(can a mod change the title in the forum itself? Longitudinal instead of Lateral)

Hi there.

My old thread about acceleration with spinning wheels and the examination of this problem: http://www.lfsforum.net/showthread.php?t=7840

This was positive longitudinal acceleration. Some pointed out that it could be something about the missing feeling for tyre spin when accelerating.

I now tested again longitudinal acceleration. This time negative longitudinal acceleration ... braking. I claimed often since years, that something is wrong about the braking in LFS.

Here is my test to prove what I want to say. Longitudinal acceleration with locked wheels is way to high.

A car with locked wheels should use a significantly longer way to stop, than a car which brakes with spinning wheels. (at least on tarmac. In snow and in sand its a different matter ...)

In LFS both are nearly the same. In most cases locked wheels result even in a better braking performance in terms of the way ...

I will add 3 graphics with F1perfview.

I did the test with the RB4. First test (blue lines) with default setting. Second test with way to strong braking pressure and locked wheels ... (red line)

1. Both times cars start to brake at 141 km/h. Both need about 73.6 meters to get to 0 km/h. If there is an slight advantage, then with locked wheels.

2. The longitudinal acceleration. In some parts its better with good braking, in some parts its better with locked wheels. In average it is slightly better with locked wheels ...

3. The average wheel speed. You'll notice, that the red line doesn't go to 0 in the beginning. That is because the rear wheels didn't lock up immediately but soon after start of braking. Front wheels did lock at once.

________________________

It's not about, that in some cases your way from 140 to 0 could be 1m better if you would brake with feeling.
There should be a significant difference between fully locked wheels and a good ABS like braking ... locked wheels have a longer way to brake. You may try go get much better results But 73m from 140 is quite ok. But to good for locked wheels ...

I think that it could be really the same problem in physics as with acceleration. Longitudinal acceleration simply is wrong in LFS in some way. I don't know how much that influences the cornering behaviour, but it does for sure at least when accelerating out of a corner or braking with shortly locked wheels ...

Except this obvious problem I really like the LFS physics as they are now. RWD street cars really behave good now if an open diff is chosen. Perhabs preload should be modelled ... but that's another story ...

At least I didn't find any more severe problems

So I hope that the next incompatible patch solves the tyre problems, as I think tyres should be done right (well at least 95%, because it is damn hard to get them perfect) before moving on to other things.

Greetings
RIP

EDIT:
Exact values of way:

73,57m with "careful" braking from 141 kph to 0 kph
72,03m with locked wheels from 141 kph to 0 kph

But this 1,50m may be the reaction time or some other small differences ... basicly its the same.

Lateral acceleration? I don't know what you mean. I would understand if you'd said longitudinal acceleration.

Oh sure. My mistake I meant longitudinal ... will fix it ...

Oh, do be quiet.

Without having ABS isn't it hard to see what the absolute best braking w/o locking the tires would be? How did you do the careful braking?

That's a good point. It is really hard. But try it yourself. Use the same setup one time with to strong brakes and one time careful.

If you are really good you get nearly exactly the same way. But you won't get a really visible shorter way ...

IRL race drivers get better ways. For sure some meters from 140 to 0. It isn't possible in LFS ... as far as I tried and some others too.

But feel free to prove me wrong. If noone gets a really better way on street tyres:
- Either there is no talented racing driver playing LFS
- or physics are still wrong in some way

Plz show data to see the reaction time and use the same! setup. No brake balance or weight change. And not only 1 or 2m from 140 to 0 plz

I think you'll get the best results if you tweak the brake power and bias exactly so that full braking power does not lock the wheels. Then just crank up the power and that will be your locked brakes test.

Quote from Forbin :

I think you'll get the best results if you tweak the brake power and bias exactly so that full braking power does not lock the wheels. Then just crank up the power and that will be your locked brakes test.

Agreed, test it on a flat surface like the skidpan also. Very interested to see what you come up with.

Well I did one test about that point.

I did a fully lock brake with RB4, full fuel, no passengers and default setup except full brake pressure (>1900 Nm)

The maximum of longitudinal acceleration was -1,2g at wheel speed 0. That's the peak and I didn't get anymore nowhere.

