I think what's basically killing the fun of driving the FOX is the body downforce.
At slow speeds the car is nice and fun. But even around Fe Gold the FOX gets so much downforce from the body alone that the engine can barely get the rear tyres loose.

Does anyone have numerical data of how much downforce such a chassis would cause?

Vain

Thanks for that thought. This makes a lot sense when you look the pictures undoz posted (thanks a lot for those, undoz! The 'Operational Theory' section of that linked page is great!).

Vain

Last edited by Vain, Fri 20 Apr 2007, 19:37.

Quote from jtw62074 :

What's important here is that preload is not effecting the differential operation at all except in these situations where there isn't very much throttle or braking being used. Anything beyond that and the locking torque from torque bias ratio will win. The triangular shapes devour the preload area underneath it. They are not cumulative.

Hi.
Where does this come from mechanically?
I can believe what you say and try to memorize it, but I'd like to understand how it works mechanically, but technical drawings of LSDiffs are seldom.
If I press two (or however many) clutchplates together to create preload they should always resist torque between them by exerting an opposing torque. Regardless of the torque difference. So if I have set up my preload-less differential to stay locked in situation B in your diff2-picture, and then put additional load on the clutches I can't help but believe that now the differential should be able to resist a higher torque-difference between the driven wheels than before.

Vain

Hidden mode:
The menu items hang on a wooden wall with many stilts. Shoot a cube between the two rightmost stilts by clicking there. If you hit the correct one of those concrete plates that you can hit by shooting between those two stilts you'll enter the hidden mode.

If that's too complicated, attempt to hit all concrete plates (the breakable ones) on the right hand side of the menu screen.

Vain

Quote from J.B. :

For example if the preload is the amount of tourque the clutch plates can take before slipping then once they start slipping the torque transfered should be a bit smaller than the preload value as dynamic friction coefficient is < static friction coefficient.

I can only judge by the technical drawings and texts linked here. Those say that the torque exerted by the differential due to preload is constant over the whole range of slipping.
From looking at the drawing I think that preload is also active while power is transferred through the diff. After all, all you're doing is pressing clutch plates together, they don't gt loose because power is transferred through the diff.
So when the diff is slipping the exerted torque should be 'preload + torque at input shaft * locking_factor', wether the torque at the input shaft is big or small, the clutch plates we press together with preload shouldn't mind. Or should they?

In that case the diagram above would be wrong.
Now you made me draw diagrams with paint with a touchpad.

Vain

Quote from Forbin :

Honest teams need not apply?

Just like in real racing.

Vain

Nice game. The after-match struggling is silly. Look what my highly skilled sumo-machine managed to do while trying to stand up.

Vain

Let's put it straight:
The car with LSDifferential with preload drives down a straight. The driver leaves throttle and brakes alone and starts turning in left.
The front wheels start turning the car, which causes forces on the rear tyres. The front wheels try to make the left rear wheel slower than the right rear wheel.
These forces mean that there is a torque at the differential.
If this torque is smaller than what you set preload to the differential resists the torque and acts like a locked diff.
If this torque is bigger than what you set preload to the differential starts slipping. The torque it exerts is exactly what you set preload to. f.e. 100 Nm if you set preload to 100 Nm. Regardless of how much the driver turns the steering wheel, the diff won't exert more torque than that, 100Nm and no more.
If the driver starts accelerating now the differential is loaded and exerts additional torque from the power-lock setting, and pretty soon preload is neglectible.

That is my current position. No warranty though.

Vain

The first video was definitely not staged. But indeed, during the first three laps both Hyperactive and me tried to be careful, we were driving a 15 minutes race after all and you don't want to spoil that with an overambitious move in lap 1 .
The second video's quality is, as said, caused by google. Does anyone have some good pointers for the input-quality for upload to video.google.com? The recompression at google is a real problem.
And about the driving in the second video: Those three drivers were fighting for entry into heat 3A. The last of the three would go to heat 3B. From heat 3B the best achievable grid position for the feature race is 13th (the first 12 are finishing positions of heat 3A). And there were merely 10 minutes to juggle it out before the heat ends, so there was quite some pressure there.
Glad you liked it. I'll keep uploading such videos from the heats, depending on wether I see good action somewhere.

Vain

W9 and W10: Is the behaviour that occurs when you exit LFS via the windows exit button while refreshing the server list desired? It takes over 15 seconds to actually exit and shows a blank backgroundscreen in the mean time.
It might be that way by design, but it isn't optimal so I mention it.

Vain

Here's another onboard video.
That's Heat 2A of STCC#7.
Drivers are:
Silva (white helmet, grey/white car)
Horvath (yellow/grey helmet, yellow/black car, appears a bit later)
Me (Black/red/white helmet, yellow/black car)

Google pretty much destroyed the quality. I'll have to try around with the next videos to find something google accepts without completely shredding it.

