engine to become 100 times more shaky.
     
  17. Josh Grindheim

    Josh Grindheim
    Expand Collapse

    Joined:
    Feb 19, 2016
    Messages:
    57
    Edit: I'm a dummy who didn't realize I already made a similar post. My bad. At least I have more to add in this one. (Sorry in advance for the long post. Long story short: BeamNG turbos act like superchargers instead of turbos. Customization and better models would be nice, too.)
    Turbos. I've been meaning to make this post for a while, since the turbos in BeamNG sort of bug me because of how much they don't actually behave like I'd expect a turbo to behave. There's almost no weight or drag/resistance to the turbine spooling and that results in the turbos having very weird behavior. So much so, you can rev a Sunburst in neutral and hit over 15 PSI - Even when the wastegate itself is set to 14. Realistically, you'd be lucky to see 7 or 8 PSI from the engine revving without being under a load. (This does vary from car to car, but for comparison, I'll see around 6.7 PSI in my 2010 WRX STI - and that's with a built motor, aftermarket intake, intercooler, etc.)

    In modern performance cars (usually with a tune if the wastegate is limited by the factory tune), the engines act a lot like a modern GPU does in a computer (in the sense that they can boost more automatically). The ECU detects air quality (usually it just goes by intake temps, but there's other factors, too) and fuel quality (this is mostly just octane, but like with air, there's all sorts of things the ECU checks - including fuel temperature) and other various factors, then allows the turbo to boost more or less, depending on what it determines is safe under the current conditions. Then, even after all of those factors have been calculated, most turbos usually run out of boost in the top of the rev range. That could be for several different reasons, including: Turbine and/or compressor size, intercooler and/or other airflow restrictions, the engine itself not being able to take/use that much boost at that RPM, etc.
    To simplify and recap: Turbos in real life need a lot of exhaust to actually spool them up, and usually fall off at higher RPMs.

    In BeamNG, turbos seem to do this: Throttle + RPM = Boost. Or Throttle + RPM + Sportiest Drivemode = More Boost. There's no RPM fall-off, no ECU adjustments* and almost no turbo lag.

    Now to finally get to the point: It'd be really nice to see the turbo logic changed in BeamNG. Depending on the turbo, you should expect to make full boost starting anywhere from 3rd to 5th gear (sometimes 2nd, if it's a really tiny turbo), not just by revving the engine in neutral. You should also easily be able to get up to highway speeds at a decent rate (and even climb hills) without making boost, yet that is nearly impossible in BeamNG. The intakes should also be under more vacuum in many cases. Sorry for comparing to my car so often, but I'll see around 10.7 to 11.2 (usually 10.9) pounds of vacuum at idle and anywhere from 7 to 0 pounds of vacuum while cruising at highway speeds (usually 5 to 2). During engine braking, I'll regularly see 11.8 to 12.7 pounds of vacuum. Meanwhile in BeamNG, almost every car seems to idle at 3.5 pounds of vacuum, then reaches at most, 4 pounds of vacuum under engine braking. Cruising at highway speeds will result in 3.5 pounds of vacuum to 1 pound of boost or even more. (I was going to send this as a video to make it easier to consume, but I never got around to doing so.)
    To recap this mess of a paragraph: Due to the lack of turbine drag and/or weight, BeamNG turbos spool too easily. This also results in a lack of vacuum when otherwise not making boost.

    In terms of ECU adjustments, this could get complicated (if I haven't already done that). To try and keep things simple, it'd be really interesting to see modern turbos in BeamNG boost more or less, depending on atmospheric pressure and temperature when the wastegates are unrestricted (as in, Arsenic, ttSport or Race mode). This active system could also allow the ECU to pull boost and enable limp mode (usually in this case, cars still make boost since you can't fully remove the turbo; but it's only 2 - 10 PSI, depending on the car). The reason I put an asterisk earlier is because this could get messy and complicated if done incorrectly, so it's up to you devs on how you want to implement it, if at all. You could add controllable sliders like "targeted boost" and "maximum allowed boost" or go the extreme simulation route and add intercooler thermals, fuel quality and other annoying, but realistic things that turbo cars have to deal with. It's just a matter of how much of a hardcore simulator BeamNG is supposed to be. But then again, this is just some feedback on what I think could be improved.

