Non-Drag Racing Transmission tuning guide

Discussion in 'GT6 Tuning' started by Otaliema, Nov 25, 2015.

  1. OdeFinn

    OdeFinn

    Messages:
    2,643
    Location:
    Finland
    This pops out one physic feature of game, yes there is mass counted for Drivetrain, clutch(flywheel) and tires.
    How much torque goes on clutch depends several things, and rpm/speed of car.
    Places where game counts slippage are clutch and tires, if you have time to test with data logger you'll see logical reasons for speed advantage. On certain point slipping changes to other way, just because of torque used on clutch vs gear(on box). There is few percentages slippage on clutch, 12000 rpm engine it easily make differences of 300rpm between that type gearing.
    Yes longer final (2.500) will give higher top speed, but it also makes cornering ability weaker/fragile when tires aren't giving good grip (i.e. Comfort grade tires). (But not on all gears giving higher speed..:) )
    That inertia has a role in game, and it's counted in game physics engine.

    One "prophet" said at there is absolutely zero difference between clutches... Dah.. Approx second per clutch when tried Caterham 7 on Tsukuba, carbon drive shaft gets half second also.. About Caterham, it's good for testing low torque engine, but should do test on opposite too and pick high torque low rev engine, engine without turbo.


    Engine brake is applied soon as speed is declaring, I'm not speaking about braking engine, I'm speaking about engine brake. Amount what engine resists Drivetrain+wheels when i.e. Lifting throttle, that's activating brake side lock, always.
     
    Last edited: Dec 29, 2015
    Otaliema likes this.
  2. Rotary Junkie

    Rotary Junkie Premium

    Messages:
    9,809
    Location:
    United States
    So the game simulates the exact phenomenon I'm talking about, but the reason it's faster is that it breaks the game and somehow causes clutch slip under acceleration to become a negative value??? k.

    Does not only increase top speed, increases acceleration at all speed ranges. Also makes throttle response sharper in both directions. With reduced grip this can be detrimental (you can only accelerate or decelerate as quickly as tires allow) but I would say fixing the lack of traction is of greater importance than reducing effective power.


    Right. The amount of deceleration from engine braking is fixed for any given RPM and is multiplied by the overall gear ratio. If it reversed as you say, engine braking would be several times stronger in higher gears, which is not the case. Complete throttle lift is where decel lock will be strongest, under braking it becomes slightly weaker. If the overall gear ratio is 3:1, the effect of engine braking at the tires will be triple what it is at the crankshaft, not 1:3. However, the effect of brakes being applied against the engine will be 1:3, that is my point here.
     
    Otaliema likes this.
  3. OdeFinn

    OdeFinn

    Messages:
    2,643
    Location:
    Finland
    It allows engine rev bit higher with slight slip on clutch, depending what gear, what torque etc, but higher rev= higher power/more torque multiply. But it's crossing advantage to other way when gearing overlaps that slip due needed torque.
    You should spend time with your example style gearbox and playing with tire compounds, grip real vs normal, different clutches and carbon and normal drive shaft.

    Amount is not fixed, it follows torque multiply curve and engine torque curve. (Not exact to that, but gives some direction)
    Hmmm.. Gonna try to find some material about engine brake..
     
    Otaliema likes this.
  4. Rotary Junkie

    Rotary Junkie Premium

    Messages:
    9,809
    Location:
    United States
    Will look into it more when I can actually play.



    I said "fixed for any given RPM", not fixed. Engine braking is a constant based on friction and pumping losses. It will always be the same at any given RPM, although it does get stronger at higher RPM, perhaps with some added bias towards where peak volumetric efficiency is.
     
    Otaliema likes this.
  5. OdeFinn

    OdeFinn

    Messages:
    2,643
    Location:
    Finland
    Don't forget to apply engine rotation inertia/mass what reduces it on higher rpm :)

    Edit: laughing for "internet wisdom" about engine braking.. Lol. Seems like everyone forgets exhaust/intake valve overlap..
     
