Physics The Physics discussion thread

[Michael Hornbuckle]

rF2 physics set the bar too high, so it's normal to get disappointed when they sell you a game with the simulator word on it.

Weird that a handful of people actually think this way when in reality rf2 physics can't replicate a modern LMP1 car, or any car with rear steering, or even an H pattern transmission.

 

[Gevatter]

I think when people say rF2 physics they strictly mean rF2 tire model. Seems to be complicated enough that Studio 397 consider reintroducing the rF1 tire model to help out the modding scene.

 

[Jim Liak]

I truly hope that this thread will continue in a more civilized manner from everyone.
Please @RC45 when you return chill.

This thread can hide a treasure hidden.

We have a modder with real car and measurements.
We have the elite of AC's modders.
And we have the attention of Lord's himself.

Why don't let this thread, re-spin and become the physics thread that will fill the gap of somewhat document lacking?

Please @RC45 , just post your actual data and settings and let everyone that has the knowledge to provide instructions so that your car becomes as best as possible.

I read whole thread and what I understood is that you can adjust geometry as real car and then apply tyres.
If you done 1st part correct then you only need correct tyres.
If you have wrong tyres you have the chance to get helped to put V10 tyres.

I would really like to drive your mod.
Please make it happen. It's up to you. Don't let those hundreds of man hours goes to vain.

P.S. Sorry for my bad English

 

[garyjpaterson]

Well in order to bring this thread back to life in a whole new way, I have a physics question!

It might be a stupid one but its a question nonetheless, and I'll actually appreciate help from those with greater knowledge than me (almost everyone) rather than just throw it back in your faces

Anyway, I'm a bit confused about tyre loads, specifically the FZ0 value; From what I can gather, this is the reference load on each tyre that all the other parameters use to calculate from, so it makes sense (to me) for it to be a fairly typical load that the tyre would be expected to be subjected to.
Please correct me if this is wrong^.

My question is about finding out this number - first what unit is the FZ0, NM? And then is there a way in-game to work out typical tyre loads from some of the various apps? I believe there are apps, (telemetry app, maybe car physics app or setup app, I can't remember just now) that show tyre loads or something that looks like it, but again I'm never sure what units they are in.

And should I be looking at static loads, or peak loads or something in between?

Again, I apologise for a perhaps silly question

 

[mclarenf1papa]

Well in order to bring this thread back to life in a whole new way, I have a physics question!

It might be a stupid one but its a question nonetheless, and I'll actually appreciate help from those with greater knowledge than me (almost everyone) rather than just throw it back in your faces

Anyway, I'm a bit confused about tyre loads, specifically the FZ0 value; From what I can gather, this is the reference load on each tyre that all the other parameters use to calculate from, so it makes sense (to me) for it to be a fairly typical load that the tyre would be expected to be subjected to.
Please correct me if this is wrong^.

My question is about finding out this number - first what unit is the FZ0, NM? And then is there a way in-game to work out typical tyre loads from some of the various apps? I believe there are apps, (telemetry app, maybe car physics app or setup app, I can't remember just now) that show tyre loads or something that looks like it, but again I'm never sure what units they are in.

And should I be looking at static loads, or peak loads or something in between?

Again, I apologise for a perhaps silly question

It's just the reference load (load, not torque, so it's just newtons) for the tire; the actual load doesn't matter provided that the grip coefficient used is correct for it. If you don't have the equation for mu (grip coefficient) vs. load I can provide it a little later (not on PC at the moment).

 

[Associat0r]

Weird that a handful of people actually think this way when in reality rf2 physics can't replicate a modern LMP1 car, or any car with rear steering, or even an H pattern transmission.

It's majorly well known that hybrid stuff, and a new drivetrain model among many other things is in the works for rF2. These devs don't take physics shortcuts, they take the time to do things properly, instead of building on shaky physics foundations.

On the other hand, AC lacks basic essentials, such as gyroscopic effects that influence vehicle dynamics, something basic that even GPL had, almost 2 decades ago. It also lacks chassis flex and suspension compliance, torque steering among other things, and these issues alone means you can't take any AC car seriously, let alone any historic car, FWD car, or cars that get airborne. It also lacks an upgrade system we had since SCGT.
AC also can't simulate moderately stiff dedicated race cars, all its open-wheeler suspensions had to be softened from real world values to avoid exaggerated transmissibility, or even explosions, even its official Ferrari F1 cars. And its tire model lacks thermo-mechanicality among other many other things. And don't get me started on its flawed aerodynamics, lack of engine temperature, radiators etc.

Also don't forget AC's flawed underbody collisions. AC can't be taken seriously for racecar simulation, let alone dedicated racecar simulation.

I think when people say rF2 physics they strictly mean rF2 tire model. Seems to be complicated enough that Studio 397 consider reintroducing the rF1 tire model to help out the modding scene.

rF2's physics is much more than just its tire model, and its complexity comes from the fact that you can use real physical tire data as input, resulting in proper dynamic outputs. Major Simulation Value right there.

