Maybe he means warm pressures since our tires are prewarmed to 40 degrees or so?
Otherwise this would be really counterintuitive. It would be nice if there was a pressure display on the F9 readout to see how temp is affecting pressure. I recall Scawen saying once that the effect isn't that much; but based on what Tristan said I'm wondering if it should be?
Pressure differences between hot and cold on DOT street tires when used on a race track can be as much as 10 PSI. I typically see 7-8psi on my car (mostly on the front left which is the tire getting worked the hardest).
I would call that pretty significant. I would suspect that full race tires are at least if not more sensitive than my street tires.
If we are setting hot pressures, then LFS has the pressures really, really low compared to the real world. Typical hot pressures are going to be in the mid 40psi range for most street tires.
I did check out the F9 views a bit more last night. The surface view does show a rapid gain and loss of temperature, just like it should, at least on fresh tires. The inner rubber temp just kept getting hotter and hotter and the relationship between the surface and the inner temperature seemed odd at times. I just can't exactly explain why though.
I did notice too that tire deformation didn't seem to change as the tires got hotter. You would think that because the tires are getting hotter, the pressure would be going up, and the deformation would be less. It is probable that the difference is just to small for me to notice though.
Scawen has said the difference is small but then tyres in LFS are preheated to 20 degrees below optimum already, so it's not like there is this massive temperature difference anyway. Warming R4s up from 20°C to 120°C should certainly have a noticeable difference, we just don't have to do that. Agreed we use low pressures in LFS, moreso in some cars than other; I think the road cars are the worst. IRL you'd pump them up to improve response and prevent excess deformation/wear. Without much in the way of wear consideration (not on the side at least) or any danger from tyres falling off rims, there are immediatly some reasons for the differences. I'd imagine it's difficult to get a universal model where low pressures result in handling like a bag of excrement yet high pressures handle nice.
Without more precise telemetry on the tyres or looking through the code, all we can do is guess. Although Android is getting very proficient at extracting obscure tyre data.
Well... That's an odd way of implementing things isn't it? I mean if it's hot pressure, then at what tire temp is it talking about?
I would call it significant as well; so I think this is a problem in LFS.
Yes I agree. What I don't get is that temperature X of inner rubber temp greatly affects grip while the same temp or quite a bit higher of surface temp doesn't affect grip the same way. Maybe it should be that way, I don't know...
I know for sure that Scawen did post and say that the effect was not very significant, certainly not 7-8 PSI... I'll see if I can dig it up when I get time it was a long time ago.
Shotglass - check the first pic I've attatched, that's what I was trying to articulate from memory, but I've got a proper version to look at now. As for the nova video - examples are hard to find so I took the best one I could find, you try putting anything to do with tyre physics into youtube! :P
I found another graph - its pretty self explanatory.
I note a 7psi increase from 33psi having gone in on understeer and out on oversteer for about 8 minutes. Whatever that means.
z3r0c00l: Do NOT believe graphs that do not show measurements plotted, and a line of best fit drawn based on the plots. Also, it's best if the source of the data is shown. These graphs don't even have values on the axes, so they're just an impression of "what happens". The thing is, these impressions are usually suggested by racing drivers whose experience might be mislead by a few things. Firstly, up to the slip angle at which lateral force peaks, that force is rising rapidly and it suddenly flattens out. It seems like there is a drop of grip but there isn't. Secondly, lateral force doesn't mean force perpendicular to the car's frame of reference but that of the tyre. The tyre is rotating with steering and perpendicular distance to the lateral force of the outside tyre (which generates more force as it is more loaded) from the center of gravity of the car is decreasing as the driver steers more. This decreases the rotational torque generated. So these curves are misleading.
Real tyres plateau after the peak only barely dropping off right up to some ridiculous slip angle. There is a lot of bogous data floating around because of the above-stated reasons. Don't buy it.
