Physics improvements to BF1

[DaveWS]

Ok, i think that the big difference between a real F1 car and the BF1 car is:

(this applies to 100 mph plus speeds, when downforce is a big factor)

A real f1 car is normally stuck to the road, however when the limits of grip are reached, the car becomes VERY unstable and suddenly becomes VERY difficult. When a real f1 car skids, the levels of grip are FAR lower than when the car is not skidding. This is why a VERY fast steering correction is required from the driver to prevent loss of control.

The BF1 car in LFS however, when limits of of grip are reached and the car begins to skid, is much easier to cope with. The levels of grip are not much lower when the car is skidding as opposed to when it is not. The BF1 car is too easy to cope with slides, and has too much stability and grip when in a slide. This is why it is possible to drift quite easily in the BF1 car.

Basically i believe that the physics should show a more severe loss of grip and stability, when the car begins to skid, for a more true to F1 simulation.

I think that these characteristics of the real F1 car are due to the tyres used, which give very little grip when sliding, but lots of grip when within the limit.

[Batterypark]

Perhaps this is a setup issue? I'm not saying you're wrong, just wanted to make sure you've considered the possibility.

[DaveWS]

Yes i have tried different setups. I am saying though that the car is too easy to control when in the middle of a slide. If you watch a real onboard clip of an F1 driver losing the rear, they have to frantically play with the steering wheel to correct it, as the car suddenly loses all stability.

If you mean that i should try with less rear wing so i have less rear grip, the car slides more often, but is still controllable easily with a simple correction.

[al heeley]

I think I know what you're getting at. It's the point at which the car reaches the edge of the envelope and very suddenly 'snaps' into a slide or spin, very little warning or chance to correct, steer into the spin or recover the situation. LFS has now a more gradual feel to the edge of the envelope which is more forgiving and recoverable, which makes the game more playable.

[Gabkicks]

i've seen f1 cars slide around and recover. also the tire skids are like that so that we can tell when grip is being lost. irl you can feel it through your seat. f1 cars dont handle like they are on rails

here's a vid of schumi throwing his ferrari around a corner at imola a few days
ago :p
http://rapidshare.de/files/18937332/tires.avi.html
the change from "grip to no grip to grip" is a bit too gradual sometimes...

[XCNuse]

heck.. when i slide mine, my arms are going everywhere to get it back which i do... i mean sometimes its easy, sometimes its just.. well.. i lost it, thats when i spin out, and it usually happens really quickly and im on my way, otherwise i have to make quick reactions..

[DaveWS]

I think I know what you're getting at. It's the point at which the car reaches the edge of the envelope and very suddenly 'snaps' into a slide or spin, very little warning or chance to correct, steer into the spin or recover the situation. LFS has now a more gradual feel to the edge of the envelope which is more forgiving and recoverable, which makes the game more playable.

Yes that is exactly what i mean. I do understand that LFS has to be slightly more forgiving, as in a computer game you have less feedback and feel from the car, but it is much easier to catch the BF1 car in LFS than say in F1C. But if they want to provide the most realistic physics, then that does not mean forgiving physics.

As for your comment "Gabkicks", an F1 car is obviously not going to drive like on rails when going over large curbs at imola.

[al heeley]

Weren't there a lot opf complaints when S2 first came out, that the new higher powered rear wheel drive cars were just a bit too unforgiving? Maybe they've swung back a little too far the other way now. Me, I like more grip. Makes it easier for me to stay on!

[Blowtus]

Al, the problem that has been fixed seems more related to the seperation of lateral and longitudinal forces - when this was wrong the cars would slide too easily under throttle, this is great now. But the 'sliding' and 'static' type friction differences that are what Dave is talking about don't seem to have changed much. The tyres seem to bite a little harder, but still quite 'gently' compared to real cars.

[DaveWS]

I think that the game should have cars for all skill levels. So it is ok for most cars apart from BF1 to have the new forgiving physics. However, the BF1 car is a simulation of a F1 car, and should therefore be challenging to drive. If people cant cope with a difficult car like the BF1 car should be, then there are other more forgiving cars in the game. Then if they improve they should be able to cope with a more challenging BF1 car.

Edit: So for the BF1 the friction difference between 'sliding' and 'static', should be greater, for a more realistic drive.

[tristancliffe]

Also bear in mind that part of the problem is the wings. Right now they work at the speed the car is going, even if it's doing it sideways. Thus at 180mph you have the same downforce sideways as you do going forwards. But a real F1 car is much more yaw (and pitch) sensitive, so any yaw massively reduces the tyre grip - maybe it's more this than the tyres being 'wrong' or too forgiving. Also the underbody ground effect isn't modelled completely, so the car won't lose downforce as it bottoms out or goes too high. Combine the two and I'm pretty certain it'll be a much harder beast to control.

Maybe for now just put S2Q tyres on it ;)

[Hankstar]

+1 grip :nod: :)

I'd actually like more grip in 2nd gear in the BF1. At the double-apex hairpin at KY GP you really have to nurse it and it's quite precarious (you can't even accelerate a tiny tiny bit because you'll lose it), especially considering that you can practically chuck it through almost every other corner on the track up to that point...having said that I'm in the 1.49 club so I guess I'm not doing too bad.
So anyway, the old grip thingy seems to still be there, but I'm glad it's not as severe as it was :thumbsup:
The BF1 was such a lovely surprise though, I haven't tested any of the other cars yet :D

[DaveWS]

Also bear in mind that part of the problem is the wings. Right now they work at the speed the car is going, even if it's doing it sideways. Thus at 180mph you have the same downforce sideways as you do going forwards. But a real F1 car is much more yaw (and pitch) sensitive, so any yaw massively reduces the tyre grip - maybe it's more this than the tyres being 'wrong' or too forgiving. Also the underbody ground effect isn't modelled completely, so the car won't lose downforce as it bottoms out or goes too high. Combine the two and I'm pretty certain it'll be a much harder beast to control.

