Corti
Registered
Maybe a little reference about track temperatures
https://www.myweather2.com/Motor-Racing/United-Arab-Emirates/Yas-Island-Circuit.aspx
Released July 2022 Release Candidate Now Available
- Thread starter Paul Jeffrey
- Start date
Simulation_Player
Registered
This is excellent, make sense to my logic.
Haven't tested new rc yet
basil_fawlty
Registered
I have low end PC. It seems to me that CPU load has drastically increased by this update. Previously, I had some car stuttering/high purple bar with a lot of cars on some tracks. Now I have a lot with a smaller number of cars.
Is this correct?
On Spielberg mod, my Tatuus F3 is rammed from behind on the straight every start and pitches up absurdly into the air.
Is this correct?
On Spielberg mod, my Tatuus F3 is rammed from behind on the straight every start and pitches up absurdly into the air.
I'm not sure this is the right question at this stage, Paul.
Am I pleased that S397 is trying to improve the sim via these features of track temps, new ABS/TC, etc.? Yes, I'm pleased about that. But the proof of the pudding is in the execution. And to decide that I have to determine if these new features "work". Which means in turn, I have to decide how I expect these features to operate, and then, compare actual to my expectation; and maybe then if there are differences, adjust what I expected (my expectation was wrong, I didn't understand) or try to analyze what is wrong with the implementation that means my expectations are not met.
After just a couple of hours and just 2 runs, largely static conditions, I'm not likely to know this yet for such complex systems and the full array of conditions I might encounter in say a 24hr event.
When iRacing added all this, they did a very detailed explanation of what behavior they were trying to replicate, and then the mechanisms (solutions) they'd built to try and achieve that. We don't have any of that here. Whilst there is some information - temps depend on where cars run, affected by shade etc. - there's still a lot we don't know. For example, there is "track temp" single value displayed - but yet, this is logically different over the entire lap distance, on different parts of the track. So is this measured at a single point? Is it averaged over some portion of the track? Do all players in a multiplayer event all have the same temps (i.e. updated and published by the server, or is it all just local calcs and there's a chance each user sees a different temp?). Has there been any adjustment for the AI for laptimes running on a hot track versus a cool one, compared to the player? What about the effects of accelerated time?
Heat transfer for tires is one hell of a difficult solve. Generally users understand a few mechanisms for how heat gets generated (sliding, or less likely for internal rubber-band like stretching, transferred from other hot components like brake discs) and how heat gets lost (through the air - convection, through the ground - conduction) but rarely do drivers have a really good sense of how that works. For instance, if you take a racing tire that has been heated with a tire blanket, take the blanket off and leave the car stationary on the ground, you'll see those temps leach away and be coolest where the tire touches the ground, until all the temps reach an equilibrium state. Is this what happens now? Not sure it does. We generally understand that, track temps are usually higher than air temps and some tracks because of their surface types get hotter than other types; but maybe also cool faster as well. The above graph being a good example. What track temp (or equivalent) was used before? Was it just equal to the air temp?
So then, you have to factor in how a track rubbers in, and as we've already seen in some prior discussions in this thread, its not fully understood that some tracks get greasy/lose grip and others gain grip as rubber gets added, and it'll change whether a track starts in "green" condition or already fully rubbered in. It baffles even the best engineers and game designers. I'd argue iRacing still don't have this right; or at least, the end result sometimes seems realistic when at other times it doesn't, which means there's some factor or feature of the model that isn't doing what you'd think it should.
So the answer to your question is: I've no idea if I'm pleased with how its executed in this sim, as of yet. It's got some basic behaviors that seem reasonable directionally.
I have a 6 core CPU, no hyperthreading. Typically I'd been seeing 30% CPU usage (i.e. 2 cores heavily used) and some spikes higher, both when driving and also when watching replays (where the fixed 60 fps limit does make a big difference, and especially also to GPU usage). In this build, where before I'd seen some spikes to 60-70% and some stuttering, I'm now seeing spikes to 90% plus, and all caused by the game and not any other external processes. No idea why this is happening, and it'll have periods where this doesn't happen and then periods when it does. Further, I'm getting drops from 60fps for replays, which had dropped a few frames down to 56 or so, now dropping as far as 50. This doesn't seem to be correlated to extra CPU or GPU utilization, either. This is all with standard S397 content DLC, like GTE cars at Sebring, Spa or Le Mans. Each of these high utilization/low performance periods has to increase the chances of stutters, no?
Try Tinypedal, it's a great alternative for now.
