RR98ITR

The Z being for Erik Zapletal, here: http://fsae.com/eve/forums/a/tpc/f/1...3010668131/p/1

There is some direct relavance to where our KRC / n-lines / etc wind up on the lowered DC2 and related chassis'.

Also, great fun to follow the referenced archival links to some well animated discussion.

Scott, who isn't looking forward to a future of hyper-electronic-ferromagnetic fluid-smart-active-levitating suspension no matter how authentic the force feedback nor how high the increment of additional speed..."Can you believe it - they used to have these "barz" that actually physically linked the road wheels."..."Inconceivable!"..."That's Nothing - they actually used to get completely worked up over the geometry of the connective mechanicals"..."INCONCEIVABLE!"...Ah, the good old days...

azian21485

Lots of head scratching going on. I tried to understand all these different roll centers before and the more I read the more controversial it became. Then, as I read some more, I started to notice more and more people not caring about them at all and they actually only cared about jacking forces. But, if you keep the RC low or below ground, jacking forces are a non issue. Ortiz had a long article and he basically came to the same conclusion as the rest of the gang on that forum. So in the end it's just a big ball of headache that doesn't mean much and Chapman's suspension philosophy still holds true.

chargeR

Good thread and some good references mentioned and linked. Thanks for the link! It is good to get the the end of a thread like that and find the participants agreeing on an understanding that roughly agrees with your own.

descartesfool

Funny thing, but just reading the article called "Hydraulic Roll System" in the September issue of Race Tech Mag, and that is for an FSAE car from Coventry. Nicely manufactured titanium components too. And they designed and built it to control a roll centre height problem which arose from packaging decisions which didn't allow space for an anti-roll bar to cure the problem. If its' good enough for the new McLaren, well....

Kozy.

A good read, haven't yet followed all the links but will get round to it.

Seeing as this is the only current thread on anything roll center related, I thought I would pose this question to you guys, and see if you've noticed anything similar.

I was playing around with my load transfer spreadsheet this week, trying to add in the effects of preload and whatnot, and I noticed something about the wheel deflections for various setups. It seems a common misconception that altering the roll resistance via the springs and bars at one end, makes one end want to roll a different amount to the other end, the end result being that the total roll angle is the average of the two. Now, obviously this is completely false, what infact happens is the load transfers at each axle match the roll resistances, and both axles roll by the same amount. That bit's simple.

Now, it would seem that there is a roll axis (I know, it's a horrible term, but bear with me) for which the total front and rear deflections become a pretty close match throughout the total range of lateral accelerations. If one end is lowered or raised a different amount to the other (in fact it seems even if they are moved the same amount it upsets things), then the resulting jacking effects alter the roll stiffnesses but not the load transfers and the result is differing deflections between the axles. Now, we know if the chassis is assumed to be a perfectly rigid object, that this is not possible, however the chassis is not perfectly rigid, and so this difference must be taken up by chassis twist.

So, if you've followed that, my question would be, would mismatched roll centers (purely in terms of geometric anti-roll, or jacking effects) create a torsional force in the chassis? Is it possible that the wizards in the Honda R&D department worked hard to set the rates and roll centres just right, so as to create the minimal possible twisting moment on the chassis? Or, a more likely scenario, am I talking out of my ****? Sorry, ***.

RR98ITR

"Oh Dear...this IS serious business indeed. Yes, definitely, incorrect roll center relativity is, as any reader of Bob Bolles knows, a very grave mistake. The Correct setup will result in the car's chassis being under Zero torsional load at all times giving "Balanced" "Handling"".

Scott, who is talking out my ****? My sorry ***? So sorry ***, putting words in my ***...it's so hard sometimes not to *** a question...

Kozy.

Funny you should mention him, I was only last night reading a thread in which his methods were being discussed, by none other than Dennis Grant:

http://www.eng-tips.com/viewthread.cfm?qid=76824

Seems his theory is somewhat new, and also a bit under wraps with a patent?

descartesfool

Oh no, not Bob Bolles. His latest book made my mind twist so much I put it down and never finished it. I much prefer Z. Much funkier diagrams with all those thick arrows.

Kozy.

Is this in a good or a bad way? As a circle track racer, are his methods as skewed as the cars appear to be?

Having only heard his name pop up this week, I'm not sure how much attention I should be paying to him...

RR98ITR

Pay him very little beyond reading what he writes and rolling your eyes (one pair of things I do believe should roll symetrically)...

Kozy.

Haha, you guys are a gas!

I guess I haven't been around these pages for long enough yet, as I honestly cannot tell if you are being serious or not right now.

Fastom

I have always paid close attention to "Roll Axis"
Could it have something to do with imaginary or effective moment arms acting through the 'Roll axis" from each end of the vehicle ?

Serious comments please, Yes even from you Scott.

RR98ITR

You know how all this starts out for most of us? We want to inform ourselves, to extinguish our ignorance, and learn a useful framework of thought. Many of us read Puhn and Smith, and got 90% decent information if we knew enough to know when was enough which we didn't. Many gadzillions of brain impulse firings later, and with the benefit of the work of people like Z, Millikens, Olley, Ortiz, and Novak, it's possible to hold the reins of a framework of thought loosely and inhabit "enough".

