11/2 - I'm bumping this up to get more input from others. I just completed my first season of autox, and with Ksport generic spring rates (10kF/6kR - 560f/336r). I'm looking to go GC for next season with ots Konis. I'm thinking I'd like to get a small front swaybar and larger rear, with stiffer front rates than rear. I know stiffer rears is more popular on this site. But I have all winter to do my research, so hopefully you guys can share some knowledge. I do plan on NOT going rcomps next season, because I believe I need better feedback from tires to get my driving better (this was proven when I went back to street tires for the last event of the season; I loved the feed back). Rcomps (if you see in my other videos) had me taking less ideal lines because I felt like once I was on that line, I was stuck with it. Also, the rcomps gave me nearly a split second of when it 'screeched' and the point where it just gave out, I am not good enough to know this 'point' yet. - is there a FWD specific tuning book?

I'm going to post one of my runs, I think my turn-ins can be a bit sharper. http://www.youtube.com/watch?v=hL7JK0_jT8w


<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by PIC Performance &raquo;</TD></TR><TR><TD CLASS="quote">
OP: You're not going to have very much luck making something like this on H-T, simply due to what H-T is: a public web forum where anyone can sign up and post whatever they want, totally unmoderated and unchecked for the most part. Anyone who looks to this thread for solid information is probably going to run into a mish-mash of un-researched and biased personal opinions, hearsay and suspension myths. Rather than try to make this a reference, OP's should just google/research the topics that they see discussed and get more answers from "cleaner" outside sources. This applies to most of H-T as well.</TD></TR></TABLE>

I have been searching (ht) for the last few hours and read various threads, and always come up with multiple answers. The latest, an argument over swaybars.


Modified by Kusai.Nihonjin.Desu at 1:27 AM 11/2/2008

 

I suspect yours might be a rather unwieldly concept for a thread, I know if started posting all my theories on this topic it would get rather long rather quickly!, though it might not be quick to actually write.

As to your suggested initial question, if you really want to know; higher rate rear springs will cause a greater % of the total weight transfer to occur at the rear of the car than at the front. As a consequence a lesser % of the total weight transfer will occur at the front, resulting in more equally loaded font tyres and less understeer. The affect is quite similar to fitting a stiffer rear ARB.

 

Possessing a comprehension for geometry will help to determine other choices, spring rates for example. In particular, understanding static and dynamic instant centers, roll centers, and motion ratios will help to guide spring rate selection and damping needs...as well as camber settings. The fanny dyno isn't always so accurate, so a little math is helpful.

 

I forgot to mention, stiffer springs in the rear lessens front weight transfer as already described, but since it does this it improves traction at the inside front wheel on corner exit. This is really the main reason to have higher rear than front roll stiffness on 'performance' FWD cars. Any benefits to reduce understeer is icing on the cake.

 

An understanding of roll centre behavior and the resulting affects on weight transfer etc is useful (necessary!) when designing a suspension from scratch or making substantial modifications to an existing suspension, but probably outside the scope of most enthusiasts modding their road car or building an occasional track day car.

Perhaps more important for a typical enthusiast to undestand might be the affect of say lowering the chassis on the locations of the instant centres of the virtual swing arms, and how this might affect camber curves etc.

The affects of geometric roll stiffness caused by roll centre location(s) are very complex, and the subject of much disagreement even amongst proffessional suspension engineeers. For the typical enthusiast it will be much easier to limit changes in roll stiffness to 'mechanical' changes, i.e. ARB and spring rates etc. This will more often have a much more noticable affect in any case, unless very large changes to geometric roll centre locations are made.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by johnlear &raquo;</TD></TR><TR><TD CLASS="quote">I forgot to mention, stiffer springs in the rear lessens front weight transfer as already described, but since it does this it improves traction at the inside front wheel on corner exit. This is really the main reason to have higher rear than front roll stiffness on 'performance' FWD cars. Any benefits to reduce understeer is icing on the cake. </TD></TR></TABLE>

The amount and rate of weight transfer is not reduced when going with stiffer springs. The same amount will be transferred regardless of what spring rate you run.

OP: You're not going to have very much luck making something like this on H-T, simply due to what H-T is: a public web forum where anyone can sign up and post whatever they want, totally unmoderated and unchecked for the most part. Anyone who looks to this thread for solid information is probably going to run into a mish-mash of un-researched and biased personal opinions, hearsay and suspension myths. Rather than try to make this a reference, OP's should just google/research the topics that they see discussed and get more answers from "cleaner" outside sources. This applies to most of H-T as well.