 

[Associat0r]

It does surprise me when people expect their $40 game to be as precise as the Red Bull Formula 1 simulators. I don't get it.

You would be surprised to hear (I also just learned) that RBR uses regular rF1 (not even rFPro) for its F1 driver-in-the-loop simulator, and that's majorly cheaper than $40 now.

rF2 physics set the bar too high, so it's normal to get disappointed when they sell you a game with the simulator word on it.

The bar was already set very high with rF1 physics, and rF2 physics majorly raised that bar by at least half an order of magnitude.

 

[RC45]

You genuinely think that F1 simulators have realtime CFD running? Oh dear. Why would they? Once you can get the data from a real wind tunnel why bother? And even if you had a perfect CFD solution (that obviously won't run in realtime) why bother trying to use it real time?

At the end the guy is a bit (ok.. a lot ) bitter but I understand where he is coming from. He's expecting AC to be this "modular physics engine" (whatever that might mean in his head lol.. I can only try to guess what he is trying to express) to allow hack together a couple of numbers from a real car and get the perfect a virtual representation.

This is a common and widespread misconception but it isn't really something to be surprised about. As humans we tend to vastly underestimate the complexity of tasks we don't understand very well; interestingly that is also the reason why programmers get time estimations always wrong and are taught to always "double whatever time you think a task might need".

He managed to argue for 5 pages without bringing on the table one single concern at a level that could have been a starting point of a serious analysis or brainstorming.. nothing at all, which really should say everything you need to know about this whole thread.

So combine genuine and innocent ignorance (we can't be all expert in everything right?), misunderstanding and wrong expectations about a software, macho ego (I got this, I got some numbers but the software isn't doing what I want ergo.. the software is obviously wrong cause I can't be) with an underlying bitterness towards us (the devs and whoever tried to explain stuff to him) and you end up like this. It's all quite predictable actually.

Does your C7Stingray model (using what I assume to be GM provided suspension geometry) run car specific massaged tire data to get the in-game feel it has?

Yes or No.

If the answer is No, then the tire data file should be able to be taken and plugged directly into another car and used "as-is" to represent a set of Hankook (or what ever tire you have simulated) in those sizes and compound.

If the answer is Yes, then this throws into serious question the validity of claims of the accurate nature of the cars simulated in AC.

 

[RC45]

Not at all. He'd broken the rules multiple times, been warned both verbally and officially. I stand by the ban 100%

Yet its ok to be called ignorant, stupid, arrogant by other posters. Have my real life car and driving insulted.

Double standard any one?

I did not say the other bloke was a janitor. And if I did, why is that an insult? Do you and the rest of the RD forum view Janitors as beneath you?

I went out of my way to not insult or use foul language. Management just needed an excuse to ban me and created one from nothing.

 

[Associat0r]

Yet its ok to be called ignorant, stupid, arrogant by other posters. Have my real life car and driving insulted.

Double standard any one?

I did not say the other bloke was a janitor. And if I did, why is that an insult? Do you and the rest of the RD forum view Janitors as beneath you?

I went out of my way to not insult or use foul language. Management just needed an excuse to ban me and created one from nothing.

I've been insulted and personally attacked here many times too and most of the time nothing was done.
I also got my voting rights removed, yet those same people who reported me for haha'ing them always haha me and they even mass haha'ed me in this thread. I should get my voting rights back (I can't even agree or like something anymore), or else those others should get theirs removed as well.

 

[mclarenf1papa]

Does your C7Stingray model (using what I assume to be GM provided suspension geometry) run car specific massaged tire data to get the in-game feel it has?

Yes or No.

If the answer is No, then the tire data file should be able to be taken and plugged directly into another car and used "as-is" to represent a set of Hankook (or what ever tire you have simulated) in those sizes and compound.

If the answer is Yes, then this throws into serious question the validity of claims of the accurate nature of the cars simulated in AC.

It's not massaged. We've told you this time and time again but for whatever reason, you refuse to listen. Kunos doesn't get manufacturer tire data on every car they produce (nor do they advertise that they do). The values could simply be flat-out incorrect (just as tires in rF2 can be incorrect because of the real life data they might lack - in that case inaccuracy is much easier to produce as the variables required for tires are numerous and very hard to acquire under any circumstances).

It's majorly well known that hybrid stuff, and a new drivetrain model among many other things is in the works for rF2. These devs don't take physics shortcuts, they take the time to do things properly, instead of building on shaky physics foundations.

On the other hand, AC lacks basic essentials, such as gyroscopic effects that influence vehicle dynamics, something basic that even GPL had, almost 2 decades ago. It also lacks chassis flex and suspension compliance among other things, and these issues alone means you can't take any AC car seriously, let alone any historic car, or cars that get airborne.
AC also can't simulate moderately stiff dedicated race cars, all its open-wheeler suspensions had to be softened from real world values to avoid exaggerated transmissibility, or even explosions, even its official Ferrari F1 cars. And its tire model lacks thermo-mechanicality among other many other things. And don't get me started on its flawed aerodynamics, lack of engine temperature, radiators etc.


rF2's physics is much more than just its tire model, and its complexity comes from the fact that you can use real physical tire data as input, resulting in proper dynamic outputs. Major Simulation Value right there.

