Physics The Physics discussion thread

[RC45]

I mean that's not strictly true....It's better to make an excel (or similar) spreadsheet to calculate it all so you don't have to go ingame to check all the time...plus it allows for more accuracy; the only real possible way to match aeromap data closely is by doing so.

Wow - you couldn't even allow me to compliment AC for a tool well done without having a go.
But then again I never expected the resident 'experts' to allow anyone to just post without tearing into the comment and poster.

At least you are consistent. I'll give you that much.

And I maintain you are not correct. Without loads of actual testing using your spreadsheets of data, s you will never know which numbers to fudge to get the right "feel". Spreadsheets have no feel - hence the need for actual testing as well .

I have about 2500+ in game miles testing various iterations of this chassis - and going strictly by the numbers it never feels quite right. With over 100,000 miles in a C5, about 10,000 of which are on track I know exactly how they should feel at all stages of modification.

Tyres (I was going to fudge the chassis but was previously convinced that the tyres need the attention if you recall) and aero needed fudging to get the feel right.

 

[mclarenf1papa]

Wow - you couldn't even allow me to compliment AC for a tool well done without having a go.
But then again I never expected the resident 'experts' to allow anyone to just post without tearing into the comment and poster.

At least you are consistent. I'll give you that much.

And I maintain you are not correct. Without loads of actual testing using your spreadsheets of data, s you will never know which numbers to fudge to get the right "feel". Spreadsheets have no feel - hence the need for actual testing as well .

I have about 2500+ in game miles testing various iterations of this chassis - and going strictly by the numbers it never feels quite right. With over 100,000 miles in a C5, about 10,000 of which are on track I know exactly how they should feel at all stages of modification.

Tyres (I was going to fudge the chassis but was previously convinced that the tyres need the attention if you recall) and aero needed fudging to get the feel right.

"Having a go" by providing a more accurate alternative to the aerodynamics which are objectively correctly modeled for what they are (really you shouldn't be "fudging" them at all - unlike the tires there's nothing that you can even argue should be "fudged")? Wow.

 

[David Dominguez]

This guy is hilarious.

 

[William Levesque]

He's not even fast so how can he tell how the freaking car drives ?

 

[RC45]

"Having a go" by providing a more accurate alternative to the aerodynamics which are objectively correctly modeled for what they are (really you shouldn't be "fudging" them at all - unlike the tires there's nothing that you can even argue should be "fudged")? Wow.

Ok - I'll play your silly little 'word' game.

A C5 Z06 presents a certain aerodynamic model. In stock form. The car was placed in a wind tunnel and the preeminent body supplier of the era assisted in the exercise.

The default lift behavior of the car was noted and documented.

Then, based on GM's advice and input based purely on theoretical input made some changes to the car - and tested the results in the wind tunnel to gauge the result.

Results where NOT as expected, and other physical changes where made based on the operators and teams experience.

The desired results where obtained. And results documented.

So, based on practical experiment, certain, wing, hood, fender, splitter and diffuser changes garnered specific results.

I have been in direct dialogue with the people who designed the cars race body work and managed the iterative changes to the C5 Z06 race program and am well aware of what these iterative changes where and the results on the race car and their wind tunnel tested cars - as well as the results when applied to a street car (closed windows and popped head lights contribute another series of results not applicable to the race car but of interest to street drivers.)

Pay attention to this important bit: Certain components and changes where NEVER modeled or theoretically tested beyond practical application. So their specific contributions to this day are unknown - but what is known is the total accumulative results of their application to the chassis.

In other words - you don't have the raw numbers to put in your pretty little spreadsheets, but the actual results are known and documented.

So what you do is take the data you have (in your pretty little spread sheets) and then the CHANGE them until you get the feel and results you know they (the various added components) provide - all the while knowing and understanding the LIMITATIONS of the GAME ENGINE you are working with.

After all YOU are one of the people who INSISTED that there is no practical way for a GAME ENGINE to dynamically and in real time calculate ALL possible aerodynamic components - besides which we only have a defined number of aero component tables to manipulate.

YOU are one of the folks who admitted that aero in the end is a fudge factor. And here you are claiming it is not - that in the context of a GAME ENGINE it is a very precise and accurately modeled component that is fed with pages and pages of spreadsheet calcs.

All your spreadsheets are pointless if the result does not FEEL correct, besides the fact that the actual per component aero contributing values of specific hoods, splitters, diffusers and fenders is UNKNOWN. Only the cumulative RESULT is known and only the cumulative result is applied .i.e. the total wind tunnel down force of front and rear weight sensor results.


It is after all, only a GAME ENGINE right?

