Track surface resistance

MaXyM

I want to initiate debate abut resistance for track surface materials.
Maybe some one has precise data about that.

Common question is, should resistance be set to zero for tarmac or not.

Preface
I have a friend who works as racing engineer irl. He's also an author of very well know Skoda Octavia Cup mod for rF. While building SOC mod, he put real life data into and figured that the car goes too fast.
After some verifications and calculation he managed that there must be something wrong with track resistance. After increasing it to level about 2500 car started to behave correctly (with tyres based on real data taken from manufacturer and from real car telemetry).

Additionally he did measurements of resistance of Brno track. It confirmed calculated values. New tarmac has resistance about 2250. 5 years old tarmac has about 2700.

Additionally in real life, longitudinal resistance is different than transversal one - it should be clear because of direction of particles in tarmac. This sentence is very important for next case.


Tyre characteristic
When tyre characteristic is measured, there some default resistance must be taken into account. The one is produced by measurement device: resistance of object against which tyre is tested.

Now question: is this resistance equal to tarmac resistance? Is it equal to new tarmac resistance or old one? Or what is its value?

Finally, is it correct to put whole resistance into tire characteristic while tarmac resistance is set to zero?
If we would have longitudinal and transversal resistance for tarmac and tyre, it is impossible to describe characteristic of reaction one against another putting all resistance into tyre data.
That is what make me wonder that putting tarmac resistance into tyre definition is correct for rFactor engine. Even if rF1 supports only longitudinal resistance of tarmac. But it is how all tracks are prepared (including reference ISI tracks). .

I'm curious what do you think about it (hope you understand my explanation)

best regards

PS.
We talked with Michelin but got no information about reference resistance of measurement device.
 
By resistance do you mean static friction?

I cannot see how longitudinal resistance could be different of transversal resistance. Even the particles are not equal or isotropic, they are layered on the floor randomly. In average, the direction of particles should not be relevant.

But it's a interesting point about the balance of tyres / surface coeficients. Let's wait for some clarify from experts.

By the way, check this out.

http://www.tech.plym.ac.uk/sme/MECH115-web/115tut_friction.htm
 
The more testing and work on car physics I do, the more I see that you have to provide proper surface resistance in order to get everything right. Of course, you can keep proper aero drag, proper mechanical friction, proper engine power/torque curves.... but either you get a car that is too fast (all the way - from accel 80-180kph on 4th gear, accel 0-200/300kph to top speed) or you have to provide unrealistic tyre rolling ressistance, like 3-3,5%... which is 3-4 times more than should be.... or you just provide ressistance for tarmac to have everything as should be. Will it be perfect? Of course not! I don't mind if my car accelerates from 80 to 160kph on 4th 0,1-0,2s slower or faster than what I have found on the internet. But with no ressitance you will get values like 0,5s better or even more!
When I finish, I will provide some data for a few road cars I'm working on, to prove the point.

Some time ago I got replay from ISI about that. They said that in the end, everything is about having final effect spot on... but I think, it would be better if they do some research about surface ressistance and provide more or less correct values already in default tracks we will have in rFactor 2. With real world data about road cars aero and performance relativelly easy accessable, there is not that hard to get the values done right. Problem starts, when someone is doing straight line tests on a track with ressistance=0. Will have timings way off better than expected (with proper tyre rolling ressistance and car's mechanical friction) and starts to wondering, wtf? ;-)
 
I've tried messing with a few tdf's after reading this thread and it's a bit like rF2 come early, especially with cars that are modelled accurately.
 