I've already explained this further up the thread -
they can't show original graphs and measurements because the tyres companies will never let them publish them, and the only people who can afford the blindingly expensive tyre testing equipment such as the Bridgestone MTS Flat-Trac LTR tire test system.
They closest they were allowed to come was best fit superimpositions.
The capabilities of the Flat-Trac allow plus or minus:
30 degree slip angle
10 degree inclination
road speeds of 200mph
lateral forces of plus or minus 6,744lbs
the same amount of vertical load
aligning torque for plus or minus 2,213lbs
longitudinal force of 4,496lbs
490hp net drive power
Second to that, what is wrong with taking information from racing drivers as to how a car feels when driving it? I personally would love to drive a racing sim that racing drivers say "feels right", irrespective of the means by which that happens.
I just pointed out that racing drivers will be mislead into thinking forces drop off when they don't. And there's real data floating around online. Avon publishes some of their tyre data (albeit up to a low slip angle) for one. LFS' tyre curves are the very reason that it actually drives somewhat like a car unlike ISI stuff. The harder challenge is doing combined forces. Correct me if I'm wrong but that book suggested the tracion circle approach in one of the pages you posted? I rest my case.
I'm out of my depth here, I am a driver, but I am not a good mathmatician. I'm also exhausting from work today, aching like a corpse from being on the kart track for an hour an 15 minutes yesterday (the track has one straight, roughly 2 seconds long... and many, many corners).
I thought the traction circle simple pointed out how close to the maximum g-forces attainable by the car a drive is getting over the course of a turn or sequence of turns. Opporating at the cars lateral force limit would mean using the slip angles that generate the most grip, irrespective of where they fall on a graph, there could be a massive drop off after 11 degrees, right down to almost nothing, but if you are at the edge of the traction circle you'd be avoiding using those kinda slip angles?
Could anyone else clarify the correctness of what axus is trying to point out?
What I'm getting at is a model suggested for tyre force combination: you take the lateral and longitudinal forces as per some curves, plot the vector on a graph and if it is outside some traction ellipse, you trim it so that it's on the edge. This is a bad model (but what ISI's engine uses from what I've heard).
I would go further and say that 'fair' amounts of drop off after the peak in the base curves could still result in a controllable car; wrong force combining is probably worse than a curve that drops a bit too much.
Chris, this is hard work, its tricky stuff that can't be explained easily.
There are a few rFactor mods that aparantly have had real driver input. The result was somewhere between Mars and Pluto physics. But it felt right.. Real drivers, most of them, don't know what they really feel. Is it peak slip occuring really fast or grip dropping really fast after the peak? Most drivers are poor engineers. The better drivers with an interest in physics have probably read books wich showed them curves made up by drivers in the past, so they adopt this as well..
Cars are bits of metal and rubber, its all measurable and able to model decently on a modern computer.
When you let 5 different Porsche Supercup drivers 'tune' the physics of a car in a game, you'll probably end up with 5 very different ones, with at least one having the engine in the front.. (this actually happened..).
Feel is not the way, decent physics models and real (proven) data is!
What is wrong? GTR, GTL, rFactor Mods, GTR2, Race.... The list of mediocre sims that have had 'real driver input' goes on and on. Either the drivers don't know how cars work, or the people that make the games listen politely, then ignore them.
lol, probably the latter.
But seriously, a lot of drivers have no experience with sims. They have a hard time translating the real world g forces and speeds to a strictly visual experience like it is in the sim. Their feedback is not something I would rely on heavily if I were trying to build a sim.
Well, apart from the fact I'm really not that talented in real life, I do have the advantage of knowing and enjoying racing sims. I'm using to not having forces acting on me, and I'm also interested in the physics behind reality and sim. Therefore, in some [but not all] ways my opinions are better than the hugely vast majority of professional drivers
The rFactor PCC07 mod physics is the most realistic physics ever in any simracing game. Apparently the modders have got input from a real life Porsche Cup driver and it shows.
Try it on a track with lots of bumps.