Physics The Physics discussion thread

[RC45]

Well, I'm willing to help people who wants help. You either don't read or you just have your agenda and you don't want to move from your position. I'm not going to repeat myself for the third time, good luck with whatever you want to achieve.

It does not matter what my position is, either the engine can use other data files with expected results or it can't - if it can't then I guess I need to fudge the tires per car.

So be it.

I don't need someone to tell me what changes to make to each tire on a per car basis to get the feel right, this trial and error method is the same no matter who does it.

I was simply hoping the engine was mature enough to use as-configured tire files where possible and maybe even as-configured wing files if need be

After all, if I use the same 72" API 3D wing element on a Lamborghini or a Ferrari or a Corvette in real life - the wing brings a specific set of down force and drag data with it, that when applied to the 3 different cars will have 3 different results, but not because the WING changed, but because the chassis responded to the SAME wing differently.

If I need to tweak the WING file to get the results I want then the entire simulation is not really modular - and if that's the way it is then that's the way it is

 

[RC45]

You are trying to make cheese on a toaster, it's just not going to work. Use rF2 or Modelica for proper physics simulation.

I see now - this mod was actually started way back in 2009/10 using rF1 - but the visual appeal was a little lacking so the mod stalled. I actually got some very valuable input from Niels back in the day - as the idea was to maybe be able to use the C6 mod as an extension of the C5 mod.

Perhaps I should revisit rF2 as an option - but man I do like the visual element of AC LOL

 

[Fat-Alfie]

Yes, maybe you should. So much anger at Kunos Lighten up feller, this is supposed to be fun

 

[Stereo]

After all, if I use the same 72" API 3D wing element on a Lamborghini or a Ferrari or a Corvette in real life - the wing brings a specific set of down force and drag data with it, that when applied to the 3 different cars will have 3 different results, but not because the WING changed, but because the chassis responded to the SAME wing differently.

That's not true IRL, why would it be in the game? Different bodies are going to have different airspeeds where the wing is, different methods of keeping the laminar flow attached all the way to the rear of the car, different deck angles (and thus different ideal wing angle)... in general terms it'll perform the same, but that can only get you 90% of the way to an accurate simulation.

 

[RC45]

Yes, maybe you should. So much anger at Kunos Lighten up feller, this is supposed to be fun

It is fun - has been and always will be. Life in general is fun.

How can it not be fun? Here we have PC's and sims, semi-sims and sim-cades that allow us to simulate our real life 600rwhp track toys and take the cars around the world on virtual race tours without leaving the house

And to be sure I think will just stick with AC for the mod for now, just to upset folks that might otherwise think running me off will change the fact that AC is not a modular sim engine after all

Now as to Kunos as an organization being rude arrogant and ungrateful for their customer base - nothing I can do about that - you need to address that anger with them

 

[RC45]

That's not true IRL, why would it be in the game? Different bodies are going to have different airspeeds where the wing is, different methods of keeping the laminar flow attached all the way to the rear of the car, different deck angles (and thus different ideal wing angle)... in general terms it'll perform the same, but that can only get you 90% of the way to an accurate simulation.

No, IRL the same wing does get used on different cars with different results. The results are different but the wing is the same.

I believe you are totally missing my point about how the sim engine should handle the file data.

I am in fact using the VERY SAME wing on my car IRL (a C5Z) that has been used on many C6Z's and even at least 2 Ferrari 575's I know of..

The SAME WING. The wing works on all the cars - just differently. But the WING is the same.

So if the sim engine was truly a modular sim engine then I should be able to use the same WING file on 2 different cars.

The wing brings a set of raw down force/lift and drag data based on AOA with it - data that is separate from the changes the car the wing is mounted on will have.

It should be up to the sim engine to APPLY the WING file data to the car and then have the appropriate different results based on the CAR the wing is applied to.

In real life we didn't take a syringe and inject C5 Corvette DNA Data into the wing, the same way C6 DNA Data or Ferrari 575 DNA Data was not injected into the wing to get the results the wing gave on the each car - each CAR the wing is used on caused the wing/car combo to behave the way it did.

The WING is the same wing. But the results are different due to the car.

Therefore the sim engine should APPLY the wing data to the car and then compute the resulting wing/car combo

If it doesn't and the WING data needs to be tweaked for each car then again this is NOT a modular sim engine.

To be sure - the wing doesn't bring different data to the each car - it brings the SAME data to each car, the DATA just has a different impact on the car - this is what the sim engine is supposed to handle, the WING file should not need to be changed to correctly reflect the change.

If it does need to be changed then the sim engine is either broken or deficient in that area.

The only time I would expect to have to edit the WING file is when a change is made to the wing itself, perhaps a temporary change like AOA (which really should be handled by an in-game slider to make changes at run time) or a more permanent one, like in REAL LIFE when I added a Gurney flap and changed the side plates on my wing. Now the wing has a different set of DATA due to these physical changes.

Later the Gurney flap was removed and the side plates changed back and the behavior of the car went back to the previous behavior because the wing was changed back.

Again, this behavior change is what the sim engine should be handling, not the WING data file.

But either way, I have already conceded that if AC requires me to tweak per-file per-car all wing, tire, engine etc. data then so be it. It just means its not a modular sim engine, that's all

 

[David Dominguez]

I think even f1 simulators would disappoint you.

