GT4 - Redline Specifications

Aleksi wrote:How to count power to weight ratio?

you just divide the weight of the car (measured in kg) by the power (measured in hp). So a car with a weight of 1200 kg and a engine with 300 hp, it's
1200kg/300 hp = 4 kg/hp

Of course, you can use which ever unit for weight and power you want, kg/hp seems to be the norm in Europe.

ok, question for the interested participants about 2 possibilities when it comes to gear ratios:
do you want free gear ratios or do you prefer stock gear ratios? I know of some cars who could really need some custom spec gears....

Now I know that we had a strict selection when doing the TCC cars, with GT3 cars doing with whatever seemed best.

I guess we have 3 options here:

1. We go custom spec with free final drive ratios and free gear ratios.
   
2. We use the production cars final drive ratio and free gear ratios.
   
3. We use the production cars final drive ratio and gear ratios from Hewland LSG-A gearbox: available in H pattern and it can hold up to 600 Nm (is that enought, NNB?). The gear ratios are a bit fuzzy to find, but this might hold it all. If somebody else figures out how to read this site, please say so. It's too late for me.
   
4. We use production car final drive ratios and gear ratios.

So that's 4 options. What will it be?

re: gear ratios

It's clear that a number of the cars in real life are using custom gear boxes. The C6 and Mustang both, for instance. I find it interesting that the C6 has a lower power/weight ratio than most of the cars and seems to be doing well this season. I guess handling counts for something...

Ok guys, time for some action. I prepared a test mule for the upcoming GT4 class, a sort of benchmark car. The problem I'm having is that it is a bit too fast .

I used the old and trusted 968 platform, more specifically the 968 turbo S (without the R).

Changes from the road car include

• lower weight (-70 kg)
  
• slightly tuned engine (+35 hp)
  
• sticky tires (0.17)
  
• adjusted downforce (lowered from -0.125 and -0.15 to -0.04 and -0.05)
  
• new gear ratios using the Hewland LSG ratios and final drives (source)
  
• wheel dimensions close to road car (235/40 ZR18 and 265/35 ZR18 to 235/645 and 265/645)
  
• adjusted loadsensitivity to -0.015 and -0.025
  

Comparing to the qualifying times of this years GT4 round at Silverstone, the test mule seems to be some 4 seconds to quick. This year, a Ginetta G50 got pole position with a time of 2:02, I manage that time in just two laps and most of you are easily some 4 seconds quicker.

Important: You will find the car under "GT4 Test Mule" in the car selection screen.

I also started on a GT4 logo, an idea I briefly discussed with NoNameBrand. I thought we could all use the same door logo, with a space for the start number and a sponsor or team name.
Based on this logo, I also created a logo which we could use in front of the car names, for easier distinction.
All this is included in the zipped GT4 logo file. Of course, I'm no graphic designer, so don't expect beauty. If you come up with something different or you don't like the idea at all, just let me know.

And here we go:
GT4 Test Mule
GT4 Logos

Let me know...

Tomte wrote:
Thanks for the clarification, I feared that the car is just too fast. I had a look at what times we achive with the GT3 compared to real life qualifyung times.... we're a tad too fast there as well, as Dandy pointed out a while ago.
Now it's the question if we ignore it or if I slam the brakes on.

Well, I think, cause we are in early development yet, we can slow down the car. It's not a good start to have the Test Mule already breaking the "speed wall" of other real GT4.

DonaemouS wrote:
Well, I think, cause we are in early development yet, we can slow down the car. It's not a good start to have the Test Mule already breaking the "speed wall" of other real GT4.

We noted for quite some time that cars are quicker in Redline than in real life, with some notable exceptions. While top speed can be matched somewhat easily, the acceleration is mostly too good.

After doing some quick laps with some road cars (968 turbo S and CS, Z3 M Coupe, 911 CS, XLR). It appears to me that the turbo S is a class above the others on this track, probably due to the higher torque.

So I thought I have a go at another test mule, based on Andreu's BMW Z4 M Coupe. For this car, some real numbers exists.
Still, I managed a 2:00 on the third lap....

So I guess we need to go down with the tire stickyness (0.15?) and the downforce (0), but I'm not sure if this will make a lot of difference.