Paul McC
Registered
You are right, I had to cut the video short for certain reasons. Anyway today I did some testing over much longer stints in cold and hot conditions to see if the heat effects tire wear as you would expect and, yep, it does! Also found that tempreture does effect pressures aswell, quite different between the two conditions. I made a video of the results to show what happened. Also during the video I realised something interesting. Should Donnington Park really be hotter than Death Valley, as in should it be hotter than the highest tempreture ever recorded there, or on earth for that matter? And is the ambient tempreture setting actually the track tempreture setting, because I can imagine the road getting up to 60o.C, but the ambient, really? We'd all be dying! Hope this video answers some questions for anyone wondering about tire wear anyway:The testing in this video is not long enough.
The tire flexibility and hysteresis is constantly changing during a race. If the tire is cool then there is little flexibility and hysteresis -> less grip. If it's too high then the compound tears itself apart --> less grip. Both resulting in less grip for different reasons. That's why we have predetermined ideal tire temperature windows where tire flexibility and hysteresis is at maximum.
Tire surface temperature affects the tire inner temperature which changes much slower and goes up during a stint. The tire core temperature also changes according to the tread thickness. The thinner it is the faster the temperature drop.
RFactor only shows the tire surface temperature but not the inner temperature.
Back to the video: at higher ambient and road temperatures the grip is better initially but as the inner temperature rises (with higher road temperature to higher degrees) the tire wears off faster resulting in grip loss.
With low temperatures what should've happened in a longer test is the tires unable to reach the optimal temperature so the grip is constantly lower than ideal but at the same time the tire wear is much slower.
So in the video the guy should've done more laps to see this effect taking place.
That's the theory irl if I remember correctly.
Has anybody tried the wind function in this game yet? I've never even thought about it but now that weather can affect things I ran some tests.
Indycar IR-18 @ Road America, measured top speed at the end of main straight, the direction of this straight is north to south
default setup / rear wing -11
no wind: 286 / 299
northern wind 72 kph: 286 / 299
northern wind 201.6 kph: 286 / 299
southern wind 72 kph: 286 / 299
southern wind 201.6 kph: 286 / 299
This doesn't seem to work.
Indycar IR-18 @ Road America, measured top speed at the end of main straight, the direction of this straight is north to south
default setup / rear wing -11
no wind: 286 / 299
northern wind 72 kph: 286 / 299
northern wind 201.6 kph: 286 / 299
southern wind 72 kph: 286 / 299
southern wind 201.6 kph: 286 / 299
This doesn't seem to work.
Last edited:
Paul McC
Registered
Very much so Paul, absoloutely fantastic news! Works brilliantly! Track and grip levels feel so different at extreme ends of the scale, tire wear works as expected in hotter conditions. Its amazing, been missing this feature for about 9 years now so this is awsome! A question though, when you set tempreture in session settings before going to the track, should that setting read "Track Temp" rather than just "Temp" (sujecting ambient temp), cause yes tarmac will get up to 60o.C, but 60o.C ambient is hotter than the highest temp ever recorded on earth in Death Valley!
D
Deleted member 58016
Guest
Don't get me wrong, but this is release candidate, not the release itself - from my experience (I am an IT guy) a release candidate gets published in order to get such replies: what has been changed, are there significant impacts which haven't been before, bugs which have not been noticed by the devs and so on.
From my experience every release candidate (not especially rF2 RCs) needs some optimization - and best you can do (as a dev IMHO) is to publish a build in order to let people test it and give feedback.
Paul McC
Registered
P.S. How about the humidity setting, is that working? May have to try it out!
MileSeven
Registered
I don’t think it ever has worked, no.
Maarten Nauta
Registered
Im a fairly negative person but this is really nice and needed if it works well for the CP system:
Improved calculation of infringement during time off track, taking into account, time gained, distance travelled, throttle application, time off track, speed differentials, positions gained during race, wall riding and more.
With rf2 its always a good step but still a long way to go. At least this is stuff people want unlike all the broadcasting stuff.
Then why it's in the game?
Remco Majoor
Registered
For such a massive difference, I would expect a lot worse performance of tires. 100deg is insane and would kill tires very quickly, yet it doesn't. I need to test this with a car with multiple tire compounds too
mantasisg
Registered
I have no idea, but there might be plenty of content, especially from mods, which might not have proper configuration for temperatures when it now comes to wide range. There might be cars that will have something wrong. Understanding WLF parameters in TGM tire model might have been less important than it will be from now. I personally never understood AbrasionCurveWLFStartStep at all, but also never put effort to understand WLFParameters shift factor curve. Having values from existing tire models from ISI cars that aren't encrypted pretty much was good enough.
For example some cars perhaps might have glass transition temperature rather off, and it is quite critical for shift factor WLF equation. It will be the parameter that in general would have biggest influence in separating winter tire from summer tire, or ultra soft tire from hard tire.
We might experience actual glass transition at low temps for some tires. And master curve will be getting shifted much more, resulting in differences for the way tires will slide, and also the way abrasive wear will occur, which also does difference for how car grips while sliding.