Maybe for now just put S2Q tyres on it ;)

I never thought of that. That certainly makes sense. I assume that you are saying that an F1 car has max downforce when facing the direction of travel, but when it is no longer quite facing the direction of travel (sideways) , there is a significant loss of grip and stability.

[XCNuse]

you loose downforce as you go sideways? because when going backwards, you have... no downforce.. ..still.. we need lift when going backwards ARG!

(i've tested it when forbin was doing the whole lift thing going upside down off a ramp.. i found out that downforce decreases to nothing from sideways to being completely backwards)

[Blowtus]

Also bear in mind that part of the problem is the wings. Right now they work at the speed the car is going, even if it's doing it sideways. Thus at 180mph you have the same downforce sideways as you do going forwards. But a real F1 car is much more yaw (and pitch) sensitive, so any yaw massively reduces the tyre grip - maybe it's more this than the tyres being 'wrong' or too forgiving. Also the underbody ground effect isn't modelled completely, so the car won't lose downforce as it bottoms out or goes too high. Combine the two and I'm pretty certain it'll be a much harder beast to control.

Maybe for now just put S2Q tyres on it ;)

I asked about this a few days ago, someone tested and said downforce drops to 0 when sideways...

[Tyrion]

I think I know what you're getting at. It's the point at which the car reaches the edge of the envelope and very suddenly 'snaps' into a slide or spin, very little warning or chance to correct, steer into the spin or recover the situation. LFS has now a more gradual feel to the edge of the envelope which is more forgiving and recoverable, which makes the game more playable.


I haven't done alot of BF1 driving since the patch was released, but while practising on AS Grand Touring I had a few occurances of that type of 'snap' exactly.There are a couple of high speed esses (?) in the latter part of sector 2 where this happened to me a few times - a very sudden loss of grip and before you know it the rear end is gone and your futile attempts at correction only end up making the situation worse :)

I guess it could be attributed to something else though, I honestly dont know. :shrug:

[Gunn]

I have had both moments that were easy to catch and those where lightning fast reflexes were barely enough to save a spin. Perhaps the ragged edge is just a little further away than people expect?

IRL F1 cars are set up to go as fast as possible for that given track and conditions. They don't set them up to be nice and comfy to drive and easy to control in a spin. In LFS many racers are not hitting WR times in their BF1 perhaps because they are not set up to go that fast. Once you engineer a very fast setup for the BF1 I find that it is indeed harder to catch a spin, or at least faster reflexes are required. I have also experienced fishtailing when barely saving a spin, which also was quite a handful to control.

It's hard to sit here and say "an F1 car is this, an F1 car is that" without having had the experience of driving one, but many people who have driven one say that they found it surprisingly easy, but very scary under brakes due to the enormous braking power and resulting G forces.

Other factors effecting spins in F1 include the condition of the track. Marbles of rubber, carbon fiber debris, dust and gravel all await the driver who strays offline, and in a spin situation this is often all it takes to make the rear end overtake the front. We don't have these things modelled into LFS but I strongly suspect we will in future as it has been discussed before at length.

I think people are focusing too much on trying to find fault with the existing version when really there are many factors that may be considered yet. I would urge racers to try tweaking those sets to see how fast they can make this thing go. It certainly has its fickle moments when pushed hard.
I notice many people using a fair bit of downforce, in real F1 they use only the minimum they think they can get away with, there is no comfort zone. On circuits with a lot of straight track it is not uncommon to reduce the amount of downforce so much that cornering is very difficult. There is always some compromise of course and in LFS I think many of us dial a lot of compromise into our setups. :)

Yes, sometimes the BF1 seems a bit easy to catch, but I can't say I feel any major issues really with how it is working.

[tristancliffe]

Okay, so perhaps LFS does model the loss of downforce with yaw. My bad. Can anyone get some rough figures for the drop off? Cos I'd say it's too linear, and needs to lose much more downforce much earlier. Real F1 cars stuggle with more than something like 7 degrees of yaw.

[axus]

Okay, so perhaps LFS does model the loss of downforce with yaw. My bad. Can anyone get some rough figures for the drop off? Cos I'd say it's too linear, and needs to lose much more downforce much earlier. Real F1 cars stuggle with more than something like 7 degrees of yaw.

Even if the LFS model does have loss of downforce with yaw, I doubt they modelled the slide's different effect on the front and rear wings correctly. Surely the rear wing would loose a lot more because most of the air doesn't even hit it at this point, whilst the front wing is still producing a reasonable amount of downforce. This would make the car more unstable when modelled. Also remember that the F1 car is on grooved slicks. This is the problem with full-on slicks and slides. The whoe tyre thread twists, rather than in seperate parts which is what makes full-on slicks less forgiving than grooved slicks.

[DaveWS]

... It's hard to sit here and say "an F1 car is this, an F1 car is that" without having had the experience of driving one, but many people who have driven one say that they found it surprisingly easy, but very scary under brakes due to the enormous braking power and resulting G forces....


Yes, people do find driving an F1 car relatively easy, that is until they have reached the limits of the cars grip, when a real F1 car becomes very difficult and sensitive. That is how F1 tyres differ to road tyres, because road tyres give more consistant levels of grip when sliding and when not sliding, whereas F1 tyres suddenly let go in a big way.

[axus]

Yes, people do find driving an F1 car relatively easy, that is until they have reached the limits of the cars grip, when a real F1 car becomes very difficult and sensitive. That is how F1 tyres differ to road tyres, because road tyres give more consistant levels of grip when sliding and when not sliding, whereas F1 tyres suddenly let go in a big way.

