Re: Car's physics and setting...

C14ru5 Thu May 14, 2009 6:38 am

Tomte wrote:Maybe I forgot to say that when I moved the CoG of the Atom from 37 cm down to 10 cm above the axles (in the spirit of testing 'your' coordinate system), I didn't find the driving behavior natural at all.

Yes, that's the same 'sliding cardbord' experience that I had. Values below 0.2 didn't work well in any car that I've tried.

Tomte wrote:The suspension modelling is fairly crude in Redline, with the wheels moving only on the vertical plane and in parallel with the body roll.

The vertically restricted suspension may be a reason why low values don't work. It could be that the low value causes to the load transfer to be transferred directly to the wheels instead of being transferred to the suspension first. Or my findings that led to this theory may simply be due to the abnormal dimensions of a monster truck...

It seems we may have a problem with my theory. According to the inertia and CoG datasheet that I provided earlier, many road cars have a CoG 20cm above the wheel axle. Race cars would then have much less, but it seems that those values don't work. Are we back to square one, then? Neutral

Tomte Thu May 14, 2009 7:31 am

Back to square one? Perhaps. In a way I don't mind the current theory that all the dimensions are related to th model's origin. I might not fully understand it in the light of your findings, but it worked well for of in the past.
Maybe the reason that the cars topple over has less to do with the CoG wanting to move downwards once the car leans over a bit but more that it wants to move further outwards dvd to the centrifugal forces.
If the forces, which act on the tires, are similarly simplifyed and act on a single point (or a contact patch that does not change in size when cornering), then the contact point works simply as a joint.
Mercedes' delivery van had a similar problem: the tires where too wide, so they offered too much grip for the high CoG, the van was easy to topple.

Tomte Thu May 14, 2009 9:38 am

Ok, I just tried to get the Monster with the CoG at the bottom of the frame around the H1 test track.
I did not manage to get the truck to topple over when I stood sideways on one of the 45º slopes. It just slid down.
The version with the CoG at the transmission split however is plain weird. It rocks like a boat in heavy sea when you stand still, only stabilising once you move. It does not topple over as well

C14ru5 Thu May 14, 2009 10:54 am

Tomte wrote:Ok, I just tried to get the Monster with the CoG at the bottom of the frame around the H1 test track.
I did not manage to get the truck to topple over when I stood sideways on one of the 45º slopes. It just slid down.

Everyone move along, there's nothing to see here. Sorry if I gave some of you false hopes that we were about to undergo a paradigm shift in the world of understanding Redline physics. The most obvious way to test the CoG has proven that (what I suspected all along) my theory was just a consequence of me fooling around with unusual proportions for a car. Surprised

Tomte wrote:
The version with the CoG at the transmission split however is plain weird. It rocks like a boat in heavy sea when you stand still, only stabilising once you move. It does not topple over as well

Tomte wrote:
Mercedes' delivery van had a similar problem: the tires where too wide, so they offered too much grip for the high CoG, the van was easy to topple.

Hm... reducing the contact patch doesn't solve the problem, but reducing both (not just one) wheels.loadSensitivity and wheels.stickyness does solve the problem. The car feels a bit weird though, and it doesn't grip very well at all. Lowering the CoG to a point below the wheel axles was still the best fix that I've found so far. I guess the problem with my test car lies in some connection that the sway bar settings have with another variable in the suspension. I'll put away the file for now, and not worry too much about figuring out exactly what's going on. Most monster truck enthusiasts aren't that concerned about hardcore realism anyway Wink

DonaemouS Thu May 14, 2009 12:54 pm

early this morning, Jonas was reading the forum. I just can't understand the reason he didn't spent a couple of minutes to explain how some "uncertain" values work... Neutral

See Flat Thu May 14, 2009 12:58 pm

Seems a shame that everyone has to play forensics with these parameters just to figure them out.
I've seen Jonas lurking in here.

djpimley Thu May 14, 2009 3:54 pm

C14ru5 wrote:Most monster truck enthusiasts aren't that concerned about hardcore realism anyway

Excuse me?

See Flat Thu May 14, 2009 4:17 pm

djpimley wrote:
Excuse me?

"most"

djpimley Thu May 14, 2009 4:26 pm

See Flat wrote:"most"

C14ru5 Thu May 14, 2009 4:37 pm

hehe

dj, you're what I would consider "above average" king

Any other secret fantasies of yours that I should know about? Tractor Pulling? Laughing

EDIT: Gotta make this post on topic... ehm... "I don't believe tractor pulling is possible/sensible in Redline"

djpimley Thu May 14, 2009 4:48 pm

C14ru5 wrote:Any other secret fantasies of yours that I should know about? Tractor Pulling?

