Cars Skoda 130 RS 1.9

[Jugulador]

If there don't exist curves and good pictures for the suspension, it can be a little difficult to make it very good, but luckily most STA (That I have seen) are very similar in design.

Can give all specs:

http://tech-racingcars.wikidot.com/skoda-130-rs

Blueprint:

Spoiler: blueprint

https://drawingdatabase.com/wp-content/uploads/2014/04/skoda-130rs.gif

And image:

Spoiler

It's from a replica, but the first link I shared have the precise information and it's not far from this picture.

you don't need super accurate measures to have fun with a car.

No, you don't. To tell the truth, the mod author do what he want with his car... but this, still, is a simulation forum and it's traditional to discuss about modeled cars. I you are offended by the simple act of discuss something, than I think that you are posting in the wrong place.

But, by experience: When the modded car authors put a work on fidelity of their models, those became the best experiences on sim. I don't know the author of this 130, but if I spent my time on something so demanding as a mod, will want to go full-precise on that, specially that is a mod for a simulator, not for a simcade or arcade videogame. But that is me. For the quality of the 3D modeling of this 130, I really believe that the author already put a lot of time on this.

 

[iwn]

Thanks for the update. CTD also upon exit with CM.

 

[Leonardo Ratafiá]

I know but there's some different approach to physics from different guys. The author seems to be fine with his cars. Then you have Arch who measures per pixel and he can get great results on some cars but sometimes you get really bad ffb feeling (specially on latest m3 S1 and S2 update)
I'd rather have some decent and coherent ffb. I find every kunos car giving good ffb effect and information even though if you ask some physics modders they will say most of the cars are completely wrong. (go figure)
I agree on getting the most out of the simulation available from the engine.

here you can also see some homologation docs (don't know if the same car though but from the gearbox is this one I guess)

https://historicdb.fia.com/sites/default/files/car_attachment/1486656367/homologation_form_number_1676_group_2.pdf

 

[Fuzo]

I see that the debate is still going on. So firstly, our sources of information are not only from google, these cars are quite well documented in former Czechoslovakia and I even own some books and publications about this car, so posting here some info, which in fact is from rally version, which is obviously slightly different to track version has no real value to us. Secondly, we always try to build the physics with real data, but to get the desired result we have to make some compromises. If you fill the data with strictly acurate numbers you are not guatanteed you will get the result you want. If someone thinks different, it is his personal choice and he can do his own mods and see how his audience will like it. We dont tell people how to do their mods or that their way is wrong. And since we have good contacts to people who own the car we always consult the handling characteristics with them and we modify the physics to make it as close as possible to real version. The person which is doing physics puts the same effort to every project as do I with the model and I fully trust him. We do our mods in our free time for free but we try to do it as good as possible.Who isn't? There are people which claim this car is one of the best historic mods ever, then someone thinks it' off, we can't satisfy everybody...

 

[Kyuubeey]

Alright, look. Just to be honest with people.

Skoda 130 Camber and toe 5deg? Low angle

132, 583 = 0.05, 0.33
82, 533 = -0.01, 0.00 REF
32, 483 = -1.13, 1.83

E30

Rear RC = M3 111.58mm, STD 116.88mm, DTM roughly 122mm

E30 Camber 15deg

229, 495 = -0.275, -0.60
204, 470 = -1.15, -1.25
179, 445 = -2.0, -1.99 REF
154, 420 = -2.875, -2.87
129, 395 = -3.725, -3.83

E30 Toe 15deg

229, 495 = 0.31, 0.27
204, 470 = 0.24, 0.24
179, 445 = 0.23, 0.23
154, 420 = 0.26, 0.26
129, 395 = 0.36, 0.36

DTM camber 15deg -> 12deg (Not taking offset into account)


339, 495 = -0.22, -0.64
314, 470 = -0.92, -1.07
289, 445 = -1.6, -1.61 REF
264, 420 = -2.3, -2.24
239, 395 = -2.98, -2.98

DTM toe 15deg -> 12deg (Not taking offset into account)

339, 495 = 0.248, 0.18
314, 470 = 0.192, 0.18
289, 445 = 0.184, 0.18
264, 420 = 0.208, 0.21
239, 395 = 0.288, 0.29

Left side is ride height, then rod_lenght in setup just for myself, and the E30's have real value on the left vs simulator value on the right. Skoda is just camber on left, toe on right.