I just explained to you why the grooved slicks make a big difference there. Grip recovery with grooved slicks is much easier - still not as easy as road tyres but its not in LFS either.

Its much more difficult to twist a tyre like |____________| that back in line
than a tyre split into 3 pieces like so |___||___||___|
Go and try the FO8 and you'll see that while you can correct slides, you have to be on the ball a lot more than with the BF1. Also remember that the BF1 has throttle control to limit the longitudinal slip of the rear tyres while you are catching the slide because if you had high lateral and longitudinal slip, you'd have a much harder time.

[DaveWS]

But its the tyre compund of an F1 tyre which gives most grip when its not sliding, and much less when sliding. Thats why an F1 car should have a much larger stopping distance when the wheels lock up.

[axus]

But its the tyre compund of an F1 tyre which gives most grip when its not sliding, and much less when sliding. Thats why an F1 car should have a much larger stopping distance when the wheels lock up.

Who said we get more grip while sliding? Martin Brundle actually said that it is best to slide juuuuuuuuuuuuuuuuust a little tiny bit, that's when you get the fastest time - 3-5% side-slip. And, I agree with you about the stopping distance - while it is currently greater when the wheels are not locked, it should be much much worse when they are.

[DaveWS]

Yes a tiny tiny bit a slide does give maximum grip, because the compound is so soft, that a tiny slide stretches the tyre across the tarmac, giving a larger contact.

Imagine stretching an elastic band, lots of tension until it suddenly snaps

[tristancliffe]

This behavious has nothing to do with the compound - all tyres, indeed all rubbery things, generate maximum 'grip' with a small slip ratio. When you accelerate you always have wheelspin, but it's best if the slip ratio is less than 5%. Same with braking and turning.

The soft compound just decides how much ultimate grip there is, and the operating temperature range (in a simplified world, at least).

[loopingz]

May the aero is not enough complex at the moment because it does not simulate part of the cars that stall and create such instability !

[tristancliffe]

Well indeed, that's what we're saying about the downforce loss at yaw - the aerofoils stall or the airflow detaches, rendering the wing extremely or completely inefficient. So as you yaw you get less downforce, which gives you less grip, which makes recovery even harder.

Obviously modern Formula One cars are much more prone to it than say a Formula Renault or even an F3000, but it's an effect that isn't fully simulated. But give it time, and I'm sure something will be done. In the mean time we have the luxury of a relatively easy to drive F1 car.

[Gargantula]

You people are talking about the slip angle of the tires, correct?

[Forbin]

What about the rudder effect of the side boards that hold the wing elements on the rear wing? Surely they reduce they car's tendency to yaw.

[axus]

You people are talking about the slip angle of the tires, correct?

Not quite... yaw is the angle at which the car is sliding or side-slip angle. It refers to the whole car and the direction in which it is traveling compared to the driection in which it is pointing. What we are talking about is that when the car is sliding, it is much more aerodynamically inefficient, creating more drag and less downforce. When you get into a slide the rear wing also is a lot more ineffective than the front, increasing the difficulty in controlling the car. Furthermore, we were discussing the fact that grooved slicks regain grip much quicker if they hit an extreme slip-angle because it is easier to straighten 5 smaller bits of rubber than one big piece.

[Hyperactive]

What about the rudder effect of the side boards that hold the wing elements on the rear wing? Surely they reduce they car's tendency to yaw.

Their size is really not big enough. The air just goes "past" them, because of the shape of them doesn't block the airflow enough to make enough air resistance. And even in higher speeds the speed adjacent to the face of these side boards don't get very high. The only reason that they are big is probably some FIA rule (or to provide as big sponsor ad place as possible...?).

The air just goes "past" them, because of their shape doesn't block the airflow enough.

[Gunn]

That is how F1 tyres differ to road tyres, because road tyres give more consistant levels of grip when sliding and when not sliding, whereas F1 tyres suddenly let go in a big way.The cars have a long narrow wheelbase, so typically they would let go in a big way once the rear reaches a certain angle. This would happen irrespective of what tyres were on the car. I don't think any real comparison can be drawn between F1 tyres and road tyres, they are so different.

[Turbocharge]

ive seen the video on lfs movies pit i think of this BF1? being run on a South City track...i dont know much about downforce...but it would handle like a real f1 if...it has less downforce....the BF1 and FO8 are the same car almost?...only different hp and engine configurations??

[DaveWS]

Ok, so it probably is more to with the aerodynamics of the car, rather than the tyres which causes the sudden loss of grip in real life, but surely it would be easier to change the tyre physics in LFS rather than having to make complicated advances to the aero phyics.

As for "Turbocharge", yes the BF1 car does seem to handle remarkably similar to an F08 with TC and much more power, in my opinion.

[Blowtus]

changing one thing to fix something unrelated is not the way to do a good sim :)
But I think it's got a lot to do with tyres as well - all the cars exhibit very soft transitions into and out of a slide.

[Karlitos]

Yes, it´s a good idea. :nod:

[Funnybear]

I'm assuming that LFS doesn't exactly model air flow over a car more rather it tells a car how it should react in any given situation. Each car as an 'air flow' box around it that will shrink or grow exponentially as the car accelerates or deccelerates. Thus giving the car and the surrounding vehicles something to reference against to indicate proper behavour. A fairly complitcated bit of programming without adding in lateral slide movement through air. I'm alos assuming that LFS has 'attack' fronts from which the air will be travelling. I can only assume that this doesn't take into account a vehicle travelling sideways or backwards at very high speeds. So what it does do is give teh vehicle a general effect which ever way it is pointing at what ever speed, thus you will have the same downforce acting ont he car going sideways at 150mph as you would going the conventional way at 150mph. Meaning, and this is my point, that slides and out of control situations are easier to catch and control.

