Deep Forest Tunery [CLOSED]

Adrenaline · Apr 11, 2011

Yeah, race softs.
I'm absolutely horrible the first 5 laps.
With Jorday's tune at Indy I went all the way to 51 seconds behind the leader, but finally got my **** together and pulled off the win. His actually seemed to have more understeer in general, but on the other hand, was much easier to control on corner exit. Which is what he says he focused on most, so I guess that's to be expected. Lead me to a new best lap of 1:09.0xx.
I don't have a clue as to how I'm gunna get around Cote 'De Azure. I swear this series is a choice selection of my weakest tracks.

budious · Apr 11, 2011

Adrenaline
I don't have a clue as to how I'm gunna get around Cote 'De Azure. I swear this series is a choice selection of my weakest tracks.

Did you install the Chassis Reinforcement? I'm curious as if his tune works as intended with or without it and whether mine works without it.

I pulled off Cote D' Azur fairly easy, again, found best control on Race Hards and my stiff setup. Dampers 8 and stabilizers 6 with race hard was 1:19" and that was just like 3 lap attempts in practice mode testing setups, sure you could push best lap far lower.

You own any of the other formula cars? If you want to borrow a Ferrari Formula and do the seasonal event for them just send a PSN request.

elric718 · Apr 11, 2011

I used this tune in B-spec, and whaled on them. Quit the last race and still got gold.

(Level 29 Bob, with newly overhauled engine helped too.)

I also noticed severe problems with it in the rain. I killed them on Monza in the rain, but only because the AI starts their first rain race on the wrong tires, as we know.

My best hot laps could not keep up with the best AI hot laps. I wonder if it is because the ride height is so low? Either the car might have been hydroplaning a little, or the puddles were like little walls of water.

budious · Apr 11, 2011

elric718
I used this tune in B-spec, and whaled on them. Quit the last race and still got gold.

(Level 29 Bob, with newly overhauled engine helped too.)

I also noticed severe problems with it in the rain. I killed them on Monza in the rain, but only because the AI starts their first rain race on the wrong tires, as we know.

My best hot laps could not keep up with the best AI hot laps. I wonder if it is because the ride height is so low? Either the car might have been hydroplaning a little, or the puddles were like little walls of water.

Yeah, it would be nice if bob would not auto-pit and insist on rain tires, the softs definitely do better in some rain conditions despite bob's judgment. There is room for improvements in the rain though, I just have gotten around to it yet.

budious · Apr 12, 2011

budious
Wet Setup:
--------------
@ Extension: 4 / 4
@ Compression: 4 / 4
@ Anti-Roll Bars: 2 / 2
@ Camber: 4.5 / 2.5
- Monza (Wet) w/ Race Intermediate (FR) & Race Medium (RR) - 1:36"
- Suzuka (Wet) w/ Race Intermediate (FR) & Race Medium (RR) - 1:56"

Suspension tweaks and setup for rain at Monza and Suzuka. You still need to be a little gentler than usual on the throttle than with the dry conditions, especially around the south tip at Monza, and in the final chicanes at Suzuka. The tires will warm up and be a little more consistent after the first two laps. You can run the times I have indicated fairly easily if you don't push the car too hard and do them repeatedly.

Race Soft adds more grip to the rear, but in a long race, the tires may begin to overheat, Medium tends to fair better and is a good balance with the Race Intermediate tire equipped up front. The Race Intermediate tire tends to show the best use up front for cornering and braking on wet conditions. I did not find an opportunity during my testing that the Rain tire showed any advantage, it was a disadvantage in light to medium rain conditions, though if heavy rain conditions do appear you can try equipping one up front and switch to a Race Soft in the back.

If you are B-specing you need to babysit Bob because he'll want to switch to those darn Intermediate and Rain tires on the rear of the car every lap. I figure you already know where the pit indicator light is and how to tell him to cancel his pit stop.

elric718 · Apr 12, 2011

Excellent work, this tune is already becoming legend.

I wish I kept a more detailed log of my B spec adventure, but I was also busy for much of it. The tune is so good that I mostly used Racing Hards, got a huge lead early, and snoozed through the rest of the race. No pit needed on Monaco, but I did try to pit just after half way on the other tracks and add extra gas.

P.S. If you could put a tune for the 25th anniversary Countach on your "to do" list, it would be much appreciated.

budious · Apr 12, 2011

elric718
P.S. If you could put a tune for the 25th anniversary Countach on your "to do" list, it would be much appreciated.

