Interesting discussion on "sliding the rear"

[yoshimura]

To understand the geometry of the bike, I would also hear moddeleur configuration of their case and their opinions.oui beta 6 is not perfect, but the geometry has to be perfect, on some models, c is far from the case.(just a parenthesis in this discussion)


What is geometry?
These are all dimensions of a motorcycle that influence its behavior: hunting, caster angle, wheelbase, the offset of the tires, suspension travel, the positions of the swingarm pivot, of axis gearbox output, the center of gravity of the aerodynamic center (longitudinal, transverse, diagonal), the moment of inertia of the whole of the motorcycle with respect to its center of gravity, but also the front forks, wheels, the gyroscopic effect of the wheels, crankshaft, the characteristics of suspensions (flexibility, amortization) for the most part.

The designer plays on these parameters to meet the specifications imposed on it: sport bike, trail, tourism, basic, etc.

Influences:
- Hunting: it determines the return direction, and thereby affects handling (when it is low) and stability (when significant) of the motorcycle.
- Hunting angle: it determines hunting changes for a given change in attitude (pitch)
- Wheelbase: when he believes it limits the variation in attitude, so hunting and thereby increases stability.
- The offset from the point of ground contact of the front tire: it conditions the induced turning when the bike is on the corner, especially with a Simultaneous action of the front brake. This induces steering conditions the handling of the bike on corner entry.
- The contact point of the offset of the rear tire on the ground: it determines the speed setting angle.
- Suspension travel: It determines the variations in attitude (pitching), and consequently, game variations, as well as the value of the wheelbase in reverse.
- The position of the axis of swing arm: it determines the variations in attitude, with the increasing length of the oscillating arm, but also the motor provided to take account of the position of the axis of the transmission output.
- The position of the gearbox output shaft: It is related to the previous, affecting the compression of the rear suspension related to the chain traction. By extending the swing arm and the passage of the chain as close as possible to the axis of the oscillating arm, it is limited compression torque of the rear suspension.
- The position of the center of gravity:
Above, it increases the mass transfer during acceleration and braking and cornering decreases the apparent angle, but it increases the time-to-corner.
Lower: it requires to deal with the ground clearance because it causes a visible corner taken very important corner, but it decreases the longitudinal mass transfer.
Foremost: it limits the rotation, but decreases traction and braking stability.
Further back: it gives more traction, but it makes the front less directive.
- The positioning of the aerodynamic center: Its height determines longitudinally transfer of the weight to the rear of the bike (under the effect of a growing push to the square of the speed), decreasing the guidance, so the stability, but increasing traction.
Laterally, it determines the crosswind sensitivity. It takes the same time to enter these settings the aerodynamic qualities (S.Cx) of the bike settings that one seeks to minimize.
- The moment of inertia: Its reduction has many advantages in competition for bets on the corner faster, more vivacity in general (faster mass transfer).
- The gyroscopic effect: It increases with the speed of the bike and with the weight of the rims, even the tires, and with the weight of the crankshaft (but also with the engine speed) and considerably slows angle variations (in input or output turn when the pilot noted the bike). In contrast, it increases the stability.
- The characteristics of suspensions are prépondérentes for adhesion, changes in geometry, and mastery of all parasitic phenomena (bars shake pumping sway).

credit:http://forum-mgp.niceboard.com/t65-geometrie-et-comportement-d-une-moto

[HornetMaX]

i read about 4-5 times i cant understand what your saying either your using the wrong words or like bobr6 im not a technical guy too .

Did you read the book ? Do you know what an effective steering angle is ? If not, one has zero chances to understand what I'm saying.

Interesting discussion. I was about to bring up that topic back with beta4 when I was trying to do controlled power-slides at Brno, but I thought it we had other problems to solve first. Now that the topic is on the table, I might as well join the discussion 

I was hoping to wake you up Stout 
If you can bring C21 into this too we're all here.

The virtual rider tries to calculate if the bike is under or over-steering. In case of under-steer, it steers more, the opposite in case of over-steer.

For over-steering this would read "In case of over-steer, it steers less". As a rider steers left in a left corner in a neutral state (no over-steering) and assumed the bike behaviour would change to over-steering the rider would have to counter-steer (to the right) --> which would be equivalent to steering 'less' (to the left)  -  albeit most riders would describe it as counter-steering to the right).

So, technically the behaviour of the virtual rider is correct.

