Cars (DATA REPLACEMENT) BMW M3 E30 Improved Physics by Arch

[Kyuubeey]

Kyuubeey updated (DATA REPLACEMENT) BMW M3 E30 Improved Physics by Arch with a new update entry:

DTM ARBs, tires, throttle

27.09.2019 Version 1.4

- Reverted DTM rollbar links to "rubber" from "solid" due to new evidence
- Tire changes
- Changed throttle graph

Read the rest of this update entry...

 

[Kyuubeey]

Kyuubeey updated (DATA REPLACEMENT) BMW M3 E30 Improved Physics by Arch with a new update entry:

Dampers, rollbars

13.10.2019 Version 1.5

- Adjustment for damper oversight
- Calculated rollbars
- Changed rollbar setup design for tuned cars

Read the rest of this update entry...

 

[schrep1964]

Hi, like what your doing with this mod. Could you maybe do a DTM Nordschleife suspension version since you did a representation of the 1992 car ? DTM did race overthere in 1992, Klaus Ludwig won in a Mercedes Evo. In a racereport about that race many drivers spoke of big suspension changes for that race alone. I find the stiff suspension as it is in your mod not suitable to race there, even in the most soft settings, the car is struggling to keep on track, eg skitting off in high speed corners under full load.

 

[Kyuubeey]

Hi, like what your doing with this mod. Could you maybe do a DTM Nordschleife suspension version since you did a representation of the 1992 car ? DTM did race overthere in 1992, Klaus Ludwig won in a Mercedes Evo. In a racereport about that race many drivers spoke of big suspension changes for that race alone. I find the stiff suspension as it is in your mod not suitable to race there, even in the most soft settings, the car is struggling to keep on track, eg skitting off in high speed corners under full load.

Hey,

If you could provide me the homologation specs for the springs, dampers, ARBs etc. used in the Nordschleife version, perhaps I could. Otherwise there's not much I can do; the current DTM car is based on official specs and some anecdote.

 

[schrep1964]

Hey,

If you could provide me the homologation specs for the springs, dampers, ARBs etc. used in the Nordschleife version, perhaps I could. Otherwise there's not much I can do; the current DTM car is based on official specs and some anecdote.

I understand that such version can only work with real data and appreciate your efforts as such with the several car physics you have done. Just a pity that the data is so hard to find. I will do some research.

 

[Akinastyle]

With your data I no longer have any feeling in my steering wheel, I no longer feel the oversteer or understeer and the car really stalls for nothing. It's confusing because I feel like I have a Logitech, it's really ridiculous.

Kyuubeey
You can press "+" on the numpad if you want higher FFB.


You didn't understand correctly or I misspoke, it's not a FFB power problem.
By trying to do the right thing (which is quite legitimate, by the way) I think you have completely distorted it, which no longer allows it to correctly transmit its behaviour in the wheel.

The problem I have is that the steering wheel loses all information beyond 90 degrees, or during an oversteer it does not follow the road and remains motionless.
It doesn't make sense.

I think the problem comes from the WBTYRE_BOTTOM on the Y axis, I haven't opened the data yet, but I'm willing to bet yours is closer to point 0.

I myself wanted to reproduce a friend's E30 drift, I went to his house to measure the coordinates.
It turns out that once reproduced in play the car does not have at all the same behaviour as in reality especially in terms of changes of direction the steering wheel does not follow the road correctly.
However, I have the real measurements, so I slightly shifted the whole geometry to the front of the vehicle and it corrected this problem.

This is the difference between reality and simulation.
I repeat, sometimes it is necessary to "cheat" to transmit the right feeling at the wheel.
Isn't that the most important thing?

 

[Kyuubeey]

Well, if your friend has extended LCAs and wheels with less offset, and/or wheel spacers, of course the scrub radius will increase, and steering feel and response will change dramatically. The forces IRL become very violent if you go even 100mm into positive scrub. Although I doubt it, let's assume you put in the correct track and geometry IRL, and it still didn't work right.

Then, by how much did you shift the geometry, and in what axis?

 

[Leonardo Ratafiá]

what similarities could have an e30 drift to an e30 m3 stock? the car feels great on ffb iirc and it's one of the best ffb (other than the S2 which is weird imo)
you can also try the other e30 325 available here.
I'd like to know what other car mods you gave 1 star

I have a Fanatec CSL Elite wheel btw, haven't tried on a DD wheel but it should be just quicker from lock to lock and a little stronger

 

[Kyuubeey]

I also have a suspicion, if the car is serious, that there is some stuff going on with the steering axis maybe being uncoupled from the LBJT by modifying the strut, but I don't know so much about those setups myself.