    Last edited: Dec 29, 2015
    Otaliema likes this.
  6. Otaliema

    Otaliema

    Messages:
    5,489
    Location:
    United States
    Thank you for keeping it civil gentlemen. Engine braking is odd IMO in the game. just my two bits.
     
    TheInfamousJEW6 likes this.
  7. OdeFinn

    OdeFinn

    Messages:
    2,643
    Location:
    Finland
    @Rotary Junkie BMW 2002 shows when torque between engine and gearbox (clutch+flywheel) is changing advantage of long final to disadvantage.
    Did testing with these, and many other power levels, it's enough to test few low power variations and maxed power, maxed power has enough torque for Drivetrain and still having enough maneuvering capability when using shorter final gear, longer starts to fail on maneuverability and acceleration, top speed staying same(or lower) as short final.
    Track used Tsukuba, grip real, noABS. All other car parts settings as stock, not even oil change, I was lazy.. Tires comfort soft, stock as everything else what's not mentioned/on Pictures.
    On maxed power short final average lap was "unbeatable" to long final, tried to catch it 10 laps without going under, .3## was closest, average .5 after short final.
    IMG_20151230_044312.jpg
    IMG_20151230_044253.jpg
    IMG_20151230_044327.jpg
     
    Last edited: Dec 30, 2015
    Otaliema likes this.
  8. Rotary Junkie

    Rotary Junkie Premium

    Messages:
    9,809
    Location:
    United States
    It does to an extent, yes, but friction outruns it considerably. Have a this:
    [​IMG]
    Motoring HP is the sum of pumping losses and friction, and this curve is based on wide open throttle. Under closed throttle the ramp-up will be more severe.

    Yes, because we have total control over the cam profiles used in GT6 and the game doesn't simply take a basic power curve for a given car and edit it... Oh, wait. Funny thing, big angry cams have slightly weaker engine braking... They also pull less vacuum at idle, I wonder if the two could possibly be related? ;)

    ABS 0. Reduced drivetrain inertia. Can't possibly see where this could cause a handling issue.

    Wait, yeah I can. Having fun fighting to avoid rear tire lock under braking are we? :lol: Stop complaining about band-aids on other people's tunes while using them yourself. :lol:
     
    Otaliema likes this.
  9. OdeFinn

    OdeFinn

    Messages:
    2,643
    Location:
    Finland
    It includes engine peripherals as well? What timing on valves? Pumping losses measured on rack is one part, but it lacks a lot of overall engine resistance during engine braking.

    ?

    ABS 0 is for good reason there, physics suffers from ABS model what PD uses, it's just for average Joe's handicap. True nature of PD's physics engine can be seen only without it, testing physics with ABS is nonsense.
    What complaining I'm doing with other people's tunes? I'm not using any "others tune" what is made for ABS.
     
    Otaliema likes this.
  10. Rotary Junkie

    Rotary Junkie Premium

    Messages:
    9,809
    Location:
    United States
    I'm not complaining about you using ABS 0, I'm saying that your short FD preference is handicapping you because of it. Rear brake bias (in a RWD car) has to be reduced with the numerically lower FD, otherwise you end up with rear tire lockup under braking that wasn't there prior. Sadly, we do not have particularly precise bias adjustments. If we did I'd probably run without ABS a lot more often. (don't even need it in GT Legends which is a MUCH more realistic sim, even on a keyboard)

    The loss map shown included water pump and a solid steel engine-driven cooling fan. If you think PD's engine braking simulation is anything more than a list of numbers correlated to RPM points you're sorely mistaken though. :lol:
     
    Otaliema likes this.
  11. OdeFinn

    OdeFinn

    Messages:
    2,643
    Location:
    Finland
    Test that BMW, problem isn't rear lockup caused by brakes(bias), whole test idea was to show that PD has more than correlated numbers from rpm.
    It shows modeled several things more than just rpm. It also shows that long final worse than short after certain power/torque level, making shorter a beater against long, even on stock setup, after tuning it wipe long off from table.
    Were you joking about that vacuum thing earlier over there?
     