I got a couple but they are sport tuned level to still be fun and not overkill for the Lambo seasonal. If you are interested in that I can post it later.

budious · Apr 12, 2011

-------------------------------------

-------------------------------------

-------------------------------------

-------------------------------------

budious · Apr 18, 2011

RX-7 Tuning Equation Demonstration

OTI equations provided by budious

If you have been following, or have attempted to follow, the second post in this thread then you already know about the Open Tuning Initiative. I know at the moment it is very disorganized, ambiguous, and in some cases inaccurate. However, I have prepared a short demonstration of the techniques that I do know work fairly well at this point. I hope after you try the following guided demonstration you can appreciate and better comprehend the equations I have published to your own benefit.

Test Car: Mazda éfini RX-7 Type R (FD) '91

Premium car, purchase at new car dealership.
Equipped at stock weight with upgrades to approximately 300HP.

Additional Purchases:

GT Auto: Oil Change
Tune Shop: Sport Soft
Tune Shop: Fully Customized Suspension
Tune Shop: Low RPM Turbo
Tune Shop: Sport Exhaust

Deep Forest Raceway Testing:

Equip the car with all purchased parts except return to factory suspension; drive a few laps around Deep Forest while noting the handling characteristics of the car.
Equip the car with all purchased parts including the fully customized suspension; zero out the rear positive toe, then drive a few laps around Deep Forest while noting the handling characteristics of the car.

Examining the defaults on the Fully Customized Suspension for the RX-7:

Damping: For the default values on the RX-7, both extension and compression are set to values of five. These values have multiple roles in the overall physics engine, but for the particular sub-component determining overall suspension balancing, we need to compute the LOG of each.

Dampers:

LOG(1) = 0
LOG(2) = 0.301
LOG(3) = 0.477
LOG(4) = 0.602
LOG(5) = 0.699
LOG(6) = 0.778
LOG(7) = 0.845
LOG(8) = 0.903
LOG(9) = 0.954
LOG(10) = 1

RX-7's Damping Efficiency Factor, or DEF = LOG(5) + LOG(5) = 1.398

Stabilizers: For the default values on the RX-7, the anti-roll bars are set to values of four. These values have multiple roles in the overall physics engine, but for the particular sub-component determining overall suspension balancing, we need to compute the Natural Log, or LN, of the stabilizer bar.

Anti-Roll Bars:

LN(1) = 0
LN(2) = 0.693
LN(3) = 1.099
LN(4) = 1.386
LN(5) = 1.609
LN(6) = 1.792
LN(7) = 1.946

RX-7's Stabilizer Efficiency Factor, or SEF = LN(4) = 1.386

Spring Rates (and Ride Height): As for the defaults on all cars in the game, ride height is zeroed. The spring rates on the RX-7 default setup are 8.4 front and 6.3 rear. As all variables in the suspension setup, the following calculation is not the single condition nor only contributing factor for what the maximum spring rate should be; it is only a determination of what the minimum spring rate should be.

RX-7's Spring Rate Stiffness Factor, or SF(SR) is calculated as follows:

(Front Spring Rate + Rear Spring Rate) / (Car's Weight KG / (Base Ride Height + Ride Height Change))

(8.4 + 6.4) / (1260/(100+0)) = 1.167

This poses a problem because the spring rate stiffness factor, or SF(SR), is below the ~1.4 threshold we are looking for. So how do we achieve a SF(SR) of 1.4? For the simple purposes of testing our theory we can start with a ride height adjustment.

(8.4 + 6.4) / (1260/(100+(+20))) = 1.400

Perfect, our formula reveals that adding +20 ride height to the car will achieve the target SF(SR) we were looking for. Let's test it on the track, watch the tire temperature indicator to realized how much grip has been optimized on your first lap. Things will probably get a little sloppy on the second lap but settle down if you keep driving. This setup achieves the intended effect we were looking for by increasing SF(SR) to 1.4 but using this method has increased lateral weight transfer on the car with the increase in ride height.

So how do we achieve a SF(SR) of 1.4 but closer to the normal ride height, or any desired ride height, for our car? There's an equation for that... and hopefully soon enough, an equation app for that... in the meantime, break out the calculator and some scrap paper.

First, we need to calculate the relative distribution of supported weight on the car's springs; this should not to be confused with the actual weight distribution of the car. This is a fairly simple process.