That's exactly why I asked Piboso to confirm his definition of under/over-steering and of "steer more".
I'm confident his virtual rider does the right thing to (try to) maintain a neutral behavior (whether this requires to "steer more" or "steer less", in Piboso's definition of it).
What I'm more skeptical about is the goal of maintaining a neutral behavior: I'd tend to think that, in principle, this should be left to the rider to decide.
If he "wants to over-steer" he must be able to do so.

I can't confirm that it will work (only Piboso could do experiments with that), but I think we would be better off if the virtual rider was not taking care of trying to always have a neutral steering behavior.

Basic idea would be: I'm leaning left, mid-turn, I open the throttle too much, rear gets lose and slides out (right). If I do nothing else, the rear goes too far and either it washes out (low-side) or it suddenly regains grip and I go high side. If I push the stick a bit more to the left (asking to lean more to the left), the rider will have to turn the bars a bit more right, maybe allowing a better control of the slide. Same could be applied entering a turn (unload the rear, initiate the lean, rear will slide out, control it by asking for more lean).
It's just a hunch, but ...

• Imo, there needs to be some sort of anticipation of the virtual rider, meaning there should be a reaction time of ~0s for some cases. From what Piboso stated on the VR-behaviour it seems like the VR always reacts, maybe Piboso calculated some delay/reaction time?

Hmm that's dangerous territory: you can't anticipate that as you don't know when it will happen.
The virtual rider will not be able to know that in that specific situation you want to "over-steer".
Right now it just assumes you *never* want to over-steer.

* (On a sidenote: Not entirely sure about this, but imo the counter-steer into a steady-state (like in the Stoner Video) would result in a situation where |rear sideslip| = |front sideslip|, but we would still call it 'over-steering', which would contradict your definition MaX. But again, it is also a steady-state - so it is some sort of neutral behaviour, which would be in sync with your definition. )

It is possible, but IMO it would be harder to control if the sideslip angle is large. That was the text from the bike dynamics book:

".. with an over-steering motorcycle, if the force needed for equilibrium overcomes the maximum friction force between the tire and the road plane, the rear wheel slips, but the expert rider, with a counter-steering maneuver, has a better chance of controlling the vehicle equilibrium and avoiding a fall"
(page 113, paragraph 4.3, you have it on google books: http://books.google.fr/books?id=rJTQxITnkbgC&pg=PA105&source=gbs_toc_r&cad=3#v=onepage&q&f=false).

Better video still (video quality is crap), from 0:39 to 1:15:

https://www.youtube.com/v/rAm1msxnMfo

To me the front is always pointing towards the outside of the turn in order to keep the slide under control.
This makes the front sliding less, which is probably more controllable.

MaX.

[TFC]

I am not 100% sure on the specifics of what pib said, but I know most obvious one is under steering.

Johnson I am not saying you are wrong whatsoever (disclaimer but I am not sure that is what is happening with oversteer. I can say that with understeer in mx bikes I loose the bike and end up on the floor because of the potentially un confirmed 'auto correct' in moments where I know I can slow down and correct the understeer myself. Instead- if understeering left, the rider seems to attempt to correct this by sharply steering left causing the frond end to loose traction. If oversteer works the opposite then 'steer less' could potentially mean steer in to the apex to level the bike, instead of out more to control the slide, as with understeer it definitely tries to level the bike instead of control the understeer..

sorry if that doesn't make sense. On my phone and it keeps moving the cursor to the top line which I only notice after typing some lol might do a couple of test vids

[HornetMaX]

I can say that with understeer in mx bikes I loose the bike and end up on the floor because of the potentially un confirmed 'auto correct' in moments where I know I can slow down and correct the understeer myself. Instead- if understeering left, the rider seems to attempt to correct this by sharply steering left causing the frond end to loose traction.

I think the very same (and it happens in GPB too): you go in too fast in a left turn, the bike under-steers (trajectory widens), the virtual rider turns the bar more to the left, you lose the front.

This would be consistent with Piboso's statement ("In case of under-steer, it steers more") and the fact the virtual rider tries to re-establish the neutral steering behavior.
That's why I not only think that this "go for neutral steering" prevents us from controlling a slide, but it may also be responsible of these cases of losing the front when going in too fast (too fast for the turn, it can happen at relatively slow speed).

MaX.

[teeds]

Basic idea would be: I'm leaning left, mid-turn, I open the throttle too much, rear gets lose and slides out (right). If I do nothing else, the rear goes too far and either it washes out (low-side) or it suddenly regains grip and I go high side. If I push the stick a bit more to the left (asking to lean more to the left), the rider will have to turn the bars a bit more right, maybe allowing a better control of the slide. Same could be applied entering a turn (unload the rear, initiate the lean, rear will slide out, control it by asking for more lean).
It's just a hunch, but ...

