A-850 Porsche 996 GT3 - tune evaluation please

[feuerdog]

tires 29/29
camber -1/-1
toe 0/-.2
caster 4.5
arb 16/16
springs 410/640
rebound 5/7.6
bump 4.1/6.4
brake 50% @ 100
diff 40/40

Racing Block upgrade only
ALL Racing Handling except for Sport Weight Reduction
Full modified width, Bridgestone, Sport tires
57c Gram Light wheels

FD 3.4
1 -2.74
2 -1.88
3 -1.44
4 -1.18
5 -1.0
6 -.86

The gearing may be a bit tall for non-auto-shift drivers, but at least it will be easy to drive.

Like all my tunes, this is a generic tune, tested for Tsukuba and Nurburgring, drive it anywhere you want.

The camber is a little high for good benchmark values(-.8/-.8 gives better benchmarking numbers) but I like the heating characteristics better with the slightly higher camber.

It should feel responsive on turn-in, very stable mid-corner, with a progressive break to throttle oversteer. It corrects quickly with throttle lift and/or feathered countersteer. Braking should be a touch early but hard and aggressive, with a late turn-in to late apex with aggressive acceleration. I expect some in understeer comments, but if you drive aggressively you can drive around it without losing its stability.

Thanks for any feedback.

[Fit4aking]

May have to give yours a shot seeing as I have one of my very own tunes on the same model.  Completely different tune and build though so it may be apples and oranges.

http://forza-tuning.net/index.php/topic,648.msg13244.html#msg13244

Several differences in tune slightly lighter springs with higher damper rates.  Reversed sway bar values. Completely different differential values.  My diff values seem to be out of whack from the norm.  I like off throttle oversteer so I tune it into the cars using the diff.  I'll see if my Arcade box has a FM2 account with some extra cash.

[Fit4aking]

Built it and drove it around Tsukuba for a while.  Understeer is evident but manageable, seems like a very safe tune to putz around on but requires a lot of aggression to make fast.  Personally I don't like the lack of oversteer but its purely my preferred driving style. Hard to maintain a solid line and hit apex's repeatably.  Just my thoughts.

[feuerdog]

Thanks Fit, would you mind doing a a few more laps for me?

Just change the Rebound though, to 5.5 front, 7.0 rear, or something close to that that will work better for you.

I don't recall, do you use a wheel or controller?

[bimmerlovere39]

Just glancing over this, I'd try a lower or higher diff decel - say, 25% or 90% - to get a little extra bit of rotation going into the turns. 

911s are strange beasts when it comes to diff settings.  I've run everywhere from 0% to 90% decel, the extremes of which seem to work, the middle of which never seemed to work for me in the 911s.  If you're really driving on-edge and trying to get a lot of rotation out of her on turn-in, run really low decel, but if you're running a 911's textbook line (STRAIGHT line braking, coast to apex, hard throttle) a really high - think 90% - decel setting can prove really useful, as it lets you throw a lot of weight on the front wheels on the brakes without having the rear dancing around on you.

The other thing that looks a little odd is your bump/rebound ratio... seems a bit close, IMO.

[feuerdog]

I've always used relatively basic rebound calcs, generally in the bump x 1.1~2 range.

IRL Porsche 911 sport damping ratios are similar to this low ratio 1:1.2'ish. Of course FM is far simpler a suspension model, but the settings work for the most part. I strongly feel that it's the ratio of values from front to rear that are more important than the values themselves.
The modified values I mentioned for Fit will make a huge differences in rotational stability IMO, making the 911 chassis generally looser than I am comfortable with(especially on speedy tracks like Nurburgring).
Rebound values of double the bump are common for me when I do on-the-fly tunes, so like I said, it's not the values themselves that are critical, just the ratio relationships.


Your diff. comments are interesting to me Bimmer. I tend to notice the midrange diff settings are more useable than thier extremes. I used to run a 60% decel diff. on this car, but for it got a little drifty on aggressive trail-braking turn-in for me(I know, trail braking 911's is bad). The drop to 40% decel helped tame that a bit without degrading braking performance(a 911 strongpoint). High speed throttle lift sweepers with high rear decel values(75%+) tend to slip me off into lawnmower land,....especially in a car with no additional aero.
The same low-to-midrange values work for accel IMO. 60-80% will definately break the rear around on exit if you are not careful.
Then again, my setting preferences are probably based on my aggressive ham-fistedness with the controller. Which might expalin why you guys tend to have more "knife edge" tunes than I prefer.