% Weight Distribution Front Springs = Front Spring Rate / (Front Spring Rate + Rear Spring Rate)

% Weight Distribution Rear Springs = Rear Spring Rate / (Front Spring Rate + Rear Spring Rate)

8.4 / (8.4 + 6.3) = .571 (57.1%)

6.3 / (8.4 + 6.3) = .429 (42.9%)

You really only need to do one to find both, simply subtract the one you do first from 1.00 to get the other. Next, you need to multiply those figures by the weight of the car to find the supported weight in kilograms on each axle.

KG Weight Distribution Front Springs = (Front Spring Rate / (Front Spring Rate + Rear Spring Rate)) x Car's Weight

KG Weight Distribution Rear Springs = (Rear Spring Rate / (Front Spring Rate + Rear Spring Rate)) x Car's Weight

8.4 / (8.4 + 6.3) = .571 x 1260 = 720 KG = 1.00 SF(SR)

6.3 / (8.4 + 6.3) = .429 x 1260 = 540 KG = 1.00 SF(SR)

The resulting figures represent the weight supported at a spring rate stiffness factor, or SF(SR), at 1. This is because the weight you multiplied against the percentage of distribution was the car's actual weight. You could have compiled the previous step with the following step, but for comprehension it was broken down into an extra step. To bump SF(SR) up to the desired 1.4 all we need to do is multiply the previous results by the new desired factor.

8.4 / (8.4 + 6.3) = .571 x 1260 = 720 KG x 1.4 SF(SR) = 1008KG

6.3 / (8.4 + 6.3) = .429 x 1260 = 540 KG x 1.4 SF(SR) = 756KG

Now we are ready for the final step. We need to divide these figures by the desired ride height to determine the final spring rate at that ride height. This can be a bit frustrating to do with a standard calculator. However, there are shortcuts for a graphing calculator and I will update this post and the OTI post with these at a later date. If your standard calculator supports returning to the previous line and simply overwriting the ride height value then this following task is much easier. Note: Base Ride Height is not always 100; a few cars may have a non 1:1 motion ratio on one linkage; a few others may have non 1:1 motion ratios for both linkages (ie. Formula cars are 2:1 wheel rate to spring compression). However, for the vast majority of cars in the game, a motion ratio of 1 for front and rear, and a Base Ride Height of 100 can be safely assumed.

Stiffness Factored Weight / (Base Ride Height + (Change in Ride Height))

720 KG x 1.4 SF(SR) = 1008KG / (100 + (+5)) = 9.6 kgf/mm

540 KG x 1.4 SF(SR) = 756KG / (100 + (+5)) = 7.2 kgf/mm

These are your new spring rates at the indicated ride height. I prefer to attempt to get these figures as close to but under a whole value to a tenth of decimal accuracy as this is the limitation of the spring rate tuning allowed in the game. I also attempt to get them with no to as little rake as possible involved because rake involves shifting weight transfer and I have not yet produced a variant on the formula to account for these other variables, among many others. Basically, what I attempt to do is divide by ride heights until I find one that will make both the front and rear at the same ride height level no less than five hundredth under or one hundredth over the next closest settable value. (ie. 12.95 - 13.01 would be set to 13.0 kgf/mm)

The final suspension configuration for SF(SR)=1.4 @ +5mm Ride Height is as follows:

Ride Height: +5 / +5
Spring Rate: 9.6 / 7.2
Extension: 5 / 5
Compression: 5 / 5
Anti-Roll Bars: 4 / 4

Why did this work? Let's look at the much larger, but still limited, view I have assembled so far (as theorized) for the internal functioning of the suspension balancing and grip determination mechanisms of the physics engine.

Stiffness Factor of Spring Rates ^ Natural Log of Anti-Roll Bar
----------------------------------------------------------------------
Sum of the Damper LOGs ^ Natural Log of the Anti-Roll Bar

or

(SF(SR)^LN(ARB)) / (LOG(Extension) + LOG(Compression) ^ LN(ARB))

1.4^1.386
---------------------
(.698+.698)^1.386

1.594
------- = 1.002 (100.2% optimized)
1.591

Now keep in mind that all this does is establish the minimum stiffness factor of spring rates, SF(SR), to be used on the car you are tuning. You can increase the final ratio anywhere above 1.000 (~100%) and the car remains drivable but there is probably another such equation to determine upper range efficiency and it is probably track specific or natural frequency of the course specific to determine the new optimization. Though, I think with the above rationalization, that the car will achieve its maximum grip efficiency while increasing the stiffness factor of spring rates beyond that threshold will continue to improve lap times in a trade off for less grip. At this point much remains speculative, I only have what works on one hand, what doesn't on another, and a whole lot at my feet I haven't gotten around to just yet.

