New Tire Model

Vehicle physics, suspension, tires etc

Re: New Tire Model

Postby Chronus » Sat May 10, 2014 5:11 pm

Thank you, Eugene. If you find it interesting and when you have the time, for sure I will show my physics calibrations to you. I certainly have quite a bit to catch up with XMR and a bit to learn about the MF2 method.

---

I'd like to take the opportunity to ask two questions:
- how can one define the diffuser and front winglets/canards?
- I changed (copied first to a new physics profile .phs, and updated as I changed things) the AMG DTM and the Nissan GT-R35 base physics (phs). Namely, I changed the friction coefficients, aero drag coeff and frontal area. Afterwards, I launched XMR and tried these cars with the profile (after a quick physics change to the steering wheel rotation), and the cars seemed stuck. Is there something I may have done to cause this?

---

I too prefer MF1 to MF2 due to the complexity of the latter (and, often, lack of proper data).

But I will invest a lot of time into MF2, I want to be able to model any tire I wish (with either the right data or the proper extrapolations, I believe I can get there) via MF2.

---

Initially, what surfaced in regards to AC's TM was an evolution on Stefano's model for NetKar Pro. I liked it - he joined two concepts (magic formula and several physical characteristics of tires) so it seemed promising. I started to read bits here and there that showed other things being added. Right now, I ignore how deep he/they went, I doubt they've gone after a full physical TM.

I know for sure (well...I don't know the source code, but read the main physics dev speaking about it) that pCARS has gone after a full physical model, which is interesting because iRacing did the same and they still have problems with it. Talking to tire engineers, I became convinced more and more that physical tire models are not quite what we need (and certainly not what part of the tire industry needs, reason why they use a mix of MF-physical models).

From the point of view of simracing developers and modders, with the data we have at our disposal, it's better for us to use a semi-empirical TM or even an empirical TM than a physical TM. But, hey, that's me talking, and it seems some developers disagree with this.

---

Your gfx engine has improved every year. Right now, my system cannot handle a FULL MAX settings XMR, so I can only imagine how impressive it can look with everything maxed out.

---

One question: is there a pdf/article explaining in good detail what types of suspensions and aero configurations we can model in XMR? (Hope you don't mind my asking this, especially if you already have answered this one before.)
Chronus
 
Posts: 18
Joined: Fri May 09, 2014 1:38 pm

Re: New Tire Model

Postby Eugene Cojocar » Sat May 10, 2014 6:59 pm

I will answer your question regarding aerodynamics coeff later because I'm out of my place at the moment.
Did you choose the changed .phs profile before testing in the menu? Maybe there are some errors in input data also. Try to enter new data step-by-step. Enter value 1 then test it. Enter value 2 then test it etc. That can help to avoid errors.

I'm working on the new web site that must include article about physics in XMR, its tweaking and much more. The current pdf is out of date.

I'm skeptical about a full physical model. I think It's very hard to implement it on modern PC.
XMR Developer
User avatar
Eugene Cojocar
 
Posts: 937
Joined: Thu Oct 18, 2012 2:49 pm

Re: New Tire Model

Postby Chronus » Sat May 10, 2014 7:33 pm

Eugene Cojocar wrote:I will answer your question regarding aerodynamics coeff later because I'm out of my place at the moment.

Chronus wrote:I'd like to take the opportunity to ask two questions:
- how can one define the diffuser and front winglets/canards?





No problem. I'll wait. :)

Did you choose the changed .phs profile before testing in the menu? Maybe there are some errors in input data also. Try to enter new data step-by-step. Enter value 1 then test it. Enter value 2 then test it etc. That can help to avoid errors.


Tracing back my steps, I did this:
- launched physicseditor
- copied base.phs to mine1.phs
- altered drag coeff, altered lat and long friction coeffs, altered frontal area
- updated (UPDATE button in editor) mine1.phs, saved it
- exited editor
- launched XMR
- chose the track, chose the car, chose mine1.phs
- altered steering wheel lock from default (for that car) to 21
- GO!
- cars were stuck, no matter how hard I accelerated

Will try again in the next hour, even more carefully.

---

By the way, tried the BMW E92 DTM (downloaded through master update) at Suzuka, with SPORT.phs (unmodded by me).

Darn good car, sir! :)

---



I'm working on the new web site that must include article about physics in XMR, its tweaking and much more. The current pdf is out of date.


That's fantastic, Eugene. On my part, can't wait to go through it and learn.



I'm skeptical about a full physical model. I think It's very hard to implement it on modern PC.