We've worked on LMP2s and LMP1s with real data and have had no issues at all with stiffnesses...also, "flawed aerodynamics"...I'd be interested in "getting you started" on that point because I'm not sure you have even the slightest bit of a valid argument for it.

 

[Associat0r]

We've worked on LMP2s and LMP1s with real data and have had no issues at all with stiffnesses...also, "flawed aerodynamics"...I'd be interested in "getting you started" on that point because I'm not sure you have even the slightest bit of a valid argument for it.

Well, those Kunos F1 cars have issues with stiffness (workarounds don't count), and so does the excellent Formula Renault 3.5 mod (at least 25% loss of stiffness) and the RSR F3 mod that had to be majorly crippled, by taking away the full range of damper valving options as used over a season by real race teams. How stiff is your car compared to the cars I mentioned, and how much damper valving options did you need to sacrifice to get your mod to work?

also, "flawed aerodynamics"...I'd be interested in "getting you started" on that point because I'm not sure you have even the slightest bit of a valid argument for it.

I just leave this here http://www.assettocorsa.net/forum/index.php?threads/oval-tracks-and-ac-physics.33739/
And this is just the tip of the iceberg.

 

[mclarenf1papa]

And this is just the tip of the iceberg.

Go on.

 

[mclarenf1papa]

Well, those Kunos F1 cars have issues with stiffness (workarounds don't count), and so does the excellent Formula Renault 3.5 mod (at least 25% loss of stiffness) and the RSR F3 mod that had to be majorly crippled, by taking away the full range of damper valving options as used over a season by real race teams. How stiff is your car compared to the cars I mentioned, and how much damper valving options did you need to sacrifice to get your mod to work?


I just leave this here http://www.assettocorsa.net/forum/index.php?threads/oval-tracks-and-ac-physics.33739/
And this is just the tip of the iceberg.

Oh and as to not evade your question, one car has nearly 9 hz frequency capability at the front with 2 mil+ ARB and no damper limitations at all.

 

[RC45]

It's not massaged.

Are you Kunos?

Simple Yes or No answer will suffice.

Kunos doesn't get manufacturer tire data on every car they produce

What is your obsession with "manufacturer' tire data any way? It doesn't matter where the data is sourced from.

The question is the same and very simple.

If the tire data is NOT massaged then it should be able to be used AS IS with any other car to represent tires of the SAME size and compound.

If not then the game engines is fundamentally flawed.

If you CANNOT use the SAME data to represent the SAME tire then how can ANY OTHER attribute be taken seriously?

If you are CHANGING tire data to match each car then SOMETHING must NOT be represented correctly.

This is not rocket science, it is simple reality.

And as I conceded previously, if the reality is that every car is uniquely massaged for in-game feel, then so be it - that is fine, just don't pretend this is a proper simulator that can use appropriate input data then.

I wonder if each of the Porsche's in the DLC has been "uniquely tweaked" for feel?

 

[mclarenf1papa]

Guys I give up...

 

[RC45]

Guys I give up...

Why, because my question is valid and you have no proper answer?

Hey, don't feel bad that you have had to tweak all you cars for "to feel right" - embrace the reality.

 

[David Dominguez]

The question is the same and very simple.

It's not because your whole dilema starts from an assumption that you make.

 

[RC45]

It's not because your whole dilema starts from an assumption that you make.

Are the Kunos tires tweaked on a per car basis - even for tires that are the SAME size and compound as others?

Here is the question rephrased in even simpler terms for you:

If we have 2 cars - a Porsche 911 Coupe and a Porsche 911 Cabrio that use IDENTICAL tires/wheels in real life (weight and gearing and engine power might not be the same), would YOU expect to be able to use the SAME tire file for both cars or would YOU expect that the tires would need to be tweaked for each car - even though they use IDENTICAL tires in real life?

Simple question here fella, try answer it please.

 

[PsyaNyde]

Are you Kunos?

Simple Yes or No answer will suffice.


What is your obsession with "manufacturer' tire data any way? It doesn't matter where the data is sourced from.

The question is the same and very simple.

If the tire data is NOT massaged then it should be able to be used AS IS with any other car to represent tires of the SAME size and compound.

If not then the game engines is fundamentally flawed.

If you CANNOT use the SAME data to represent the SAME tire then how can ANY OTHER attribute be taken seriously?

If you are CHANGING tire data to match each car then SOMETHING must NOT be represented correctly.

This is not rocket science, it is simple reality.


And as I conceded previously, if the reality is that every car is uniquely massaged for in-game feel, then so be it - that is fine, just don't pretend this is a proper simulator that can use appropriate input data then.

I wonder if each of the Porsche's in the DLC has been "uniquely tweaked" for feel?

Sorry i'm a big assetto fan but what he says makes sense to me, infact the handling in R3E feels more real and consistent to me once you tweak settings (pity its so expensive in comparison), but what do i know, i'm just a race sim gamer.