Final Note: If you are intending to reproduce this exercise on other cars keep in mind the following things.

See the OTI post on the Supra example for how to deal with motion ratios if the car you are tuning uses one.
Chassis rigidity factors into this equation somewhere; attempt with new dealership purchases without chassis reinforcement for most normalized results.
Stiffness Factor of Spring Rates >= Damping Efficiency Factor >= Anti-Roll Bar Efficiency Factor

Rotary Junkie · Apr 18, 2011

I'm going to smack you for saying 100mm ride height and 1:1 motion ratios can be "safely assumed". Hard.

I know for damn sure that the '91 FD's stock ride height will be 118, 124, or 135mm (depends on whether racing suspension inherently lowers the car, haven't really bothered to do a proper visual check though it seems to on certain cars). NOT 100mm.

Also, far too much math, not enough feel based tuning.

budious · Apr 18, 2011

Rotary Junkie
I'm going to smack you for saying 100mm ride height and 1:1 motion ratios can be "safely assumed". Hard.

I know for damn sure that the '91 FD's stock ride height will be 118, 124, or 135mm (depends on whether racing suspension inherently lowers the car, haven't really bothered to do a proper visual check though it seems to on certain cars). NOT 100mm.

Also, far too much math, not enough feel based tuning.

Yes, I'm well aware of this. I have tried other motion ratios and ride heights. For the most part I'm sure this process was normalized to an extend to reflect only certain types of linkages and not to reflect real life travel. You can simply adjust my my use of 100 for any you think is more appropriate and tell me if your results are better than using 100. The formula should work the same in this regard. The only two that I have actually "confirmed" (to mean it loosely) on cars so far are 100mm and 50mm with good results.

As for the equation itself, it could be an entirely separate process from the actual equation that tracks the movement of the wheel up or down. Using a variable in one equation in one instance doesn't account for its overall influence in the system. It might be best for me to re-term my approaches since everyone is so hard set on only accepting them in their real world form.

Base ride height as I define it is the amount of available spring compression travel, not the actual height off the ground. I only term it that way because raising the ride height in the game has a proportional effect as extending the spring compression travel by the same percentage. Read the MR2 example in the OTI post for more context and proof if you want to recreate the scenario for yourself.

To determine motion ratios in the game I construct test scenarios around raising ride height and making comparable changes to spring rate at given base rates to produce similar handling qualities. This is how I identified that 100 is quite safe to assume, and 50 on the Formula cars. That's not to say others don't exist in the game.

Also, if the tuning system in this game was anywhere near close to representing reality then any of the alternative tune calculators around here that rely on real world equations for their basis should have become the norm and able to produce stable tunes a long time ago. My willingness to think outside the box is the only thing that has gotten me this much headway into deciphering its non-conventional approach.

As for feel based tuning, the whole demonstration was only for the maximization of grip. It had nothing else to do with what combination of sway bars and damping settings and spring rates actually produce a faster car on particular track. This is still very much a trial and error system. The particular sub equation I am singling out has only to do with the oscillation and squat resistance in relation to damping in achieving maximum grip on the tire contact patch. You can clearly see the fruits of my labor by comparing the pre-tuned settings with my tuned settings if you look at the tire temperature/wear indicators. The wheels will retain darker shades of blue for a longer period after reaching this harmonic intersection, much as if they now had the warming characteristics of a hard compound tire instead of a soft.

budious · Apr 19, 2011

Mine's BNR34 Skyline Efficiency Evaluation

OTI equations provided by budious

Let's try this again for the doubters on my process. This time I have chosen to break down a stock tune on a high-performance pre-tuned premium model. The stock setup on the Mine's BNR34 Skyline laps Tsukuba smoothly and consistently at around 58 seconds. So let's examine the stock setup using my efficiency equation.