Then I did several brake tests, where I tried to be a good braker. I saw that data and there are very different g forces longitudinal at different wheel speeds. But I didn't get more than -1,2g.

That means :

Either I never hit the correct point in 10 braking tests. Nowhere in time just for a short period.
Or the physics aren't correct and you can't reach really more than -1,2g with that car, that tyres and exactly the same setup. Which is the peak value with locked wheels

Also note that I deactivated steering axis. This is why I took an AWD car to get a straight start with no steering. I also ran always into the limiter in 2 gear to have the same speed all the time when starting to brake.

A good driver should get about 10-20% better performance on dry. On wet or icy even a lot more. Which would be about 10m less in a 73m brake way.
Or at least -1,3g or more ...

Well, didn't you just define that the maximum 'g' the tire is capable of is around -1.2 (although there's a -1.29
peak there...) at which point the wheel starts locking up ? I don't see anything wrong with that. Obviously the
maximum a tire grips is also the point where it starts sliding. According to those graphs, locked wheels didn't
increase in acceleration and threshold braking seems to be able not only to reach the same grip level as locked
wheels, but also be able to peak slightly higher before locking up.

How much less grip do locked wheels have compared to threshold braking irl ? Try locking up all wheels on your
car and it stop rather violently. Anyone wanna sacrifice their tires and do some runs in their real car ?
The main problem with locked wheels is when trying to steer them. You can't. LFS does that. Also, i would think
that the contact patch of a locked tire would overheat earlier which is when it would start losing grip , although
seeing race sportbike throwing the rider in the air sometimes hints otherwise...

Fonnybone, rolling friction coefficient is much higher than sliding.

I have sacrificed my front tires (they blew up a week after the test) by locking them up going ~100kph. No significant braking. On good tarmac, setting aside the fact that the tires will not lock as easy, i can do threshold braking. Once it was so successful that it hurt my passenger's chest from the seatbelt. (on fresh laid asphalt and the new tyres i bought after the previous once blew up LOL)

Also, being a biker as well, i can tell you that the reason for doing 'endo' (AKA 'stoppie' that is, the rear wheel lifts off under heavy braking) is not the very powerful brakes, but the weight of the bike. Of course almost any bike has enough powerful brakes to flip it over if you are not careful, but you can do that in almost any bike, with almost any brakes. It's the weight distribution of a bike that is the determining factor in the 'endo'.

Quote from Fonnybone :

Well, didn't you just define that the maximum 'g' the tire is capable of is around -1.2 (although there's a -1.29
peak there...) at which point the wheel starts locking up ? I don't see anything wrong with that. Obviously the
maximum a tire grips is also the point where it starts sliding. According to those graphs, locked wheels didn't
increase in acceleration and threshold braking seems to be able not only to reach the same grip level as locked
wheels, but also be able to peak slightly higher before locking up.

How much less grip do locked wheels have compared to threshold braking irl ? Try locking up all wheels on your
car and it stop rather violently. Anyone wanna sacrifice their tires and do some runs in their real car ?
The main problem with locked wheels is when trying to steer them. You can't. LFS does that. Also, i would think
that the contact patch of a locked tire would overheat earlier which is when it would start losing grip , although
seeing race sportbike throwing the rider in the air sometimes hints otherwise...

Höh? You are right about that 1.29. Didn't see that ... have to examine that Don't know if there were different conditions involved, but I also tested it with slicks and I got -2.35 with locked wheels and -2.32 with careful braking.

Hmm, perhabs it is really really hard to get that degree. Have to continue testing

EDIT: Hmm really strange. I can reproduce better values now. It seems to be very very hard to hit the right setting. But then it is way way harder to hit the limit in LFS as in real life. Even one of the best german drivers (n1lyn) couldn't reproduce good values. At least in last version before last update. Perhabs this really was fixed this time, but I doubt it.
In real life nearly every racing driver should be able to produce better performance than locked wheels ...

Shouldn't this post have its own thread called "physics discussions" or something? Its a bit to heavy for "general" in my mind

Well I think it is solved.

I did again a few test runs and it seems I was wrong!

I really reproduced much better braking ways by releasing the brakes. I did a 70m now

It just seems to be harder to get the right braking in LFS ... so more challenging.