 

Pic Perrformance said:
"The amount and rate of weight transfer is not reduced when going with stiffer springs. The same amount will be transferred regardless of what spring rate you run."

The degree of weight transfer that occurs from the inside to the outside as a sum of total weight transfer at both the front and rear of the chassis is determined only<U></U> by the strength of the lateral acceleration (i.e. 'G' force), the CG height, and the track width. Insofar as this goes you are correct.

However, if you alter a spring rate (or ARB rate) at one end of the car this will alter the % of the total weight transfer that occurs at that end of the car, and also at the other end of the chassis.

I.e. if you start with a hypthetical car with equal weight distribution and equal front / rear spring rates etc then with a purely lateral acceleration the weight transfer will be equal front / rear (i.e. 50% will occur at the front and 50% will occur at the rear). If we now increase total effective spring rate (i.e. springs + ARB) at say the rear only so that the rear now has 75% of the chassis' total roll stiffness, then we will increase weight transfer at the rear to 75% of total sum weight transfer, and decrease weight transfer at the front to 25% of the total sum of weight transfer.

These % numbers don't take into account any weight transfer that also occurs through the vector of suspension geometry (i.e. roll centres), and thus are merely approximations. Since I'm ignoring geometric roll stiffness at the moment, all mentions of roll stiffness in this post are only referring to 'mechanical' roll stiffness, i.e. roll stiffness created by spring rate etc and not geometry.

This difference in front to rear roll stiffness is called the 'roll couple', and is of fundamental use in tuning grip and handling balance. Note that changing the roll couple has no direct affect on total sum weight transfer unless it results in a higher lateral acceleration being achieved, in which case it has an indirect affect. Also note that the greatest % of total weight transfer will always occur at the end of the chassis with the greatest roll stiffness. The rate (i.e. speed) of weight transfer will also be faster with a higher spring and / or ARB rate than with a lower rate, so if we have a higher roll stiffness at say the rear then rear weight transfer will not only be greater but als be more rapid than weight transfer at the front.

 

My effort was a broad stroke suggestion and in spirit provided a few clues for the OP; when my knowledge of suspsnion theory surpassed that of simply selecting spring rates based upon the next guy, I was amazed by the depth of information, and, by how complicated - or a mess - a car can be as it moves along the road.

Most folks do not understand the connection between instant centers and camber via virtual swing arms. The same is true with RC locations - in a symetrical layout - relative to CofG; ride hieght can affect weight transfer since the length of the moment arm (difference from front to rear) will have an affect on weight transfer. This knowledge can be enlightening, and, help one to achieve better results...and to make better use of springs...


Modified by meb58 at 7:54 AM 1/22/2008


Modified by meb58 at 7:55 AM 1/22/2008

 

Meb58,
I don't mean to imply that the pursuit of such knowledge is pointless or not to be enouraged, just that I'm not sure how many people on a forum such as this would actually understand, participate or be interested in such hard core discussions!

Personally I'd welcome it if some serious theoretical discussions started up, we'll see! Perhaps what's needed is a 'Suspension / chassis - hard core theory' sub forum?!

 

I would love that! I'm not a suspension engineer nor did I major in physics. But I do love this as I'm sure the OP does???

And, once we get into all of the different vectors acting on a car thru all of the suspnsion links and bushings - as the RC moves around - I get lost keeping track of what happens. I don't have enough fingers to hold the various pages open in my books all at once This just isn't my field. Milliken is helpful, but I sense curious gaps...

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by meb58 &raquo;</TD></TR><TR><TD CLASS="quote">I would love that! I'm not a suspension engineer nor did I major in physics. But I do love this as I'm sure the OP does???

And, once we get into all of the different vectors acting on a car thru all of the suspnsion links and bushings - as the RC moves around - I get lost keeping track of what happens. I don't have enough fingers to hold the various pages open in my books all at once This just isn't my field. Milliken is helpful, but I sense curious gaps...</TD></TR></TABLE>

I've done quite a bit of theorising on and fairly in depth discussion of chassis dynamics, mostly in relation kart chassis, but the physics are still the same for cars if applied in different ways for different affects. I've read a fair bit in books and on line (less good stuff on line than you'd think, unless on quite obscure sites somewhere I haven't found yet), but most of what I've read is more or less a springboard for independant thought since most of the published work I've come across to date only takes you so far (particularly in the area of weight transfer theory which once you start thinking about seriuosly turns out to be far more complex than any books I've read makes it out to be).

