Differences in aero calculations in CarFactory vs rFactor telemetry

LesiU

Recently I got information that CarFactory does not calculate aero properly... that sound a bit strange at first because I always thought CF is a tool, the only tool for ISI physics engine, that everything's got right, but...
I did some tests. Took SkidPad track from rFC, made tdf file with clean surface (default track doesn't have tdf file at all, so probably has sort of default fake bumps generated by rF) and took my car for a straight line speed test with telemetry recording.

What I did, is compare front/rear downforce values from CarFactory with downforce I get in rF from telemetry, which is: total tyre load for each axis, minus static load, minus fuel load at that time (having in mind fuel tank position, which has influence on that fuel load). Test was done at steady 300km/h speed (so there was no weight shift - no damper/spring work at that moment).
Of course I took ride height from telemetry (undertrays are set to 0.0 so no corrections needed), put it in CarFactory and took recalculated downforce values.

So, what I get is:
- CarFactory: +173.0N/+647.4N (total 820.4N of downforce), 21.1% front
- rFactor: +68.7N/+730.6N (total 799.3N of downforce), 8.6% front

That's 2,6% total downforce difference, which is not that bad, but... downforce distribution is totally different! Calculated 21.1 vs real 8.6% is a big difference, isn't it?

Because Kangaloosh has never answered on any on my e-mails and his thread at RSC does not exist for a long time, maybe someone else know CarFactory enough to tell where such an error came from?

Is there any better tool to calculate aero for rFactor?
 
Last edited:
Is there any tool that could help in aero calculations?
I did some more tests with CarFactory and it do it all wrong - body, front wing, rear wing... and who knows what else? Maybe aero drag too?

If there's no such tool, then can someone from ISI explain how aero is calculated? At least for drag, because downforce can be done with telemetry.

For total body drag, there's a general formulae (found here: http://koti.mbnet.fi/tspartan/gp1975/airoopas/index.php?id=functions.php ):
BodyDragBase + BrakeDuctSetting*BrakeDuctDrag + RadiatorSetting*RadiatorDrag + BodyDragHeightAvg*ARH + BodyDragHeightDiff*Rake

But what is that BodyDragBase? How to deal with real world drag coefficent, frontal area (or combined Cd*fA coefficent, which maybe is more likely to be used by ISI) and of course speed (air density I guess is constant in rF 1).
 
I started work on an excel based calculator to compare many cars, but got busy with other things...

If I remember correctly, carfactory doesn't include some of the small effects.

For example:

there is a moment caused by offset drag center - push back on the top of the rear wing, and the front end will lift up

tilted vectors - the lift vector tilts forward and back if the car does, so a wing can produce 'lift' which is directed forwards or backwards, causing drag or thrust. yes, massive rake can cause extra top speed

a tilted and yawed wing produces lift and drag from vertical, sideways, and forward motion

It's been a while and I don't remember exactly what carfactory does, but I didn't find any errors - just stuff missing.
 
By saying errors I mean, that values I get in telemetry, are not what I get in CarFactory. Are there things missing or CF does it wrong... the final effect is the same - the forces are not the same :(

Will you find some time to explain basic calculations for drag?
From the document I provided in the previous post it looks like, the drag in rF is basically Fd = C * v^2, where C = 0.5*Cd*fA*q:
Cd - drag coefficent
fA - frontal area (plantform area)
q - air density

If I understand that correctly, all those body drag coefficents:
BodyDragBase + BrakeDuctSetting*BrakeDuctDrag + RadiatorSetting*RadiatorDrag + BodyDragHeightAvg*ARH + BodyDragHeightDiff*Rake

after summed up, should give that C coefficent from Fd = C * v^2 formulae.
If that's how it works then we have some basics :)

What about that offset drag center? It would be hard to get that from general downforce/lift data... or maybe there are some common values for it, based on drag value?


What about body lift "BodyFore" coefficent and related to it "BodyCenter"?