I've tried messing with a few tdf's after reading this thread and it's a bit like rF2 come early, especially with cars that are modelled accurately.
So what exactly did you change?
Would like to try that out myself :)
 
I can only guess, but it should be something like this:

// Roads
[FEEDBACK]
Dry=RoadDryGrip Wet=RoadWetGrip Resistance=2250.0 BumpAmp=RoadBumpAmp BumpWavelen=RoadBumpLen Legal=true Spring=0.0 Damper=0.0 CollFrict=0.4 Sparks=1 Scraping=1 Sound=dry
Reaction=tiresmoke Tex=smokewhite.tga Max=1024 Scale=(1.0,1.0,1.0) Growth=(4.0,3.0,2.0) ASDEnvelope=(2.0,0.2,3.5) DestBlend=InvSrcAlpha SrcBlend=SrcAlpha
Reaction=skid Tex=skidhard.tga Max=2500 Pixel=NoReduceDetail Particle=Plane+Deformable+SingleSided DestBlend=InvSrcAlpha SrcBlend=SrcAlpha
//Reaction=wetskid Tex=skidwet.tga Max=1024 Duration=0.40 Pixel=NoReduceDetail Particle=Plane+Deformable+SingleSided DestBlend=InvSrcAlpha SrcBlend=SrcAlpha
//Reaction=spray Tex=rainspray.tga Max=1024 Scale=(0.3,0.05,0.6) Growth=(0.3,0.3,0.6) GrowthVel=(0.08,0.09,0.30) Power=0.41 RampSpeed=80.0 OffsetVel=0.10 ASDEnvelope=(0.05,0.07,0.90) DestBlend=InvSrcAlpha SrcBlend=SrcAlpha
Materials=road
 
I have made a lot of dirt oval tracks and I can not get that resistance in the TDF to do anything. I can make it 0 or 100,000 and the lap times are exactly the same on the track. Seems like it does not work to me, if I am wrong somebody please tell me what I did wrong.
 
Difference between 0 and 100 is very small. too small to feel anything.
By default we racing using 0. In relation to our knowledge, correct values should be about 2500 (2250-2750 on Brno circuit due to tarmac age). And even with 2500 set, the difference is very small. It is noticeable in telemetry (for example max speed). Also with RFFB I can feel weight of a car better (but it is more subjective)

I believe for dirt resistance may be greater. Maybe some about 6000 or so (have no data for it)
 
With 2500, the difference is noticable mostly in lap times and also Vmax. For a GT class car, on typical race track, lap times are about 2-3s worse than with 0 ressistance.
 
Someone cares to explain what it exactly does?
Guess it has something to do with the grip a tyre can produce on the tarmac?
Higher value means less grip?
 
It is friction. More friction -> more resistance -> more power must be used to work against it so max speed will be lower. I guess it may also affect thermodynamics of a tyre as well as its wear.
 
Is it an addition to tire rolling resistance (torque per unit of deflection) perhaps?
 
I have been searching for a few hours for info on this, is there anything concrete available?
 
Only our and simracing.cz (Skoda Octavia Cup and BMW1 authors) comparisons to real data which are going to proof that.
SR.cz was the first who noticed that there is something wrong. Because putting real data into mod returns wrong results. After that friction of tarmac of Brno circut has been measured and numbers put into tdf file makes car behaviour correct.
We (sr.pl) confirmed it after making our experiments and calculations using real data for tyres and cars.

ISI is mute like a fish.

Note, that missing friction may be compensated in various way. Most modders use aero to do that because aero data is most hard one to get. So they believe that increasing aero drag get correct results. Unfortunately it increases draft to unreal values (most mods have unreal draft effect). Another way is to add some coefficient into tyres physics. It may work, but it means that tyre no longer behaves as real one. Moreover, this approach will work only in rF1 in which tarmac has only one friction (the same in all directions). In reality longitudinal and transversal friction are different which may not be simulate by just adding coefficient into tyre.
 
Last edited:
Ok I had to see for myself how effective it is. I made a straight track for 2km and drove it.

Resistance=0.0 - Feels like normal.
Resistance=2500.0 - I can feel it pulling me back, found it hard to accel in 6th gear. Very unrealistic to me tbh. Felt like I was driving my old Datsun 200b up a hill.
Resistance=200000.0 - I cant move.