I can see what you mean but this would be completely counter intuitive, as the normal reaction irl would be to steer the opposite way to what you suggest in that situation. It all boils down to there is no steer control while using the lean control option.
Also with what you have stated with under-steer

I think the very same (and it happens in GPB too): you go in too fast in a left turn, the bike under-steers (trajectory widens), the virtual rider turns the bar more to the left, you lose the front.

So are you saying that the rider predicts an impending wide line and adjusts? This seems odd as under-steer is a physical effect of loosing grip or pushing on the front end as over-steer is the rear loosing traction and going out of line with the direction of travel.
It would seem more appropriate to consider that a player goes into a bend too fast and realizing it tries to ask for more turn or braking from the bike, eventually exceeding the grip level of the front tyre causing a low side. Seems pretty real to me as loosing the front (under-steering) is not normally a recoverable event on 2 wheels as it happens so fast. And if someone did have time and try to recover from such a situation the same would apply as with the rear, forget the corner and steer more with the direction of travel and sit up(reducing the lean angle) but the aim being keeping it rubber side down.
Not sure if Piboso meant it as you describe as I find it hard to believe that the virtual rider is trying to compensate for an event that has not yet occurred.

[HornetMaX]

Basic idea would be: I'm leaning left, mid-turn, I open the throttle too much, rear gets lose and slides out (right). If I do nothing else, the rear goes too far and either it washes out (low-side) or it suddenly regains grip and I go high side. If I push the stick a bit more to the left (asking to lean more to the left), the rider will have to turn the bars a bit more right, maybe allowing a better control of the slide. Same could be applied entering a turn (unload the rear, initiate the lean, rear will slide out, control it by asking for more lean).
It's just a hunch, but ...

I can see what you mean but this would be completely counter intuitive, as the normal reaction irl would be to steer the opposite way to what you suggest in that situation. It all boils down to there is no steer control while using the lean control option.

Right, but that's already the case: to lean left irl you steer right, while in GPB (except DST) you push your stick left.
The goal is not to correct this (there's DST, if one wants that). The goal is simply allow to control (somehow) a rear slide.

I think the very same (and it happens in GPB too): you go in too fast in a left turn, the bike under-steers (trajectory widens), the virtual rider turns the bar more to the left, you lose the front.

So are you saying that the rider predicts an impending wide line and adjusts? This seems odd as under-steer is a physical effect of loosing grip or pushing on the front end as over-steer is the rear loosing traction and going out of line with the direction of travel.
It would seem more appropriate to consider that a player goes into a bend too fast and realizing it tries to ask for more turn or braking from the bike, eventually exceeding the grip level of the front tyre causing a low side. Seems pretty real to me as loosing the front (under-steering) is not normally a recoverable event on 2 wheels as it happens so fast. And if someone did have time and try to recover from such a situation the same would apply as with the rear, forget the corner and steer more with the direction of travel and sit up(reducing the lean angle) but the aim being keeping it rubber side down.
Not sure if Piboso meant it as you describe as I find it hard to believe that the virtual rider is trying to compensate for an event that has not yet occurred.

As far as I've understood Piboso's explanation, there's no prediction: it just measures the steering ratio (i.e. measure if there's under/over-steering) and react in consequence.

The problem is that a little front slide is manageable (just like a little under-steering), but if the virtual rider tries to return to neutral behavior, he ends up steering more, causing the front low side (without you doing anything). If he didn't try to keep it neutral, maybe you'd just go a bit wide, with a chance to lift the throttle and recover, or just go off-track ...

MaX.

[Phathry25]

Are we playing the same game MaX?  If I go in too fast the bike will just track wide.

The way I read how the rider works is that we control lean angle. Really I just hold my control stick all the way down and try to hit the right speed. If I'm too fast I run wide, but the rider certainly does not try to turn too much and end up causing under steer. If I press the front brake I can make him crash if I want. But really IMO it all feels pretty natural at the front of the bike.

At the rear of the bike something seems a miss. It feels alright on corner entry, enough that I would say my crashing when trying to back it in is usually my own fault. There is some awkwardness when the bike gets into a wobble under braking, you might as well just lock the brakes and get it over with. Lol. Although due to the exaggerated nature of video games this could well be the case in real life. We just think we should be able to get away with it here even though it would be a huge moment on a real bike.