The real F1 car doesn't have the luxory of having uniform Downforce. Once a F1 car goes squirly at nearly 200 mph the lateral forces acting on it are immense. Most of the time, if an F1 car goes lateral at these speed even someone like Schumey is just a passenger.

Until LFS can fully model airflow from any direction in any circumstance at any speed you will never get the BF1 to handle exactly like a F1 car. There is to much design and effect going on to really get a true handle on it. Not unless you could download wind tunnel test info and implement them. But to do that you would need the entire LFS community computing power to simulate real world airflow.

I think LFS is as good as it can get for now. Scawen might get tighter on the code and implement more ideas but it's never gunna be any better than a best guess untill we have computers on our desks that can accuratly model and simulate real air flowing over a real car . . . . in Real time.

[Turbocharge]

ive seen the video on lfs movies pit i think of this BF1? being run on a South City track...i dont know much about downforce...but it would handle like a real f1 if...it has less downforce....the BF1 and FO8 are the same car almost?...only different hp and engine configurations??man....did you even look at the different aerofoil shape of those two cars?!?!?!?!:pillepall

[DaveWS]

...I think LFS is as good as it can get for now. Scawen might get tighter on the code and implement more ideas but it's never gunna be any better than a best guess untill we have computers on our desks that can accuratly model and simulate real air flowing over a real car . . . . in Real time.

Yes, I can see that modelling real airflow to significantly increase realism would be very complex and required vast amounts of processing power, but maybe there is an alternative more simple way to give similar handling characteristics, tyres maybe?

[axus]

Yes, I can see that modelling real airflow to significantly increase realism would be very complex and required vast amounts of processing power, but maybe there is an alternative more simple way to give similar handling characteristics, tyres maybe?

That will lead to problems at speeds where aerodynamics have no effect. And AFAIK X-Plane already has a CFD model so it can't be that heavy... maybe low-poly car models can be used for a basic CFD model at some point to get this right and not too CPU-intensive.

[DaveWS]

I personally think that F1C newest (ish) mod RH2005 has the best F1 physics of ANY F1 sim yet, including LFS BF1, much better than 99 - 02 season and RH2004 in F1C. However what lets it down a bit is the relatively poor force feedback, which is much better in LFS, and is why my new favorite place to drive F1 is in LFS. If they could combine the physics of RH2005 with the force feedback of LFS, they would have the F1 sim to end all F1 sims.

[petrichor]

I think I can tell a difference between fast and slow corners. When I lose it in the BF1 in fast corners, it snaps so fast I'm in the wall before I can rotate the wheel. This usually happens when I hit a bump wrong or get the slightest amount of rear tire on the grass. This may be due to loss of downforce, but it certainly is unforgiving.

In slow corners, the slides are gentler, and I can let the rear step out a bit for turn-in. I have plenty of time to correct, and nothing weird happens. I don't see how anything is wrong.

Maybe you guys are running too much understeer or too soft of a tire compound. :shrug: I usually end up running the hardest compound to keep from overheating, and they end up being hard to get up to temp and very unforgiving.

[DaveWS]

I think I can tell a difference between fast and slow corners. When I lose it in the BF1 in fast corners, it snaps so fast I'm in the wall before I can rotate the wheel. This usually happens when I hit a bump wrong or get the slightest amount of rear tire on the grass. This may be due to loss of downforce, but it certainly is unforgiving.

In slow corners, the slides are gentler, and I can let the rear step out a bit for turn-in. I have plenty of time to correct, and nothing weird happens. I don't see how anything is wrong.

It seems odd that you say that, because having driven the car again recently (on KY GP) I took a curb too much on one of the high speed corners, and it went into a slide. However I find that the car is still too stable whilst in a high speed slide, and i managed to correct it without too much difficulty.

*Edit. (on KY GP) At the slow speed double apex hairpin however, the car is very twitchy and unstable.

[Funnybear]

Then use that setup you have for fast tracks and find another more suitable setup for slow corners . . . .

I don't see how you can get anything more accurate out of LFS. After all, hace you driven an F1 car to corroborate your supposed discrepancies?

[DaveWS]

No, but I watch F1 on TV very closely, and the onboard shots give a very good idea of real life handling. Compare a real life replay to a LFS replay, comparing driver inputs.

Plus, have you driven RH2005, the physics simulate the loss of grip and stability in fast corners very well. Like i said its the rubbish force feedback in RH2005 which ruins the driving sensation.

[jtw62074]

That is how F1 tyres differ to road tyres, because road tyres give more consistant levels of grip when sliding and when not sliding, whereas F1 tyres suddenly let go in a big way.

How do you people know this? Ever seen F1 tire data?

[tristancliffe]

No, he is assuming it because F1 cars are either gripping madly or spinning off very suddenly. What he fails to take into account is the complexity of real life compared to LFS, the ability of the F1 drivers, and how close each is to perfection and the limit. If you're running within 0.2 seconds of the cars potential every single lap then it's bound to go wrong sometimes, and usually very quickly.

[DaveWS]

How do you people know this? Ever seen F1 tire data?

It was a guess, and I admitted several threads later that the problem was more to do with aero physics. However many people think that the tyres in LFS have too smooth a transition between grip and slide.

Since there are obviously no drivers on this forum who race F1 cars, we have to rely on educated guesses.

*Edit: spelling

[petrichor]

I*Edit. (on KY GP) At the slow speed double apex hairpin however, the car is very twitchy and unstable.

But it is easily corrected, right? That's my worst corner on that track, so I'm not really sure. Part of the instability might be the bumps there.

I just thought of another factor. F1 cars have more area in sideview in the back than in the front. If you are sliding, this would put more sideways aero force on the rear, tending to stabilize the car. I think this would be very important in real life. At low speeds, you have a car with little downforce and a rear weight bias, so it could be very twitchy.

