Hiya,

There haven't been too much threads about it but this one (old!) http://www.lfsforum.net/showthread.php?t=4214 shows a plot from LFS. I heard (blame Axus ) that load sensitivity was not or not modelled to a significant extend. I quite like LFS again with the patch; especially the 'race' cars have improved. I wonder how much of the remaining problems I have are down to load sensitivity not being modelled or not being modelled to a great enough extend.

There is lots of AVON excel data available, measuring lateral force at quite a few loads. It would be pretty easy to come up with load sensitivity curves for f3000 / f3 / fford tyres based on this data. I also have some numbers from Goodyear rubber from the Z06 Corvette. So should loadsensitivity not be modelled in a detailed way; I don't see why as there is lots of data around based on which you can realistically implement it (at least for lateral mu vs load).

Driving on South City yesterday, There is that drifty righthander, short straight, followed by what basically are 2 lefthanders, taken as one ish, the last one followed by an immediate right. I find myself always in some sort of fear at that corner (just one example) as it is so easy to initiate a sort of constant slide, where the effect of lifting or giving it a bit of throttle doesn't seem to change the balance of the car enough to stop this slide. In all cars this can happen and you can sort of leave a long skidmark, even though you're not *really* going sideways or giving driving inputs that could sustain this sliding.

Tyres are complex and it could be many things. However, thinking about it, load sensitivity might be part of it. During a long slow lefthander, your grip is mostly mechanical. Accelerating or lifting / braking a little mostly changes the load on the outside front and rear.

Now if load sensitivity where not modelled; the effect of lifting (which is what I'd naturally do) would not change the friction coefficient on the tyres. Load would change (weight moves to front) but the grip increase is linear. Hence the balance of the car not necessarily changing.

Now if load sensitivity is modelled, and I do the 'natural lift', weight would go to the outside front, making its friction coefficient drop a bit, causing some understeer. The rears have more lateral grip in this case as there is less load on them, plus you take away longitudinal force. Lifting would, I imagine, have a pretty siginificant effect on mild slides, straightening the car out, after which you can try again and apply power.

I also notice in LFS the general risk of oversteer on corner entry. Hard turn-in would load the outside front a considerable amount, loosing just a tad of efficiency and grip there. Washout some call it, is lacking, imo, in LFS. Braking into a corner is the 'worst case' for the tyre; lots of sudden load on it. It seems to have no effect on grip in LFS. Take T1 at Blackwood (the right hairpin) for example. Turn-in is unnatural to me, how there often is almost no risk of washout yet easy lift oversteer can happen very late into the corner, and once the slide has started, lifting or acceleration (change in front/rear load) doesn't seem to do enough.. I find myself somewhat unrealistically just choose to sustain the slide which costs far less time than trying to stop it and drive smoothly out of the turn.

Still, tyres are complex, this may be down to force combining all again but I think load sensitivity might be part of it. Since, as said, there is plenty of real info available to derive it from, I see no reason why it can't easily be improved, SHOULD it not be right at the moment.

The thread I linked to at the start of this post; perhaps someone can try to plot lateral grip at various loads with X30/31? Say the F08; which we can compare to the AVON formula 3000 data..

Whadda y'all think?

I think you're mixing up "weight moving to the front" and "mass moving to the front" a bit. If you lift mid-corner, "weight" moves to the front, but "mass" certainly doesn't move at all, in the sense that the CoG is still at the same place in your body frame. You get increased load in the front, but zero increased mass there, so lifting mid-corner creates oversteer considering only this increased load, with or without load sensitivity.

Only with severe load sensitivity, the available grip doesn't rise as quickly as load does, and you get less oversteer from lifting. But you still get oversteer, not understeer. To get understeer when lifting you'll look at other sources, like coast locking on the differential.

Personally, I think the whole "weight" shifting, "weight" distribution, "weight" this and that, is really lousy terminology. It makes you think there's mass flowing around the car when there isn't. "Dynamic loads" would be much better.