Akinastyle, if you've driven a bone-stock E30 M3, it'd be more helpful as a base to judge from. The steering forces are just forces @ rack ends and I have a serious doubt they're incorrect to a noticeable degree, if the geo is right.

 

[Akinastyle]

I'm not talking about the steering angle but how the steering wheel follows the road on a production car, a kart or a truck the steering wheel follows the road very well.
I have moved the whole geometry on the longitudinal axis (Y on 3dsmax) I do not know the placement of the axes on AC
example: WBTYRE_BOTTOM=X.XXX, X.XXX, X.XXX.
I have shifted the red value of my example each geometry point STRUT_X, WBCAR_X etc.

 

[Kyuubeey]

So Z. You've just increased mechanical trail. I'm not talking about steering angle, I'm talking about steering geometry.


Okay, well, for one, there isn't a "road" in sims to follow. There's no real groove or anything in the physical mesh, just bumps. If you're talking about tramlining and following the groove and minute undulations in the road, I think sims' physical meshes don't have enough precision for that to be represented like IRL.

The laserscan doesn't give you mm accuracy, I don't ever recall seeing even cm accuracy. Think of it like several cm wide triangles, when exported into a poly engine. It'll just be undulations. Someone who really knows can correct me if I'm full of ****.

Also, roadcars like the E30 will have a shitton of slop anyway before anything gets translated into steering and vice versa, if anything the sim should follow the road better than the real car.

You said something about understeer, oversteer feeling. Increasing mechanical trail "fixed" it. Okay, maybe, but I don't have enough info to know what you're actually comparing it to. The main thing is the tire anyway, if the tire doesn't produce the same forces as IRL, they won't be the same @ rack ends.

 

[Akinastyle]

@Leonardo Ratafiá
I'll try the 325 and give you my feeling.
everything has done the wheel is more lively the strength is another debate and it remains personal, in any case I said goodbye to clipping.
Kyuubeey
Unfortunately I didn't drive an e30 but I have an e34 and an e36.

 

[Kyuubeey]

Of course the wheel will have higher forces subjected to it with more scrub and trail. But like I said, there's no small grooves and stuff for it to move from in simulators, so the comparison is a dead-end. If you're talking about the rack end forces increasing or decreasing with slip, or the lack of, well, that's not what other people have reported nor what I feel.

Specifically, which version of the car you have trouble with? You said something about DTM earlier. You've driven a 90's DTM style car on slicks to compare to? Or do you mean the standard car?

 

[Akinastyle]

yes I have moved the pivot point of the link arm with respect to the wheel hub.
The only thing to know is whether the measurements are taken by car on a lift or by car on the ground.
I'm not talking about the way the car oversteals or understeers, but the response in the steering wheel.
To compare this I simply base myself on a KUNOS car.
moving on an axis will not change the basic geometry if you do it on all the elements.
I'll show you a picture so it'll be more explicit.

 

[Akinastyle]

I'm talking about the dtm version

 

[Akinastyle]

on my example of photo you can see that I moved the whole geometry without touching the kpi or the caster angle for as much as I moved the whole geometry.

 

[Kyuubeey]

But that geometry's not even remotely accurate at all. It's completely wrong.

I wouldn't compare any roadcars to the DTM version. Kind of pointless, especially because the tires probably have 30%+ error due to lack of data for them.

EDIT:

Here's the DTM car geom I made, for example.

I don't know if your geometry is measured, or KS or what that is, but your caster is like 1/2 of what it should be, so there is no wonder the steering feel is lacking in yours. Don't compare to KS cars, just forget it. Almost none of those are accurate.

 

[Akinastyle]

it's an example to compare the two photos of the same geometry with the application of the offset I don't care if this geometry is true or false I just use it for my example.

 

[Kyuubeey]

Well, I still remain unconvinced that I should put in several cm of neg. mechanical trail because some guy who drove a drift E30 thinks my DTM car feels off. Need some more basis or examples.

 

[Akinastyle]

I don't drive an e30 drift it's a friend's, don't mix everything I have an e34 stock and an e36 stock.
I just reproduced a friend's e30 geometry like you do for that one but it didn't give the desired result.
that's why I think that putting real data doesn't make it any better.
After all, you didn't ask me for anything, so I'll stop bothering you, but I think you completely distorted the final feeling in the wheel.