You can say that loss of downforce due to side slip will make it unstable, but that depends on which end loses more downforce. If both ends lose downforce equally, it should just slide more, but not necessarily spin out. If the airflow is directed slightly to the side, the rear seems to lose more air because of the large side elements, but it is also getting cleaner air, because less of it is coming from the driver/airbox area. So I don't think it is obvious what will happen.

Maybe the fact that I'm getting dirt on the outside tire when I lose it is causing a dramatic loss in traction, and I'm confusing that with aero effects. So I guess I don't really know which part of the handling is aero and which is other stuff. But I do lose it plenty when I hit a bump wrong while totally on the asphault. One bad place is the fast, downhill right-hander (T6?) after the 3rd hairpin at the beginning of Aston North.

As for the tire comments, F1 tire data is in the Ferrari Formula 1 book (p. 190), if anyone has it. It looks pretty forgiving to me, especially at low loads (not much falloff in lateral accel for high slip angles). The measurements here don't go past the peak slip angle for high loads, so I don't really know how much it falls off, but the peak slip angle for high loads is very high, much higher than at low loads (10 vs 5 deg). So you should feel like you are sliding more at high speeds and probably should try to keep the car straighter at low speeds. Who knows if LFS's tire model looks anything like this.

[Gunn]

Ok, so it probably is more to with the aerodynamics of the car, rather than the tyres which causes the sudden loss of grip in real life, but surely it would be easier to change the tyre physics in LFS rather than having to make complicated advances to the aero phyics.LFS development doesn't work like this. Scawen isn't about to change the tyre physics to compensate for some other area. He's trying to make a realistic tyre model and a realistic aero model, not fudge the two until people are happy.
If the aero is not right then yes, he will make advances to the aero physics. This is what drives a wedge between LFS and its so called rivals, here somebody is actually trying to make a simulator rather than just throwing together a crowd pleaser with no real teeth.

[DaveWS]

It was just a suggestion, i don't know how difficult it would be to make advances to the aero.

Also the rear wing of an F1 car gives a lot more downforce than the front wing normally, so in a slide there would tend to be less rear than front grip. (going by the loss of downforce when yaw occurs)

"petrichor" I not really sure that the large side surface area at the rear of the car would siginificantly straighten out the car in a real slide, otherwise F1 drivers would be able to drift F1 cars or correct slides with ease. It may help a little though.

[skiingman]

. And AFAIK X-Plane already has a CFD model so it can't be that heavy...
Uh, hell no.

X-Plane uses blade element theory. That might be appropriate for car wings, but car wings are in a realm of aerodynamics simply not seen in airplanes, so the model would probably have to be rather impure to be "realistic". At low angles of attack with single element wings theory is very accurate. Modern multi-element car wings operating at the verge of stall are not going to be accurately aerodynamically modeled by any reasonable theory AFAIK.

CFD=not going to happen. Calculation for a single moment with a single set of conditions would take a typical PC minutes to hours, depending on the complexity and amount of iteration. Not only that, CFD over a "low poly model" is far more worthless than table lookups or things like blade element theory.

As to suggestions that the advertising endplates are going to give the car longitudinal stability, consider that any vertical surface forward of the CG does the opposite. The moments are small and the areas fairly small. On the other hand, taking a wing that is on the verge of stalling and yawing it even a couple degrees (like, less than five) away from the flow will dramatically effect its performance. Anyone that has ever kicked a rudder pedal in a Cessna thats about to stall knows this.

[Funnybear]

Reading this thread and thinking over somoe of the comments again I can't help but think that some people want the cake and too eat it too.

I mean, maybe LFS is ALOT more like real life than anyone suspects. You can't expect any car, even a F1 car to corner like it's attached to rails at greater and greater speeds. There will be a limit at some point. But it's the nature of cars and tracks that nothing will ever be perfect. They may be trying with the new F1 tracks to give them Snooker table surfaces and huge run off areas but even those tracks still have imperfections. Ripples, bumps, huge **** off holes, barriers, change in tarmac, off cambers, blind corners, bad setups, bad driving; The list is endless. All these things conspire to make driving, let alone racing driving, difficulty at the best of times. Driving a car is an insanly complicated thing, that even the most experianced get wrong sometimes.

So when people critizise the handling of the BF1, or any car in LFS for that matter, they need to take into account so many factors before berating the cars physics and aero. Take that double apex hairpin in Kyoto GP. Maybe it's difficult to control a car through there because it's difficult to control a car throught there. The problem is as simple as that, it's up to you to come up with a fix. Test and test some more. Prehaps lay off the critism untill you can drive and setup the car properly.

In regards to the comparision of Flight sim (Xplane was it?) aero and the aero needed for something like LFS, they are very different things. A car has so many things to deal with than just 'air'. Air behaves differently at ground level than it does 'up there' espeacially when you add in some serious F1 speed. You also have ground effect, lateral movements, rapid changes in direction, vertical movements, rapid decceleration and raped acceleration and basic mechanical grip.

Anyway with this new patch, although not perfect, it's still such an improvement and the theory behind everything so evident that how anyone can seriously critisize at this moment in time is beyond me.

[Blowtus]

That one came across pretty fan-boy ish mate. Finding ways to convince yourself / others that lfs is real, instead of looking at it objectively. LFS is a partially completed simulation - every physics update is a great leap forward. To assume that anyone commenting at all negatively / constructively lacks all feeling for vehicle dynamics, but then when the next patch comes out somehow the hordes saying how much better it feels is valid, seems pretty backwards to me :)

[DaveWS]

Sorry to bring this up again. Earlier in this thread i heard that real F1 cars lose considerable downforce at high speeds with tiny amounts of yaw (when the car starts to slide), and that LFS did not yet simulate this, and you actually had uniform downforce based on speed, but not which direction that you were facing.
Someone else, claimed that LFS DOES actually model this loss of downforce. Can anyone clear this up? And if it does, how large is the downforce drop off ..... etc

[tristancliffe]

iirc the downforce is full at zero yaw and zero at 90 degrees yaw, and the change is approximately linear, and applies the same to all cars with downforce.