 

[RC45]

This guy is hilarious.

Rather than debate the facts presented you also go for the personal attack I see.

He's not even fast so how can he tell how the freaking car drives ?

Keep up the personal insults there. You don't have much else to contribute, so why no yeah?

Personal attack posts duly reported.

And as for Kunos joining the fray to pile on yet again, I was actually congratulating you on a providing some decent tools, but I suspect that due to your frequently demonstrated antagonistic nature and desire to alienate your customer base you probably wouldn't know a compliment from a hole in the ground.

Pity really, you have had so many opportunities to try be civil with your customers yet still choose not to be.

 

[mclarenf1papa]

Ok - I'll play your silly little 'word' game.

A C5 Z06 presents a certain aerodynamic model. In stock form. The car was placed in a wind tunnel and the preeminent body supplier of the era assisted in the exercise.

The default lift behavior of the car was noted and documented.

Then, based on GM's advice and input based purely on theoretical input made some changes to the car - and tested the results in the wind tunnel to gauge the result.

Results where NOT as expected, and other physical changes where made based on the operators and teams experience.

The desired results where obtained. And results documented.

So, based on practical experiment, certain, wing, hood, fender, splitter and diffuser changes garnered specific results.

I have been in direct dialogue with the people who designed the cars race body work and managed the iterative changes to the C5 Z06 race program and am well aware of what these iterative changes where and the results on the race car and their wind tunnel tested cars - as well as the results when applied to a street car (closed windows and popped head lights contribute another series of results not applicable to the race car but of interest to street drivers.)

Pay attention to this important bit: Certain components and changes where NEVER modeled or theoretically tested beyond practical application. So their specific contributions to this day are unknown - but what is known is the total accumulative results of their application to the chassis.

In other words - you don't have the raw numbers to put in your pretty little spreadsheets, but the actual results are known and documented.

So what you do is take the data you have (in your pretty little spread sheets) and then the CHANGE them until you get the feel and results you know they (the various added components) provide - all the while knowing and understanding the LIMITATIONS of the GAME ENGINE you are working with.

After all YOU are one of the people who INSISTED that there is no practical way for a GAME ENGINE to dynamically and in real time calculate ALL possible aerodynamic components - besides which we only have a defined number of aero component tables to manipulate.

YOU are one of the folks who admitted that aero in the end is a fudge factor. And here you are claiming it is not - that in the context of a GAME ENGINE it is a very precise and accurately modeled component that is fed with pages and pages of spreadsheet calcs.

All your spreadsheets are pointless if the result does not FEEL correct, besides the fact that the actual per component aero contributing values of specific hoods, splitters, diffusers and fenders is UNKNOWN. Only the cumulative RESULT is known and only the cumulative result is applied .i.e. the total wind tunnel down force of front and rear weight sensor results.


It is after all, only a GAME ENGINE right?

You'll find I took no part in the conversation regarding active cfd within a game engine...that was others....you'll also find I haven't argued that the aero is a fudge factor...you're just making stuff up at this point.

"In other words - you don't have the raw numbers to put in your pretty little spreadsheets, but the actual results are known and documented."

Anyway if you have the loads on the axles at a given speed (which you've said you possess multiple times in your quoted post), you make a "wing" in AC at each of the car's axles and use yes, an excel sheet, to match the loads at a given speed. Then you simply plug in the height/angle tables you've made into the AC lut files and the car will be the same as your real life data. If you have the net result, the pieces and parts don't matter. Think of it like the implementation of an engine in any simulator; it doesn't matter how many cylinders it has or the quality of the parts it uses, you're just interested in that final power curve. It's the same thing with aero.

And if you're saying that the result feels incorrect when you use the correct numbers, then we're back to the concept that aero is objectively 'un-fudgeable', if you have the correct data it will work correctly, there's no grey area as there is with the tires (which you aren't really fudging in the first place because you don't have proper data for them either way... so really you could be making them more accurate).

 

[David Dominguez]

Rather than debate the facts presented you also go for the personal attack I see.

Hilarious isn't an insult, look:

hilarious
adjective
1.
arousing great merriment; extremely funny:
a hilarious story; a hilarious old movie.
2.
boisterously merry or cheerful:
a hilarious celebration.
3.
merry; cheerful.

About making a contribution, here is one: if aero loads are right, you can uncheck one box more from the list of potential trouble makers. @mclarenf1papa explained it perfectly.

 

[RC45]

You'll find I took no part in the conversation regarding active cfd within a game engine...that was others....you'll also find I haven't argued that the aero is a fudge factor...you're just making stuff up at this point.

