Is Suspension Tuning Backwards? - A Test with RX-8

  • Thread starter Maturin
  • 458 comments
  • 42,984 views
Pungent
However, a spring pushes the force acting against it toward its center axis for the system, in this case the car. The stiffer back spring pushes this force easily against the center of balance, dumping a lot of force toward the front. The front spring is not strong enough to cope with the force. Therefore the front tires lose grip, but now the car is not turning inward. To steer you must have grip in the frontend. The result is understeer.
Click to expand...

this is an incorrect explanation.

First of all, as long as you stay off the bumpstops, there is no such thing as "the front spring not being strong enough to cope with the force." Either you run out of travel or you don't.

the stiff outside rear "dumps a lot of force" as you put it to the inside front. More accurately, it limits the transfer of weight off the inside front. This results in a more even weight distribution on the front axle, resulting in better grip.

I will draw this out again. example car has a 50/50 weight distribution.

500 500
500 500
STATIC

700 300
700 300
right turn, evenly sprung

600 400
800 200
right turn, stiffer rear spring

again, note how the stiffer outside rear spring "dumps" its force onto the inside rear, reducing the weight transfer in the front.

For what its worth, I bought a Pontiac GTO in the game, and equipped it with the semi-racing suspension. I moved the sliders around a pretty good amount. A stiffer rear spring combined with a softer front spring (within the limits of the semi-racing sliders, which don't go up very high... only to like 8 kg/mm) produced more rotation in all corners under all conditons.

The effect was not very drastic, but then again, neither is the range of the slider on that car's semi racing setup.

The shocks behaved properly also - stiffer rear produced more rotation when the car was in lateral transition - in all corners, under all conditions.

it would seem the GTO behaves properly. The Prelude does not. I will compare the prelude to other FWD cars, and I will also buy the racing suspension for my GTO to try more extreme settings.

(FWIW, the semi-racing suspension does not alow adjustment of sway bars, and the shocks adjust with one slider, not individual ones for bound & rebound - just incase anyone was wondering).

I still maintain that the physics engine is flawed, but only on some cars. Perhaps just the FWD models. the testing I mentioned above should be interesting.
 
As near as I can tell, the tire models are hosed, and without reasonably realistic "traction circle" and load effects, you aren't going to see accurate response to suspension changes or much else. It's still a great sim for a console game, not on par with good PC sims, but a lot of fun.
 
FUNKED1
As near as I can tell, the tire models are hosed, and without reasonably realistic "traction circle" and load effects, you aren't going to see accurate response to suspension changes or much else. It's still a great sim for a console game, not on par with good PC sims, but a lot of fun.
Click to expand...

regardless of the "shape" or "size", so to speak, of the traction circle, you would think that the developers tested each vehicle for A FEW MINUTES and made sure that basic tuning principles were accurate.

I had to win some more money last night (research funding :P) instead of doing my test, but the theory I'm going to run with now is that many cars react to setting adjustments as they should, and many do not.
 
Now I do no have extensive tests to proof this. But I think the suspension react based on the car's weight distribution. I used a Skyline and the Nissan R92CP to test it. The Skyline respond to the reversed suspension settings. But the R92CP respond to the suspension settings normally. (soft front produce more oversteer) This lead me to believe the suspension react to tuning differently based on the weight distribution. You need to soften the lighter end to gain more oversteer. Anyone want to try it out? :crazy: This is not based on any physics, just some observation.
 
mugen_s2002
You need to soften the lighter end to gain more oversteer. Anyone want to try it out? :crazy: This is not based on any physics, just some observation.
Click to expand...

hmm, intersting theory (completely FLAWED! LOL.... but good possibility).

I need to broaden my garage and test with more cars. you might be right. Is there a specs screen that shows the F/R weight percentage of the car?
 
No, I just go by the real car data that I can find. The Skyline is 52/48, idk about the R92CP. But it's mid engine, so it should be rear biased weight. If you want to do an extensive test, you can use a RSX (something like 65/35) then use a RUF RGT (something like 35/65) then use a Z4 as control, (50/50). I don't have extensive testing, to back this up, it's just my own experience.
 
Greyout
I still maintain that the physics engine is flawed, but only on some cars. Perhaps just the FWD models. the testing I mentioned above should be interesting.
Click to expand...
You are not one of those guys that believes there are little Skylines and Rufs running around in your Playstation right? There is no way that Polyfony weighed each and every bolt, every swinging arc and dimension of every piece of every car in the game. No Way. So, what we are left with is a simulation, a group of algorithms or functions, which are essentially mathematical equations. If the equations map the numerical values of your inputs to what you expect to see in the car on the screen, you have a good simulation. I feel for you Greyout, in that the programmers left you wanting in terms of FF simulation...
 
aarque
You are not one of those guys that believes there are little Skylines and Rufs running around in your Playstation right? There is no way that Polyfony weighed each and every bolt, every swinging arc and dimension of every piece of every car in the game. .... I feel for you Greyout, in that the programmers left you wanting in terms of FF simulation...
Click to expand...

no, I didn't expect that for a second.