Skoda has insane toe change, which it should not have, if the sweep angle is a few degrees like it seems to be, then it should have very little toe change and very little camber change.

Yet the Skoda gains camber closer to what a 10+ deg assembly would gain, but the toe is massive on droop and rebound. I would imagine on small accelerations and small braking, the toe-in gain is enough to really stabilize the rear compared to IRL. Something like probably 3 - 4x too much on rebound and probably 10 - 20x too much on compression. Also there is basically no positive camber gain on rebound, which might be closer to reality if the angle really is as low as it looks, but I would still expect at least something like +0.50 or so, which would decrease grip on braking and lift off.

Then here are the roll centers from inside KS' suspension viewer.

It's like I said; Skoda is super low. My E30 is a little closer to where it should be IRL, only a few mm off at static.
I can't find good pics of the Skoda rear, but it sure doesn't look like it's close to the ground; the angles are all mild, and trailing arms have very little RC migration, so I can't imagine it moving much more from the 100 - 150mm region they usually are. This means the rear rolls too much and has a better load split, so it will have too much grip, like I said.

Read it and think about it a little. And please don't lecture me about real numbers not working when your car probably doesn't have any real numbers for the thing we're talking about. It's the 2nd time this month.

 

[iwn]

This discussion reminds me of flight simulation circles where often a simulated aircraft flies like magpie and gets much praise as "realistic". As a real world pilot I often wonder that as absolute majority of aircraft fly better without a pilot. A gusty weather requires a bit pilot input but otherwise they are quite on the rails. Now if I imagine to drive a race car in my AC I expect it to have grip and to behave predictably in order to take it to the limit. This Skoda gives me that sensation.

 

[Kyuubeey]

Well, we need to admit that it's very very hard to determine anything from feel in sim. So talking about feel is really unproductive unless we have a good understanding of the mechanics of what is going on exactly.

I'm talking about the pure numbers, and unless I am being mislead seriously by the pictures I could find, the numbers are wrong. There is no discussion about it; it's clear. A trailing arm will not change much at all if it's on a racecar, or made with IBM computers or what. At most, stiffer links will mean less lateral toe deflection compared to a roadcar one, and less camber and toe variance from flex.

 

[Fuzo]

@Kyuubeey why do you still continue to do this? I told you couple of times that we don't want your help, why aren't you focusing on your own work but you have the urge to dissect the work of someone else and try to prove that our numbers are wrong? I will tell you again, our modding ideas are very different, if you don't like it be at least so kind and don't try to push your agenda to us. We are respecting all other modders and their ways, please do the same with us. HOWGH!

 

[Kyuubeey]

I posted some actual real numbers because you claim to have a "different idea" and at the same time claim you're trying to make it as realistic as possible. Which is what I try to do.

Instead of being upset that I take a little of my time to solve problems for you and start talking about philosophies and agendas, it'd be more honest to just say "Thank you, but we are happy with it how it is right now".

I just posted a more honest representation of the kinematics of the suspension, so that users like @Jugulador who have a problem with it can understand why exactly they think there is one.

A criticism like "It's too good for an old car" is a little annoying, I understand, and often not true. Some cars are very good. However the trick here is to try to understand why it is being said. In this case, it appears to be incorrect RC and curves for a semi-trailing arm rear suspension. It's your choice if you want to do anything further.

Reminds me, I think you are the same modding team to whom Leonardo and I posted a Skoda Octavia manual with damper curves in them, and I explained that the reason your real drivers say that your 100% incorrect dampers feel good is because the car should probably ride on the packers and it doesn't.

Anyway, like I said I understand if your team doesn't understand how to fix it/you don't want to due to time or whatever. Both are okay. But if your team wants to study a little and become better, I posted some stuff above which is good, and I can PM more help if they want. Or maybe they are talking behind my back about how stupid I am, who knows.