I can't give a detailed answer on this, but I'm sure load sensitivity is modelled in some form or another. During testing of tyre curves I was perplexed to find that the tyres produced much more grip with a sharper edge on the plateau on light loads than it did on full cornering loads. Later I learned that it was the load sensitivity I was observing. The higher the tyre load, the worse the grip and the smoother the change from rising curve to max grip is. I'll try to post some screenshots of RAF data when I come home from work.

Of course.. 'weight transfer' is often used; and what's wrong with that? If you put large weighing scales under each wheel, which would be kinda tricky to do, they'd show more weight.. But thats not about load sensitivity anyway. I suppose a couple % of non linearity could already do a fair bit, as it happens on the front and rear.

Cool AndroidXP, looking forward to that! Could you plot friction i.e. (lat force / vertical load)? I'll do the same on the F3000 excel thingy that should be on my HD somewhere at home..

Quote from Niels Heusinkveld :

Now if load sensitivity where not modelled; the effect of lifting (which is what I'd naturally do) would not change the friction coefficient on the tyres. Load would change (weight moves to front) but the grip increase is linear. Hence the balance of the car not necessarily changing.

This part is wrong because you think that you have increased load->increased grip for the increased weight, and in the end not change the balance. But, it is in fact increased load->increased grip with constant mass there. You will change the balance.

You can put a scale under the front wheel, and it will read increased "weight". It does not mean there is increased mass. A scale never measures mass, it measures a kind of force which we call weight. Mass vs grip (force) is really what determines if you understeer, not weight vs grip.

I'm finding that the engine inertia improvements have changed car handling dramatically - far more than I envisaged. With that though far more realistic diff setups are possible while still having a quick car unlike before.

Would be interesting to get Bobs original road going setups and port them into X30 to see how they handle now compared to similar cars in RL

What you don't have to forget is that while additional load decreases the tyre grip efficiency, the net lateral force the tyres under load produce is still higher than the one at unloaded ones (kinda obvious, but anyone who hears about tyres giving better grip under lighter load for the first time might be confused about this).

So for example if the tyres under 2000N load produce 3000N lateral force (1.5G) and under 4000N load only produce 4000N lateral force (1G), the grip is severely reduced (1.5 down to 1), but the output is still 4000N on the loaded tyres and only 3000N at the unloaded ones. If you lift off in mid corner moving all the load to the front tyres, they will indeed lose lots of efficiency, but they will still provide more lateral force than the rears.

Yeah good point... I'm far more sure that something is wrong than I am sure WHAT exactly is wrong..

Won't hurt to compare LFS to the Avon data though.

Edit:

Quote :

If you lift off in mid corner moving all the load to the front tyres, they will indeed lose lots of efficiency, but they will still provide more lateral force than the rears.

Absolutely. Still, I think my argument would stand; as we're talking relative changes. Perhaps nothing is wrong with LFS's load sensitivity; but if there is, the result from lifting mid corner would mean more understeer compared to how LFS currently is; even though on an absolute scale the fronts can develop more lateral force than the rears. The diffrence will become smaller with more pronounced load sensitivity.


Edit2: Pic attached of simplified straight line load sensitivity from an unknown Avon single seater tyre; more later..

I got one of Bobs old road going sets for the XRG and modded it abit for how I think a RL standard street car would be. Changed gear ratios even more than Bob had and fiddled with the suspension slightly to what I think would be a common settings for a street car

The end result is attatched try it and let me know what you think.

XRG_Road Going.set

So, here we go. Surprisingly the load sensitivity turns out to be a lot less severe than I first thought it was.

On the first graph we see a typical slip angle/lateral grip graph, with the data normalised to combine the data of both rear tyres and positive as well as negative slip angle into one curve. I colour coded the data points to be red under light load, green on medium and black on heavy load.

The second graph shows the same normalised data in a normal load vs lateral grip fashion.

I might have to repeat that data with the current X31 tyre physics and in the FBM instead of XRT. Now, should I take R1, R2 or rather R3 tyres...

