improving terrain physics

[karlwb]

the bike and rider physics models,together with the setup adjustments available are really superb, the major limitation of the sim are the terrain physics,with only a 2d hard/solid surface to manipulate.
one solution would perhaps be to make an entire 3d track surface in blender,or cover the entire surface in mini 3d statues which i guess is completely impractical or it would have been done by now

but could it be possible to duplicate the 2d surface with another invisible plane 2" above whose only component is a resistance factor,this would perhaps produce the desired "give" of dirt etc. surfaces.

[Kawasakis]

Yeah we definitely need this. Because now riding in rough tracks is pretty much terrible.

[Armadijl]

Not if you got good suspension kawa

[Kawasakis]

Well yes but its not so realistic. Asphalted tracks? lol

[ddmx]

Would it be possibly to adjust the plushness of the terrain? i.e. Muddier tracks have much softer dirt, and blue groove tracks have extremely hard dirt.

[MiNi]

Oh yea. I hate the blue groove feel that goes around. The problem is that it is on every type of track except for sand tracks. There is a reason why Fort Winton is my all time favourite track.

[jlv]

I could make the tire to ground contact joint have adjustable spongyness. It could easily introduce other problems though.

[Mr. Snrub]

Hello... my name is Mr. Snrub, and I come from some place far far away.

I was just thinking about the current friction model. I was thinking perhaps it would seem more appropriate to vary the friction coefficient with lean angle, rather than with "slip speed"? (Or perhaps both models apply?). Here is my justification for it:

When riding in sand, mud, loam etc, the ground material will build up a small berm at the tyre, giving a 'banked turn' effect. If the ground was stable enough to form a normal force in the direction of the bike's lean angle, you would get the following equation:



Where theta is the lean angle from vertical.

Of course, this is not possible, as the ground material will eventually collapse/slip as the lean angle is increased. So here is my suggestion.

Implement some sort of angle coefficient, which will multiply with whatever the current surface friction is (defined by the current slip-speed friction model). Have it set up similar to the 'friction' statement in the tileinfo file. Eg:

angle 1 0
angle 1.1 10
angle 1.25 20
angle 1.25 25
angle 1.1 30
angle 1 35

Where 'angle' is the header of the line, the first number is the angle coefficient, the second number is the lean angle (in degrees). The coefficient is linearly interpolated between each defined value. This example could be for some shallow loam, where the friction increases until the loam becomes unstable and ineffective.

Hardpack would be defined like this:

angle 1 0

Please think about implementing this, as without any physical berm buildup, we will never get that proper sand/mud feeling.

Snrub out

[jlv]

I don't think lean angle makes any difference. A vertical tire with a lateral force applied would be almost identical. Once the tire slips a little, dirt builds up on the other side and slows it down regardless of the lean. You can model that with the current system by ramping up the friction at high slip rates. It might be worth separating the lateral friction though, since the tire has a smaller radius in that direction which could affect the rate the sliding friction increases at.

(BTW, 10,000!)

[FMXIAN]

I think DJ has had a few other posts under different users than "Mr. Snrub" haha. But still, congratulations on 10,000+

[Phathry25]

Good ideas here, although jlv shot down mr.snruby's, he has a point though.

I think the biggest hurdle in improving the terrain physics is that there's very little available information, and is more or less a no-mans land in terms of racing simulation physics. So there's nothing to compare to or to bounce ideas off of.

The current system feels fairly good to me, as the front end washing out feeling is something i ve always struggled with on a real mx bike. I don't know if its the tires or the higher ride height than a flat track bike, but irl when I lay my flat track bike over the back end steps out first, on the mx bike the front starts sliding first.

I will say this, I have no ideas at all to improve on what we have now.

[karlwb]

I could make the tire to ground contact joint have adjustable spongyness. It could easily introduce other problems though.

who knows? how could that be implemented? perhaps as "tyre air pressure adjustment" in the bike setup menu,might be worth a beta test at a convenient time!

grip gradient i think is a key issue,when a bikes leant over a knobbly on dirt or other loose surface digs in,slides digs in,you balance that feedback,and on harder surfaces you can feel the side knobs give when you're pushing near the limit but in game most often its a very short feedback rate.

[karlwb]

grip gradient i think is a key issue,when a bikes leant over a knobbly on dirt or other loose surface digs in,slides digs in,you balance that feedback,and on harder surfaces you can feel the side knobs give when you're pushing near the limit but in game most often its a very short feedback rate.

it's similar to braking in game,without anti lock,the rate at which the bike will rotate lifting the rear wheel is too fast,the feedback duration is too short because the grip is too instant,with anti lock on the rear still lifts but the rotation is slower and its more realistic.
i don't know how the anti-lock method for braking is applied in the game, perhaps something similar could be implemented for slide,if this improves the feedback duration this would make the washout moment more progressive when leaning down and the high side moment more progressive when lifting back up, it might also make the feet off when landing more progressive.