[DaveWS]

I am just gessing here, but surely the change must be almost almost exponential? i.e half downforce levels at say 5 - 10 degrees?

[tristancliffe]

Possibly even more agressive than that, but yes, it should be in that region.

[AndroidXP]

Also I don't know if the 8° optimal slip that we have on all LFS tyres are that realistic. :shrug:

[DaveWS]

Ive just watched a replay of a lap where i was purposely flicking the car into slides, with the forces view on, an didn't really notice any loss of downforce until i span the car fully.

How many other people out there think that this is a realism problem which should be fixed? ... Devs..?

[Gabkicks]

at what speed?

[DaveWS]

140-150 mph (220 - 240 km/h) or so

[tristancliffe]

Ive just watched a replay of a lap where i was purposely flicking the car into slides, with the forces view on, an didn't really notice any loss of downforce until i span the car fully.

How many other people out there think that this is a realism problem which should be fixed? ... Devs..?

Interesting experiment, I'll have to try that. I won't say I will, because I did that with Shotglass regarding flatspots or something and STILL haven't got round to it.

[Nick_ll]

No idea if that can be of any use, but I think the main differences between a real F1 and the BF1 are that in LFS some things about the differentials are not implemented (I'm sure Scawen want to do them) and the suspension doesn't have a 3rd spring, which I'm sure Scawen wants to have too.

[DaveWS]

I was watching the Turkish Grand Prix today, and one thing that i noticed while the cars were on the parade lap was the way they were warming up the front tyres. They would be driving at slow speeds and then turn the wheel to full lock, causing the car to understeer ahead and heat up the fronts.

I thought that i would try out this technique on LFS in the BF1. However when i tried it out the car didn't understeer at all, in fact the car was turning far too much, and even sometimes oversteering, not what i had observed on the TV.

I guess that this is a tyre physics issue, so i carried out some more "tests":

This first test is on the skidpad, where i have driven to the outer ring, and at the limits of the cars lateral grip, proven by the green arrows on forces view (f key). I have then turned into the turn with FULL steering lock, expecting the car to understeer wide and take a wider path (as on the TV). Instead the car began to oversteer and take an ever tightening line!

The second test is on the dragstrip. I have driven to the limiter in 4th for a constant speed, and given my best attempt at "threshold" braking. I have then repeated the test with fully locked wheels, which gave a shorter braking distance! (marginally). This test was with the RACE_S setup but no downforce.

The third test is on the dragstrip but accelerating with NO tc. I set a time of 9.5 seconds. I then tried again with tc set to 5%, and set a time of 9.24 seconds. I am certain that if this test was carried out in real life the results would be several seconds out, not 0.26 seconds out.

The forth test is my attempt at a parade lap in the BF1. A total disaster.

Unless otherwise stated the tests have all been with the standard RACE_S setup.

Some tests may have been crude and inaccurate, but still give a conclusion.

I recommend you view the tests in the forces view.

Here is a link to a typical parade lap, where you can see the technique that the drivers are using to warm the tyres: http://www.youtube.com/watch?v=1XFWJIDzQJo

[DaveWS]

Braking test

[tristancliffe]

They have the advantage of a) being better drivers than us b) being in the car and c) not having to make do with crappy force feedback and tyre sounds for feedback.

If your car was oversteering doing it then maybe you weren't doing it right? Maybe the F1 drivers aren't just 'weaving the wheel about', but are doing so in direct relation to what the car needs to promote understeer. Of course, LFS's physics aren't going to be 100% (that is impossible).

As for the braking, who is to say your threshold braking is any good. They say the last thing a driver learns how to do well is braking. Try (using Shift-U) braking from 100mph in any car, using the gears and brakes, using the brakes only, and locking the wheels (don't do this bit in real life). When you stop put a marker by your front wheel, then repeat. Do it about 30 times and see how consistent you are.

Tip: Use a different colour cone for each type of test so you can see at a glance how well you did in each. I'd suspect that all these braking tests are let down by poor threshold braking.

[DaveWS]

If your car was oversteering doing it then maybe you weren't doing it right? Maybe the F1 drivers aren't just 'weaving the wheel about', but are doing so in direct relation to what the car needs to promote understeer. Of course, LFS's physics aren't going to be 100% (that is impossible).

As for the braking, who is to say your threshold braking is any good. They say the last thing a driver learns how to do well is braking. Try (using Shift-U) braking from 100mph in any car, using the gears and brakes, using the brakes only, and locking the wheels (don't do this bit in real life). When you stop put a marker by your front wheel, then repeat. Do it about 30 times and see how consistent you are.

Tip: Use a different colour cone for each type of test so you can see at a glance how well you did in each. I'd suspect that all these braking tests are let down by poor threshold braking.

I have thought about all of these things, but i did have 30 odd attempts at getting the braking right (no i was not consistant, but never managed to beat locked wheels), and all wheels were green in forces view on my best attempt...

I have some experience in Karting, and if you lock the wheels, you will not stop until you see a barrier. Why have ABS? because locked wheels do not stop quickly.

Any real car understeers in real life with full lock compared to the correct amount of lock required for the turn.

Just look at the replay of the skidpad test, then please make another comment.

[tristancliffe]

I haven't watched the replays.

Why have ABS? Because it enables you to turn. In fact I'd go so far to say that with road ABS the stopping distance would be shorter with locked tyres than the ABS stopping distance. But you'd just drive into the accident rather than drive around it safely. That is the major benefit of ABS.