[bimmerlovere39]

Interesting.  I've always run bump * 1.5 to 2 for my baseline shocks, then worked from there (usually 4.0 to 4.5 for bump and 6 to 8 for rebound, with fine tuning from there.  Then again, I'm a big believer in shock tuning, and after getting all settings roughed up, the vast majority of my tuning is in minor ARB (<0.5 from baselines) changes, diff tweaks, and a lot of time with shocks (and ride height - I'm one of the few people I know who actually raises cars with any degree of regularity).  I've found that these changes can utterly change the nature of a car (E30 M3, I'm looking at you), and I'd even go so far as to say that the values make more of a difference than the F/R ratio (on almost all of my cars, front and rear are within 1.0 of each other on any given shock setting.

Then again, the majority of my cars are front engine, rear drivers.  On my 911s and my FWDs, there's a significant gulf between the axles.

Interesting little difference in philosophy here.

I have a gut feeling that, if I were to drive this thing, it would feel like the wheels weren't really working as hard as they could, and I'd likely lower the bump rates a little and boost the rebound a little.  That said, these shock settings could increase the feeling of stability which, I suppose, is an admirable function in and of itself.

There's no doubt that if you run a high decel, you have to get in to some creative driving in high speed turns, but sometimes a little lift before the turn can help.  This is, admittedly, the weak point with the really strong decel diffs.

As for accel diffs, my settings tend to be 15-35%, 50%, 60%, 65%, or 75%, depending on the car/track.

[Fit4aking]

Ran it with your damper changes.  Immediately the car felt better.  Loved the extra turn in and predictability of off throttle oversteer.  It feels faster but is only about a tenth or so faster in lap times on Tsukuba.

[fndrbndr]

A tenth is huge on Tsukuba, though, if you're consistently one tenth faster.  I'm guessing you're running comfortably under one minute per lap?  (Never tried an RR on Tsukuba, so I'm not sure where the bar should be)

[Fit4aking]

So I decided to run it once more now that I have a bit of time to devote to just racing and not kid sitting as well.  I shaved another 1.100 off of my previous fast time in the F-dog modified tune.  For the record I'm sure its a mediocre 56.580.  The car is becoming familiar to me and I'm consistently running a 56.5xx or a low 56.6xx.  While a hotlapper may hunt for the elusive perfect lap the racer in me sees repeatability, consistency, and predictability.  All very good qualities.  Very stable car, may take it to a different track for a bit and compare notes.

Qr'd the lil monster for the sake of it.  Turns out to be near the QR's of my beloved TVR's so its got to be a good tune. 
RA2 57.579
SSh 73.648
Tsk 56.580
Total 187.807

[Choccy]

Curious about something.

If I tune a car for the Nurburgring, meaning properly biased for that venue only, that tune demonstrates some competence at other similar tracks. Road Atlanta and Maple Valley for example.

If I run a Ring tune at any predominantly flat circuit, the car is, quite clearly, overdamped and handles accordingly. Stiff. Less compliant to turn in, etc.

Likewise, a tune for Tsukuba tends to work well for Sunset Infield or Motegi. But is not well equipped for maximum attack at any circuit with crests and dips.

How are you determining a single tune that works well for two wildly different venues? I'm not saying it's impossible to find a compromise. The sport level spring tune would be such an example.

The final turn at Tsukuba has nearly no surface imperfections(bumps, dips or camber) that would upset the traction of the tire, save the lateral slip that causes wheel hop.

The final turn at Nurburgring is practically imperfect with camber change and minor undulations.

Having said that, I'm trialing your 911 GT3 tune at The Ring right now.

[feuerdog]

- Thx Fit, sorry to keep asking, but do you use a wheel?
I know when Ske used my tunes with his wheel they ended up being way to understeery, but that was mostly due to the more direct control a wheel can give you.


- sChoc, I'm using a 911 as a test platform because it tends to be an unstable and unpredictable platform. I'm working on a new tuning methodology that revolves around integrating the amounts of work the axles have to perform and then distributing that work back into the components equally. The difference with this kind of calculation is that its even more generic than my old calculators, but also accounts for more of the work attempting to be done. A key component is how ARB balance, caster, and accel/decel forces interrelate. Instead of using calculations, this method uses more reference tables.

As far as track use is concerned, I undertand the difference in fine tuning needs for tracks like Tsukuba or Nurburgring, but I think a generic tune is far more capable than some people give it credit for. In some ways, I think Tsukuba and Nurburgring are quite similar. It's flat tracks like Sebring and Silverstone that will change more settings IMO.

I use Tsukuba to tune chassis rotation, oversteer/understeer balance, because it sees some of the hardest, tightest, and diverse entry and exit features of almost any track we currently have. It's also an easily repeatable short track, so its test, tune, retest, retune, cycle is simplified.
Nurburgring is good for high speed stability cornering, but because it's such a unique track I find it's only useful for really finalizing a tunes capabilities. How a chassis reacts to high speed elevation changes, the jumps, and high speed corners being key features here.