Where I feel things are going sideways are under acceleration. When you initially accelerate you can see the rear tire slipping laterally just a little as it should, maybe even a bit too much. But that's not important. Where it goes wrong is when the tire spins up.  Since we don't know for certain what the virtual rider does it's hard to hypothesize what's going wrong there. It could very well be the tire simulation though. It's hard for me to get into it typing from my phone without being able to check the in game behavior. I'll try to make time to make a video soon to describe what I am seeing.

What stout says all makes sense to me. If the rear tire has a smaller turning radius than the front you counter steer. If it's larger you're just steering the bike like you would a car. Not that that helps with a solution to the problem at all since we're not sure what it is yet.

[teeds]

Right, but that's already the case: to lean left irl you steer right, while in GPB (except DST) you push your stick left.
The goal is not to correct this (there's DST, if one wants that). The goal is simply allow to control (somehow) a rear slide.

But I think, as was correctly pointed out by yourself, that the input for counter steering is more a torque input with very little actual steering rotation. To correct a slide the input would be way in excess of this and it's not to affect the angle of lean but to correct a mistake of to maybe much power at the rear maybe oil on track or similar.
Also the corrective input irl is very instinctive, which involves keeping the front wheel tracking the direction of travel which in turn has now changed due to the rear breaking traction. The virtual rider seems to imitate this pretty well from what I've seen, the control is then in the players hands via the throttle. Too much and you low-side, too little and you straighten up and loose the rear steering effect, but if you could hold it just right you hold the slide. Currently I think that sweet spot is just too small hence your complaint and I would attribute this to other factors of dynamic grip of the rear tyre once it starts spinning and moving sideways. Same thing in MXB.

The problem is that a little front slide is manageable (just like a little under-steering), but if the virtual rider tries to return to neutral behavior, he ends up steering more, causing the front low side (without you doing anything). If he didn't try to keep it neutral, maybe you'd just go a bit wide, with a chance to lift the throttle and recover, or just go off-track ...

I agree with Phathry25, the front seems ok to me, I've learnt the limits now and if I go to fast into a corner and therefore too wide bail out by sitting up, slowing and going straight on just like I would irl.

[HornetMaX]

Currently I think that sweet spot is just too small hence your complaint and I would attribute this to other factors of dynamic grip of the rear tyre once it starts spinning and moving sideways. Same thing in MXB.

Could be as well: a while ago I pointed out that in the Pacejka model for bikes, there's a slef-admitted "bug" that may lead to underestimating the lateral force generated by the tire at large camber and sideslip angles ("large" in the math sense of validity of linear approximation). Original post: http://forum.piboso.com/index.php?topic=1320.msg17177;topicseen#msg17177

The way I read how the rider works is that we control lean angle.

That was our general understanding of it. But with the discussion on the MXB forum, Piboso has added this:

The virtual rider has a control for yaw, too.

And this:

The virtual rider tries to calculate if the bike is under or over-steering. In case of under-steer, it steers more, the opposite in case of over-steer.

To me that's more than "we control the lean angle": trying to control the lean angle and the over/under-steer at the same time may not make a lot of sense, as at steady turning under kinematic conditions the two things are tied.

Anyway, don't know, maybe I'm just misinterpreting Piboso's statements (hence the request to for clarifications addressed to him) or maybe I'm just rambling (hence I should stop and go play cod4 for a change).

The two facts however stands:

• sliding properly (both entering a turn and power-sliding mid-turn) seems a bit too difficult in GPB and a lot too difficult in MXB.
• in the cases where we lose the front in an unexpected manner we see that as soon the front slides a bit, the virtual riders steers in (steers right for a right turn) way too much, making the situation worse and leading to an instant low-side. That is consistent with what Piboso is saying I think (In case of under-steer, it steers more).

Just wondering if the new information we have about what the virtual rider does relates to the two facts above.

MaX.

[teeds]

This is what i'd call classic under-steer but this type of recovery is rare, pretty sure Colin thought he was going down there too.

https://www.youtube.com/v/vjcR3KQgklQ


I agree that the description of the virtual riders reaction to under-steer by Piboso seems confusing. I've have seen while playing MXB (i'll try get a video of it next times it happens) the front beginning to tuck and have occasionally got out of it by reducing my lean request, but mostly it's time to hit the reset button. This happens when the front wheel hits negative camber, which strikes me as normal.
Maybe Piboso was talking about the opposite of the rear stepping out, ie the rear stepping inwards?  Does that ever happen? Then his statement makes more sense but is maybe a bit irrelevant. But I suppose if you get a tank slapper the rear could swap sides and the rider will try keep the front wheel in line with bike travel, as one would attempt irl.