In LFS I can routinely get understeer from steering to much. I do it a lot on outlaps in qually. But I don't drive the BF1 much, as I'm not either talented enough, or just because only about 10 people are in LFS, so the racing tends to be awful. And when I do drive it I'm more concerned about everything else than warming up the tyres like real life drivers.

Don't get me wrong - I'm not saying you are wrong in your analysis, just that any old person can't necessarily do in a sim what talented racing drivers can do in a car, no matter how realistic the sim.

[tomylee]

Sometimes when I turn very fast the wheel to the left and right, the BF1 does lose a bit the control but very often I manage to get it back in to control.
I never drove a F1 car :D but what I see on TV it does not look like that when somebody is losing the control of the car at 300 km/h that he will get it back on track :scratchch

I think there is some improvement to make :nod:

[Jamexing]

Acyually Tristan, a well caliberated ABS will definitely stop FASTER than non-ABS if the driven surface was suited to the ABS settings. A great example is the Lotus Exige's ABS. It's designed for thereshold braking and only really activates on panic applications. It's also well known that Kinetic firction of a tire can be about 30% less than rolling friction.

The ultimate stop is generated only when the tires are slipping a finite amount. That's why an over-active ABS would easily increase braking distances, especially for someone who has no idea how to threshold brake. An extremly skilled trail-braker would or well tuned ABS would generate the best stops.

However, I do agree that the BF1 could use better physics. In fact, I don't drive the downforce cars much because of their incomplete aero physics. Can't believe the aero has NO ptch sensitivity. Maybe Scawen somehow programmed the ULTIMATE active suspension system for all the downforce cars....

The biggest aero issues are definitely in the GTR class. Can't believe 3 massively different body styles have identical Aero! No wonder the XR GTR and FXR are suffering unfairly, since their aero packages are horribly mismateched to their mass distributions.

Another point is the fact that F-1 cars have a 3rd spring/damper setup that connects the spring/damper stes of both sides at the rear suspension. This gives 2 ENORMOUS advantages:

1. Stiff on vertical compression. Reduces pitch sensitivity and squat under aero load. Additional damper also better controls heave motion

2. Forces the inside wheel more to the ground whilst cornering, improving traction.

My point is, unless both aero AND F-1 suspension simulations are completely worked out, don't expect near perfect physics. For best realism, I recommend the non-downforce cars for now.

[tristancliffe]

Read my post again, and you'll see I said road car ABS. Whilst the Exige is a road car, it's primarily track orientated - not a comparison. Besides, who wants a track day car with ABS???

Sometimes you speak sense James, and other times you read what you want to read and make up the bits in between ;)

Can't believe the aero has no pitch sensitivity? No ther sim has yet either, but LFS is a work in progress. You always seem to forget that. Maybe Scawen just hasn't got around to improving the aero model completely yet?

And now we come back to the balancing issues, which you seem incapable of writing a post without blathering on about... You've made the point in more appropriate threads, why not let it lie until Scawen has time/inclination to do something about it.

[DaveWS]

Look there's no point in argueing. I made this thread, so we could find out areas of LFS physics which could be improved in the future. tristancliffe, it seems to me that all you do is defend LFS. Look i believe that LFS is one of, if not the best sim out there at present. But it still needs work. Why was this sub-forum made? So people could pick out areas for improvement. Thats why i started this thread. You can't expect LFS to be improved unless people post (constructive) criticism.

[Jamexing]

That's what I've been trying to do since I've got here (useful critism). And some still think I'm uttering crap... Hah. :(

Oh, well calibrated ABS is a boon for track racing, especially when racing gets long and consistancy under driver fatigue REALLY counts. Last time I checked though, the ABS stopping distances tend to consistantly BETTER than non-ABS on TARMAC. However, ABS DOES increase braking distnces if you're on a low traction off-road surface. That's why Mitsubishi Pajeros come with Multi-Mode ABS. :)

And yes, my chief reasons for getting the S2 license are:

1. Fundamentally good physics.
2. Constant improvements.

Which again explains why I tend to point out flaws that I find too obvious. :)

[Blowtus]

we had a fairly extended discussion of abs pretty recently in this thread here if interested
http://www.lfsforum.net/showthread.php?t=9929&page=2

continually pointing out the same flaws, that have been pointed out over and over before you arrived, does get a little repetitive ;)

[Jamexing]

Great Thread! The slip ratio graphs explain my points in a simple manner. Note that race ABS such as that of the Lotus I mentioned is meant to gain maximum braking power and only activates in panic applications for directional control. :) However, note that if an excellent EBD (Electronic Brakeforce Distribution) system was used in conjunction with an excellent driver or well tuned Race ABS, the stopping distances would be amazingly short. It's simply a matter of making the most of your tires' grip. :)

Anyway, let's get back on topic.

[DaveWS]

I never drove a F1 car :D but what I see on TV it does not look like that when somebody is losing the control of the car at 300 km/h that he will get it back on track :scratchch

Exactly, this was my point on the first post of this whole thread. The fact that you never (hardly) ever see an F1 driver rescue a slide at those speeds. Look at turn 8 in turkey last year, so many people were spinning the car.

The BF1, almost corrects itself at these speeds in the game. Good for gamers, but not for realism. I would buy GT4 if i didn't want a challenge heh heh.

[Hyperactive]

Basically the cars stopping quicker with locked wheels is not a huge problem in LFS because you can't lock your wheels without tire damage. So by locking your wheels you gain nothing. The bigger problem is that you get a better start by stabbing the loud pedal and breathing the smoke. Also the aero problems are known, and we can be sure that this will be updated. But it is up to Scavier when it's done. Ground effects are one thing which should be in LFS.