I just expected that they spent a few minutes with each car, testing the settings to see how their model behaved, as part as normal QA.

If they did do that, then the above flaws I mentioned with the prelude would not exist.

The reason that I said that I think that it doesn't apply to all cars is that my tests with the GTO suggest that the suspension adjustment results are NOT the product of a perversion of the basic laws of tuning by PD, but some kind of an error when certain variables exist in a certain way (what we see as a FWD car depicted on the screen).

I honestly don't think its a problem with just the FWD car simulation method, because the original posts of this thread suggested that the problems existed with some FR models.... and it probably is not all FWD models, as I have only messed with one.

I am not sure if your post is sarcastic or not, but I am very aware what I should, and should NOT expect a car video game to be that makes an attempt at realism. I KNOW that the game is simply strings of code that, in the end, enables my TV to display different colors in the proper places at the right time, which my eyes transmit to my brain, which then decodes and associates with tangable objects that I have interacted with in the real world, which really were just impulses to my brain to begin with.

blah.

something is wrong with the simulation, when using many cars. My S4 didn't respond properly. My Prelude DEFINATELY didn't. But so far, my GTO seems to respond as it should.

I wonder what the big gaming guide will say, if it ever comes out.
 
Not so much pointed sarcasm as sympathetic cynicism was my last post. I respect your thorough investigations and explanations, Greyout; perhaps I was trying to offer perspective.
I did see an image of the guide when I accidentally clicked to the Japanese GT4 website; the cover, rife with kanji or katakana, also had the English words "The Bible" on the cover. For some reason it makes me think of Celicas that resemble Mustangs and Yamahas that look like Harley Davidsons. The Bible indeed.
 
Greyout
regardless of the "shape" or "size", so to speak, of the traction circle...
Click to expand...

With R tires it's not even a traction circle, more like a traction hyperbola or something. You can have massive longitudinal slip but the tires still have lateral grip.
Thankfully they did OK on the N tires.
 
FUNKED1
With R tires it's not even a traction circle, more like a traction hyperbola or something. You can have massive longitudinal slip but the tires still have lateral grip.
Thankfully they did OK on the N tires.
Click to expand...

so its more like a traction square then, huh? :)
 
Greyout
this is an incorrect explanation.

First of all, as long as you stay off the bumpstops, there is no such thing as "the front spring not being strong enough to cope with the force." Either you run out of travel or you don't.

the stiff outside rear "dumps a lot of force" as you put it to the inside front. More accurately, it limits the transfer of weight off the inside front. This results in a more even weight distribution on the front axle, resulting in better grip.
Click to expand...

Ok, I'm trying to figure out exactly what you mean here. It sounds like you are saying the same thing as I am but in different words.

"Running out of travel" is what I mean by overloading the front spring. A soft spring can only push at a certain amount. Exceed this amount and the spring compresses to its maximum point, effectively the travel is gone, and the force from the body goes straight to the shock, which then compresses quickly, and the body transfers directly to the rotor/tire. The effect is no suspension. It is now stiffer than the back end suspension, force is going directly to the tire from the body.
 
Greyout
so its more like a traction square then, huh? :)
Click to expand...

Yeah that's what it seems like.
BTW your weight transfer explanation with "I'm not an engineer" disclaimer was pretty good. I am an engineer and you got it right.
 
Pungent
"Running out of travel" is what I mean by overloading the front spring. A soft spring can only push at a certain amount. Exceed this amount and the spring compresses to its maximum point, effectively the travel is gone, and the force from the body goes straight to the shock, which then compresses quickly, and the body transfers directly to the rotor/tire. The effect is no suspension. It is now stiffer than the back end suspension, force is going directly to the tire from the body.
Click to expand...
Please, I know it's a small detail, but if you could try to express it this way:
"Exceed this amount and the spring compresses to its maximum point, effectively the travel is gone, and the force from the road goes straight to the shock, which then compresses quickly, and the body transfers directly to the rotor/tire."
Most loading, except weight (which is static), actually comes from the road forcing the car up, not from a dubiously massive car pushing down onto the road.
 
My brain is sore from reading all this stuff, and trying to make some sense of it.