But, you completely lost me with your last post.
 
Ok I had to see for myself how effective it is. I made a straight track for 2km and drove it.

Resistance=0.0 - Feels like normal.
Resistance=2500.0 - I can feel it pulling me back, found it hard to accel in 6th gear. Very unrealistic to me tbh. Felt like I was driving my old Datsun 200b up a hill.

You feel like normal at resistance=0 because you are using mod which is made for this friction. That's all. (but I didn't noticed such problems to run on 6th gear)
But if mod would be done by copying real values into it (without tuning/cheating to get correct behaviour), cars will go too fast.
 
OK, I get what your saying and it makes a lot of sense, but it seems like it will be an infinitive task to get every track ever made set up like that, so I don't see what can be done about it now.

Although it would be easy to do if it was added to rF2 and the standard started off with the track resistance being an important part of track building. It wouldn't make it too hard for beginners because most race track surfaces are made from the same special compound so a standard would be set and it would work well.

Come to think of it, you know how the tires have always been a challenge in rF? maybe this is the reason why, I'm no physics guy so I wouldn't know where to look first, maybe someone who does tire physics could enlighten us on how resistance is added to a tire, is it added into the formula or does it have it's own special setting?
 
Well, for someone who has been struggling with tyres for years, I've found this to be the answer to a lot of questions. The fact of the matter is, to put it bluntly, ISI got this wrong. I've done new tdf's for all of the long tracks on my harddrive and without exception, the difference is amazing. It's like I've found the last piece of the puzzle and it's not just me. I've got a neighbour I use as a test pilot, he can drive a car but not a sim, when I put him on the Ring he did several clean laps straight away. All the time asking me, "What have you done?" and comment's like "I can feel the road!"
 
Check your car's tyre rolling ressistance. If that's a normal road or GT racing car, it usually should be between 600 and 1400. Also, check aero coeff in that car.

To have realistic Vmax and acceleration to higher speeds, mod creator probably put higher rolling ressistance and more aero drag, to overcome that missing surface ressistance.

That's why ISI should provide some generic resistance values in rF 2 tracks so that would be assumed as standard and all mods will be modified for that.
 
The results with adjusted tyres (rolling resistance) are even better, more lifelike. The rest of the car is OK, being one the creators, I know.
The main problem with doing it the ISI way is, it's the same all round the track. Trackmakers often add a lot of different track textures, by varying the resistance as well as the other variables, you make make a huge improvement to the driving experience.
 
Last edited:
I prefer putting real world values, assuming that if sim is good, then those values should work (and if not, then tweak the values, but only then!). Slicks have rolling resistance of about 1.0-1.2%. Road tyres have less, about 0.6-0.9%. That will give you values like those in my previous post.
Of course, you can tweak everything to what you want. That's why we see way to much downforce in most of cars, too much aero drag, less powerful engines, high friction torque on unpropelled axis etc. I suggest that you should start with as much real world values, really focus on how that works (I mean, you can think at first that something might not be right, but do some more tests and think again, really!) and if you don't have some, then adjust them to match the rest. That's my suggestion.

Oh, and notice one thing - tyre's rolling ressistance works only in directon of tyre roll. Track ressistance act, on car no matter in what possition it is. Think about that :)
 
OK, got it :)

Yes, that's also one of those niuances that makes the difference. I guess that ISI did it, because 5 years ago things were a bit different. Maybe they didn't expect, that rF community will be so much into trying to make as realistic content as they can?

That rolling ressistance is not something that must be provided. At first, few guys almost convinced me that tarmac ressistance is not needed, because it's already in tyre roll ressistance.. but then I realized that tyre rolling ressistance is just about, how it changes with different load. And if I'm not wrong, it will change in the same way between different load on tarmac, concrete and other solid surfaces... but overall ressistance between those surfaces is different.
And when I did complete physics for a few cars recently, it was clear for me that tarmac ressistance is needed to have all that work (and believe me or not, value of 2500 gave me almost the results I was looking for).
 