"In other words - you don't have the raw numbers to put in your pretty little spreadsheets, but the actual results are known and documented."

Anyway if you have the loads on the axles at a given speed (which you've said you possess multiple times in your quoted post), you make a "wing" in AC at each of the car's axles and use yes, an excel sheet, to match the loads at a given speed. Then you simply plug in the height/angle tables you've made into the AC lut files and the car will be the same as your real life data. If you have the net result, the pieces and parts don't matter. Think of it like the implementation of an engine in any simulator; it doesn't matter how many cylinders it has or the quality of the parts it uses, you're just interested in that final power curve. It's the same thing with aero.

And if you're saying that the result feels incorrect when you use the correct numbers, then we're back to the concept that aero is objectively 'un-fudgeable', if you have the correct data it will work correctly, there's no grey area as there is with the tires (which you aren't really fudging in the first place because you don't have proper data for them either way... so really you could be making them more accurate).

No wait - you must be correct and the inclusion of the aero and suspension apps when inAC dev mode are pointless tools, testing to validate data is just a waste of time when you have spreadsheets.

Spreadsheets without testing is where its at right?

Or are you just out to argue every post people make?

 

[mclarenf1papa]

No wait - you must be correct and the inclusion of the aero and suspension apps when inAC dev mode are pointless tools, testing to validate data is just a waste of time when you have spreadsheets.

Spreadsheets without testing is where its at right?

Or are you just out to argue every post people make?

I'm not arguing anything...you said that "It is very evident that developing any car in AC is not possible without judicious and regular use of the aero and suspension applets." But while the apps are useful, they're are limiting in terms of matching data. That's why I provided the option of using a spreadsheet, a more accurate way of doing so (and provided your calculations are correct, it eliminates the use of the aero app entirely -the values in the sheet will be the same as what AC returns, just easier to read). I'm not debating that the apps aren't helpful additions, because they are (especially when not matching data - they make it easy to model simple aero), just that there are better/more efficient ways of doing aero within AC.

 

[RC45]

I'm not arguing anything...you said that "It is very evident that developing any car in AC is not possible without judicious and regular use of the aero and suspension applets." But while the apps are useful, they're are limiting in terms of matching data. That's why I provided the option of using a spreadsheet, a more accurate way of doing so (and provided your calculations are correct, it eliminates the use of the aero app entirely -the values in the sheet will be the same as what AC returns, just easier to read). I'm not debating that the apps aren't helpful additions, because they are (especially when not matching data - they make it easy to model simple aero), just that there are better/more efficient ways of doing aero within AC.

You are arguing.

If your spreadsheets provide a LUT entry that when applied to a modeled chassis in this GAME ENGINE does not feel correct when driving then your spreadsheet computed value has added no value to the entire exercise.

If your app and the AC tool return 12.5kg of lift at 108km/h at 1.25* AOA but the car feels planted around a corner as a LeMans prototype, but the car was supposed have a wallowy or disconnected feel, then the entire exercise was a waste of time.

I have driven AC game content cars that once you observe the wing data it is evident that the in-game car has been fudged to 'feel' right. Whether the fudging was in the tyre file, the aero LUTs or the chassis geometry is not even important.

The reality is the fudging is either to compensate for limitations of the AC physics engine or it is to make up for lack of actual data.

But in the end the result KS was going for is the same and the car either 'feels' right or does not 'feel' right.

But it is evident your sole role here is to defend all things AC and attack anyone that doesn't share your unwavering devotion to the game engine.

 

[mclarenf1papa]

You are arguing.

If your spreadsheets provide a LUT entry that when applied to a modeled chassis in this GAME ENGINE does not feel correct when driving then your spreadsheet computed value has added no value to the entire exercise.

If your app and the AC tool return 12.5kg of lift at 108km/h at 1.25* AOA but the car feels planted around a corner as a LeMans prototype, but the car was supposed have a wallowy or disconnected feel, then the entire exercise was a waste of time.

I have driven AC game content cars that once you observe the wing data it is evident that the in-game car has been fudged to 'feel' right. Whether the fudging was in the tyre file, the aero LUTs or the chassis geometry is not even important.

The reality is the fudging is either to compensate for limitations of the AC physics engine or it is to make up for lack of actual data.

But in the end the result KS was going for is the same and the car either 'feels' right or does not 'feel' right.

But it is evident your sole role here is to defend all things AC and attack anyone that doesn't share your unwavering devotion to the game engine.

Well of course now I'm arguing, but only because someone decided to write a 15 paragraph antagonistic response to a valid suggestion I made...but oh well.

"The reality is the fudging is either to compensate for limitations of the AC physics engine or it is to make up for lack of actual data."