I've got the Mitchell program, and I toggle iterations and watch the GRC move around, and I watch the n-lines. While I watch that in the back of my mind I've got the benefit of knowing what ideas get weighted how much when and where. And of Knowing that fast geometric weight transfer might be put to practical use at the rear of a front wheel drive car - that's why I lowered the mounting points of the inner ends of my rear UCA's a bit to get the GRC up to loosen the rear end generally. Course I do that and I have to check the bump steer again and find that I've screwed up Brian Slames outstanding work a little bit and will still be relying somewhat on Chapman's Theorem of Immobility. Fine.

So put me in the camp where they believe in the roll axis and its inclination AND sway barz AND (some) movement of the wheels up and down AND AROUND (eventually)...

Scott, who reserves the right to discover, however painfully, that I am again Wrong!...which I don't mind if behind Door Number-2 is Right!...but it's rarely That easy...

Kozy.

Well for now, I think I am going to give credance to the incorrect roll axis/stiffness thing causing chassis twist. My admittedly crude measurements suggest that the standard EK9 is designed in such a way that this effect is minimised, as does my current suspension, which despite being soft, high, and slightly front biased, feels absolutely superb.

I have no idea how much this 'twist' actually affects things, but given the levels we go to to improve chassis rigidity to reduce twist, it seems logical to try and stem the problem at the root. The less monkey barz you have to run through the cabin to keep the thing true, the less weight you run, the faster you go, right? So, I think any future tweaks should go as far as practically possible to avoid inducing excess twist through badly balanced upgrades, so long as it doesn't result in running horridly wonky RSDs, RCHs or anything else...

RR98ITR

Normally I would make a cutting remark about how setup can be varied depending on the drivers "style" and note that some drivers have a style that produces fast laptimes while others have a style that produces slow laptimes. But I'm not gonna do that this time. I've seen you on several boards earnestly asking good questions. You look to be taking a wrong turn here, and that's fine, you can go down any wrong road you like (I always do!), and you can always turn yourself back around, though sometimes it's tough doing it Austin Powers style.

Important thing: you saying that your car feels absolutely superb prompts two reactions: first, I'm very happy for you, second, it is unfortunately meaningless without some kind of context as to what the potential of the car is and how much of it you are getting out. SO Many times you hear this kind of comment from somebody who is a couple of seconds off the most that can be extracted AND at That speed it's no longer superb but terribly uncooperative and unsatisfying. Not saying I KNOW that's you, just saying I don't (and can't) know it's Not you.

Scott, who LOVES the pic Scott Lear took of Meris in the latest GRM...

Kozy.

Interesting that you should say it is the wrong road, how do you come to that conclusion? Attributing the feel of a car to the position of some arbitary points in space? I can see that I guess, a lot of hocus pocus involved, far to much to simplify it that much.

I certainly wouldn't base an entire setup around minimising a number that may or may not have noticeable effects on the cars handling, clearly the contact patch loadings are of far more importance, however I think it is a point of interest, enough that if after the running the numbers I get a 'twist number' that is severely out of whack with the stock car, then perhaps I should have a look at making a few tweaks to decrease it.

Intrigued as to where else you have seen me... and under which pseudonyms...

Kozy.

You are entirely correct about the feel of my car being somewhat irrelevant. It's a little difficult to quantify due to relaxed nature of our regulations (you are either modified, or not), and everyone else having a far bigger budget. I know I do not drive my car to its absolute maximum, as a few times I have let friends drive it who have managed to pilot it more skillfully than I, but only to the tune of maybe half a second maximum. The car last year was a contender for a class win (despite the only regulation being that it must be over 1400cc), I lost out on the championship class win at the last event by around 3 points to a VX220.

I believe that ease of driving trumps outright speed for me, I tend to be a spin or win type (minus the winning so far), so a stable predictable platform rewards me better than the technically faster, but harder to drive one. I found this out the hard way this year, when some unbalanced upgrades that should have made the car faster, infact made it slower. This is why I like to chase up details like this, in an attempt improve sometimes unquantifiable intangiables that improve my confidence in the car, and thus minimise that 'spin' window as far as possible.

RR98ITR

Eng-tips as Komodo86?

Here's the thing - to get the contact patch loadings you want on a typical fwd, you necessarily have to put some torsional load into the chassis. It ain't a crime. What is a crime is too many "monkey barz" runnin' thru the cabin space, unless you need something to hang your dry cleaning on.

Scott, who has some experience with "Hmmmm, it was supposed to go Faster"...

Kozy.

Seems you were 100% correct when you said I was heading down the wrong road. Having cast doubt over my math, I went back and checked it and sure enough, found an error which was creating the 'twist' effect. Thanks for suggesting, whether intentional or not, that there was something amiss, I probably wouldn't have caught that otherwise!

Ah yes, that would be me.

descartesfool

Found this beauty from John Deere. Front roll centre is not moving around. You can just imagine how the couple between the CG and the roll centre is going to control the roll angle. If you could just get the CG down to that rolling point, no roll at all. Anyone think you can generate a jacking force with this suspension?

RR98ITR

Note that all the front weight is unsprung weight! And it's completely undamped in roll as well.

Scott, who is unfamiliar with the sporting world of tractors...but imagines they have guru's in their world too...

Kozy.

Hmm, I would assume with CG=RC there would be no jacking, however I think lateral behaviour is irrelevant. Tractors are straightliners...



Now, about the anti-squat...