Can you drive an F1 car on the ceiling?

If car is going uphill, you need to apply downforce straight towards surface, not straight down. Otherwise downforce would start acting partially like drag.
So physics engine cannot apply downforce always vertically - this force's vector must change with car's orientation.

And if physics engine is capable of applying force at different angles, then it shouldn't matter what angle it is. Unless there's a bug in software somewhere.

I believe this is rather a limitation in track technology itself - where racing surface is stored in form of heightmap, rather than full 3D mesh. If you recall - rF1 had HAT data for that.

 

I think best car for this type of testing might be 60s Formula. It has fairly large tires and no overhangs, also suspension to dampen out some of the surface inconsistencies.

Of course, it has not enough downforce to drive on ceiling. But seems like 90 degree is already a big challenge.

Loop should be accomplished with 60s formula.

 

https://dl.dropbox.com/u/7364491/Stuntworld_devmode.rar

New version for dev mode.

https://dl.dropbox.com/u/7364491/Stuntworld.rfcmp

And component for Single play.


Edit: Much better now!!!

 

I've been testing the tube and the loop and the road seems to only offer support in the upper face. So going from the 90º, car wheels begin to pass through the tube and it goes crazy. The same happens in the looping at passing 90º.

I think changing polys isn't going to change something.

 

Direction of force shouldn't be an issue. If the program looks at wing angle and decides from that which direction the force needs to be applied, then it won't matter what angle the car or track is as it should still work. There is still supposed to be things like gravity and dirty air that would affect how the wing surfaces perform, but their ability to create lift should remain constant regardless of the angle of the wing in relation to the ground and gravity.

 

If the road is calculated using a height map, then you would be correct because we would be on the wrong side of the poly once we achieve 90 degrees and past it in the tunnel and loop. The track would expect the car to be on top of the tube and loop instead of inside.

 

I just uploaded new versions, you should try it. It gives better results, seems like the simulation is consistent indeed. It just requires A LOT polygons to pull it off.

Or then again... I am not sure if it is that good in the end...

 

sorry .......... this sounds totally nonsensical to me.
i dont have a physics major at all, but .... surely air doesnt have a direction............whatever angle you attack it from.......

again i havent googled it, im purely going on what you hear on tv.
but its constantly quoted , that an f1 car can travel upside down because of the amount of downforce it produces.

this totally makes sense to me. If its upside down....its still generating downforce because air is getting to its wings. There is no 'up' and 'down' in air..............is there?

 

I have tried new version too, and I see from outside camera the same results... when I reach 90º, my wheels go through road and car goes crazy.

 

Definitely this. I also took closer look, I found out that Clio does pretty well inside the tunnel when trying to barrel roll there. But also, just after 90 degree tires fell through surface thus losing their grip, making a barrel roll impossible. I don't think this is a bug, it is just the way this physics engine have been set up. But surely, these gimmicks *should* work in any decent physics engine...

 

This what rF2 beta forum was meant to be, threads like this.

I think ISI need to make a totally seperate main section off-topics fourm.

So Threads like .........." RF2 vs "whatever " or " Your favourite uboob clip" can go there too.

1/2 of them have nothing to do with the beta.

 

These topics were perfected in simulation(?) game called Stunts:
Jump to 1:50

At least this point we can't turn rF2 into a new Stunts.

http://en.wikipedia.org/wiki/Stunts_(video_game)

 

Hehe cool to see a test track but results dont look promising by all reports. If there is such a 90 degree limitation in the way the tracks are constructed would it take much of a change by ISI to overcome this if they wanted to? I will dl the track and have a little play with it just to see what happens when the game fails.

I still remember the old stunts game that could easily handle loops, but this may need a completly different approach and the ISI engine maybe just cant do it without a major overhaul.

Unfortunatly with such a limitation we dont get to test the aero simulation because we are limited by the upside down surface itself befor the aero even gets properly tested.

 

Hey, even at "Your favorite uboob clip" usually there are physics or graphics discussed, as yourself has taken part.
Are you sure you want everything organised as a strict company procedure?

I wouldnt want any of this, but hey, whatever, Ive seen a lot of this to be surprised by now.

 

True. But I think we could still test if we can make a car to fly. Build such aerodynamic parts for it that when driving required speed it would lift off from road. I would like to see, or experience that made in rF2

 

I was clearily talking about car going uphill, not air going uphill
Car going uphill will be differently oriented and therefore downforce will not be acting the same direction gravity does.
Car driving on ceiling is also differently oriented and therefore produces downforce acting in different direction - in this case it's opposite to gravity.

Probably the same I wrote at the end of my previous post

So you get normal "tyre vs surface" physics almost to 90 degrees and just collision physics above 90 degrees, which doesn't give you any kind of control.


If you want to make this kind of aero test, you will need to make kind of a bobsled track - very narrow, just for one car with walls on both sides.

Road will go straight for some time, so you can get up to speed and then there would be half a loop and a straight line ceiling made of material with minimum friction.
You would loose control half way up, but that's why those walls are there - to keep you in line.

 

As it doesn't look possible with current physics but racing on an upside down straight would have an interesting effect. As you approach another car in a draft taking air off their wings then its possible one or both will fall off due to the reduced downforce.

Lift created by a fixed wing is always relative to the wing whichever way it is pointing, you could easily make a race car take off if aero is modeled realistically but it will not stay up there as you have will lose all power once the drive wheels have left the ground. You would also have no control of the wings so it will just go up and end up in a Weber type crash.

Those crashes happened because the car was designed for minimum downforce and coming over the crest would lift the front up off the ground and air got underneath the car. The car didn't have enough downforce at this point to counteract the lift and so it got airbourne twice. This could have been avoided by lifting off coming over the crest or even a quick dab on the brakes to lower the front as it comes over the crest as rally drivers do over some crests.

If you consider an A380 can produce enough lift to take off at around 270kmh with 650 tonnes, an f1 car at 600kg could easily take off at much lower speeds and the wings are there also to avoid this happening. If they didn't create enough downforce then going over crests AKA Webberstyle would make them end up in the scenery. The F1 wings are basically just a reverse airplane wing creating reverse lift, upside down in theory it should have enough lift to keep it on the roof. The wings are fixed so they produce lift pushing the tyres down to the road, it shouldn't matter which direction the car is going relative to gravity as long as enough lift is generated to easily overcome it.

This is the same as military jets that can fly in any direction relative to gravity because they produce so much lift at the speeds they travel gravity is a very small factor relative to the total amount of lift. The main difference is they use a jet engine and not a driven by wheels on the ground so the car on the roof will be much less stable then it is on the ground having lost its weight which is now working against the lift.

600kg car with 1200kg of downforce going upside down would be 1200kg - 600kg working against it giving it only a 600kg weight holding it on the roof. Driving a f1 car on a flat road with 0 downforce would be completely unstable at any speeds much over 100kmh and a slight bump would send it flying, also you would lose traction and then speed which intern would reduce lift and down she comes. Even if you take the wings off an f1 car the body will create some downforce so zero downforce would be completely un-drivable. You would need at least 3x the cars weight in downforce to make it driveable upside down and I can see a real test driver raising a hand for this test lol.

 

Here is the first completed loop in the GTR its far from pretty and more down to luck as it landed back on its wheels after losing the loop near the top. But I drove out of it

The Big loop and upside down tunnel is just not possible with current in game collision physics going weird anywhere beyond 90 degrees.