10 replies to this topic

[Greg Locock]

A few threads back I said I was going to write a 3DoF (well, 4 as it turned out) vehicle model to investigate the effect of PMI on turn-in and spin characteristics.

Well, having prevaricated for a bit I found a much better solution

http://www.ing.unipi...73/meccveic.htm

has the educational version of carsim, and, rather more importantly the manual. If you are at all interested in whole vehicle simulation then this is worth the longish download (10 Mb all up). There's also a paper describing the basis of the model.

Importantly it has a closed loop steering model. Sadly it doesn't have a closed loop throttle, I think, in which case it won't be much use for circuit sims. Nonetheless, even a closed loop double lane change gets pretty exciting, if you drop mu.

[gug]

if anyone follows that link and gets freaked by the italian, dont worry. the program is in english, and so is the manual.

[Greg Locock]

I strongly recommend this program. It turns out that there is a lower limit to the PMI, below which it makes no difference to the handling of the car. The base car is a Hummer, so the numbers are rather meaningless. I am going to build an F1 car up and repeat the experiment - has anyone got a good estimate of the PMI of an F1 car? here's my guess:

wheelbase = 3.3m mass= 600 kg. PMI guess= m*r^2=600 *(3.3/2*.5)^2 where the .5 is a guess because the heavy stuff is near the CG. 408 kgm^2

I'm assessing the handling by looking at the steering inputs needed to do one lap of a 40m radius circle, with straight entry and exit. The closed loop steering attempts to follow the track given, the oscillations as it does so are indicative of the ease of control of the car (easily checked by setting the CG height to zero). Half way round the circle I apply throttle to see how it copes with a disturbance.

[Paolo]

In the Inertia thread :
http://forums.atlasf...&threadid=61372
it was suggested an inertia of 600 kg m2 was in the ballpark.
I agree with the estimate : my results (based on a model in wich the car is represented by superimposed parallelepipeds) are from 600 to 700 depending on the weight of components, with pilot aboard.
The central tank doesn't have much influence, wether it's empty or full.

[Paolo]

By the way, the scarse response to lowering inertia limits could be well attributed to certain mathematical problems involved with simulation I am sadly aware of (I'm building one myself).
There is the necessity of cutting friction forces when they are beyond a certain limit, depending on the chosen time step, to avoid certain instabilities.
It could be that lowering the inertia limit just triggers this mechanism.

Of course I don't know how the particular simulator you refer to works, but I'd try a confrontation with an analitycal solution for a simple rotation case : spin the car around its baricentre and check the time it takes to stop.
For easiness, block the suspension.
Do the analytical and numerical solution agree ?

[Greg Locock]

Good idea, I'll look into how the tyre model works - but it doesn't seem to spin, usually the sim just bails out.

Re your sig: move to Australia!

[Paolo]

Bailing out ? Well, the sim probably uses linearization of dynamic equations : that means it cannot cope with small radius turns. Even our University simulator, a very good one, has this problem.
My personal one is not based on the same concept , so it hopefully won't have this problem. But, boys, it's an hard task....

BTW : nothing against moving to Australia : in 2 years my current job expires and I'll have to look around... but I didn't know they had plenty of RWD there....

[Patrice L'Rodent]

Not just rear wheel drive cars Paulo, but very good rear wheel drive cars. Australias top 2 best sellers are Holden (GM) and Ford, and their top selling models are all RWD....and have BIG engines. The smallest engine in the Holden Commodore, Aus best selling car, is a 3.8 V6 but there is a 3.8 supercharged V6 and a 5.7 V8 if you want more.
In Australia the good lord dictated that them back there are for drivin' and them up front are for steerin', and the back wheels on FWD cars are there to stop the bumper draggin' on the ground.
Pat D'Rat

[Paolo]

Ehm, Patrick... 3.8 L ???
I said AFFORDABLE....

I can buy a BMW here, if I want to face economical catastrophe.

[Greg Locock]

AFFORDABLE? Well, $US 19000 buys you a 3.8L dunny door, or a little more will get you its incredibly sophisticated competitor with rather more power.

Not that I'm biased...

So how much is decently fast beemer in the states, say 9 sec 0-60?

[Patrice L'Rodent]

What Greg is alluding to is that the Ford Falcon has an OHC 4 litre inline six with variable valve timing as a starting point, and ends up with a 5.4 litre OHC V8 rocketship. I wasn't playing favourites, just started with the smallest engine.
They build em big and strong in Aus =]
PDR
PS Forget anything you remember about 60s era Falcons in the US, the Aussie bird is a very different creature!