The two facts however stands:

• sliding properly (both entering a turn and power-sliding mid-turn) seems a bit too difficult in GPB and a lot too difficult in MXB.
• in the cases where we lose the front in an unexpected manner we see that as soon the front slides a bit, the virtual riders steers in (steers right for a right turn) way too much, making the situation worse and leading to an instant low-side. That is consistent with what Piboso is saying I think (In case of under-steer, it steers more).

You're bang on with point 1, but on point 2 are you sure it's not this? -

https://www.youtube.com/v/yTiXqFyTW3I

Just before he goes down the wheel turns towards the apex suddenly, but to me this is an effect of under-steer and not the rider turning in suddenly causing it?
However I do agree there are some mysterious front end slides now and then in GP and MX bikes.

[HornetMaX]

This is what i'd call classic under-steer but this type of recovery is rare, pretty sure Colin thought he was going down there too.

Yep. Now imagine the same thing and, as soon as it starts sliding toe front, he tries to save it steering more to the right ...

You're bang on with point 1, but on point 2 are you sure it's not this? -

Just before he goes down the wheel turns towards the apex suddenly, but to me this is an effect of under-steer and not the rider turning in suddenly causing it?

Hard to be sure about anything in this discussion. One possibility is that the virtual rider, trying to recover the under-steering, actually causes an even bigger problem. The other possible explanation could be that the large steering (to the left in the GSXR video) happens when it's already too late (front gone) and hence is not responsible of it.

Re-reading all this today, maybe the intended interpretation of Piboso statement ("In case of under-steer, it steers more, the opposite in case of over-steer.") is that if turning left and over-steering, the virtual rider will steer less, i.e. steer less to the left, which is steer to the right, counter-steer. That would be fine then, so maybe is just a matter, as Stout said, of not doing that quickly enough (i.e. the reaction to an over-steer is not strong/quick enough).

MaX.

[TFC]

if turning left and over-steering, the virtual rider will steer less, i.e. steer less to the left, which is steer to the right, counter-steer. That would be fine then, so maybe is just a matter, as Stout said, of not doing that quickly enough (i.e. the reaction to an over-steer is not strong/quick enough).

MaX.

It's pretty ambiguous in both examples. Playing today I have definitely noticed the 'steering into apex when understeering', but it still doesn't make sense to look at. Plenty of speed and the right angle through a berm and half way around he just ditches the front end..

What strikes me as a bit odd however, is surely an oversteer in bike and car terminology involves the front wheel(s) pretty much staying on course (which would be treated as correcting / maintaining / controlling)..  When loosing it at high speed, the only way to bring an oversteering mass into alignment is to ease off and control the acceleration and try and get the back end to bite.. Unless you are very skilled, trying to keep the front end on course is the only thing that will keep your car on the track..

Am I wrong?

I would have thought if anything that there is far more going on here than just the rider steering.. Or maybe there isn't, which is why it's so hard to control. I can't deny it - the rider in MXB does an excellent job of correcting oversteer, if your throttle control and rider lean is right you'll be back in a straight line in no time, it's just a shame if you end up correcting it in a way that leads to understeer lol 

[HornetMaX]

Playing today I have definitely noticed the 'steering into apex when understeering', but it still doesn't make sense to look at. Plenty of speed and the right angle through a berm and half way around he just ditches the front end..

Show it to Phatry25

MaX.

[Phathry25]

Going to assume Fat was playing MX Bikes.  The behavior is different between the two games.  MX Bikes I can get the front end to wash out by using a little too much throttle exiting the corner, that's not right.  I could understand if it was the back of the bike I was losing.  That would at least be acceptable, IMO.  GP Bikes, I can't do that unless I start riding like an idiot.  There's something else at work in MX Bikes causing it.  Like I said, feels right to me in GP Bikes.

When the front end does go it doesn't really matter what steering input you use, you're not gunna have a good time.  The one thing GP Bikes is missing is the rider picking up up with his knee.  That is something that it looks like PiBoSo is looking to add to MX Bikes, no reason why it wouldn't be ported to GP Bikes in due time as far as I can see.  Then again, they might have asked for it to work at low speeds only, which is ridiculous.  Better go call them all idiots!

[HornetMaX]

Trying the stk1000 nikeset, after 2 laps I got this (same fall, 2nd video slightly more zoomed in and with better angle):

https://www.youtube.com/v/uycr1NNPVxk https://www.youtube.com/v/Y3Ix8M9cUOA

You don't get the impression that the front is gone *because* the virtual rider steers in ?

MaX.