Vain

I fraps'ed a nice fight in Heat 2A of round 7. Click here. First appearing XRT is Silva (from the start, white helmet), who during the first lap starts an argument for positions with me (red/white/black helmet, yellow/black car), and later we meet up with Horvath (yellow/grey helmet, also yellow/black car, my teammate). Pickard also makes an appearance as he blasts past us in his FXO on the first lap.

Google-Video pretty much ate the quality (the video was already at the target bitrate and resolution, but was transcoded to flv *argh*) but we stayed so close to one another it doesn't matter.

Vain

I read about that in those linked articles.
So the preloaded LS diff behaves like a locked diff until the torque difference reaches preload?

Vain

@evilgeek:
That explains well how a LSD diff works under load, but I'm aware of that. I'm talking about: What kind of torque does the LSD diff exert under a given difference in wheel speed (cornering), with no input torque and a given preload setting.
Either the unit of the setting is wrong or I lack a fundamental understanding of what happens in that situation.

Vain

@J.B.
Let's imagin input torque at the diff is 0 Nm. Preload is 100 Nm. The car is going in a straight line.
Using your calculation for the case preload > input torque I'd get to a torque at the left wheel of 50 Nm and torque at the right wheel of -50 Nm.
This is obviously not true. What didn't I understand? This is the point why I assumed a clutch to operate, which requires a constant in the unit of Nms/rad.

Vain

I also have a question.
Why is the unit of preload Nm?
If I assume a linear clutch the torque exerted is some constant by the difference in angular velocity between the two wheels.
M = c * (omega1 - omega2) or in units:
Nm = ? * 1/s
So obviously the constant's unit must be Nms, or Nm*s/rad, if you measure angular velocity in radians/second instead of 1/second.
Or is the c in the formula calculated from normalization as in
c = M0/deltaOmega0
with M0 being the setting in the setup and deltaOmega0 an example difference in angular velocity (usually 1 1/s)?

Vain

The rule is "If you cross the line and you're involved in an incident it was your fault."
So if you cross the line and there is no incident it's all fine. But if there are other cars around we stay clear of the line.

* I'm uploading a few onboard scenes from Heat 2A, one of the qualifying races.

Vain

Quote from Jakg :

fair enough - care to show some of the "n00by moves" your talking about?

Didn't you see the broadcast?
Self-criticsm isn't your strong point, is it?

None of the drivers make mistakes when they drive on their own. Half the field does laps within .5% of the WR. Sub-WR times aren't seldom at all. There is a stiff competition, the safety margin is small and pressure is high. On a track like So1 that's bound to cause some bent panels.
I was a victim of my own pressure in lap 2 and I admit it. But that's the point of the STCC race format. All in all I had a lot fun driving that race even though p5 is a bit underwhelming, but my battles during the race were a lot fun (unfortunatly you didn't get to see much of it).

Vain

Quote from hackerx :

Heh OK, it just seemed to me bit like random pickup race on public server, given the number of unsuccessful chicane exits.

That's true. That's mostly down to the pace and tension in the race.
My 'unsuccessful' exit at lap two was caused by never successfully settling into a rythm, other drivers were driving just inches behind the tail of another car, and more drivers were just driving on the absolute edge.
If LFS simulated damage to the side mirrors I guess almost every car would have lost it's left side mirror at that chicane. If you leave more space you're off pace, at least if you want a top 15 finish.

[Edit]
And Jack, stop claiming it's because driving is too difficult. You drive a FWD!

Vain

Can you perhaps use an LFSW script that displays the best current online-pbs for a certain selection of benchmark-tracks to compare the different cars?
Say you take the best 50 XRT pbs from Bl1, Bl1R, So1, So4R, etc. etc. and compare them to the best 50 FXO pbs. that should be a proper measure of the relative pace of the cars.
Then take into consideration that the FXO pays for it's improved handling by a horrible start off the line and you should be able to track down a very good balancing within a few weeks.

Of course there are a couple of tracks where people cut a lot. e.g. KY3, KY3R and BL1R can't be chosen for such a comparison.

Vain

Watch STCC.
(Most names are easily remembered because they are derived from the track. The SO-corner names e.g. are derived from the street names that are on the signs on the So tracks.)

Vain

The LFSW-messages lagged behind a lot this evening, but shortly after 21:30 GMT+2 it worked well again.
I don't think that the number of users dropped significantly around that time, so I wanted to report that.

Vain

I've had it yesterday evening for a while on several servers. The LFSW-messages lagged behind for minutes.
Today around 14:00 GMT+1 the problem was gone.

Vain

Finally I came up with something worthy.

Vain

Quote from SlamDunk :

Yeah, the kind of "pure", nothing but in-game stuff ones are the best, in my opinion and I'm very interested to see how other people would do a cockpit gameplay video.

Just look here.

Vain