Stiffness Factor of Spring Rates ^ Natural Log of Anti-Roll Bar
----------------------------------------------------------------------
Sum of the Damper LOGs ^ Natural Log of the Anti-Roll Bar

or

(SF(SR)^LN(ARB)) / (LOG(Extension) + LOG(Compression) ^ LN(ARB))

Mine's BNR34 Skyline GT-R N1 Base '06 (1340KG)

(Stock Setup)
Ride Height 0 / 0
Spring Rate 12.0 / 11.0
Extension 7 / 7
Compression 7 / 7
Anti-Roll Bar 4 / 4

Stiffness Factor of Spring Rates, SF(SR)

(12.0 + 11.0) / (1340/100) = 1.716

Damping Efficiency Factor, DEF

LOG(7) + LOG(7) = 1.690

Stabilizer Efficiency Factor, SEF

LN(4) = 1.386

Suspension Efficiency Evaluation

1.716^1.386
----------------
1.690^1.386

2.11474
----------- = 1.02157 (~102%)
2.07008

Simply enough right? This isn't a tuning process, it is a tuning validation of the efficiency of certain combinations of settings. Don't confuse it as such. This will make your tuning more efficient if you comprehend and use it but determining the combinations that produce the best results on a given race course is still a manual process, though it could be easy to test and define some presets in the future. However, you can use this equation to test the efficiency or the complementary recommended settings to the desired spring rate or stabilizers you choose to run on the car.

VelociRaptor · Apr 20, 2011

Really interesting ideas.
It would be fine to be able to evaluate a setup.
Big work to be able to develop an abstract numerical approach, you should have tested a lot of cars !!!

Trying to understand the approach, i have some questions, if you could spend some minutes to analyse them.

Generic formulas on the first post:

budious
Stiffness Factor of Spring Rates; SF(SR):

The formula taking motion ratio into account

( ( FR SR / FR MR ) + ( RR SR / RR MR) ) / ( WT KG / ( ( Base RH ) + ( RH / ( FR MR + RR MR ) )

You may have forgotten a "/ 2" on the sum of the MR otherwise i can't find your +20 RH on the RX-7 application (it should be +40).

------------------------

RX-7 Tuning Equation

budious
This poses a problem because the spring rate stiffness factor, or SF(SR), is below the ~1.4 threshold we are looking for.

Can you explain how you selected 1.4 ?
Is it because it's the good number to reach 100% on the global efficiency formula ?

budious
8.4 / (8.4 + 6.3) = .571 x 1260 = 720 KG x 1.4 SF(SR) = 1008KG

6.3 / (8.4 + 6.3) = .429 x 1260 = 540 KG x 1.4 SF(SR) = 756KG

Why are you using 1.4 as a factor knowing that the basic setup has already a stiffness of 1.167 ?
It seems that you consider the basic setup as a base 100 value and not 1.4/1.167.

budious
For most cars, if you performed weight reductions and the car was 12% lighter afterwards, then reducing ride height by 12mm would equate the same relative spring rate.

Considering most cars with MR 1:1, i suppose the formula will be modified in order the integrate a ratio between Stock Weight and Reduced weight (assuming weight distribution is not affected by weight reduction operations and excluding change in Dynamic weight transfer)

Great Job 👍

budious · Apr 22, 2011

VelociRaptor
You may have forgotten a "/ 2" on the sum of the MR otherwise i can't find your +20 RH on the RX-7 application (it should be +40).

------------------------

RX-7 Tuning Equation

Can you explain how you selected 1.4 ?
Is it because it's the good number to reach 100% on the global efficiency formula ?

Why are you using 1.4 as a factor knowing that the basic setup has already a stiffness of 1.167 ?
It seems that you consider the basic setup as a base 100 value and not 1.4/1.167.

Considering most cars with MR 1:1, i suppose the formula will be modified in order the integrate a ratio between Stock Weight and Reduced weight (assuming weight distribution is not affected by weight reduction operations and excluding change in Dynamic weight transfer)

Great Job 👍

There was no need to use "/2" on the RX-7 because front and rear were both MR of 1 so I eliminated the extra step from the problem. Sorry if that caused confusion.

Edit: Actually you probably meant I should have ( ( ( FR RH / FR MR ) + ( RR RH / RR MR ) ) / 2 ) in the denominator I am solving for. I probably should have included that but it slipped my mind. The jest of the solution is to set the numerator to a common base of which I convert to the standard 100mm base; then I do likewise for the components in the denominator that have not yet be converted to a common base, in this case the changes in the ride height front and rear divided by their respective motion ratios yields the stiffness factor of those incremental changes even if offset for rake. However, for the time being, I try to avoid rake as using it shifts the natural weight balance forward or back which should at least partially influence the supported weight distribution depending on the rake angle.