Tomte wrote:
So I guess we need to go down with the tire stickyness (0.15?) and the downforce (0), but I'm not sure if this will make a lot of difference.

I have tested a .car for Nissan 350z with numbers close to real life (when I could find them) and when I reached a handling that pleased me it was faster than the real thing up to 8 secs on all the tracks where I found real lap times. I checked the length of the tracks in reality and RL - they matched
I have tried a higher front air resistance (and lost about 2 secs ?).
I have lowered the torque but this significantly changes only the 1st lap time ...
Downforce was already 0.
Tire stickyness is 0.17 as you told me to do.
I weakened the brakes and lost some time again (2 secs ?) since I have to brake (and go slower) earlier before the turns.
I'm still too fast even though with such settings I now have difficulties reaching the real top speed on most tracks (I can't reach it on "official" GT4 european challenge tracks).

Of course I'm no expert in .car files but the 350z is fine to drive (for me at least) and these are conclusions I made 2 weeks ago ...

Cu this evening on IRC to discuss the matter Tomte

there is a value we didn't touch in all the .car file.

the wheel friction. May this value change how the car handle? (maybe making slower, increasing tarmac resistances, and so on)

DonaemouS wrote:there is a value we didn't touch in all the .car file.

the wheel friction.

I don't understand wheel fricion enough to be willing to change it at this moment. Right now I leave it at 10, but I remember that there where already some discussions about it while balancing the performance of the TCC cars.

On another note, I rechecked the downforce/lift values given on Mulsanne Corner. The model 2000 Porsche 911, the similar aged Audi TT and Mazda RX7 produce lift in the area of 90 to 270 kg at 240 kph, which would translate into lift coefficients between 0.30 and 0.92 in total.
Given that most cars in GT4 don't have any wings which would cancel out or reduce the lift, my assumption of using -0.1 to 0 was probably a wee bit off
Now, we also know that in Redline, gravity is not up to what it used to be, so I wonder how much this influences our discussion here.

In any case, if we don't find a good and comprehensible way of bringing the GT4's performance down to real life, I'm prepared to accept a standard which is reasonable in comparison with Redline GT3, TCC and road cars.

Tomte wrote:I don't understand wheel fricion enough to be willing to change it at this moment. Right now I leave it at 10, but I remember that there where already some discussions about it while balancing the performance of the TCC cars.

How about looking at the test that I did on the matter a while ago?
Car's physics and setting: wheels.friction
Result: a road should have wheels.friction close to 50, if that value is to represent all the power loss after the gear box.

Changing wheels.friction for a GT4 car will have little effect on lap times, because it only is noticable at low speeds. At high speeds, the air resistance becomes the most important factor.

I think the better-than-real-life performance comes from the suspension, sway bars, load sensitivity and tire slip ratio. No easy way out of that mess, I'm afraid. But I agree - test driving a car in any simulation game should result in slower laptimes than in real life, as we're not as good a driver as the race drivers that set the lap record we're comparing it to.

Another idea I'm having is to limit the power to the wheels to say 98% or so (by setting wheels.powered to 0.49 instead of 0.50). This mainly acts on the top speed and a bit on acceleration and costs me some 2s per lap.
But then again, what's the top speed of a GT4? That depends on the gearing obviously.

For now I have 3 test mules in the works, 2 with 6 gears, one with 5. I spaced the gears so that the cars reach about 270-280 kph on a flat road, but on most tracks these cars hardly reach 245 kph. As this mostly equals 5th gear, I wonder if a top speed of 250 - 260 isn't sufficient.

EDIT: I just tried the 98% thingy on the two other test mules. It really cripples the top speed, but not in a consitent manner. Some cars are just a bit affected, others seriously. Seems to depend on the amount of torque available... It was probably a crap idea....

Tomte wrote:I just tried the 98% thingy on the two other test mules. It really cripples the top speed, but not in a consitent manner. Some cars are just a bit affected, others seriously. Seems to depend on the amount of torque available...

The consistency has to do with the differentialLockCoefficient (which has to do with difference in torque). I found that out last year in the early stages of the Ford RS200 development. If the diffLock is above a certain value when the sum of wheels.powered < 1, top speed is crippled.

