Rollovers since 1.3? (take a look, Aristotelis)

[chaostheory]

Hi.

This is my first post here, but I want to make known that (as a real world kart racer) I love and enjoy the feeling that the physics this sim has.

I'm not sure if this is a bug and I couldn't find it posted before, so I think this could be a deep issue so I opened a new thread.

I've noticed a problem in tracks with relatively high curbs (aviano, newbury... almost everyone). When I hit them with the inside front wheel, while set in the corner, I rollover if I dont opposite-lock.

To some extent, I think it could be reasonable with the abarth, but in a F2000 for example (with its wheeltrack and CG height), it is a bit exaggerated.

The vids are not mine, but are examples of what I mean.

http://www.youtube.com/watch?v=lnuRtS3rGpA

The chance for this to happen in real world is very low (I never saw it in any class) because for a single seater, the CG height/Wheeltrack coeff. is very low.

It'd have to be rolled around 70 deg to turn unstable, and before it happens, the CG is responsible of supporting the torque needed to return the car to the natural position (see fig. 2.3)

I've analized the problem, and while I'm not an expert, I think the bug could be related with the position of the CG or more possibly with the excessive grip the sim gives to the tires when only the flank is in contact with the track (inmediately after the contact with the curb, the car starts the roll, but as it is set in the corner, with the outside tires loaded, it should start sliding also. It is more loaded on the outside tires, but the area is much smaller)

Fig.1:

Fig. 2:

Without more calculation, I think a reasonable maximum roll angle in real world could be about 30 deg (picture 2.2). At this point, common sense tells the car would lose the grip momentarily (or not have enough grip to continue rolling) and it would mean it returns to 2.1.

Not comparable, but for example at 0:30 here:

http://www.youtube.com/watch?v=dyaZ0SagRec

Or 0:04 here:

http://www.youtube.com/watch?v=_wXIAZ10-10

As you can see, the result of a curb hit in real world is usually a 4 or 2(inner) wheel jump of the car and a loss of grip, resulting in a momentary stright line in the trajectory (the only fact of going straight means no more centripetal force, so no rollover). The car looses part of the grip in the outside (loaded) tires and recovers the previous mass distribution (not without desestabilization)

In the sim, it seems that happens the oppsite, the tire gains absolute grip instead of loosing it.

I bought the sim when it was 1.0.3, I think, but I noticed the problem only in 1.3. I think everything have been improvements with the last version, but this annoys me.

I hope this helps to improve the best sim to date.

Anyway, for now I try to avoid the curbs and I enjoy the soft like a kid. Keep the good work.

Alberto.

Edit: clarification and new example

Edited October 3, 2011 by chaostheory

[marcocale]

Even if not supported from a detailed examination comparable with the one exposed from chaostheory, I have to agree that I had the same impression a couple of time with the FTarget in Prato.

Edited October 3, 2011 by marcocale

[marcocale]

Update. happened yesterday a couple of time also with F2000 at Newbury and Snetterton.

Edited October 4, 2011 by marcocale

[Guest CAR]

I don't think it's a new problem (if it's even one at all). In 1.1 if you took too much curb @ turn 1 Snetterton you rolled it, same as in 1.3 now. Formula Renault UK cars stay off the curbs at certain circuits, Thruxton for example:

(ok it's the new car but they did the same with the old).

Edited October 4, 2011 by CAR

[marcocale]

I don't think it's a new problem (if it's even one at all). In 1.1 if you took too much curb @ turn 1 Snetterton you rolled it, same as in 1.3 now. Formula Renault UK cars stay off the curbs at certain circuits, Thruxton for example:

......

(ok it's the new car but they did the same with the old).

I can't tell you if it's something already in 1.1, for sure it's something that doesn't convince me. If you look at this video (0:38 3:04 4:14)

You can notice that the car (a FRenault, very similar to FTarget, same place of first video of chaostheory) doesn't seems to suffer too much even if the driver passes totally on the internal curb at high speed.

By the way, at Snetterton, I rolled in the last chicane, first gear at very low speed. For me not very credible.

I'm not sure to have saved the replay, if I found I'll post it.

By the way, not so big issue, but surely something worth to have a look when Aris and kunos will have time.

Now we can go back to enjoy the great new multiplayer of version 1.3

Edited October 4, 2011 by marcocale

[Guest CAR]

Looks to me like he's using a lot less curb than the two examples in the original post (in the places they had their moments). The curbs are pretty big @ Snetterton =)

[chaostheory]

I don't think it's a new problem (if it's even one at all). In 1.1 if you took too much curb @ turn 1 Snetterton you rolled it, same as in 1.3 now. Formula Renault UK cars stay off the curbs at certain circuits, Thruxton for example:

(ok it's the new car but they did the same with the old).

I think no one can affirm that a F2000 is more prone to rollover than a clio cup, or a Vauxhall HP Firenza. They both use(d) the curbs @ Thruxton.

http://www.youtube.com/watch?v=0Q4OY69zk54

The suspension geometries are different, but what would explain now why the abarth and the vintage do the same?

I think the reason because FR2.0 and other sport cars don't use curbs is that it could case damages or unstabilize the vehicle, or both, specially if the ride height is set with a fast corner in sight.

Do you really think this could cause a rollover to a 30cm-CG-height-openwheel-car(if not 25cm), even with the most unfavourable setup?

Edit: subestimation of the data (corrected)

Edited October 6, 2011 by chaostheory

[Guest Tee]

In my opinion,

It might be a connected behavior with my video I uploaded late last year, flying abarth.

Once the tyre went off from the ground a lot the car got easier to fly, even though it was fixed later.

Edited October 6, 2011 by Tee

[Guest McSaltens]

for me i agree that in netkar we have too much tyre grip, and in ks2 also way too much downforce, but in the first examples, taking those kerbs at that speed with a high lateral g load can rollover a formula,

but also i think that the problem here is not only too much tyre grip, also like the weight of the car or the grabity physics are not correct

[Guest CAR]

I think no one can affirm that a F2000 is more prone to rollover than a clio cup, or a Vauxhall HP Firenza. They both use(d) the curbs @ Thruxton.

The Vauxhall's look like they're gonna roll with normal cornering Current touring cars use the curbs @ Thruxton, but often suffer punctures from doing so.

The suspension geometries are different, but what would explain now why the abarth and the vintage do the same?

I think the reason because FR2.0 and other sport cars don't use curbs is that it could case damages or unstabilize the vehicle, or both, specially if the ride height is set with a fast corner in sight.

Do you really think this could cause a rollover to a 30cm-CG-height-openwheel-car(if not 25cm), even with the most unfavourable setup?

Edit: subestimation of the data (corrected)

When you do see a Formula Renault car take one of those curbs (by accident) it completely unsettles the car so they lose a load of time, & I think they could launch the car as the original nKPro video's posted showed. Take for example the Singapore sling, during a recent KS2 event we saw the same behaviour as in F1, with many people doing a Kobayashi when taking too much right hand curb. I have had some strange rolls in 1.3, but can't say for certain it was the sims fault, as normally when you know that's gonna happen you don't make the same error again or if you do your ready for it =)