Corrected form then for Stiffness Factor of Spring Rates:

( ( FR SR / FR MR ) + ( RR SR / RR MR ) ) / ( WT KG / ( ( Base RH ) + ( ( ( FR RH / FR MR ) + ( RR RH / RR MR ) ) / 2 ) ) )

That's one of those things that I can keep straight in my head but I lose in translation when I try to explain it to others. Furthermore, after an additional half dozen edits, the number and order of the closed parentheses may even be correct now...

I selected 1.4 in the particular example because the general rule outside of any equation is a three-way comparative relationship. My attempts to construct formulas has been to uphold this rule though my approaches may be flawed, the general concept appears fairly solid.

Stiffness Factor of Spring Rate greater than or equal to Damping Efficiency Factor greater than or equal to Anti-Roll Bar (Stabilizer) Efficiency Factor.

SF(SR) >= DEF >= SEF

Follow that general rule of thumb and you are on the right path. Remember this approach is assigning the minimum boundaries for the relationships between the suspension settings, not the maximums. The values themselves are proportional to the other components in the equation. Example, you may want stiffer springs with strong dampers but softer anti-roll bars on a flat corner and smooth course like Tsukuba; but at Deep Forestyou may want softer springs with near equivalent damping and anti-roll bars strong but not exceeding the factors of the other two components.

Tsukuba SF(SR)=2.0; DEF=1.9; SEF=1.0 (random figures)
Deep Forest SF(SR)=1.5; DEF=1.5; SEF=1.45 (random figures)

For weight changes just plug in the current weight of the car after any level of weight production you have done. For simplicity of the examples I mentioned before I had used stock figures and minimally tuned variances to keep control factors within a narrow range. Proportional is always proportional because it has to do with the length of the suspension travel available, a change in the weight will not affect the percentage of support provided by every mm of spring rate.

Keep in mind this was an open demonstration on my current approaches and should not be implied that my current equations are perfect; especially the overall efficiency calculation. However, the relationships between the three variables I have identified and equations for calculating each respective sub components' factor seems to be pretty accurate. My intention was get more people to review my work and build upon it because I don't have the time to invest into it further at the moment.

Think of this task as defining physics for a virtual world in which you cannot see the source code; your task is much like that of Sir Isaac Newton when he could only make observations about how gravity worked and produce sets of experiments based on control conditions and then extract equations that can reproduce those observations. In both senses you are blind and cannot confirm your accuracy other than by peer review and repeated testing.

VelociRaptor · Apr 23, 2011

Thank you.
I understand some more things now.

I feel this new formula :

budious
( ( FR SR / FR MR ) + ( RR SR / RR MR ) ) / ( WT KG / ( ( Base RH ) + ( ( ( FR RH / FR MR ) + ( RR RH / RR MR ) ) / 2 ) ) )

much better.
Especially, i have a courage Judd that should use different motion ratio Front/Rear.

I had not understood this:

budious
SF(SR) >= DEF >= SEF

Anyway, you should expect to end with SF(SR) close enough to DEF has their ratio shall be close to 1 (stabilizer has no influence on the last formula for global efficiency as x^n/y^n=1 means x=y). The former one with sum of LOG divided by LN seemed good also except for extreme values of 1 (i adjusted a bit some of my tunes to this with good results).

---------------------

Not been a tuner as you (i had good results with some cars but it's much trial and error), i have to work with existing Garage tunes to check what your model says (it's another way).

Anyway, from some weeks, i have done several tests that are questionning me.
I don't race cars in real life so i have no barrier in the game to compare with.
I have read your answer to some questions relative to real life methods compared to your approach.
To go in details, i started with oppositelock Quicktune sheet and i think that the game does not reflect so much FR vs MR real life behavior (considering oppositelock tried to simulate this).
I start always with FR configuration in the sheet now and i have done comparisons with different Garage Tunes, it's the closer solution with smaller deviations (sometimes need to mix Damping and Right Height stiffness but it works).

Finally, i also read somewhere that in real life you should not use Ride Height to adjust balance but my 3 prefered tunes from Garages are using 2 or 3 mm higher Rear than Front (For now i stay level on my tunes).
Those cars are really giving you smooth feedback and gracefull rear grip lose even on race soft (i hate thoses grip cars leaving you on the sand in a heart beat).
So, i'm not so far to think that it's not so bad in the game to use a little bit ride height balance to tune (we are looking for a game model that give an efficient answer to PD algorithms).