 

[Fuzo]

OK I have to specify, we try to make our mods as realistic as possible handling-wise not number-wise, mostly from the feedback of real drivers. And we know that this approach has it's pros and cons but we can live with it and people driving our mods are OK with it as well. At least majority of them. We decided to go this way mostly because in our Czechoslovak community, everybody knows those cars, they know how they drive, what to expect from them, so we didn't want to argue with them that the mod is accurate in numbers but does something else what would they expect. And we are still learning of course. But debates like this are quite discouraging to be honest.

 

[Kyuubeey]

Okay, sure. That kind of view comes from not understanding the feedback / output relations. Real drivers' thoughts, video footage and even your own feeling can trick you sometimes into thinking that the car is right when it's not at all. It just has to be right enough in the right areas.

My view is that if the numbers are right, it's right. Not the input numbers necessarily, but the output.

The challenge here is that you can achieve "almost" the real feeling behavior in probably 10 or more ways. There is enough unknowns that you can tweak this and that in response to a feeling you would want the car to have.

So the goal is to find as much of the real input data as possible, put it in correctly, then check to see that the output is making sense. From there, it goes into things like tires where there is basically no data; but we have some rules, or some knowledge about how things usually work.

Suspension is one of the reliable data inputs, where real numbers often give very good results as long as you understand what you want to achieve.

There are things like subframe flex, bushing flex and so on which change the output result IRL, and while the kinematics of the suspension might be right; the wheel is actually losing some camber and gaining some toe at 1G lateral because of flex, for example. Better understanding of the mechanics allows you to work with those facts to get it closer.

I'm 100% sure that closer suspension kinematics = more realistic car. It's never, ever been wrong up to now, and I'm nearing 3 digits in the cars I've made. That's why I think that basic suspension kinematics are a good idea to try to get closer, then you can work from there.

I hate this misinformation that "you can't use real numbers in sims". Literally no one who is any good at making physics says that like how people mean it. I literally only hear it from people who don't know what they're doing. But it still has *some* truth to it. For a car like this which has I presume solid metal links and nothing too crazy in terms of suspension, I don't think it applies very much.

 

[iwn]

Was it so that Skoda used backbone chassis vs. more common Mcpherson type? What might be this game's default?

 

[Kyuubeey]

Was it so that Skoda used backbone chassis vs. more common Mcpherson type? What might be this game's default?

The pics I've found all have semi trailing arm rear, or something that sure looks like it. Resources also list it as "semi trailing arm".

Seems to be low angle, just a few degrees maybe, but this angle isn't good either.

AC doesn't have a default, but you can use MacPherson (Which can also make Chapmann struts and other stuff well), Double Wishbone which lets you make all kinds of suspensions from almost any SLA setup to modern multilink, trailing links with traction rod and upper control arms, Honda style trailing arms with camber control links etc. if you set it up right.

Then there's Axle which has some issues and generally doesn't work very well for various reasons, and Multilink which has bugs, so I avoid it.

Trailing arms should be made with DWB, it's overall the best IMO.

 

[Jugulador]

Fuzo, my mate, don't be upset because discussion. I'm not trying to dictate how your work should be (that is free of charges and I really appreciate it). If your mod archived your goals just leave it be as it is and ignore everything I wrote below. My intention is to give you feedback in the most constructive way and help you if you believe that your mod can be improved (aways can be... a lot of very good mods here in RD are still being updated, even if released years ago), just open your head (and heart) to what community have to say about it.

I see that the debate is still going on. So firstly, our sources of information are not only from google, these cars are quite well documented in former Czechoslovakia and I even own some books and publications about this car, so posting here some info, which in fact is from rally version, which is obviously slightly different to track version has no real value to us. Secondly, we always try to build the physics with real data, but to get the desired result we have to make some compromises. If you fill the data with strictly acurate numbers you are not guatanteed you will get the result you want. If someone thinks different, it is his personal choice and he can do his own mods and see how his audience will like it. We dont tell people how to do their mods or that their way is wrong. And since we have good contacts to people who own the car we always consult the handling characteristics with them and we modify the physics to make it as close as possible to real version. The person which is doing physics puts the same effort to every project as do I with the model and I fully trust him. We do our mods in our free time for free but we try to do it as good as possible.Who isn't? There are people which claim this car is one of the best historic mods ever, then someone thinks it' off, we can't satisfy everybody...