What I have yet to see (correct me if im wrong here, and full marks to whoever mentioned it) but people keep going on about springs, forgetting their main purpose: to work in conjunction with the damper. The spring stops almighty jolts being transmitted thru the car body; the damper stops the spring from oscillating such that the ride would feel like sitting on jelly otherwise (common in old american cars - for some reason the USA preferred stupidly weak springs/dampers). However, the damper unit, containing fluid as it does, acts also like a viscous spring... and lets not forget progressively-wound springs, where the spring constant varies according to the compression/extension (i.e. not obeying Hooke's Law). No-one's said anything about castor angles, king pin inclination, the actual suspension layout altogether surprisingly has an effect... (MacPherson strut, Chapman strut, Panhard rods, swing-axles etc...)

There are so many variables involved in suspension, that separating one out is nigh-on impossible. All these attempts at explanations are all well and good, but for a lot of fans I imagine, they just make no sense/are of no help. i am utterly confused for a start.

So lets stop hypothesising and do behave like true scientists: if it cant be done by mathematical proof or logical reasoning, use experimental data instead. Never fails...
 
rk
Please, I know it's a small detail, but if you could try to express it this way:
"Exceed this amount and the spring compresses to its maximum point, effectively the travel is gone, and the force from the road goes straight to the shock, which then compresses quickly, and the body transfers directly to the rotor/tire."
Most loading, except weight (which is static), actually comes from the road forcing the car up, not from a dubiously massive car pushing down onto the road.
Click to expand...

Most cars have bump-stops (a rubber block inserted around the damper's piston rod) to prevent the spring/damper unit compressing fully, thus destroying the coil-over-shock unit. The shocks NEVER compress really quickly - they're designed to have very slow movement, especially in rebound, otherwise they wouldn't damp the oscillations from the spring. Dampers are designed for very small movements. Obviously you do reach a point where the damper is at its maximum travel, and as such acts like a solid beam - this would be your maximum bound. but you really have to hit a bump hard for that to happen, or be doing something quite insane.

Your last sentence makes no sense - most loading comes from the road reacting against the weight of the car... and fairly obviously a heavy car will have more loading than a light car. Why do you think trucks have much more stiff suspension than a car?? there are no other mysterious forces acting, otherwise the car would take off!! Simple bit of statically-determinant mechanics should prove that to you.

enough from me - i could go on all night.
 
RenesisEvo
Your last sentence makes no sense - most loading comes from the road reacting against the weight of the car... and fairly obviously a heavy car will have more loading than a light car. Why do you think trucks have much more stiff suspension than a car?? there are no other mysterious forces acting, otherwise the car would take off!! Simple bit of statically-determinant mechanics should prove that to you.

enough from me - i could go on all night.
Click to expand...
Considering the car to be at the center of the universe is a simple way of removing the confusing aspect of gravity and its relation to suspension operation. Basically the suspension protects the car from the road, not vice versa; granted the analogy begins to falter when you consider weight transfer and its effects, but for understanding basic suspension, it is simple. Bump hits wheel, spring abosrbs and stores energy, hands off to damper, damper turns energy into heat, end of bump and story.
Not to veer to skewerdly off topic, but does this have anything to do with whether suspension tuning works backward or not...?
 
*bump*

Alright..I've been watching this thread since the first post...and I'm waaaay to confused with all this crap going on in the backround.

Somebody simply answer this question with "Yes" or "No".

Is the suspension tuning backwards?
 
I concur, Bad760.

The physics engine is more sensitive to many changes in the suspension system but the Spring Rate question (dubbed the "Maturin Effect" for posterity) would seem to be settled.
 
Greyout
well you can also call it the "greyout effect" because I get it on several cars
Click to expand...
Would you be kind enough to list the cars and the conditions under which your spring behavior reverses? I would like to set it for my cars and experience it.
 
I'm with Greyout. I believe it's backwards, so I just "tune" the suspension around it to make the car do what I want it to do rather than what it should do on paper.

Basic concepts (these are real life and probably have a 50% chance of being found in the game's code:)
-Cars can be overdamped or oversprung and still suffer from the same poor handling effects, but for different reasons.
-Springs are there to control body roll in the effort of maintaining proper camber. These should be used in conjunction with swaybars to get the proper wheel rate for any particular application.
-The ultimate goal is to have a perfect contact patch all of the time. This isn't going to happen, so one compromises a suspension setup.
-Whenever one's suspension hits the bumpstops up front, there is never an actual bump. In fact it's just felt as understeer as when the car rests on the bumpstops it also spikes the springrate, overloading that tire and traction is lost.
-There is a 1% chance that the game designers accounted for bumpstops, physical shock travel (metal to metal if bumpstops were removed) and coil bind. Basically, throw anything related to these topics away.
-Whomever said that shocks move very slowly is a pour, misguided soul. Piston speed can vary GREATLY. There are so many variables, it's just silly.

Have fun! :)

-Biggly

P.S. I feel the game is flawed if I can't set it up exactly like my race car, or any of my friends' race cars and have it perform nearly in the same fashion. I've tried it before and not only is it not close (close would have been good enough for me!) but it's basically inverted on real-world handling characteristics. That's why I call it a game and have fun playing the game, but calling it any sort of simulation is far removed from anything grounded in reality.
 