I will make a numeric model and check the formulas against Garages tuned cars to give a feedback but now i am going on week-end break.

Thanks again.

budious · Apr 23, 2011

[P] McLaren MP4-12C '10 (tuning championship entry) @ 1119KG

+ all aero kits @ 0 / 13
+ all weight reductions
+ chassis reinforcement
+ lsd @ 24 / 21 / 13
+ fc tran @ 267mph auto
+ brakes @ 5 / 5
+ race hard tires

Ride Height: -3 / -3
Spring Rate: 10.0 / 13.0
Extension 10 / 10
Compression 10 / 10
Anti-Roll Bar 7 / 7
Camber 2.5 / 1.5
Toe 0.00 / 0.00

Probably a lot of people wondering about the extreme settings I used on my entry and they are very specific to particular types of courses. Since the test courses indicated for this round were R246, Grand Valley Speedway, and Cape Ring Periphery, the settings were tested there and showed good results. They may be complete crap on Tsukuba, Fuji, or anywhere else for that matter.

Quick break down using my equations as follows:

SR(SF) = (10.0 + 13.0) / (1119 / (100 + (-3)))
SR(SF) = 23.0 / 11.5360824742
SR(SF) = 1.99374441466

DEF = LOG(10) + LOG(10)
DEF = 2

SEF = LN(7)
SEF = 1.94591014906

So in the chain comparison using my general rule of thumb; SR(SF) >= DEF >= SEF
1.994 >= 2 >= 1.946

The SR(SF) was fairly close as without drastically changing the entire setup due to the increments of tuning available. For all intensive purposes we'll just round it up to 2.0 and consider it a met condition of being greater than or equal to in the comparison.

Suspension Efficiency Balance (one theoretical version)

SF(SR) ^ SEF
--------------------=
DEF ^ SEF

1.99374441466 ^ 1.94591014906
--------------------------------------=
2 ^ 1.94591014906

3.82939256555
--------------------
3.852807616

= 0.993922600657 (~99.4%)

Noting possible influences on this equation, I noticed that approximately after the 700 mile accumulation mark that my initial tune in this extreme began to exhibit a slight turn towards instability. There is definite cause to consider the GT Auto chassis maintenance service important even with chassis reinforcement installed and at some future point to attempt to compensate for those factors in the equation/comparison model. Also, on my previous tune of this car the settings were pretty similar but I had used only extension 9 / 9 instead of 10 / 10. Since I used the same base car purchased earlier and accumulated more mileage upon it, my feel for the car may have been affected as well. This may indicate some previous chassis rigidity amplified the DEF and as the chassis weakens so does its amplification of DEF. Food for further thought.

Adrenaline · Apr 27, 2011

Lexus LFA
Road Course - Indy

4th Deep Forest Tunery
budious never actually finished this tune and to be honest, it's rather noticeable. This was the... not hardest, but... took the most work, to get around the track. I felt like I was babysitting. I had to be overly focused, up on the wheel at all times and just had to try too hard. It ran some very quick laps, but in all honesty, I'd take a slightly slower tune, just for the peace of mind. Lose your concentration on this one and you'll pay for it. It has a very loose feel around the entire track, it looks like it really just wants to slide around. It's rather difficult to get the gas peddle down, it lacks forward bite and rear traction. Although I do think that 'soft' is the way to go with the springs on this car, but I think this was a bit too soft. Lack of camber and toe, no LSD or tranny gears due to the incompleteness, but I wanted to include it anyways. I'd encourage budious to come back to this tune, as he has a solid platform going, it just needs to have some rough edges smoothed out and completed.
Best 3:
1:27.831
1:27.719
1:27.536

Full comparison here: https://www.gtplanet.net/forum/showthread.php?p=5252805#post5252805

EivlEvo · May 3, 2011

Greetings Deep Forest Tunery:

I thought I'd message you to let you know that I just reviewed your RUF Yellowbird in my Tuner Review thread (see my sig).

Normally, I would post it in this thread, but since it was a 6 way tuner review the post got quite extensive.

If you have any questions, comments, or concerns... please don't hesitate to notify me.

