GPL 1967 v3 Dev blog

[PTRACER]

Bottom post of the previous page:

Dont know if this is any help @PTRACER

Maybe good for interest if nothing else. If you havent already seen it.

Lol during my searcI did find a full workshop manual at Hewland Classic on how to change the ratios on a Hewland FG or DG, but dont think that is any help
http://www.hewlandclassic.com/assets/manuals/fg_400.pdf

I have not but it sure looks useful!

The manual, not so much Maybe someone like Doug Nye will have a ratio chart for the BRM....
And I bet the Ferrari data has never been published.

If you are able to tell me a list of gearboxes used in the 67 season that might also be helpful

[mcmirande]

Hi Paul, Horacio Ostoich points me out that maybe in this forum there is info about the BRM

https://forums.autosport.com/forum/32-tnfs-archive/

[psosa]

Hello Paul!

We were checking and as Marcos said, Horacio Ostoich knows the Autosport forum well and there is a post he had read from 2002.

It has some interesting information, you might find it useful:

https://forums.autosport.com/topic/3289 ... echniques/

[PTRACER]

Hello Paul!

We were checking and as Marcos said, Horacio Ostoich knows the Autosport forum well and there is a post he had read from 2002.

It has some interesting information, you might find it useful:

https://forums.autosport.com/topic/3289 ... echniques/

Ah ha, lots of interesting setup information there, particularly about the BRM P83! Thanks Pablo!

[PTRACER]

This week's update:

I have rewritten the "Loss of grip with tread speed" code (featured in most newer mods) to match the calculation used in Assetto Corsa.

Previous calculation: Mμ = Mμ x (1 + (WheelSpeed x GripLossCoeff))
New calculation: Mμ = Mμ / (1 + ((Slip Velocity) x GripLossCoeff)

In this case, Slip Velocity is the WheelSpeed multiplied by the Slip Ratio, while Mμ is the Friction Coefficient. It's a really subtle change, but makes all the difference.

[LasseA]

How different does that make the feel of the car? Is it rather subtle or rather huge?

[mcmirande]

Thanks for the update, Paul!! I can't figure out which consequences that change has in the physics, but "makes all the difference" seems great

Cheers, Marcos.

[PTRACER]

I'm still trying to understand the differences myself, but after 30-45 mins of testing:

1. Below the limit the car is more stable and there is a feeling of more grip in general
2. Steering is less sensitive at higher speeds. Takes a little more effort to get the nose to the apex especially if you enter the corner too fast.
3. You can't just throw the car into the corners and let the friction from the tyres slow you down. That will now result in a spin.
4. Car less able to hold a slide the faster you go. E.g. If you try to powerslide at 180mph you will die.

Haven't found the optimum setting yet so I'll keep at it...

[Duncan]

Hi Paul,

Thanks, I've enjoyed reading about your work here and having the opportunity to think more about GPL's physics and how you are developing the project. Are you still continuing with it?

Thanks, Duncan

[PTRACER]

Hi Paul,

Thanks, I've enjoyed reading about your work here and having the opportunity to think more about GPL's physics and how you are developing the project. Are you still continuing with it?

Thanks, Duncan

Hi Duncan, welcome to The Fastlane and thanks for your interest!

Yes, work has resumed on the project. I took a break for a few months to work on the GPL Addons site, the Slipstream patch, 1955 mod AI patch etc.

To give a more detailed update, I was unable to find the sweet spot with the following code I had written.

I have rewritten the "Loss of grip with tread speed" code (featured in most newer mods) to match the calculation used in Assetto Corsa.

Previous calculation: Mμ = Mμ x (1 - (WheelSpeed x GripLossCoeff))
New calculation: Mμ = Mμ / (1 + ((Slip Velocity) x GripLossCoeff)

When I revisited Assetto Corsa's code, I realised that Slip Velocity was not such a simple formula as [Wheel Speed x Slip Ratio].

Actually it's SQRT((WheelSpeed*SlipRatio*Cos(LongSlipAngle)^2) + WheelSpeed*Sin(LatSlipAngle)^2). I've been doing a deep dive analysis of GPL's tyre code over the last week or two to see if those values can be obtained and I believe they can. Next I'll be writing a test patch to see if this new calculation is any better.

In addition to the above, I've requested permission to add some features in from a patch by another author. I'll let you know what they are if/when permission is obtained...

[PTRACER]

I had a 3 day weekend so you can guess how I spent it.

Firstly, as per the other thread, I learned a little more about the Bump Steer setting which is not utilised in default GPL and found the ideal values to reduce the effect, which makes the cars wander rather a lot less in all driving situations. I recall this particular aspect of GPL's suspension physics holding me back a year ago (I was turning left coming out of right handers and the car wasn't sliding) and I'm glad I managed to find a solution.