I did not tweak the car for anything yet, to be honest, so I trusted slightly in the former tunes damper ratios which I now admit were low and underevaluated by me. What i'm really looking for here is not how I can improve my tweaking, but how optimized I can get my reference tables.
What I was hoping for was a capable handling car, for use on any track, using purely untweaked settings.

You know how I tune, i'm the president of the Generic Calculation Society.


PS: For reference I tested my non-aero 700+hp 911 GT2 S-class on the ring last night, and had some high speed stability issues, so you are both absolutely right, the damping ratios I used previously will need to be revisited.
This should work better as a starting point:
Spring/100 = bump
Bump*1.5 = rebound

[bimmerlovere39]

Spring/100 = bump
Bump*1.5 = rebound

Wow.  I think you may have really stumbled across something there... 

That fits a lot of my tunes frighteningly close, from what I can remember.

[Fit4aking]

Sorry Feuer, I am a strict controller user. I have the wheel but also having two boys under the age of 5 makes having it around more of a plaything for them than me.  I use the controller only because it is easy to hide in the living room when I'm not there.

Good tune by the way. I really am beginning to like the way it feels my only gripe as of right this second is the gearing being so wide.  However as a test car, at the moment, I only tweaked what you asked to change.  Perhaps this little project of yours will satiate me until FM3 arrives in my hands.

[feuerdog]

- Fit, no worries. I am a strict controller driver, I was just curious because I think its an important difference to know.
The gearing is specifically wide(especially in lower gears) to take advantage of torque multiplication at higher speeds. It's part of the tuning I do as an auto-shift driver. It also helps keep the 911 chassis driveable in full throttle conditions. I'm sure if you bump a couple of tenths into the FD you'll find the car will be tuned far more aggressively.

Spring/100 = bump
Bump*1.5 = rebound

Wow.  I think you may have really stumbled across something there...  

That fits a lot of my tunes frighteningly close, from what I can remember.

These calcs are ancient, with the slight exception of experimenting with inverted damping in FM1 and FM2.

When I want to throw a quick tune at a car thats how I do it.
50% of weight * distribution = spring
spring/100 = bump
bump*1.5 = rebound

IRL the bump/rebound ratio is very close to this 1:1.5 ratio.

Based on the info I learned in my Milliken book(http://www.millikenresearch.com/rcvd.html)...

These examples are very rough but:
A street going luxury car(Non-police Crown Victoria) =  1:2
A street touring sedan(BMW M3) = 1:1.5
A track tuned sports coupe(911 GT3) = 1:1.2
A track specific race car(?) =1:1 (or less in some cases, depending on aerodynamics)

Keep in mind these are "ratios" not strengths. So as the cars get racier the required base damping levels increase.

To go back to sChocs question,.....The above example is why I tend to have/want 1:1.2 damper values, assuming of course that FM is realistic enough to emulate realistic suspension dynamics.
The spring values for the most part are determined to support the car for the speed at which it is capable of travelling while retaining a performance level of compliance. It is in this realm of tuning the car to it's abilities(build) and not to the conditions in which its driven(track) that I determine my base calc values.

Remember our weight multiplier calc discussion sChoc? The 50% multiplier you used worked well with my base damper calcs becuase the those base damper values were subsequently based on those spring values. More spring, more damping, less roll, etc. My only concern with the 50% stiffness had to do with the compliance of the springs in rough conditions(curb clipping). The new rough/curvy tracks in FM3 may accentuate my concerns in this situation. To me the 50% tuned cars felt like they "skipped" across bumps more than absorbing them.

The new CoG/weight transfer/tire loading physics may resist overly stiff suspension settings though. At the very least I hope they do for extreme ARB settings alone.

That said, the base weight/spring calcs of my calulator were based on multipliers less than 50% initially, typically in the 32%~40% range. They later evoled into a higher valued results, but added in a variable to soften lightweight cars and stiffen heavyweight cars at the same time. The final versions includedeven more stiffness variables, but I eventuall found my best handling results in the 38~44% range, especially with the spring/100+bump calc. The simplicity of this calc has remained becuase of the limited range and need for simplicity of the calculators. I can add varible multipliers for bump from springs but the end result usually made for some very ill handling lightweight/heavyweight cars.

My damper calcs have not yet compensated for aerodynamics, because I had a hard time realizing any benefit(in FM2) for doing so, not to mention the fact that the v5+ calcs had already gotten far too complex for "ease of use".

"Ease of use" is why i'm attempting a new approach to tuning. I'd like to tune(or offer free tuning) without having the need to open a spreadsheet calculator for every car. I want to return to basic paper and pencil tuning, but with an emphasis on driver/owner tweakage.

PS: Sorry for the walls of text, but its slow here at work.