Cheers!

nomis3613 · May 7, 2011

Re "How to select and optimize drivetrain components, and the not so obvious effects":

Thankyou so much for sharing this information. When I first read it here, I thought you were crazy! But after some testing, I gotta agree with you now. On a very tempermental MR that I am struggling to tune, I found that ripping off the drivetrain upgrades made the power delivery a lot smoother, so the car is much more predictable now. With negligable loss of acceleration and top speed.

Some questions please, on a few minor points:

budious
If the driver disengages the clutch entering a corner to downshift, the following re-engagement of an engine idling at high rpms will send torque straight to the wheels, even if no throttle is being applied. The effect once transfered to the drivetrain can be mitigated by a high acceleration value on the LSD.

Why LSD Acceleration, not Deceleration? I'd have thought when downshifting, the engine braking would cause a deceleration torque? If you are familiar with heel-and-toe braking, it is like you are suggesting that PD have programmed in a slight over-rev in the rev matching?

budious
Since the idling speed of an engine is influenced by the flywheel equipped, choosing the correct flywheel becomes part of the overall equation, higher idle is not necessarily better.

I don't see how idle speed would affect handling, in racing you're always well above idle?

budious
Symptoms of clutch engagement being counter productive can also include lurching in which a sudden change in acceleration or deceleration becomes apparent
...
following a release from full throttle to no throttle; the sudden transition will result in rapid deceleration caused by...dropping off of engine rpm's before the clutch finishes fully disengaging, a result of too much friction. This will be more pronounced on engines equipped with lightweight flywheels.

In a transition from WOT to closed throttle, you go from full engine power to full engine braking, but I disagree that the clutch is being disengaged (unless you gearshift). My experience in GT5 is that the stock clutch gives smoother power delivery through gearshifts, but there is no difference when a gearshift is not involved.
*edit* ahhhh... you're talking about AT, yeah I agree it might be disengaging the clutch. Now I get what you're talking about. IMHO it doesn't apply to sequential MT though.

Totally agree with you about the flywheel.

budious
the combination of a semi-race flywheel with an LSD acceleration value of 50, a car can power oversteer its way through corners; whereas, when paired with a stock flywheel it would just run wide into the barrier or grass. For purposes of this guide, I refer to this behavior or running wide as snagging.

Totally agree, I found myself needing to bump up LSD Accel to compensate for the stock clutch.

budious
The test car displayed a pertinacity for popping wheelies at the crest of the hill exiting the first hairpin on Deep Forest Raceway when equipped with single-plate or twin-plate clutches.

Not saying you're wrong or anything, but I found a difference in wheelspin over bumps due to flywheel changes, but not clutch.

As for Surging, again it comes back to our different beliefs about whether the clutch is ever disengaged without a gearshift. *edit* ...and whether you're running AT or MT.

budious
Barreling is evident on the test car when equipped with a single-plate or twin-plate clutch, an upgraded flywheel, and an LSD acceleration value of 50. Barreling is an extreme form of snapping that could be characterized as having the feel of Skid Recovery Force with an exception that there is only a 50/50 chance that it will be in the direction that you wanted to go.

This is gold! I think in the past, I've mistaken barreling as a suspension fault (such as uncontrolled weight transfer). From now on, I'll always check out the drivetrain upgrades, too. Thanks for the info!

budious
Carbon drive shafts...shaved almost two full seconds off of lap time alone in place of the standard drive shaft while using stock clutch and stock flywheel; there was considerable improvement in both acceleration and deceleration.

Did you notice any drawbacks? In theory, it would have a very similar effect to a lightened flywheel (reduced drivetrain inertia).

By the way, I now totally agree about jerking and veering.

budious
My theory would be that the clutch shift interval (CSi) is present, and that each clutch upgrade fills this duration with a wider range of friction values.
...
...for each upgrade the friction spectrum is wider and longer; and thus amplification of torque transference during periods of slip occurs.

Just a thought, maybe LSD Initial could be used to mitigate the sudden torque changes of upgraded clutches.

Again, thanks for the epic post, it was really helpful.

Adrenaline · May 31, 2011

Why did you delete all of your tunes...

Deep Forest Tunery [CLOSED]

Adrenaline

budious

elric718

budious

budious

elric718

budious

budious

budious

Rotary Junkie

budious

budious

VelociRaptor

budious

VelociRaptor

budious

Adrenaline

EivlEvo

nomis3613

Adrenaline

Similar threads

Latest Posts