    Otaliema likes this.
  12. Mach 5

    Mach 5

    Messages:
    22
    Location:
    United States
    Sorry, I should've been reading it more clearly.
     
    Last edited: Dec 30, 2015
  13. Mach 5

    Mach 5

    Messages:
    22
    Location:
    United States
    Well I know that GT6 doesn't give to clear of instructions on the Data Logger, or how to efficiently to use it in general tuning and fine tuning.

    Thank you in advance!
     
    Otaliema likes this.
  14. eran0004

    eran0004

    Messages:
    8,711
    Why would a taller final drive ratio result in more braking torque at the wheels?

    Having a list for each car would be a complicated solution, especially when you've got 1000+ cars. It would be much easier to just use a formula, such as: engine braking = x * engine speed.
     
    Otaliema likes this.
  15. Rotary Junkie

    Rotary Junkie Premium

    Messages:
    9,809
    Location:
    United States
    Tested it. Neat. Car is so constantly traction-limited that it drives best using a gearbox with a 5.000 final around Tsukuba, preferably without a CF driveshaft or clutch upgrade (though the omission of those very noticeably harms acceleration in 3rd gear or higher). Car as tested was totally stock on comfort soft tires with all power upgrades, FC trans, triple plate clutch, and CF driveshaft. Choosing Tsukuba as the test track makes certain the test favors this, as, again, you can only accelerate or brake as hard as your tires allow... And in this case, the difference between what the tires can put down and what it has power for are massive. Results were largely as expected, including remarks on brake bias, though I did underestimate the difference in the level of engine braking.

    Proceeded to SSRX, built identical gearsets with 5.000 and 2.000 finals, geared just long enough that kissing the rev limiter was not an issue, 0.2sec difference at 5km, which is quite in-line with previous tests, although skewed a bit by distance as the entire advantage was developed at lower speeds and then carried to the top end.

    Not so much "more" braking torque. The force exerted by the brakes is the same, but just like lighter wheels, reducing the inertia of the driveline means it takes less brake force to slow the tires down. With no ABS, this can and will mean locking the tires up more easily. In the case of the 2002 Turbo on CS tires, it also made off-throttle cornering considerably more... "interesting".

    Considering that list has to be there for engine power anyway, it's the simplest solution. And then engine modifications are simply modifiers applied to the power curve.
     
    Otaliema likes this.
  16. Otaliema

    Otaliema

    Messages:
    5,489
    Location:
    United States
    Hmm with all this debate I might just make a theory section and start putting this stuff in it. As a lot of it is subject to interpretation from the limited data we can get in the game and how you run the lap.

    Sent you a PM as to not side track this thread.
     
  17. eran0004

    eran0004

    Messages:
    8,711
    But the prop shaft has such a small radius that the change in angular momentum is minimal, especially when you consider the combined system of wheels and drivetrain. Does it really make a noticeable difference?

    It may be more accurate if you can find reference braking force data for each engine, but it's far simpler to use a single formula and apply it to all cars. Especially if you want to be able to adjust this later on. Rather than editing a thousand cars with, say, 20 values each (20 000 values in total) you can just edit a single formula.

    Of course, you could use a single list and apply it to all cars, in which case it would only be marginally more complex than the formula.
     
    Otaliema likes this.
  18. Rotary Junkie

    Rotary Junkie Premium

    Messages:
    9,809
    Location:
    United States
    Quite. Did a little more SSRX testing and the difference between max final and minimum final is actually considerably more noticeable than a CF driveshaft, which backs up other testing. The difference in rear brake values usable between them (for a RWD car) is surprising.

    Data is cheap, processing is not, so having it be a dynamic number after load-in is unrealistic. I also have no reason to believe any of the standard cars have changed from a data perspective since GT4, although tuning option modifiers have (so the vast majority of that thousand were done already). Further, you already need x data points for the base power curve, and every car has a discrete engine file anyway.