    Finally, moving onto the turbos themselves. BeamNG is lacking in different (let alone, highly-detailed) turbo models. The Vivace ones look amazing, though the I-Series and Covet ones aren't as detailed. But then all of them look the same no matter if they're factory, race or stage 3. It would also be nice to see variable sizes to target certain rev ranges, since this is something that games like Gran Turismo have been doing for decades. Last but not least: Customization. Currently, turbos and wastegates sound extremely quiet, even if they're pushing over 30 pounds. This could just be something as simple as stock, sport and race air filters (to change the turbo volume and have a slight impact on performance) and a couple different wastegate options (to change the wastegate volume), too.

    Again, sorry for the long post, but it's hard to compress this much info about something as complex as turbos. But hey, at least the turbos in BeamNG aren't just a button that you push which shoots flames out of your exhaust, so it definitely could be much worse than it is.
     
    #117 Josh Grindheim, Oct 20, 2021
    Last edited: Oct 20, 2021
    • Like Like x 6
  18. default0.0player

    default0.0player
    Expand Collapse

    Joined:
    Nov 30, 2018
    Messages:
    1,441
    You are confusing turbo boost to manifold pressure. If the turbo is running but the throttle is partially closed, the boost is positive but the manifold pressure is still less than 1atm. When doing engine braking, the turbo is not turning since there's little exhaust, the 3.5psi is the vacuum caused by non-running turbo, the 11.8~12.7psi is the manifold vacuum mainly caused by a closed throttle valve, which is not shown in BeamNG boost gauge.

    Also the turbo lag is very present, how did you ignore or not feel the turbo lag? The Cherrier FCV has very significant turbo lag, be it base model I3 or 500HP I5 engine.

    The fall-off at very high RPM is the torque fall off, not the boost pressure. The torque fall off is the engine is sucking in air so fast that the intake is not enough to pull in that amount of air that quickly, but, the turbo pressure is regulated by wastegate, it maintains the boost pressure when enough exhaust gas is present so boost remains constant.
    --- Post updated ---
    That depends
    Example 1: Ibishu Pessima GTz. The turbo from this car is not very good, it cannot build any boost at or below 3000RPM, and in higher RPM it still has significant turbo lag.
    Example 2: Hirochi SBR4 TTS2. The turbo is not laggy, but it cannot build boost at 2000RPM or high RPM but the throttle is low
    Example 3: Cherrie FCV(most if not all of them) The turbo can build boost at a fairly low RPM or high RPM with low throttle, but the turbo lag is significant even at high RPM with throttle.
     
    #118 default0.0player, Oct 20, 2021
    Last edited: Oct 20, 2021
    • Agree Agree x 2
  19. Josh Grindheim

    Josh Grindheim
    Expand Collapse

    Joined:
    Feb 19, 2016
    Messages:
    57
    Sorry about getting the few mixed-up, I was wondering if that was the case, since "Boost gauge" isn't very specific.
    As for turbo lag, it's noticeable on some cars, like you said - Like the Pessima GTz, but it's not the same with every car. The Covet, Autobello and Sunburst can spool their stage 1 turbos (or equivalent) with their smallest engines available almost instantly (I listed them in the order of turbo spool speed). The Pessima's GTz turbo might seem like it's not very good, but it's probably one of the more realistic turbos, and it still actually spools pretty fast.
    A lot of performance-oriented turbos (unless they're small and/or twinscroll) will have a lot of lag and won't make boost in the lower RPM range. I get my Subie isn't every car in the world, but it only starts to make boost around 3500 RPM, and around 4500 RPM, it can take up to half a second for the turbo to spool up completely.
    Lastly, the manifold VS wastegate measurement you pointed out earlier would also explain turbo fall-off. If measuring in the manifold, it will show boost lower as the engine revs closer to redline and out of the "turbo range," which is self-explanatory.

    Edit: BeamNG should show manifold pressure, as that would've prevented a lot of confusion, and it would match what most boost gauges do in other games and real life.
     
    #119 Josh Grindheim, Oct 20, 2021
    Last edited: Oct 20, 2021
  20. default0.0player