Only our and simracing.cz (Skoda Octavia Cup and BMW1 authors) comparisons to real data which are going to proof that.
SR.cz was the first who noticed that there is something wrong. Because putting real data into mod returns wrong results. After that friction of tarmac of Brno circut has been measured and numbers put into tdf file makes car behaviour correct.
We (sr.pl) confirmed it after making our experiments and calculations using real data for tyres and cars.

ISI is mute like a fish.

Note, that missing friction may be compensated in various way. Most modders use aero to do that because aero data is most hard one to get. So they believe that increasing aero drag get correct results. Unfortunately it increases draft to unreal values (most mods have unreal draft effect). Another way is to add some coefficient into tyres physics. It may work, but it means that tyre no longer behaves as real one. Moreover, this approach will work only in rF1 in which tarmac has only one friction (the same in all directions). In reality longitudinal and transversal friction are different which may not be simulate by just adding coefficient into tyre.

Just for curiosity, may I ask how did you measure that Brno friction?

I can't see how real tarmac friction could be anisotropic. Also, friction only makes sense in the contact of two materials, there's no friction defined for any of each alone. But I may be mixing things here, rolling resistance is not really a friction but an energy loss due to tyre deformation and hysteresis.

I don't know how rolling resistance data is usually provided by manufacturers but if the tyre is intended to be used on tarmac, to have it defined on tarmac seems logic to me. Then Resistance in the tdf should be 0.

On the other hand, not all tarmacs are the same of course, an asphalt with tiny particles will make contact patch deformation (and so on, rolling resistance) different than an old cracked tarmac with big particles. But this difference is still strongly dependant on the tyre compound, mainly on its hysteresis.

And if this wasn't enough, Rolling resistance also depends on rolling speed, and I'm not sure if this is implemented in current engine.
http://hiwaay.net/~bzwilson/prius/pri_tire_020.jpg

Interesting thread BTW :)
 
Yes, but notice that in rF, tyre rolling resistance is per load and per deflection... there is completly nothing about surface. It's just about how more or less drag you get on tyre when it is beeing deformed by load. It will on ice, it will on tarmac, right? You put more load and it will generate more rolling drag both on ice and tarmac. But surface ressistance is different between ice and tarmac itself. Of course you also have CoF incorporated into all that.... :D
 
may I ask how did you measure that Brno friction?

It's not me. it was Mirzza/simracing.cz (real racing engineer and author of famous mods). He used special device for that.

I can't see how real tarmac friction could be anisotropic.

Tarmac/asphalt is finishing only in one direction. That's why its particles are ordered in one direction making longitudinal and traversal friction different.


Also, friction only makes sense in the contact of two materials, there's no friction defined for any of each alone.

Really? so you cannot define friction of single materials? means that glass has the same friction as tarmac as well as ice?


I don't know how rolling resistance data is usually provided by manufacturers but if the tyre is intended to be used on tarmac, to have it defined on tarmac seems logic to me. Then Resistance in the tdf should be 0.

Not quite true because manufacturers test tyres in testing facilities not against tarmac.
Lesiu may give you some more info, cause he talked with them.


Finally everything starts with incorrect car performance after setting it with real values. Performance turned into correct one after using tarmac friction.
best regards
 
Yes, but notice that in rF, tyre rolling resistance is per load and per deflection... there is completly nothing about surface. It's just about how more or less drag you get on tyre when it is beeing deformed by load. It will on ice, it will on tarmac, right? You put more load and it will generate more rolling drag both on ice and tarmac. But surface ressistance is different between ice and tarmac itself. Of course you also have CoF incorporated into all that.... :D

My guess is that resistance on ice and tarmac are the same as far as both surfaces are undeformable and have the same roughness. Maybe ISI should have implemented surface Resistance paramater as a multiplying factor (default=1) instead of an addition to tyre's rolling resistance, that way we could simulate tarmac resistance variations and they would still be dependant on tyre characteristic.