    Not to mention, those x data points realistically take very, very little time to input (or zero) depending on exact approach. Even at a minute a car it's somewhere around 2 working days for one person to set all of them. Compared to what modeling takes that's nothing. :lol:

    Now, what wouldn't surprise me is if engine braking is directly affected in the same manner as the power curve. Only thing that changes the shape of the curve is forced induction anyway, and even then it'll seem close enough. The fact we can double torque output from already highly-strung naturally aspirated motors across the entire rev range is slightly suspect though. :lol:
     
    Otaliema likes this.
  19. eran0004

    eran0004

    Messages:
    8,711
    I've never noticed any difference, but I'll test that later.

    Even if you have a list, you need to process the data before you use it, because you can't have a list of items for every single engine speed from 0 to what ever the rev limit is (0, 1, 2, 3... 4000, 4001, 4002... 7500, 7501, 7502... etc). If you're using less data points than possible rpm values you need to perform an interpolation, which is more costly than a simple multiplication.

    Assuming that they used lists for engine braking in GT4. And while you need these data points for the torque curve you don't need them for the engine braking. Why use a complicated solution when you can use a much simpler method? Having to use a list for torque values is a weakness, but there are no good alternatives. There are good alternatives when it comes to engine braking.

    Why would anyone think it's a good idea to spend two working days implementing a change when there is an alternative solution that would take two seconds?

    You mean, power upgrades increases engine braking? Should be easy enough to test.
     
    Otaliema likes this.
  20. Otaliema

    Otaliema

    Messages:
    5,489
    Location:
    United States
    IIRC Priano did some propshaft and clutch plate restring when GT6 first came out using AT, at Bathurst reverse uphill straight. His results where 1% speed increase over the entire rev and gear range per upgrade, how ever double and triple plates didn't stack. So a CF propshaft and a double/triple plate adds 2% speed to the car.
    Assume that this increase applies to engine braking you should see a greater deceleration with these parts. Improved engine braking as less mass means it can't turn as long and greater contact area on the double/triple plate would improve engine breaking due to less slip.
     
  21. OdeFinn

    OdeFinn

    Messages:
    2,643
    Location:
    Finland
    did you test what was amount of power when advantage changes to other?
    Now understanding my 3:1 - 1:3 statement?

    If they have implemented real life style there, then only engine stage changes, clutch+flywheel and propeller shaft are making changes to actual engine braking amount, gearbox/final and gearing are just how much it multiplies it.
    Data for engine braking is not huge, engine size, model (v,i,b,w..), cylinder amount, strike order, compression ratio, valve timings, then value for "mass" of engine moving parts, flywheel+clutch, gearbox etc.

    Pretty much yes, propeller shaft and flywheel weight reductions also makes "power direction" changes quicker, snappier reactions from throttle liftoff.
     
    Otaliema likes this.
  22. eran0004

    eran0004

    Messages:
    8,711
    Okay, I bought a BMW 2002 and ran 70 kilometers worth of brake tests :lol:

    In the chart below from left to right:

    FD 5 = Final drive 5.000
    FD2 = Final drive 2.000
    Stock = No installed parts
    CP = Carbon prop shaft
    FW = Lightweight flywheel (triple clutch)
    FWCP = Carbon prop and lightweight flywheel
    Asp = Aspiration power upgrades (all upgrades that increase power without increasing the revs)
    Rev = Engine speed upgrades (all upgrades that increase the revs of the engine)

    The tests were done with brake balance set to 0 - 0, in order to let the drivetrain have the biggest possible impact on braking. For all tests I stayed in top gear during the braking, to make sure that engine was always running at the same speed. For the different final drive tests I adjusted the fifth gear so that the total ratio (FD * gear ratio) was the same in both tests. The gearbox was left stock in the rest of the tests, so comparisons shouldn't be made between the first two tests and the rest of the tests.

    The chart below indicates the top and bottom times (thin lines) and the times from second lowest to second highest (wide bars). Time is measured in seconds.