    default0.0player
    Expand Collapse

    Joined:
    Nov 30, 2018
    Messages:
    1,441
    Have a look at the turbo jbeam and you can figure it out
    Code:
    "vivace_turbo_i5_2.5_petrol_stage3": {
        "turbocharger": {
            "maxExhaustPower": 1500,
            "backPressureCoef": 0.0000009,
            "pressureRatePSI": 25,
            "frictionCoef": 15.5,
            "inertia":0.6,
            "damageThresholdTemperature": 800,
            "pressurePSI":[
                //turbineRPM, pressure(PSI)
                [0,         -3.5],
                [30000,     -1.5],
                [60000,     35],
                [90000,     45],
            ],
            "engineDef":[
                //engineRPM, efficiency, exhaustFactor
                [0,     0.0,    0.0],
                [650,   0.0,   0.1],
                [1000,   0.0,   0.15],
                [1500,  0.1,   0.17],
                [2000,  0.18,    0.20],
                [3000,  0.38,    0.3],
                [4000,  0.57,    0.4],
                [5000,  0.68,    0.6],
                [5500,  0.685,    0.8],
                [6000,  0.68,    1.0],
                [7000,  0.64,    1.0],
                [8000,  0.45,    1.0],
                [9000,  0.0,    1.0],
            ],
        },
    },
    The Vivace and the SBR4
    Code:
    "sbr_twinturbo_stage3": {
        "turbocharger": {
            "flutterSoundFileName":"event:>Vehicle>Forced_Induction>Turbo_03>turbo_bov_race",
            "hissLoopEvent":"event:>Vehicle>Forced_Induction>Turbo_03>turbo_hiss_race",
            "whineLoopEvent":"event:>Vehicle>Forced_Induction>Turbo_03>turbo_spin_race",
            "bovEnabled":false,
            "flutterSoundPressureCoef": 1.0,
            "flutterSoundVolumeCoef":0.65,
            "hissVolumePerPSI": 0.035,
            "whineVolumePer10kRPM": 0.065,
            "whinePitchPer10kRPM": 0.07,
            "wastegateStart":"$wastegateStart",
            "wastegateLimit":"$=$wastegateStart+1",
            "maxExhaustPower": 180000,
            "backPressureCoef": 0.00040,
            "frictionCoef": 150,
            "inertia":20,
            "pressureRatePSI": 50,
            "damageThresholdTemperature": 900,
            "pressurePSI":[
                //turbineRPM, pressure(PSI)
                [0,         -2.5],
                [30000,     -1.5],
                [60000,        6],
                [90000,       18],
                [150000,      27],
                [200000,      36],
                [250000,      45],
            ],
            "engineDef":[
                //engineRPM, efficiency, exhaustFactor
                [0,     0.0,      0.0],
                [650,   0.23,    0.09],
                [1400,  0.35,    0.15],
                [2000,  0.49,    0.22],
                [2500,  0.64,    0.35],
                [3000,  0.84,    0.68],
                [4000,  0.90,    0.83],
                [5000,  0.95,    0.92],
                [6000,  0.98,    0.97],
                [7000,  0.97,    0.98],
                [8000,  0.94,    0.94],
                [9000,  0.90,    0.87],
            ],
        },
    },
    If you read the "pressurePSI" or "engineDef" you can't get anywhere. You need to pay attention to the "maxExhaustPower" and the "backPressureCoef".

    As you can see in the SBR4, both the "maxExhaustPower" and the "backPressureCoef" are high, which means that if you floor it, the turbo will boost up very very quickly and if you only use partial throttle, the back pressure will make the boost to drop very rapidly. Thus high throttle high RPM, quick high boost. Partial throttle low to mid RPM, little boost.

    In the Vivace "maxExhaustPower" and the "backPressureCoef" are low. Low "maxExhaustPower" means even the "engineDef" shows low efficiency at low RPM, the exhaust power is already maxed out, thus as long as you use throttle, the exhaust power are consistently 1500 all the way to max RPM/throttle. Low "backPressureCoef" means if you use low throttle, the boost will drop very very slowly. The boost will only drop quickly if you use zero throttle to trigger the BOV.

    You can do the following experimentation understand it.
    Drive a turbocharged SBR4. Rev the engine at high speed, then use throttle 50%~100%~50%~100%~50% and the boost pressure rises and falls very quickly according to your throttle, and if you 5%~100%~5%~100%~5%, you'll find that the boost is zeroed very quickly when you use 5% throttle.
    Then drive a Vivace and do the same. When lowering the throttle to 5%, you'll find the boost takes a looooooooooooong time to drop, and the turbo can full boost with as low as 30% @ low RPM to as low as 4% @ high RPM. Then @ high RPM if you first zero the throttle to trigger the BOV, then apply the throttle again, you'll find the turbo lag is about the same whether you use 10% or 100% throttle.

    Thus, high "maxExhaustPower" and high "backPressureCoef" feels like almost a centrifugal supercharger, very linear and responsive, and low "maxExhaustPower" and low "backPressureCoef" feels more sticky or hysteretic.
     
    • Like Like x 4
  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