Yeah, well, uhm, okay. Did a 16 minute stint with all kinds of crazy FBM driving. Attached is first a normalized colour coded lateral grip vs. normal load graph, with the colours indicating the individual wheels (green = left front, black = right front, blue = left rear, magenta = right rear). The second graph is without normalization (with normalize I mean that an absolute value was taken for lateral grip).

It's kinda hard to draw conclusions from it, especially considering a lot of setup quirks are influencing the data. For testing I took the default setup, so downforce, high front ARB and tyre temperatures had their fair share of influence, but what I think I can deduct from this data is:

• The front tires lifted into the air due to ARB (plenty of zero-load data)
  
• The load sensitivity is not a linear function (at zero-load the grip is nearing infinity)
  
• This is hella confusing
  

It's art! :spin:

Here is the real f3000 data.

Front and rear, the results at 8 degrees lateral slip versus mu at 300, 450 and 600kg.

lol brains!

ok it's not really brains this time but it seemed fitting

Indeed :insane:

I'm on a coffee / beer / choclate diet; perhaps I'll try to make sense of your art there later Android

Quote from Niels Heusinkveld :

Here is the real f3000 data.

Front and rear, the results at 8 degrees lateral slip versus mu at 300, 450 and 600kg.

not exactly useful with just 3 datapoints ... you can plot pretty much any curve through these

@android
could you explain this one http://www.lfsforum.net/attach ... id=44143&d=1197571236 to me
either im really thick right now or theres something wrong with the graph but either crossing the 0 point on the "grip" axis with rising loads looks very wrong

Nah, it's logical. The grip is just the lateral grip, so for "right" lateral forces you have positive grip and for "left" forces negative grip.

If we look at the magenta line we see the lateral grip of the right rear tyre. It starts out with a low load (1000N) in the positive grip part, indicating that it was the inner tyre in a righthander, thus "right"/positive grip. Then we move on to the 1500-2000N part, which is what we get when we're driving in a straight line (both wheels equally loaded, so it crosses with blue). Of course, when driving in a straight line there are hardly any lateral forces, so they are around the zero mark. Then we move further down to the high load part, which tells us that here the wheel was the loaded outer wheel in a lefthander, thus showing "left"/negative grip.

that does somewhat make sense but in that case the calculation is wrong since µ has to be dimensionless and nonnegative

either way with this in mind the curves look very flat and load insensitive to me in the relevant part of the graph

Quote from Glenn67 :

I got one of Bobs old road going sets for the XRG and modded it abit for how I think a RL standard street car would be. Changed gear ratios even more than Bob had and fiddled with the suspension slightly to what I think would be a common settings for a street car

The end result is attatched try it and let me know what you think.

XRG_Road Going.set

I tried it on South City Chicane Route, and it felt really awesome! Good job! I never tried a road going set before, but it was really fun to just drive around in the sunset, looking at the great work the devs have put down on this track. I really liked the FFB your set gives -- feels just right IMO. I'll use it for some laps like that when I need to relax after a race!

µ is signless but multiplying a negative slip angle by a positive constant gives a negative force. So unless you take that into account, µ will come out signed.

Ahaha, stupid me. I screwed up the normalised FBM graph, I've re-uploaded the corrected version now. Makes a lot more sense that way.

Quote from Ball Bearing Turbo :

lol brains!

ok it's not really brains this time but it seemed fitting

damn, you stole my line

Quote from AndroidXP :

Ahaha, stupid me. I screwed up the normalised FBM graph, I've re-uploaded the corrected version now. Makes a lot more sense that way.

ive either turned even thicker than before or that graph makes no sense at all (just like your previous black xrt one which i couldnt figure out either)

im curious but this is confusing...
so load equals amount of weight of the car it has to ''hold''
what youre saying is more load = less grip
and if i turn the tires too fast it will be under too much load and loose too much grip(interrogation)
that means i must turn the wheels sloooowwwwlllyyy to get the most out of the tyre(interrogation)

PS:i got no interogation button and sorry for my english.