Well since the aero simulation isn't really limited in AC and you have that data, that's certainly not what you should be adjusting to suit how you think the car should handle...the one thing you don't have any real data for are the tires...why you wouldn't adjust those and those alone is absolutely and completely beyond me...it's not fudging if you don't have data in the first place.

And I think it's fairly obvious that my "sole role here" is definitely not "to defend all things AC and attack anyone that doesn't share your unwavering devotion to the game engine" to everyone but you...because if you recall I'm yet to say that AC is perfect, have said that the tires on specific Kunos cars most likely aren't accurate, and in the past have pointed out numerous bugs/issues with the physics (that have mostly been fixed) on the support forum.

 

[David Dominguez]

If your app and the AC tool return 12.5kg of lift at 108km/h at 1.25* AOA but the car feels planted around a corner as a LeMans prototype, but the car was supposed have a wallowy or disconnected feel, then the entire exercise was a waste of time.

It's not a waste of time, we are telling you it's the opposite. Just try to read for once.

 

[Michael Hornbuckle]

Anybody else seeing the irony of complaining about 'fudging' data while building a mod based on 'feel'?

 

[RC45]

Well of course now I'm arguing, but only because someone decided to write a 15 paragraph antagonistic response to a valid suggestion I made...but oh well.

Your initial antagonistic snide response started the entire exchange, but then again that's why you made the comment

It's not a waste of time, we are telling you it's the opposite. Just try to read for once.

No really, if your spreadsheet produces the incorrect result, you would never know until you test drive and compare the behavior to measured results.

Anybody else seeing the irony of complaining about 'fudging' data while building a mod based on 'feel'?

Other than how similar the feel and feedback of pretend driving a modeled vehicle in a game engine when compared to the known experience and feel of driving the REAL car is, please explain what other metric would or could be used to demonstrate that the GAME produces an accurate REPRESENTATION (after all its not freaking real, it is a faked) of the real vehicle?

Please - enlighten the entire known universe on what OTHER metric can possibly be used to validate the GAME representation of the vehicle?

Bearing in mind that even game AUTHORS admit that in the end if it feels right then the goal of providing an entertaining game has been achieved.

The days of "its too difficult to drive so it must be real" are long gone.

Just ask the AC creators their opinion on the "its too difficult to drive so it must be accurate" bandwagon.

 

[mclarenf1papa]

Other than how similar the feel and feedback of pretend driving a modeled vehicle in a game engine is, please explain what other metric would or could be used to demonstrate that the GAME produces and accurate representation of the real vehicle?

Please - enlighten the entire known universe on what OTHER metric can possibly be used to validate the GAME representation of the vehicle?

Bearing in mind that even the game AUTHOR has admitted that in th eend if it feels right then the goal of providing an entertaining game has been achieved.

If your steering inputs are close it's going to have a pretty similar feel to it anyway...

 

[David Dominguez]

No really, if your spreadsheet produces the incorrect result, you would never know until you test drive and compare the behavior to measured results.

And why should it produce incorrect results? The maths behind are very elemental.

 

[RC45]

If your steering inputs are close it's going to have a pretty similar feel to it anyway...

In other words it would FEEL correct.

 

[mclarenf1papa]

In other words it would FEEL correct.

Matching data to achieve a similar feel to the real car or blindly going by the concept of "feel" and changing whatever you find necessary are not one and the same...

 

[RC45]

Matching data to achieve a similar feel to the real car or blindly going by the concept of "feel" and changing whatever you find necessary are not one and the same...

Who said anything about blindly going by feel alone to arrive at the final product?
I clearly state that the final FEEL is your only true metric of success in a GAME that relies on a limited and finite input driven physics engine.

Once you have assembled all the data you have, drooled over all your spreadsheets, compared all your real world telemetry and finally driven the model in game - if it does not FEEL like the real thing then all the effort was for naught.

So along the way you ADJUST (fudge) various inputs until you get the result you need.

There is no other way to compensate for an imperfect game engine.

Period.

To say otherwise is massively ignorant.

And considering the tyre model literally drives the entire simulation - whether the numbers are fudged at the tyre slip angle level or fudged at the aero level or fudged at the suspension level YOU would never know if the car felt correct while pretend driving in game.

Granted, in an effort to appear all professional and automotive engineer-like, what we do is try make the changes as small and limited as possible (for example to ONLY fudging the tyre data) to make OURSELVES feel good about how accurate we like to claim our PRETEND virtual cars are.

If the guy or girl behind the controller or wheel reports back that the car FEELS like the real thing, the game designed has achieved their desired goal of creating a realistic simulator.