Also another point, which was already mentioned in here is that there might be additional challenge to match AI speed to players speed, but I suppose it should be not too difficult with some tinkering. Majority of races ever happens in temps from 15 to 30degs Celsius, so I suppose it is enough to focus in such range.
Unless you want to simulate something like Goodwood Members meeting, which I do, where temps are under 10C and sometimes under 5C and also wet
Very exciting.
For example some cars perhaps might have glass transition temperature rather off, and it is quite critical for shift factor WLF equation. It will be the parameter that in general would have biggest influence in separating winter tire from summer tire, or ultra soft tire from hard tire.
WLFParameters = Glass transition temperature Tg, is rather critical and changes with compound. On
the other hand, the remaining variables are pretty much the same for all rubbers, with the exception
of butyl rubber (not typically used for tread compounds). The default values fit into the following
shift factor WLF equation:
log10(aT) = -8.86 ((Tq - Tg - 50) / (51.5 + Tq -Tg))
Where Tg is the Glass Temperature, Tq is the current temperature and aT is the shift factor. This
constructs what is known as a master curve, where the shift factor horizontally shifts the curve based
on temperature, this fits into the 3 following sliding curve parameters below:
SlidingAdhesionCurve = Adhesion is defined as friction that occurs when microparticles are physically
in contact with each other. (min sliding speed (log10(aT v)), grip multiplier (for min), peak sliding
speed, grip multiplier (for peak), max sliding speak, grip multiplier (for max)). Where aT is the shift
factor (see above under WLFParameters) and v is the current sliding velocity.
SlidingMicroDeformationCurve = Same as above for microroughness component of driving surfaces.
SlidingMacroDeformationCurve = Same as above for macroroughness component of driving surfaces
which are a result of larger surface deformations in the tyre.
the other hand, the remaining variables are pretty much the same for all rubbers, with the exception
of butyl rubber (not typically used for tread compounds). The default values fit into the following
shift factor WLF equation:
log10(aT) = -8.86 ((Tq - Tg - 50) / (51.5 + Tq -Tg))
Where Tg is the Glass Temperature, Tq is the current temperature and aT is the shift factor. This
constructs what is known as a master curve, where the shift factor horizontally shifts the curve based
on temperature, this fits into the 3 following sliding curve parameters below:
SlidingAdhesionCurve = Adhesion is defined as friction that occurs when microparticles are physically
in contact with each other. (min sliding speed (log10(aT v)), grip multiplier (for min), peak sliding
speed, grip multiplier (for peak), max sliding speak, grip multiplier (for max)). Where aT is the shift
factor (see above under WLFParameters) and v is the current sliding velocity.
SlidingMicroDeformationCurve = Same as above for microroughness component of driving surfaces.
SlidingMacroDeformationCurve = Same as above for macroroughness component of driving surfaces
which are a result of larger surface deformations in the tyre.
We might experience actual glass transition at low temps for some tires. And master curve will be getting shifted much more, resulting in differences for the way tires will slide, and also the way abrasive wear will occur, which also does difference for how car grips while sliding.
Also another point, which was already mentioned in here is that there might be additional challenge to match AI speed to players speed, but I suppose it should be not too difficult with some tinkering. Majority of races ever happens in temps from 15 to 30degs Celsius, so I suppose it is enough to focus in such range.
Unless you want to simulate something like Goodwood Members meeting, which I do, where temps are under 10C and sometimes under 5C and also wet
Very exciting.The wind function I think already existed in RF1 but it never worked because I think ISI said it was disabled, but that possibility was there.
In RF2 it appears in the session weather settings in the GUI but I don't know if this has anything to do with it:
"Wind and Crowd Motion":false,
"Wind and Crowd Motion#":"If true, trees respond to ambient wind, crowds have some random motion"
Perhaps taking into account the direction of the wind to move trees, flags and other objects (clouds?) susceptible to that element requires a lot of calculations (in real time?), and I suppose that the objects should have their properties or animations, which I suppose is circuit designer work.
In RF2 it appears in the session weather settings in the GUI but I don't know if this has anything to do with it:
"Wind and Crowd Motion":false,
"Wind and Crowd Motion#":"If true, trees respond to ambient wind, crowds have some random motion"
Perhaps taking into account the direction of the wind to move trees, flags and other objects (clouds?) susceptible to that element requires a lot of calculations (in real time?), and I suppose that the objects should have their properties or animations, which I suppose is circuit designer work.
Bill Worrel
Registered
Note, he was using Fahrenheit, so ~37 degrees C.
B
Binny
Guest
Has this build affected fps ? Has anyone noticed any new bugs ? As a VR user how is its performance ?
I like what I'm reading so far.
I like what I'm reading so far.