Looking at the aerodynamic bug, the BF1 not loosing enough downforce during large slip angles. I am defenately not an expert on this area, but just by thinking a bit further, the major change that occurs on bigger slip angles is that the "rolling resistance" (the forces that slow the cars down, drag as a major one) grow expotentially. F1 cars are extremely sensitive to any aerodynamic disruptions. Too bad this isn't the case in LFS atm..

But, still physics updates that mainly improve BF1 handling sound a bit odd. Looking this from different perspective: almost non existent engine damage, locked diff better than clutchpack, no reason to use manual clutch, car interiors from 1976 video games and sounds that are just bearable. Plus a lot lot of smaller things.

Imho, BF1 can wait :)

[tomylee]

This is anyway a problem in LFS, that when you lock your wheels there is nothing to feel afterwards. When you check the F1, they always say that after a big lock you have problems to handle the car well.

[Gabkicks]

yeah i've heard the vibrationsfrom the flatspots are so bad that the driver can barely focus.

[kieran20]

i would rather see the sounds of the bf1 be improved rather than the physics, they just sound so terrible and really un realistic. they sound like a 2 stroke f1 car ;)

[Blowtus]

expansion chambers hanging out the side of an f1 car? one can only dream :woohoo:

[Tomi]

I was watching the Turkish Grand Prix today, and one thing that i noticed while the cars were on the parade lap was the way they were warming up the front tyres. They would be driving at slow speeds and then turn the wheel to full lock, causing the car to understeer ahead and heat up the fronts.

I thought that i would try out this technique on LFS in the BF1. However when i tried it out the car didn't understeer at all, in fact the car was turning far too much, and even sometimes oversteering, not what i had observed on the TV.

I guess that this is a tyre physics issue, so i carried out some more "tests":

This first test is on the skidpad, where i have driven to the outer ring, and at the limits of the cars lateral grip, proven by the green arrows on forces view (f key). I have then turned into the turn with FULL steering lock, expecting the car to understeer wide and take a wider path (as on the TV). Instead the car began to oversteer and take an ever tightening line!

The second test is on the dragstrip. I have driven to the limiter in 4th for a constant speed, and given my best attempt at "threshold" braking. I have then repeated the test with fully locked wheels, which gave a shorter braking distance! (marginally). This test was with the RACE_S setup but no downforce.

The third test is on the dragstrip but accelerating with NO tc. I set a time of 9.5 seconds. I then tried again with tc set to 5%, and set a time of 9.24 seconds. I am certain that if this test was carried out in real life the results would be several seconds out, not 0.26 seconds out.

The forth test is my attempt at a parade lap in the BF1. A total disaster.

Unless otherwise stated the tests have all been with the standard RACE_S setup.

Some tests may have been crude and inaccurate, but still give a conclusion.

I recommend you view the tests in the forces view.

Here is a link to a typical parade lap, where you can see the technique that the drivers are using to warm the tyres: http://www.youtube.com/watch?v=1XFWJIDzQJo



The RACE_S setups are quite oversteery compared to real life(saving!) setups. I bet if you prepared a setup for a 1h30 race, your front tires would be set at a much higher pressures and the car would behave more like the real ones.

I haven't had the time to look at your replays though.

[Gabkicks]

i would rather see the sounds of the bf1 be improved rather than the physics, they just sound so terrible and really un realistic. they sound like a 2 stroke f1 car ;)

i guess your seniority has kept the usual suspects from jabbing at you for making such a statement... or maybe they arent awake yet lol.

[Jamexing]

Last time I did a CFD check, an airflow vector only 5 degrees off dead straight results in as much as 5-15% increase of static pressure at the lower surfaces of the car.

In layman terms, downforce drops by as much as 5-15% depending on the area of the car. Overall loss is along the lines of 7%. The test was done on an Indycar.

Fact is, aero is extremly yaw, pitch and roll sensitive IRL. LFS still doesn't model these factors, that's why it's so weird when the car starts to slide. IRL, cars snap when yawed excessively due to both loss of downforce AND the relativiely twitchy nature of low moment of inertia MR characterisics, coupled with the fact that slicks don't tolerate excessive slip angles very well.

[DaveWS]

Precisely what I was saying a while ago on this thread. Thanks for confirming it thought. :thumb:

Downforce does begin to drop off, but only at extreme yaw angles in LFS. I am sure this wouldn't be too :shy: much trouble to have corrected.

[KiDCoDEa]

so if i do a set, make it default in lfs, where u can warm up tyres and understeer like u see in tv (even without brake balancing), then lfs physics become perfect?
no they wont, they will just counter much ignorance.
also if lfs physics were 100% spot on and perfect, these exact same threads would pop up over and over.

findings that aero is not complete when its stated it is not complete from the start, and discovering that it needs undertray angle / height and yaw affecting the produced df etc, when that exact same line is listed over and over in the TODO list... wow what a great feedback.
what next will people find, that the AI DOESNT PIT?
and omg lfs calls itself 0.5v and its not complete!!!!

[DaveWS]

so if i do a set, make it default in lfs, where u can warm up tyres and understeer like u see in tv (even without brake balancing), then lfs physics become perfect?
no they wont, they will just counter much ignorance.
also if lfs physics were 100% spot on and perfect, these exact same threads would pop up over and over.

findings that aero is not complete when its stated it is not complete from the start, and discovering that it needs undertray angle / height and yaw affecting the produced df etc, when that exact same line is listed over and over in the TODO list... wow what a great feedback.
what next will people find, that the AI DOESNT PIT?
and omg lfs calls itself 0.5v and its not complete!!!!

If you look at the date of the first post of this thread, you will find that the aero was discussed a while ago.

In LFS some downforce drop-off IS modelled, just not very realistically.

This has only been added to the "todo list" (hyperactive's log) a few months ago. AFTER the last post before yours.

[tristancliffe]

I doubt that Dave, I bet it was on Scawen's todo list long before we brought it up ;)