Precisely what I have been telling people year after year. Inevitably, "people" fall for the hype and marketing headlines.
Chronus
 
Posts: 18
Joined: Fri May 09, 2014 1:38 pm

Re: New Tire Model

Postby Towers » Sat May 10, 2014 10:36 pm

Here it is my trial for a sport road tire for the mustang. Comments and suggestions are welcome.
Attachments
mustang_tires_test.rar
(22.86 KiB) Downloaded 158 times
User avatar
Towers
 
Posts: 82
Joined: Wed Oct 16, 2013 12:13 pm
Location: Italy

Re: New Tire Model

Postby Chronus » Sat May 10, 2014 11:22 pm

Towers wrote:Here it is my trial for a sport road tire for the mustang. Comments and suggestions are welcome.



Just downloaded your tires.

Hope you don't mind my asking, but I imagine these are for the Ford Mustang GT ?
Chronus
 
Posts: 18
Joined: Fri May 09, 2014 1:38 pm

Re: New Tire Model

Postby Chronus » Sat May 10, 2014 11:38 pm

Eugene Cojocar wrote:Did you choose the changed .phs profile before testing in the menu? Maybe there are some errors in input data also. Try to enter new data step-by-step. Enter value 1 then test it. Enter value 2 then test it etc. That can help to avoid errors.


Ok. Launched the editor. With the original base.phs of the GT-R35, I did a copy called mine1.

I worked on this one.

So, changed:
- drag coeff
- frontal area
- corrected lift for body, splitter and rear wing
- steering wheel max angle
- changed tire friction coeffs

Updated the profile.

Saved.

Launched XMR. Chose the Nissan GT-R35 and went on to Suzuka. Chose the mine1.phs.

GO!!!

Everything went FINE.

Which is great...just I am not sure why previously showed probs.

I will follow this with other cars. For now, great. :)
Chronus
 
Posts: 18
Joined: Fri May 09, 2014 1:38 pm

Re: New Tire Model

Postby Towers » Sun May 11, 2014 12:13 am

Chronus wrote:Hope you don't mind my asking, but I imagine these are for the Ford Mustang GT ?


Yes sure, ford mustang gt
User avatar
Towers
 
Posts: 82
Joined: Wed Oct 16, 2013 12:13 pm
Location: Italy

Re: New Tire Model

Postby Eugene Cojocar » Sun May 11, 2014 11:46 am

Chronus wrote:.
No problem. I'll wait. :)


Thanks, I hope to be in my place in a couple of days.

Precisely what I have been telling people year after year. Inevitably, "people" fall for the hype and marketing headlines.


Seems, this is a feature of our time. I doubt it's possible to fix that.
XMR Developer
User avatar
Eugene Cojocar
 
Posts: 937
Joined: Thu Oct 18, 2012 2:49 pm

Re: New Tire Model

Postby Eugene Cojocar » Thu May 15, 2014 11:28 am

Chronus wrote:I'd like to take the opportunity to ask two questions:
- how can one define the diffuser and front winglets/canards?


I had a look at the code of aerodynamic model.
You can define only Body, Front/Wing and Rear/Wing Lift/Down forces.
It's a little strange for a first look but the aerodynamics model is simplified at the current stage.
For example, Body or Front Lift forces depend on frontal area and appropriate coefficient like Body Lift Coefficient or Front Wing/Axle Lift Coefficient. Actually I will have to add planform area.
XMR Developer
User avatar
Eugene Cojocar
 
Posts: 937
Joined: Thu Oct 18, 2012 2:49 pm

Re: New Tire Model

Postby Eugene Cojocar » Thu May 15, 2014 12:14 pm

Towers wrote:Here it is my trial for a sport road tire for the mustang. Comments and suggestions are welcome.


I just tested it and had a look at the tire model.
From my point of view the pure cases are good. But combined cases of the tires are not correct for me.
I mean under pure cases driving when:
1) Turn angle <> 0 and Longitudinal slip = 0.
2) Longitudinal slip <> 0 and Turn angle = 0;

When you turn to the left or right and skid at the same time the tire works in combined conditions when:
1) Longitudinal slip <> 0 and Turn angle <> 0.

For example: drifting, burnout, braking with locked wheels etc.

For Tractive/Braking Properties try to play with the Pacejka Slip Angle to watch the Tractive/Braking forces when Slip Angle <> 0.

For Cornering Forces try to play with the Pacejka Slip Factor to watch the Lateral forces when Slip <> 0. Seems it has not enough lateral forces when Slip <> 0. You can define combined slip behavior by changing the ABy1, ABy2, ABy3 ... AVy6 values in the Lateral Pacejka Data dialog like on the pic below:

Image

Try to copy these values from the BMW M3 E92 DTM and play with it them get better behavior.
XMR Developer
User avatar
Eugene Cojocar
 
Posts: 937
Joined: Thu Oct 18, 2012 2:49 pm

PreviousNext

Return to Physics

Who is online

Users browsing this forum: No registered users and 1 guest

cron