Let's say, I have a car with 2.355m wheelbase and 0.05/0.02 (f/r axle) lift coefficents. That would suggest I should put Z coordinate to -1.682 ( (0.05/0.07) * 2.355 ) but that won't work like that - telemetry shows more lift at front and some downforce at rear. Am I missing something in that calculation? That drag torque blends in with body lift, making all the difference?
But then, if I know lift values, how to take off from them that drag force causing partial body lift?
 
Last edited:
You can figure out the drag coefficient from a coastdown test. If you need a track with a straight line long enough to do it message me and I'll send you one.
 
Thanks for offer but that probably won't work for me as I would expect. There are still mechanical drag and rolling resitance involved into that. Of course rolling resistance is known value, as well as friction torque but how they are calculated at speed? Don't know.
 
There's a way to figure those out too. When I get home i'll post the formula.
 
LesiU said:
after summed up, should give that C coefficent from Fd = C * v^2 formulae.
If that's how it works then we have some basics :)
Click to expand...

I believe that's correct.
LesiU said:
What about that offset drag center? It would be hard to get that from general downforce/lift data... or maybe there are some common values for it, based on drag value?
Click to expand...
I derived the increase in rear downforce due to drag as follows:
DownforceIncrease = DragForce*(y/W)
Where y is the height of the element above the ground, W is the wheelbase.

The decrease in front downforce is the same value.
 
So the total downforce for the body for particular axis is what you have defined with BodyFore (keeping in mind Z in BodyCenter) and add to that DownforceIncrease (lift at front axle)?
 
That's how I interpret it (assuming you're ignoring other aero forces from moving left, right, up, down, etc).
 
Yes, I'm ignoring it at the moment (I have no idea how big they can be apart from the fact that I don't know how to calculate them... so I took them from other mods, assuming they are OK :) ).

Thanks for all the information!
 
LesiU,
In your tests where did you place the front and rear wings in relation to the axles?
Placing the rear wing behind the rear axle will lift the front wheel through leverage, and vice versa the front wing has the same effect.
 
Yes, I know, but in the end, it's about load on front and rear tyres at specific speed and that is different between rF and CF (I exclude static load of course).
 
LesiU,
I think you are missing my point.
At speed, if the rear wing is 1.5 meters behind the rear axle, it will generate downforce on the rear wheels and lift on the front wheels through leverage. This effect will showup in motec and will create the results you show above.
At 2 meters behind the rear axle the effect is even greater.
 
I told you that I know about that :) In telemetry you will see the final effect of that leverage (after all calculations), right? CarFactory simulates that effect to!
If you measure a car in wind tunnel, front and rear downforce/lift values also will "incorporate" that leverage effect.

So in the end it doesn't matter where I put that values in relation to front and rear axle, because both rFactor and CarFactory simulate that leverage effect.
Final aero load on tyres is then the only important thing.
 
Are any aero point centers below the reference plane?
Is the Front Wing, or Diffuser below the reference plane?
example:

(x,-0.1,z)
 
That doesn't matter. I also did tests with front splitter and rear wing at 0.0 (to be sure that leverage we were talking about earlier, in case of beeing wrongly calculated too in CF, won't have influence) and it was just not right. Balance f/r also is not right.
There is also issue with body rake. It looks like CF use hardcoded value of 0.3985, because no matter what you provide, you still get the same rake diff all the time (1,23N per 1mm change @ 200kph).
 
Sorry for posting this irrelevant question but i need some help for analysis section of kangaloosh. I want load my hdv file on analysis section but gives me error. It says "the folowing error occured while opening an existing mod for analysis: Unable to parse .tbc tyre at c:\.."
Engine, tyre, pm, gears and damage files is in same folder with hdv. What is wrong in tbc file? I don't do anything, it's standart rfactor trainer tyre.
 
Probably, there is "something" not right in tbc file. Maybe some missing lines or misspelled parameters... unfortunatelly, CarFactory have bug reporting like older Borland compillers - it will never tell you properly, what's not right ;-)
 
I tried my all mods in rfactor, even f1.07 mod and other isi made mods give this error messages. Is there any other process before open the hdv?
 
Last edited:
Nope. Everything should work right on, but if not, then you still have no idea is it something important or just CarFactory don't like some parameters.
 
You must log in or register to reply here.

Back
Top