...err, whatever; I'm sorry to re-ask this question which was covered about 47 pages ago, but, with so many settings unvalued or placed on a relativistic 1 to 10 scale, how do you know exactly what to translate game settings into? Can you actually measure and set your spring rate to have a range of 2.6 "kgt"/mm to 15.0 kgt/mm, then make some comparative settings in-between? Forgive me if I am mistaken, but I believe that would require a set of coils at every tension level, as pre-loading and other spring adjustments wouldn't substitute. Also, if you would, please explain what a stabilizer of "4" translates to in real life, and how that compares to one which is, say "5". Is the first one 1/5 smaller? Or is it 50% as stiff (large?) as the maximum, 7 could be? And then there is the whole damper thing. Some people believe the scale is absolute, meaning a low value of 3 would correspond to a very low energy level, not likely to be used with higher spring settings, say above 8; while some people believe the damper scale is relative, meaning it presents a range of settings appropriate for the spring tension selected, where a value of 3 might correspond to among the "least resistive", but still reasonable settings you might use with springs at 8. Some of us, apparently, have already figgured this out, pray tell, which is it please?
And as far as bottoming goes, I know that when I land off the crest at Quiddelbacher Heights and my car screeches and jerks to the side, a few clicks of ride height will cure it. Sure seems like suspension bottoming, since there are rarely sparks...
 
Well, if I understand your question right you're asking if I've tried different springrates in my racecar or how I translated damping to the game?

Springrates I've used before: 7k/5k, 7k/8k, 8k/10k, 10k/12k and with a variety of bar settings in real life. So I tried a couple of those (the ones I liked best in real life) on the GT4 car and then set the camber. I also left the shocks on the default settings at first as I don't like a particular amount of over nor underdamp in my real car unless I'm using it to try and get the rear out on a particularly odd course.

I tried a few bar settings, which didn't have a whole lot of effect overall. The springrates used in real life and their effects were quite counter to what happened in the game. On my pushiest real-world setup I found the best rotation in the game; camber not changed. Bars I adjusted a bit back and forth to find what effect they were having and then shock settings I again, left alone as I felt that in the game they were rather insignificant. (Previous tests of extreme settings on the shocks didn't produce a large amount of change like I've seen on my vehicle, so I set them in the middle and focused on where the game would allow for the most change.) As for whether the game scale is properly mated to the springrate or absolute...I'd say six of one, half dozen of another. However if I were pressed to choose one, I'd lean towards "not absolute" as the changes seemed the same in their responses regardless of springrate.

As far as your specific example of your particular vehicle setup and "bottoming out" what exactly is bottoming out? I've done this before on cars and bringing the ride height up certainly helped, but I'm not sure what you're trying to ask really. Are you questioning my statement about the designers not taking into account things like bumpstops, coil bind, and physical shock travel? I feel it's perfectly valid. I never said that it couldn't bottom out, simply that the game can't account for specific facets of typical coilover issues and I listed them as such. :)

-Biggly
 
DrBiggly
Well, if I understand your question right you're asking if I've tried different springrates in my racecar or how I translated damping to the game?


As far as your specific example of your particular vehicle setup and "bottoming out" what exactly is bottoming out? I've done this before on cars and bringing the ride height up certainly helped, but I'm not sure what you're trying to ask really. Are you questioning my statement about the designers not taking into account things like bumpstops, coil bind, and physical shock travel? I feel it's perfectly valid. I never said that it couldn't bottom out, simply that the game can't account for specific facets of typical coilover issues and I listed them as such. :)

-Biggly
Click to expand...
All I am saying is that you are asking perhaps a little much. Polyphony took some car somewhere, possibly several, and distilled it down to crystalline math; it's practically freakin alchemy when you come around the corner at Montegi with that sunglare, or catch a glimpse in your rearview of the Panoz coming through the dust cloud you left getting a little loose at Aremberg on your way down into The Foxes' Pipes.
Want to tighten up? Just a click or two of camber and it's back out on The 'Ring.

Now you want to take this math and re-materialize it into a tight ride. It's like Vger trying to make humans based on the recipe scratched on the side of Voyager, in my humble opinion. There just isn't enough clear data.
So far as bumpstops, coil bind and shock travel; in so much as they can be reasonably mapped into a function, I believe they are protrayed. Bumpstops: I can give evidential examples that can be applied in game. Coil bind: seems a little too esoteric, it would be nice, but I am guessing PD decided it didn't affect overall performance enough to warrant the trouble. Shock travel: that one seems too obvious to me so perhaps I misunderstand which aspect seems absent.
In the end, we essentially agree "enjoyment" is key. :)
 

Latest Posts

Back