You said that your mod is based on the original version, that were around 60 cars build to rally (the RS on it)... I didn't find anything about these cars racing on tracks until some years later or the replicas and the modified original cars (included the not RS versions turned into it). You said that you have books about the car and know the racing scene... can you explain exact what are you simulating? Because it can change my opinion, if you say that it uses modern components.

And we can't want a result if you never drove the real car (as you don't say that has drive it)... I think that you can put some trust on the game engine or you may not be simulating the real thing, just reproducing an idealization of it (specially that fells like you are a little emotional about this car... what I REALLY understand and don't condemn).

@Kyuubeey why do you still continue to do this? I told you couple of times that we don't want your help, why aren't you focusing on your own work but you have the urge to dissect the work of someone else and try to prove that our numbers are wrong? I will tell you again, our modding ideas are very different, if you don't like it be at least so kind and don't try to push your agenda to us. We are respecting all other modders and their ways, please do the same with us. HOWGH!

Sorry, mate, but there is no "different idea" about simulation than reproduce the reality. If we were talking about a primitive simulator, them ok, but AC have an engine that work exactly with real numbers.

Feeling is a crap idea for reference, including real drivers (remembering that folks used to racing are flawed to explain and judge sims...they tend to be to lenient. Senna, Mansell, Hamilton, Vettel... all have an historic of very crap feedback), specially if they are emotionally involved with the vehicle. I admit that even my opinion isn't something to settle in stone, but I just using it to question you mod to my use, not to judge it as "the real score". So, calm down and don't be so defensive about your work.

I just posted a more honest representation of the kinematics of the suspension, so that users like @Jugulador who have a problem with it can understand why exactly they think there is one.

+

A criticism like "It's too good for an old car" is a little annoying, I understand, and often not true

Speaking about "honesty", you seems to be ignoring my posts to wright something like this. Because I didn't just said "herp derp, too old to be good"... I posted data from reliable sources... including the picture that you post (hours before you)... that I posted and said that it is from a replica model (you posted as if it is from original model).

I'm not an expert on AC engine (because never modded on it, not because I don't understand it), but I spent some time under real car bonnets to at least have some idea from how a car may behave (or may not, in this case). This semi trailing arm, as used on this car (the backward mounted angle), deliver a very robust package, but not a very responsive one... even if you tight the springs and dampers, the arm will have a natural inertia that slows the lateral response... it's a trade off they made because the car was made for bump roads and short (tight) corners, not to race tracks with fast corners and splits (that can be raced and probably was, because in 1975 people just used to be more freely on their minds... you can use the GTAm as the very best example of that lol). There is where my "old car" opinion came from. That is why I'm questioning this mod... because the lateral weight shifting don't fell right (anywhere close to it, to be honest). If this kind of suspension was set in a so tight way the car should had a more "block" response, not a precise (that is consiliate tight and responsive)... that is why this geometry of suspension was abandoned as soon as more modern materials were available and smaller parts could sustain more weight.

 

[Kyuubeey]

The inertia you're talking about comes from the unsprung mass, which seems completely ok in this car. 70kg per wheel, it's around where it should be. It's true because of the construction of semi trailing arms, the mass ends up relatively high, but it's also durable and reliable. Something like a Chapmann strut will have generally lower unsprung mass, but it also might not have the characteristics you want.

Nowadays we consider the multilink or DWB to be better than the trailing arm, kinematics wise, because you can achieve a higher grip in the scenarios you want and generally more predictable response due to less camber and toe variation, and you can engineer the curves to be how you want. Trailing arms are simple, and the math for it is simple, but it also means they're limited.

The trailing arms have behaviors like losing a lot of anti-squat when lowered, and the roll center migration amount being lower than in other independent suspensions, and large toe deflection on G forces due to how the bushings are setup. It's also easier to set up the IC and RC in ML/DWB than it is with STA.