To really get to some sort of grips with this sort of thing you need to develop an ability to visualise forces acting through the various force vector pathways, which helps give you the in principle big picture, but doesn't take you far enough if you were say designing a speciifc suspension from scratch where you are likely to start needing real mathematics.

Milliken and Milliken are authors I've yet to graduate to, I keep meaning to order a copy but it's a realtively expensive tome and my incoming / outgoing ratio needs to improve a bit before I do!. When I do get the Milliken book I suspect I'll have to brace myself for a dose of relatively complex maths, which is a teeny bit scary considering my maths ability!

 

I get the visualization; I've been part of a 3D visual profession my whole life. So I more or less understand the migration of loads, more or less. But the actual loads as they relate to selecting the proper spring rate, ride height, swaybar dia., track and so on is a wee bit above my head. And yes, many of the formulas in Milliken are complicated...some are not. So, I guess, I use the fanny dyno, I guess more, I ask around and so on and so forth. But then, this is but one hobby

Eventually, a bigger piece of the puzzle becomes clear and helps to connect other less known aspects. Fun stuff!

 

Even with a firm handle on the theory, calculation and modelling etc only get suspension designers and engineers into the ball park. Every design / modification needs to be tested and at least fine tuned by informed trial and error on the road / track. The dynamics are hugely complex, not everything can be accounted for on the drawing board.

Probably the hardest thing to get your head around is the difference between 'mechanical' roll stiffness (I've heard it called 'elastic' roll stiffness) created by spring and ARB rates (and transiently by damper rates) and 'geometric' roll stiffness, and their interaction with each other.

These different kinds of roll stiffness behave very differently, but are always present in any suspension in greater and lesser %s of the total roll stiffnes of a particular suspension (in the real world). Where it gets really hard is when you start to realise that it must be the case that the relative %s of each type of roll stiffness in effect change relative to each other as %s of the total roll stiffness as lateral acceleration and thus weight transfer increases (or decreases).

The clue to this realisation lies in understanding that the geometric roll centre isn't actually the point around which the sprung mass actually rolls unless geometric roll stiffness is 100% of total roll stiffness (as you get when the GRC and the CG are at exactly the same height), or only at the very infinite instant that weight transfer begins to occur, after which geomtetric roll stiffness starts to decrease as a % of total roll stiffness and mechanical roll stiffness starts to increase as a % of total roll stiffness (even if the GRC doesn't actually move, which it nearly always will).

We can know this because the point around which the sprung mass actually rolls starts to migrate away from the GRC toward the outside contact patch the instant that acceleration and weight transfer begins to occur and then increases (unless of course geometric roll stiffness is 100%). This affect is present to greater and lesser degrees in any different suspension design, and changes with any modification to geometry and spring rate.

It's important because it dictates how much of the total weight transfer occurs instantaneously through the geometric vector and how much occurs 'slowly' through the vector of mechanical roll stiffness, and at what stages in the cornering process what % of weight is transferring via geometric or mechanical roll stiffness vectors. This all has a significant affects on transient handling characteristics.

Really, this stuff can do your head in! It takes a huge amount of thought to zero in on a reasonably correct in principle undrstanding, and I'm not there yet! It's difficult because to understand X you need to undertsand Y, but to understand Y you need to understand X. You end up overshooting your understandings of both X and Y then having to backtrack trying to home in on them, constantly referencing one to the other. I'm getting a mild headache just thinking about the problems of exploring the dynamic, let alone thinking about the dynamic itself!

Discussions of this nature quickly become very time consuming and taxing, the need to be absolutely precise in your language (let alone your concepts) wears you out fairly fast! Individual components of theory tend to be relatively simple in themselves, but the interactions between them are brain damagingly complex.

I think I'm at the point where I'm no longer labouring ander any major misconceptions to do with the fundamentals of suspension / chassis theory, but that's probably akin to 'famous last words'!

 

Well, I ususally refer to the RC as the kinematic RC and some of your thoughts highlight that phrase. As the RC migrates around the CG its leverage on the CG and therefore on weight transfer to every other component can be linear to (non-linear?) and back. And, throw in a bump, a little wind...and lets stand on the brakes for a second...wow. I should just buy a reallt good kinematic program and play.