MaXyM, don't get me wrong, I'm sure you know what you're talking about. Thanks for the answers.

What I meant about friction, using your examples is that glass simply has no friction defined (it's not the same than tarmac or whatever, it just doesn't exists), glass vs glass does and it's not the same than glass vs rubber.
 
Yes, all that stuff is relative, but I don't know how physics engine is calculating that. I think there must be some reference there (because you have tools to measure that resistance).

If ISI say that tarmac should have resistance set to 0 and explain, why then car's performance with resistance=0 is so much way off compared to what is expected nad it will be clearly said that it's becase rF's engine is not calculating some important things (or is doing that in wrong way) , then I will accept that :) On the other hand, why providing proper tarmac resistance provide also proper car performance with real world parameters? Coincidense? Well, that's something that I asked ISI few months earlier but... didn't get clear answer from them :(
 
Well yes, if it works this way there's no reason for not doing so. I'd just like to know how it actually works, as well.
I'd appreciate an explanation of how this proper tarmac resistance is computed and what that special device is about, if any of you knows. :)
 
Hi,
i use Pendulum Tester for tarmac resistance information, with my friend from Roads Administration - the official name is the - Determination of cohesion of bituminous binders pendulum test. And of course many, many hours of test TDF files in rF...but... 2500 - 2750 is the right numbers i think. Speed, G, lap times and others drive factors is very, very close to real data from telemetry and of course, feels right while driving ( I have a feeling from a real pilot ). We really needs an explanation directly from the ISI, exactly how it all works. For my community ( SimRacing.cz ) we currently uses these values, in our experience it is better in all aspects of the simulation than the original value
 
Thank you mirzza.
The fact is that after some experimentation, i think that to set the tdf resistance to 2500 does the same than to add this value to tyre's rolling resistance and leave the tdf to 0, or to add the tbc value to the tdf and leave the tbc to 0. So there must be an optimum distribution between the two that makes the tyre behave correctly on any surface, not only asphalt. IMO if the tbc rolling resistance is taken from real data, then we are assuming that the test rig had 0 resistance, but I don't know what kind of rig surface are they using and if it is any close to real tarmac.

I'm afraid we will need to wait for an ISI explanation :)
 
The topic starter basically said "We put in real numbers but the result wasn't quite right" , which can be due to many things. Losses in engine/drivetrain/rolling resistance and things like drag changing with temperature, etc etc. rFactor just does what it is told to do. Often are more drivetrain losses, or the real drag coefficient * frontal area is a bit optimistic, or the real circuit is at altitude and you had less real power because of that..

Rolling resistance or track resistance shouldn't be taken for more than it is. Soft surfaces like dirt, snow, gravel traps, defined in the TDF need this extra resistance parameter because they are not 'soft' in rFactor, so more resistance makes them behave a bit like softer materials i.e. gravel traps.

The tire rolling resistance is indeed only measurable on a surface, and since tires are meant to run on 'roads', the typical rolling resistance force as % of vertical load is between 0.5% for eco street tires to probably 1.5% for racing slicks, depending on a huge number of variables. When you know the tire compression due to vertical load, you can work out the RollingResistance value in rFactor (torque per meter vertical tire compression). That gives you a tire with a rolling resistance on a 'reference road surface'.

For other 'road materials' like old asphallt, new asphalt, rough asphalt, concrete, would it make sense to add resistance to the TDF? I don't see a reason. Even if someone has real rolling resistance measurment on a few different road surfaces, as long as they are 'rigid' surfaces, perhaps the resistance changes between 1% and 1.5%, but that would be quite a lot already. In the big scheme of things, this is very minor and there are always bigger problems to fix with the mod than this!

I believe but am not sure the TDF resistance is calculated the same as tire rolling resistance, i.e. a torque per meter compression. If so, if you MUST tweak road surfaces, and again these are 'solid' road surfaces like concrete/asphalt, don't go to an extreme to fix a different problem. Try adding 0.5% max, which usually is a resistance value of less than 500. If you need more, the true problem is with the real data, or the losses etc I mentioned at the beginning of this post!
 