On the other hand, you can easily get wheel hop on acceleration or braking in DWB/ML suspensions which aren't setup well, and ML suspensions especially can become simply unusable if not setup well because the tire will try to go through the wheel well! That's why you basically have to use a computer program to optimize the travel of the wheel with ML to make sure it doesn't hit the inside of the wheel well; but a trailing arm design is easier to design and test, it seems.

Then of course, there's the manufacturing and design cost thing as well. I would believe that just copying a proven trailing arm design, like for a cheap FWD economy car, is cheaper than engineering a high performance multilink for it. Packaging exists too; trailing arms are very flat, so you can have a lot of luggage space above them, while the subframe and camber arms for a DWB/ML maybe doesn't allow it.

/rant

I can tell you what it is with this car, it's the roll center height coupled with the curves. Not much more else to it. I think mainly it's the roll center height, but I can't find a good pic of the exact suspension from behind to be able to say for sure. Curves, for sure, they're wrong, no matter what angle the links are at.

 

[Jugulador]

The inertia you're talking about comes from the unsprung mass, which seems completely ok in this car. 70kg per wheel, it's around where it should be. It's true because of the construction of semi trailing arms, the mass ends up relatively high, but it's also durable and reliable. Something like a Chapmann strut will have generally lower unsprung mass, but it also might not have the characteristics you want.

Nowadays we consider the multilink or DWB to be better than the trailing arm, kinematics wise, because you can achieve a higher grip in the scenarios you want and generally more predictable response due to less camber and toe variation, and you can engineer the curves to be how you want. Trailing arms are simple, and the math for it is simple, but it also means they're limited.

The trailing arms have behaviors like losing a lot of anti-squat when lowered, and the roll center migration amount being lower than in other independent suspensions, and large toe deflection on G forces due to how the bushings are setup. It's also easier to set up the IC and RC in ML/DWB than it is with STA.

On the other hand, you can easily get wheel hop on acceleration or braking in DWB/ML suspensions which aren't setup well, and ML suspensions especially can become simply unusable if not setup well because the tire will try to go through the wheel well! That's why you basically have to use a computer program to optimize the travel of the wheel with ML to make sure it doesn't hit the inside of the wheel well; but a trailing arm design is easier to design and test, it seems.

Then of course, there's the manufacturing and design cost thing as well. I would believe that just copying a proven trailing arm design, like for a cheap FWD economy car, is cheaper than engineering a high performance multilink for it. Packaging exists too; trailing arms are very flat, so you can have a lot of luggage space above them, while the subframe and camber arms for a DWB/ML maybe doesn't allow it.

/rant

I can tell you what it is with this car, it's the roll center height coupled with the curves. Not much more else to it. I think mainly it's the roll center height, but I can't find a good pic of the exact suspension from behind to be able to say for sure. Curves, for sure, they're wrong, no matter what angle the links are at.

That is what I said, but with less acronyms lol.

Anyway, Fuzo's idea of a good handling and easy to get in car is not far from the real thing... the mod, with more 1:1 data, will still be like that, but drivers will have to be a little more careful, as real drivers seams to be (I watched "some" videos of people racing it). It's a win-win situation. But, to stay synthetic on my posts (no more wall of texts here lol), there will be small situations (as you pointed some) that will require some more care when driving it.

 

[Kyuubeey]

It's possible the car would handle better, or be faster with a realistic rear as well. It'll just be different, and you might have to change something else too, to keep the feeling.

 

[Jugulador]

It's possible the car would handle better, or be faster with a realistic rear as well. It'll just be different, and you might have to change something else too, to keep the feeling.

The car will probably become a drifting dream, as touring cars of that era still are. I don't know about be better to drive, but certainly it will be funnier.

 

[Haris1977]

Very good indeed.

I have a small issue: Wheels do turn very quick when you steer left/right (in addition to kunos other cars). Can i tweek the steer linearity of the specific car?

 

[Fuzo]

Very good indeed.

I have a small issue: Wheels do turn very quick when you steer left/right (in addition to kunos other cars). Can i tweek the steer linearity of the specific car?

the Skoda has steering wheel angle of 1040 degrees, if your wheel is configured for less than this number, usually 540 degrees, it will turn 2x faster. There is a possibility to set "car specific controls" in CM for every car but I don't know if it will help you, I have never used this feature.