So this seems to be right now. But I still think something is wrong about the acceleration thing A difference there should be easier to achieve ...

So Thread solved somehow

Quote from george_tsiros :

Fonnybone, rolling friction coefficient is much higher
than sliding.

Yeah, i understand, i just meant that at the moment it starts locking up,
basically the tire is trying to generate a higher friction coefficient than it
actually can, that's where the LFS tires seem to stay. What comes after
that? What makes the tire lose it's grip ? Heat ? Something makes the
rubber change properties suddenly. I always thought it was heat.

Quote from george_tsiros :

Also, being a biker as well, i can tell you that the reason for doing 'endo' (AKA 'stoppie' that is, the rear wheel lifts off under heavy braking) is not the very powerful brakes, but the weight of the bike. Of course almost any bike has enough powerful brakes to flip it over if you are not careful, but you can do that in almost any bike, with almost any brakes. It's the weight distribution of a bike that is the determining factor in the 'endo'.

Actually, i walk talking about accelerating out of a turn.

@Fonnybone:
A tyre that has grip - so under static friction - can bend itself into the profile of the asphalt. The asphalt has many holes as everone knows and the tyre bends into these which makes it stick to the asphalt.
When the tyre is ripped out of it's grip, say due to a bump, it begins moving relative to the ground and can't take advantage of the holes in the asphalt anymore due to it's fast movement. That's why the coefficient of friction drops when a tyre skids.

I hope that was what you were asking for...

Vain

Quote from Fonnybone :

Actually, i walk talking about accelerating out of a turn.

Also known as high-siding, when the rear end steps out then suddenly grips.

http://www.youtube.com/watch?v=CCkhAq9rztQ

@Vain: So you're saying that with a smooth surface, both a rolling tire and a skidding tire should have the same grip?

Were you watching in force view mode during the careful braking to get as close to locking as possible w/o locking?

Also good job setting up and performing the analysis,

Well I am actually now as far to say the opposite of my origin post

I just tested the "braking help" which is a very very slow and bad ABS system in some way.
And if you put the brakes on full power (19xx Nm @ RB4) you will achieve a much shorter brakeway with "braking help" than just flooring the brake without it.

I achieved -1.31g with careful braking (without braking help of course) on street tires. With locked wheels only -1.20.

It remains, that it is very difficult, but it seems to be correct.

Still strange is the acceleration topic. It would be fun to be able to start slightly better than someone else going full throttle at start. But most of it will come with a "free" start, where you can decide when to accelerate

Free starts will be the time to test it. The last time you brought this up (or was it the time before ) I did some runs in the FZ50 at the dragstrip. I found the same results as you did here, it's possible but quite difficult to launch really really well. Plus it's not really fair since a full throttle start is simple and instantaneous but a quick launch using proper clutching is hard AND has to be timed right. I could get a little 10th quicker over the quarter, but that would increase a lot with proper false starting coded. Having the analogue speedos running of the drivertrain properly now really makes launching well much easier.

If it works on braking, the code will work on acceleration to I would bet

I think I am gonna go for the peak g values of acceleration today in the other Thread. We'll see what happens.
But if it is the same, the difference is to small.
IRL proper clutch/throttle work is visible on a 1/4 mile race ... often reaction time is measured too so you can see the difference. And everyone tries to avoid wild spinning wheels ...

but -> other thread later ...

Ive read somewhere that on dry tarmac braking distance of locked wheels is shorter then with abs.

Im not sure how true is it

Um ..... With ABS your wheels dont lockn (duh!) .... they are on the limit of a pre-defined settings by the manufacturer... howverer, I do know on Bikes, that a ex racer can brake faster on a FJ1300 quicker without ABS than with, simular to cars maybe, I've never seen the exact test.

Sort of off topic, but in the chassis flex thread, there's mention of the overall flex affecting handling. Wouldn't LFS cars' flexlessness as it is now affect "realistic" handling?
It could account for LFS cars not matching real cars' feeling of grip, or is it on a smaller scale?

Exactly that was mentioned before in the acceleration topic. I did some tests with acceleration and there is a higher g force again. But as with the braking it is very very hard to get there and impossible to hold it.

See http://www.lfsforum.net/showthread.php?t=7840