Tomte wrote:Another idea I'm having is to limit the power to the wheels to say 98% or so (by setting wheels.powered to 0.49 instead of 0.50)

I would hate to see a kludge like that. Have you played with the engine's friction settings? I don't think it makes any significant difference but perhaps a very large increase in RPM and base friction would take the edge off acceleration.

I wouldn't like to play with the friction settings for the simple reason that I use these settings to control the off-throttle and coasting behaviour through corners.
And I think you would need an awful lot of friction to do that, I already tried that once. Of course, one could compentate a bit with engine inertia, but this often has other side effects.

Maybe it has less to do with acceleration but with cornering ability (or wheels.stickyness).

I think I will not follow the wheels.powered settings any further, better to leave them alone. Instead, I'm more and more thinking "what the heck, as long as the performance difference is reasonable to the GT3 and TCC cars, it will be ok."

Tomte wrote:Maybe it has less to do with acceleration but with cornering ability

That's what I believe. For instance, low sway bar settings easily adds two-three seconds to a lap time. (I'm not saying that low sway bars are correct, just that they have a great impact on performance.)

Interesting idea.
In the meantime, I updated the GT4 test pack, now including another old banger of mine and the Z4 M which Andreu was kind enough to share.

3 test mules
Please have a go.

I'd just like to know the development status of the GT4 class.

For what I know, we could have the following cars:
BMW M Coupe (Z3)
BMW Z4 M
Cadillac XLR
Chevrolet Corvette Z06
Ford Mustang FR500C
Nissan 350Z
Porsche 968 turbo S

For now, I have the .car of 4 cars in the works, still trying to slow them down.
While getting the cars down to parity (which is more difficult than I imagined), I'd like to keep them different enough, so that they all have their character. Specifically, I don't want them all to behave tomte-like.

- The BMW M Coupe should be quick, light, but not easy to master.
- The BMW Z4 M is for now very quick, with reassuring understeer.
- The 968 is again just difficult to balance. While having the lowest power/weight, it has one of the best torque/weight ratios. On some tracks it flies, on some it's just last.
- The XLR is the heaviest car with a short transmission, understeers noticeably, but is overall an easier drive than, say, the M Coupe.

On most tracks the Z4 M is the fastest car, sometimes beaten by the 968.
Actually, the Z4 M is annoyingly quick, I will need find some new different ways to slow it down some more. I'm thinking FAR at the moment, which would therefore be above factory specs.

I'd like to get a consensus on the performance vs reality level of the GT4 class:
I see the following options:

1. Lap times as close as possible to real life times, even at the cost of using lower than published specs
   
2. Factory specs all the way thought, the cars just end up where they end up.
   
3. Balance the cars that they are reasonably slower than GT3 and close/faster than TCC, depending on track)
   

Personally, I'm in favour of option three.

Apart form that, I noticed that the UV-Map of the M Coupe is not in a state to easily make a race skin for it, I will need to redo that.

NNB and I agreed that for future Redline Race Classes, the naming format should be: Race class - Manufacturer - Model Name (for example 'GT3 Laborghini Gallardo'), as these cars are meant to be driven against each other, be it LMP1, GT3, GT4 or TCC cars. We think that this should only apply for race cars, not for road cars, trucks or the like.

Feel free to say 'nonsense' or that I overcomplicate things.

Is there any built-in cars that would fit into class with some modifications? I don't have any modeling skills (yet)

Aleksi wrote:Is there any built-in cars that would fit into class with some modifications? I don't have any modeling skills (yet)

I'd say the Audi TT could fit, maybe with a tuned VR6 engine (no turbo/supercharging, just classic tuning like the BMW Z4 M). Which leads to the question if we allow AWD. I would be against it, as in GT3 (where the Gallardo and the Audi R8 had to be converted into RWD only).
Apart from that, a performance limited Viper might be in order as well, but it would need to be severely resticted, similar to the GT4 Corvette C6 (which runs with 390 hp in this series).

Tomte wrote: a performance limited Viper might be in order as well

NOOO!

Vipers should always be about raw power IMHO.