I wrote non-linear because deflection is a tuff animal to get one's hands around. tires and bushings can scew...or screw with a good calcualtion.

Love your last sentence.

 

QUOTE=meb58]Well, I ususally refer to the RC as the kinematic RC and some of your thoughts highlight that phrase. As the RC migrates around the CG its leverage on the CG and therefore on weight transfer to every other component can be linear to (non-linear?) and back. And, throw in a bump, a little wind...and lets stand on the brakes for a second...wow. I should just buy a reallt good kinematic program and play.

I wrote non-linear because deflection is a tuff animal to get one's hands around. tires and bushings can scew...or screw with a good calcualtion.

Love your last sentence.[/QUOTE]

A reply! I was starting to think I'd scared off all the horses!

Some engineers will refer to the point around which the sprung mass actually rolls as a 'force based roll centre', but I'm yet to come across a description or explanation of this that isn't at best quite woolly and at worst evasive. I have my own ideas as to the nature of this phenomenon, but actually explaining it fully is very difficult and text consuming (I've tried, and it tends to turn into a convoluted, difficult to follow and lengthy treatise). Lack of hard data tends to turn it into an exercise in ‘thought experiment’ in any case. I quite understand why attempted explanations of this tend to be woolly and evasive!

Here's an abridged attempt:

We need to understand that geometric and mechanical roll stiffness (at a single axle line) can be expressed as either a % strength relative only to themselves (between 0 and 100% of possible stiffness of that vector), or as a % strength relative to the sum of roll stiffness (i.e. the two vectors added together). The % strengths of each relative only to themselves constantly change as acceleration and weight transfer changes, and also change as %s of the sum roll stiffness.

This is also the case even if the GRC doesn't actually change location (as with a true swing arm suspension, but not with double wishbones etc where the GRC will move), except that in the case of the GRC not moving the geometric roll stiffness strength doesn't change relative to itself, but does change as a % of sum roll stiffness as mechanical roll stiffness changes strength.

This is due to the springs et al becoming effectively stiffer the more they are compressed, i.e. the more 'weight' (mass X force) transfer occurs the less weight remains to be transferred, and this remaining weight then 'sees' an effectively increased (pre-loaded) spring rate of the partially compressed spring on the outside. When the springs are increasingly loaded in a bump loading they are linear in their resistance, but effectively progressive in their resistance to weight transfer.

Geometric roll stiffness remains a constant (relative only to itself) if the GRC location remains constant, but typically it migrates from the centre line to a position to the inside of the centre line and usually lower, which both reduce the strength of geometric roll stiffness at the same time as mechanical roll stiffness is increasing, which is why mechanical roll stiffness tends to be more influential after turn in than geometric roll stiffness (though geometric stiffness tends to be more influential on dynamic behaviour of the chassis at turn in and shortly after), so long as geometric roll stiffness isn't a very high value.

This means that initial roll will begin at the GRC, but as the outside spring starts compressing the sprung mass will increasingly 'pole vault' (through the spring that is increasingly behaving more like a stiffer vaulting pole) from a point located at the outside contact patch, and thus the point of actual roll will migrate from the GRC increasingly toward the outside contact patch as mechanical weight transfer increases from 0 to 100% (of itself and as a % of sum stiffness). The end result at the extreme (as the inside wheel fully unloads) is the sprung mass rolling purely around a point that is located at the outside contact patch. This isn't a case of the sprung mass rolling around the GRC, then instantaneously switching to rolling around the contact patch point, it's a gradual progression.

I'm tired now. This is a simplistic explanation of how I understand the dynamic, (and it may not be 100% correct) and is in no way a complete explanation.

For sanity's sake it's best not to attempt to include tyre and suspension compliances in any theoretical understanding (or attempts at same), it's hard enough just to explore the in principle dynamic behaviour assuming 'rigid' articulations. Suspension bushing compliance and flexures in the suspension members etc will have real affects, but the tendency will be to make the 'hard' behaviour more of an approximation than actually wrong (assuming it's correct in the first place), more so the greater the compliances may be. A good reason to avoid such things as overly soft suspension bushes and softer tyre casings, at least for performance applications.

 

Not so much a theory, but I'm on ksports, and I've been hating the daily driving. I had them set to the softest it can go, since when I was on koni yellow/ground control, the softest setting was the best for the street, comfort-wise.