I just like variety. I'd love to see a realistic Monster truck but I know it's not possible with Redline right now. I'm very pleased with the way I was able to hack the Mack truck into shape but the Jeep has never been quite right due to limitations in suspension settings and low gravity.

IMHO the game engine is obviously designed for sedans and muscle cars. Stray too far from this form factor and things get funky very quickly.

C14ru5 Tue Jun 09, 2009 6:58 pm

Wake up, everyone! There's a working variable available that I haven't seen anybody use so far:

Code:: wheels.rollCenter

Based on my experience from testing with a Monster Truck plugin, wheels.rollCenter specifies the individual roll center for the suspension of each wheel, given in meters above 0 in the model. In other words, putting a value of 0 or below gives a boat-rocking sensation, while putting a value higher than wheels.pos gives the car a more wiggly feel. If left out, wheels.rollCenter = wheels.pos <- at least that's what I think, but perhaps the resting height of the wheels seems more reasonable? More testing needed.

djpimley Tue Jun 09, 2009 7:02 pm

C14ru5 wrote:
Code:
wheels.rollCenter

I have never seen this in a car file - not even the built-in ones. Shocked

C14ru5 Tue Jun 09, 2009 7:15 pm

Me neither, I was reminded of it when I read Ahruman's Redline Config File Specifications, and to my surprise, it had an effect! (unlike suspensionType, etc.)

Tomte Tue Jun 09, 2009 7:20 pm

looking through this documents, some interesting add-on variables seem to exists. I still wonder if one can build a trailer for a truck Smile

Redline Arrays.

EDIT: Ha, you beat me to it.

C14ru5 Tue Jun 09, 2009 7:27 pm

Yes, but I guess those were planned features that were left out of the game. It doesn't hurt to try them, though Wink

EDIT: Extreme values to the roll center don't seem to do anything!? Oh please, don't let this be another false discovery... Mad

DonaemouS Wed Jun 10, 2009 3:07 am

Code:: wheels.friction

I tried this variable on the newly released RS6. My test track was Runaway 1 mile test.

First test was with a value of 10. I got 28'17.
Second test with a value of 50. I got 28'73.

Now following C14ru5 comments, this variable should affect just the acceleration. But it's like it isn't. The friction affect how the wheel reduce (or hit) the torque in deceleration.

I made this small movie, showing how shorter is the braking space for the RS6 ƒ50 compared to 10, but also, using the first gear to kick the rear end, in the ƒ50 this behavior is widely reduced.

DonaemouS Wed Jun 10, 2009 8:28 am

reading and reading again the C14ru5 posts, I'm quite sure I want to use those formulas to wheel.friction variable. They have something right, I mean, rising the friction will lead us to reduce the power on brakes (something I need, cause the sedans are really heavy and need high values on brakes).

Also, we will not have sedans going as rocket (the real cars feature electronic speed limiter). But it's not all, cause we also reduce acceleration waiting to hit the 80kph (speed in which the air resistance start to have effects).

Having cars with "heavier" wheel friction can compensate also the infamous weight loss while we are in slipstream, and rising the grip while using lift.

I will spend some waste time to calculate how much should be the friction for front and rear RS6's wheels... scratch

DonaemouS Wed Jun 10, 2009 12:52 pm

ok here what i found:

i tried to calculate wheel.friction for the RS6. The weight of the car is 2055; this value include both the fuel, the driver, the body and the 4 wheels. The weight distribution is 60/40, resulting in 1233/822. The RS6 equip 275/35r20 Dunlop tires.

using the formula:

Code:: Nf*B

where Nf is the normal force and B is the rolling resistance coefficient, I obtain what follow:

616,5*9,81*0,00833=50,37871545 Newtons

(0,00833 is a value i got on Low-rolling resistance tires using the tire with similar sizes).

This for front wheels, while for the rear is:

411*9,81*0,00833=33,5858103 Newtons

Using these values we got these results on Runaway 1 Mile test:

Time: 28'65
Top/AVG= 272/198Kph

Now, I apparently missing the real "one mile test" to have values to compare with. If someone find something, just post here.

Actually the wiki formula could fit our needs. What u think about?

Sponsored content