Secondly, I can also say the SPEED SENSITIVTY code from Assetto Corsa has been implemented successfully. Since default GPL does not calculate slip ratio properly, I had to implement a new slip ratio formula which is having a positive effect. I've got a formula for Camber Gain from AC so that'll be the next thing I attempt if I can find out where the camber values are stored in live memory.

The one thing I would love to add is relaxation length which will make the car less eager to change direction, but I don't even know where to begin with that. The mathematical formulas presented in various publications are too complicated for me to understand and I'm not sure where I would even begin code wise. If anyone has any ideas, please let me know.

About the features I wanted to add from other patches - one of them was the dirty tyre effect from dirtgearpatch. I messeged Olaf 1.5 months ago and there was no response. (It seems people with coding knowledge don't like to help each other for some reason). Though I can always locate the code and try to implement it by myself.

On a positive note, I purchased a proper copy of rFactor 2 and have been able to decode some of the physics from the Eve F1 and Eve F2 cars. I now have a better idea of the amount of lift late-60s Grand Prix cars will produce at speed and also the weight balance, which again I had got 'right' a year ago and then changed because of GPL acting weirdly about it. Now I know why the issue is fixed.

With each change GPL is starting to feel less and less familiar and although I'm incredibly satisfied with how far I've come, I'm worried others will not feel the same way. My goal will be to release a 2 or 3 car demo publicly before I actually finish working on the full 11-race set that I have planned to gauge interest.

[ROOKIE4EVER]

I'm following your work, even though it's all far too technically complex for me.

[PTRACER]

Today I picked this up for the first time in a while and had the opportunity to finalise some of the tyre parameters and lift values.

One thing that I have found really tough until now is getting the base grip levels just right and mostly because of the behaviour under braking.

The original GPL had very low base grip values, which felt great at high speed, but like ice at low speed. Of course you can increase those base grip values to make low speed driving easier (which is what the 1967v2 mod did), but then high speed grip becomes too great (too easy) and braking distances become ridiculously short etc.

Today I partly managed to resolve the braking issue by introducing the BRAKE_DX_MOD formula from Assetto Corsa. A few further tweaks are needed but it's one step towards resolving the issue.

[PTRACER]

Today I partly managed to resolve the braking issue by introducing the BRAKE_DX_MOD formula from Assetto Corsa. A few further tweaks are needed but it's one step towards resolving the issue.

I'm well stuck into this again. Latest additions:

1. BRAKE_DX_MOD - Adds or reduces grip under braking
2. I've managed to fudge it so that braking has a separate slip curve and lower peak slip angle
3. Tyre stiffness is now dependent on speed, which also affects the radial growth of the tyre
4. I've begun working on the first steps towards a proper brake temperature patch - one where the heating rate depends on the wheel speed, rather than just the car's speed. In other words, locked brakes no longer result in increases in temp. Brake bias will also be a factor and each brake will have its own temp. More on this later!

[Michkov]

Today I partly managed to resolve the braking issue by introducing the BRAKE_DX_MOD formula from Assetto Corsa. A few further tweaks are needed but it's one step towards resolving the issue.

I'm well stuck into this again. Latest additions:

1. BRAKE_DX_MOD - Adds or reduces grip under braking
2. I've managed to fudge it so that braking has a separate slip curve and lower peak slip angle
3. Tyre stiffness is now dependent on speed, which also affects the radial growth of the tyre
4. I've begun working on the first steps towards a proper brake temperature patch - one where the heating rate depends on the wheel speed, rather than just the car's speed. In other words, locked brakes no longer result in increases in temp. Brake bias will also be a factor. More on this later!

Very excited for #4. The lack of individual brake heating is my number one gripe of the 55 mod. Out of curiosity, how does the stiffness correspond to speed? Does it go up or down the faster you go.

[ROOKIE4EVER]

Today I partly managed to resolve the braking issue by introducing the BRAKE_DX_MOD formula from Assetto Corsa. A few further tweaks are needed but it's one step towards resolving the issue.

I'm well stuck into this again. Latest additions:

1. BRAKE_DX_MOD - Adds or reduces grip under braking
2. I've managed to fudge it so that braking has a separate slip curve and lower peak slip angle
3. Tyre stiffness is now dependent on speed, which also affects the radial growth of the tyre
4. I've begun working on the first steps towards a proper brake temperature patch - one where the heating rate depends on the wheel speed, rather than just the car's speed. In other words, locked brakes no longer result in increases in temp. Brake bias will also be a factor and each brake will have its own temp. More on this later!

Interesting in its complexity.