Today I set the ksports to the middle setting, and the ride improved a lot. While I can feel more bumps due to the stiffness, the bounce cut down drastically. Feels as though the braking distance has cut down a bit as well, and I'm guessing that's due to less nose diving.

Now my question is, is this 'more comfort' due to the struts matching the spring rates better? Stiff spring rates + stiffer damper setting = better ride? I figured perhaps it was the soft damper setting with the stiff spring rates that was causing the 'bounce'.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Nihonjin.Desu &raquo;</TD></TR><TR><TD CLASS="quote">Not so much a theory, but I'm on ksports, and I've been hating the daily driving. I had them set to the softest it can go, since when I was on koni yellow/ground control, the softest setting was the best for the street, comfort-wise.

Today I set the ksports to the middle setting, and the ride improved a lot. While I can feel more bumps due to the stiffness, the bounce cut down drastically. Feels as though the braking distance has cut down a bit as well, and I'm guessing that's due to less nose diving.

Now my question is, is this 'more comfort' due to the struts matching the spring rates better? Stiff spring rates + stiffer damper setting = better ride? I figured perhaps it was the soft damper setting with the stiff spring rates that was causing the 'bounce'.</TD></TR></TABLE>

Being underdamped causes bounciness. Turning your shocks up, increased the rebound, reducing the bounciness. I'm going to guess that those shocks adjust bump and rebound together, which is why you feel the bumps more as well now.

Braking distance should be unchanged. At the same rate of slowing, you'll still transfer the same amount of weight, and the tires will still have the same traction limit.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by TunerN00b &raquo;</TD></TR><TR><TD CLASS="quote">Being underdamped causes bounciness. Turning your shocks up, increased the rebound, reducing the bounciness. I'm going to guess that those shocks adjust bump and rebound together, which is why you feel the bumps more as well now.</TD></TR></TABLE>

Rebound stiffness also has a big affect on ride harshness, the difference between full soft and full stiff with Koni yellows is quite substantial in this regard, and the Konis don't change bump stiffness when rebound is changed, being only rebound adjustable.

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by TunerN00b &raquo;</TD></TR><TR><TD CLASS="quote">Braking distance should be unchanged. At the same rate of slowing, you'll still transfer the same amount of weight, and the tires will still have the same traction limit. </TD></TR></TABLE>

Yes, but increasing damper rates (front bump rate and / or rear rebound rate)increases the speed with which forward weight transfer occurs, if not the ultimate degree to which it occurs. This tends to make the car subjectively feel more responsive to initial braking inputs (I noticed this perception after fitting Koni yellows to my car, more so at the stiffer end of adjustment), and is also responsible for at least some of the perception of increased responsiveness to change of direction, i.e. stiffer dampers also increase the rate of lateral weight transfer making the chassis feel more responsive than perhaps it really is (though this is a bit over simplified, to some degree the chassis actually will be more responsive with stiffer dampers).

Initial nose dive (with stiffer front bump rate) or rear rise (with stiffer rear rebound) will be reduced (if not the ultimate degree of nose dive / rear rise once all transferring weight has finished transferring), and the driver will subjectively 'feel' initial braking response as a car reaction that is more longitudinal than rotational (i.e. body motion around the centre of 'pitch'). Initial nose dive / rear rise won't as strongly 'mask' the sensation that the brakes have good initial bite (initial brake response will be felt slightly more as a longitudinal reaction than an initial pitch reaction), but of course objective ultimate braking performance won't be significantly affected since the ultimate degree of weight transfer won't be affected.

Vertical motion has a big affect on perception of braking performance. I used to have an old BMW that had a rear suspension geometry that caused the rear end to squat under hard braking (i.e. a torque reaction from the rear brakes through the suspension geometry into the body), and as a result of this the whole car would 'drop' front and rear under hard braking, making the car 'feel' like it was 'clamping' itself onto the road under brakes (front suspension lowering due to forward weight transfer, rear suspension due to torque reaction).

This gave a perception of braking performance being better than possibly it really was. The affect on body motion in this case being analogous to this car having an effective rear rebound rate above 100% stiff (which of course is impossible), not just preventing rear rise but actively lowering it and decreasing the sensation of pitch (by actually reducing real pitch motion).

This would also have had some real affect on actual braking performance because the front and rear lowering under brakes lowers the CG which in turn means less longitudinal weight transfer so more braking effect more evenly distrubuted over all four contact patches, though I suspect at least part of the perception was in the driver's head as much as real.

 

post season bump