 

[William Levesque]

It's all very nice and stuff but.... How do you even assess what feels "right" ?

 

[Andy_J]

To be honest, if a race game feels ok, then that's enough isn't it?

 

[RC45]

It's all very nice and stuff but.... How do you even assess what feels "right" ?

Great question - should it be lap times (assuming the tracks are accurate) real life vs sim life?
Should it be raw performance date results - 0-60, 60-0, top speed, slalom & skidpad?

More likely it is again totally subjective based on "the feel" through the wheel and visual cues sitting in ones den in your sim cockpit.

For me its been whether I can push as hard as I feel I have pushed in real life with my car and if I can feed in the same responses and get expected control of the in-game car.

Not having a C5 in AC its been a bit hit and miss and hence the mod. When I tried the in-game RT12 I was basing the feel on what I recall from the track time I spent with the real car - and granted it wasn't back to back (in fact its been 10 years since I was in the RT12 so its a 2006 AWD model I am basing the "feel" on and it was only 1 track and only about 150 laps over a weekend - but after about 100 laps you begin to 'get' the car ), I felt the steering feedback was bit numb in AC so I ended up tweaking the steering rate to get the "feel" I remembered, based on how much control I could keep over the car when pushing hard.

But again, purely subjective stuff here LOL - so what is the proper measure of "feel"?

To be honest, if a race game feels ok, then that's enough isn't it?

Yes - that is the only yardstick we have.
Although I have always looked forward to more - I have always wanted a 'game' where I could have my uniquely modded car represented for my own personal enjoyment on tracks I may never get to in real life and for testing future mods and upgrades before laying out the cash IRL

 

[RC45]

I think even f1 simulators would disappoint you.

Not if it was built on rFPro it wouldn't

 

[garyjpaterson]

Can't wait for some to post a genuine physics question to justify the creation of this thread.

 

[SportRotary]

@RC45

I'm still genuinely curious as to what is lacking in the "feel" of your C5 model when you use real-world parameters. How does the steady-state balance of the car (understeer/oversteer) compare to your experience? How quickly does the car react on turn-in? Can you feed in power on corner exit like in real life? How does the car react to curbs/bumps?

If the above characteristics don't match reality, you have a lot of knobs to turn and you need to make sure each parameter really matches reality: wheel rates, damping (don't forget to factor in the motion ratios), suspension kinematics (roll center heights, roll center migration in bump/roll, camber curves, trail/SR, caster, KPI), ARB rates @ the wheels, CG height, CG location, etc. etc. Representing each one of those parameters by lines in a text file isn't trivial, and I would venture to say that even the most skilled race engineers would spend well over 100 hours measuring, testing, and characterizing those parameters to input into a sim. As an added complexity, obviously AC isn't running an FEA in the background, so the compliance of the parts must be accounted for (the stiffness of the ARB bushings must be accounted for when calculating the stiffness of your ARBs, for example).

If you think there's really a major flaw in AC's physics engine, list out the parameters I described above and prove that they match your measurements. Measuring the length of the control arms and taking some pictures of the suspension assemblies is maybe 5% of the measurement work that's needed. You need to present more objective data if you want anyone to take your argument seriously. If you can make a case that you've accurately measured all of these parameters, and that you've accurately represented them in AC; then we'll all bash the AC physics engine together.

 

[David Dominguez]

Not if it was built on rFPro it wouldn't

You are basically asking for realtime CFD...

 

[RC45]

You are basically asking for realtime CFD...

I am pretty sure that was in the tagline in the original AC promotional video... along with real tire physics etc.

 

[David Dominguez]

You are a funny guy.

 

[AccAkut]

I am pretty sure that was in the tagline in the original AC promotional video... along with real tire physics etc.

I sorta doubt you even know what realtime CFD is... regarding tires, there is a reason all tire manufacturers still employs facilities like this (my desk was in the long white building) for their development and testing, there is quiet a lot of trial and error in use even in the development of REAL tires.. so don't lament the need to bump numbers in an entertainment product.

 

[Amir Vodokotlic]

What is your agenda?

Physics doesn't seem to be it.

 

[RC45]

I sorta doubt you even know what realtime CFD is... regarding tires, there is a reason all tire manufacturers still employs facilities like this (my desk was in the long white building) for their development and testing, there is quiet a lot of trial and error in use even in the development of REAL tires.. so don't lament the need to bump numbers in an entertainment product.

And this is a moot point in the context of this discussion - the dev of a GAME claims their software simulates what it doesn't. This is borne out by the requirement to fudge numbers per car.

If it is hard to simulate tires very well, then don't claim you can. Simple as, really.

I am not the one hawking a toy racing game as a simulator - it just turns out I was suckered into spending money on the game - which was before I got to experience the rude arrogant personality behind the name. Had I read one of Kunos condescending posts before i had bought the game I would not have.

But since I have spent the money - I will do what ever I darn well please with the game - even try my hand at modding for it.

And exactly why do I need to care about real-time CFD? I just rely on the vendor of the product I buy to do their own analysis and publish the results after the fact.

Mod Edit: Personal insults not needed.

 

[RC45]

What is your agenda?

Physics doesn't seem to be it.

My agenda? I should ask Kuno the same question - if he 'hahas' or 'agrees' to another post in this thread I might have to report him for stalking...