    2002brake.png
    So, to conclude: There is a measurable difference between FD 5 and FD 2. There is a slight overlap in the times, but that's most likely due to my own timing errors.

    Carbon prop and lightweight flywheel both reduces the braking time, the flywheel seems to have a greater effect than the carbon prop, although there is a time overlap so it's hard to tell for sure which is the most efficient. Combining carbon prop and lightweight flywheel gave the shortest braking times.

    Power upgrades does not seem to have an effect. In the chart they are both slightly above the stock setup, but the time spread is so big that it could just be due to timing errors (it was harder to keep a constant speed with the more powerful engines).
     
    Rotary Junkie, Otaliema and OdeFinn like this.
  23. Rotary Junkie

    Rotary Junkie Premium

    Messages:
    9,809
    Location:
    United States
    One data point every 200-500rpm is common, and you misunderstand what I mean by data is cheap. I should also rephrase how it's done, it's more of a minimum torque value (used at closed-throttle until down to idle) and a maximum torque value (used at WOT) with the two blended vs throttle position. Literally everyone in the industry uses this method.



    You do need them for partial throttle. Otherwise you'd end up finding redline eventually at 1% throttle.



    Accuracy and not having to alter the game engine, for one.


    As shown, not surprised that they didn't. Which is in fact a good thing.

    Not in the slightest.


    Firing order? Cam timing? Compression ratio? Ahahahaha none of these things are remotely simulated. Try a single engine inertia value, separate flywheel/clutch (probably fixed across all vehicles, would explain the Caterham's troubles), and then driveline/prop shaft.


    Yep, and with FD being as large of a difference as it is (changing driveshaft speed by a factor of 2-2.5 is not exactly small) it's near the difference between a very heavy driveshaft and an impossibly light one. CF driveshaft itself posts roughly the same "gain" at either final however.
     
    Otaliema likes this.
  24. eran0004

    eran0004

    Messages:
    8,711
    You can blend against the output of a formula, you don't need a list for that.

    It's only more accurate if you have access to reference data and that data is so complex that it's hard to reproduce it using a formula. Major changes are done to the game engine now and then, this would be a relatively small change, depending on how complex you want the formula to be.
     
    Otaliema likes this.
  25. OdeFinn

    OdeFinn

    Messages:
    2,643
    Location:
    Finland
    Who said at those are simulated, those datas should be in there if someone wants to simulate engine brake properly.
    Vacuum or compression?
     
    Otaliema likes this.
  26. Rotary Junkie

    Rotary Junkie Premium

    Messages:
    9,809
    Location:
    United States
    They aren't needed for accuracy beyond initial simulation to find said values, as the results are consistent for any given engine configuration. Unless we want to count air temperature and barometer while we're at it but I know of literally one game with that capability and it's a drag racing game.
     
    Otaliema likes this.
  27. OdeFinn

    OdeFinn

    Messages:
    2,643
    Location:
    Finland
    Serious, consistent for any given configuration.. ...well if you want so
     
    Otaliema likes this.
  28. Rotary Junkie

    Rotary Junkie Premium

    Messages:
    9,809
    Location:
    United States
    Engines don't magically change. The way a given engine acts is how it acts until you intentionally change something about it. And by configuration, I mean everything, not just "oh it's an inline four".
     
    Otaliema likes this.
  29. OdeFinn

    OdeFinn

    Messages:
    2,643
    Location:
    Finland
    So your message contains information of?
    You say at every same size i4 has identical engine braking, no matter manufacturer?
    Or every of those have own amounts of engine braking?
    You would be good on politics.

    Going to sleep, getting more answers while sleeping.
     
    Otaliema likes this.
  30. Rotary Junkie

    Rotary Junkie Premium

    Messages:
    9,809
    Location:
    United States
    The same the last few had, hoping you'd get the point, but...

    You didn't.

    Close enough. Any engine will have unique properties that determine engine braking (as well as all other attributes). These don't change unless you change them. Not sure why I have to say this twice.

    Can't lie well enough for that
     
    Otaliema likes this.