Guys, come to talk about tires

Well since the cat is out of the bag now, I would like to post this here so people can see it and I can hopefully try and understand how to get this to work the way I would imagine it would.


This is the contraption in question... and while this technically isn't "tires" strictly... I think the amount of discussion here is where I will get the best help... or perhaps we all discover something.

Alright... now there is not much trickery here... the wheels in the track are all props, but that can be seen here.


The blue nodes around the outside are the structure of the track... the "rails" if you will. That is what holds the actual moving track in place, but doesn't come into contact with anything. The purple nodes on the outside are the actual physical track itself. The track has contactable triangles on the track itself, but since it is not closed, it provides no flotation in mud whatsoever (which in a way kinda makes sense from a calculation standpoint, but in real life, not so much). To combat this I applied contact triangles on the physical track structure itself as seen here.


This provides an equivalent amount of flotation as compared to a normal tire. Normally, in mud, the massive surface area of the contact patch would keep this thing from sinking all that badly into the mud (unless it is just that mucky... but then I feel like a tired vehicle wouldn't even move). This thing however... while trying to achieve the same kind of floatation, this is the best I have gotten.


Not great... but that wouldn't really bother me, as, like I said, it could be really really soft so this 8,000lb vehicle would still sink to the frame rails. What it still should be able to do though is have MASSIVE amounts of traction in mud and sand like this. However, that doesn't seem to happen.

It takes everything this thing has... locked diffs, most powerful engine, so on and so forth for this thing to drag itself out of this mud while with normal tires, it can pull itself out of mud fairly easily with normal amounts of power and no locked diffs. It just has no traction whatsoever.

Now, in the past, I would just add surface area to the nodes so that they would act like little scoops in the mud or sand to give some sort of extra propulsion. This doesn't appear to work anymore. So I ask the question... how is traction calculated in fluids like mud and sand? I need to try to make this thing have some more traction in mud as right now, its just about useless in mud and sand while it should be amazing... like... that's the entire point of buying these things.

Edit: I feel it necessary that I tag both @Goosah and @estama in this because if anyone is gonna know how these calculations are made and how to maximize floatation/traction on something like this, it will be you guys.

 

Which tires are best on wet surface? I tested this in Mount Coot-tha wet. The Result is Race>Sport>Rally=Off-road.Only race tires can do 1.0G cornering on wet asphalt. Is that simulated realistically? Idk since I'd never driven aggressively when raining.

 

I'm going to guess no. As far as I know this game doesn't really have wet road simulation yet, so the wet road is probably done using low-grip asphalt or (maybe) ice; thus the necessity of tread won't be properly captured.

 

I don't know there being anything that would actually simulate water film under the tire.

With racing tire one would get less grip as coarse surface of pavement is smoothed out and water carries part of the load, as water is not solid, there is only so little coarse stuff to make grip to happen. Then with speed that water has no time to be pushed away as much and there is even less contact to pavement itself.

While hydroplaning is kind of there, water layer on pavement is not afaik.

I guess that groundmodel of wet pavement would need some adjustments so that thread depth of tire has greater effect to resulting friction, so maybe there would be way to improve realism of that map, but still I'm not sure if it still would be enough to actually behave like having a film of water.

To be realistic road might need more data like painted areas where water is deeper and shallower, same paintable layer for variable grip levels of ice etc. might add to realism, but not sure how impossible such would be to develop.

So for map you tested, it probably has not all possibilities to create realistic water effect exhausted, that is one variable that would need to be examined before knowing if simulation itself is simulating grip realistically or ish.

 

You did the same as I did for my prototypical tank. Difference is that I have roadwheels to support the vehicle and I added a pressuregroup to the hull to make it amphibic.

I also had the same issue with mud and water propulsion and therefore I added the flaps/paddles that @estama suggests.

This has two downsides:
1: more nodes/beams/coltris
2: you can't make those flaps too big, you can't make them too small and if you care about weight, getting them strong enough so they won't invert(and therefore lock up and mess with the roadwheels) is hard.
The benefits are traction, nice water traction and interaction

And a question: How do you drive the track? I use a jbeamed and driven cog/gear

 

Somehow I figured this would be the answer, so I hope that in time the performance of these things will improve, and perhaps adding such a thing would improve other aspects of the game as well.

Out of curiosity, for drag in air, do you use a Fluid Boundary Condition to calculate drag, or is it more of a direct calculation off of the number that is defined in the JBeam? If you do use boundary conditions to calculate drag in air, perhaps that could be used to calculate drag in all fluids? Technically the thicker fluids would just use higher viscosity and thus incur much more drag. This would then help with the drag forces on a flat surface through a liquid... and thus probably actually improve traction for both tracks and tires (as now the sidewalls of the tire would have some drag on them as well) this would also help with the issue that was mentioned a while ago about the monster truck not having any propulsion when in water as there seems to be a lack of boundary condition fluid drag in all other fluids other than air.

To propel the tracks, I used rotators... this is what the code looks like.

Code:

    "rotators":
    [
        ["name","[group]:","node1:","node2:","nodeS","nodeArm:","wheelDir"],
        {"radius":0.54},
        {"brakeTorque":4700},
        {"parkingTorque":0},
        {"speedo" : true},
        ["FL", ["track_fl"], "trkfrmfl28", "trkfrmfl29", 9999, "fsuspl10", -1],
    ],

It's really simple to propel them actually... you are basically using the built in propulsion code to drive the track section rather than the wheels... however, doing it this way I have accidently discovered what I think to be a bug.

When propelling the tracks in this manner... naturally I have 4 of them, one for each wheel, and in reality, it should be a one to one swap. However... to get them to work properly, for whatever reason, I had to add in an extra gear reduction at the hubs for these tracks. The tracks on the left of the vehicle use a 1:1 gear reduction as there is no issue with them, but for the tracks on the right, to make them turn the same speed as the tracks on the left, they needed a gear reduction of 0.705:1... and I have no idea why. The code is identical across all 4 tracks right down to the spaces and indents... but for whatever reason, both tracks on the right need that reduction to turn the same speed as the tracks on the left.

Honestly, the only reason I noticed it is because I threw in some locking differentials and the truck acted like it was pulling really really hard and binding up on pavement.