Niels, but your attempt to this topic is "you guys for sure have something wrong in the physics" without even checking by yourself all that arguments you provided (and you will see why I'm saying that, when you do that tests :) ).
Take real world data, put WHP engine curve (but then set all friction torque to 0. I assume, you know why). Put proper tyre rolling resistance and aero drag. Do some straight acceleration tests with tdf=0 and check with real world results.
Then, we can talk :)

And one more thing. Aas you said that too - tyre rolling resistance is a % of tyre load. That's all you know. Where is then information about test surface? You assume, it's "reference road surface". Do you think that they put tarmac on tyre testing surface? You know that's impossible :) What, if that surface is just a stiff rubber? Or maybe some sort of a sand-paper? Only we can assume, is that's definatelly a surface that on which tyre won't slip during such testing.

Even if someone has real rolling resistance measurment on a few different road surfaces, as long as they are 'rigid' surfaces, perhaps the resistance changes between 1% and 1.5%, but that would be quite a lot already
Oh, yes! Mirzza already said, that typically a range between 2500 and 2700 (200, so even less variation, but I guess for wider set of surfaces, the differences can be bigger).

Of couse i don't want to argue but at least you could do the homework and do some tests, before saying that we are definatelly wrong ;-)

Cheers
 
Last edited:
This is indeed a great find! We had a few tracks during the past season where there was a bit of an icy feeling, and by adjusting the surface resistance to values suggested in this thread, and compensating for it in the tyre's rolling resistance, it now feels like driving on asphalt again. The car is much more predictable in its moves and the steering is more rigid. (I have been using RealFeel since it first came out) Great detective work guys!
 
Oh, yes! Mirzza already said, that typically a range between 2500 and 2700 (200, so even less variation, but I guess for wider set of surfaces, the differences can be bigger).

As far as I remember, Mirzza said me, that common values changes from 2250 to 2700 |(for Brno circuit) depending on:
- measurement direction (longitudinal vs traversal)
- age of tarmac (difference about 500 between new and 5 years old one)

So, differences are quite significant.
 
Last edited:
And one more thing. Aas you said that too - tyre rolling resistance is a % of tyre load. That's all you know. Where is then information about test surface? You assume, it's "reference road surface". Do you think that they put tarmac on tyre testing surface? You know that's impossible :) What, if that surface is just a stiff rubber? Or maybe some sort of a sand-paper? Only we can assume, is that's definatelly a surface that on which tyre won't slip during such testing.

But note that even if the surface used in the test facilities makes a significant difference it should be easy for them to apply a correction factor. At the end of the day a rolling resistance factor which is way off from what you get on a road is pretty useless in my opinion.

I don't know your sources but personally I'd rather trust tire test data before than engine power curves which are usually taken with inertia dynamometers which doesn't take into account drivetrain inertia, and that's just to say one thing they doesn't do right. I wish I had reliable data to do the test you are suggesting myself but I can't give a verdict without being 100% sure about the other 999 rfactor parameters affecting vehicle performance.

Have a happy new year guys :)
 
The way I see it is, the problem with 0res tracks is not that the track is zero resistance, but the relationship between tyre and road is inverted. This is why slow corners are a problem, the resistance numbers get close to zero and you slide off track. A bit simplistic I know.
 
The way I see it is, the problem with 0res tracks is not that the track is zero resistance, but the relationship between tyre and road is inverted. This is why slow corners are a problem, the resistance numbers get close to zero and you slide off track. A bit simplistic I know.
I think you're confusing rolling resistance and friction force. The friction force available that determines if you slide is calculated from grip parameters. The track file has grip values (like RoadDryGrip=1.00) that I think are just multipliers for the friction calculated from the hdv grip parameters.
 

Back
Top