3-Link (Page one, from old site)

Printable View

08-16-2004
chicane67

3-Link, Page one (from old site)

ddennis68
Registered User
Posts: 264
(7/25/04 3:04 pm)
Reply 3-link
--------------------------------------------------------------------------------
Now that I've gotten everybody fired up about the way-to-anal-about-grammar-and-spelling-crowd over at c-c.com I have a question about some of the things I have been looking into over there. I'd ask them but they won't allow me into their little group. I don't have a rustang nor do I want one (even thought the 04 Mach 1 looks good), a few of the fox chassis guys were removing one of the UCA from the rear and adding a PHB, effectivly creating a 3-link. Is there anyone who has attempted this on an "A" body chassis and what were the results? I really like the 3-link (not everything over at c-c is that bad) and the reason I am leaning toward that over the truck-arm now is the weight factor that I had not thought of before. I was thinking some nice adjustable links with johnny-joints to eliminate any deflection from the rear, a good PHB, and some bracing to the rear frame section. I know, I know-not another list with a what do you think at the end, this is actually a well researched and thought out plan of attack with no bolt-on package available anyway. Just looking for input in case there is something I missed. Thanks-
Dennis-

check out progress of Bondobucket

Mean 69
Registered User
Posts: 81
(7/25/04 3:25 pm)
Reply No!!!!
--------------------------------------------------------------------------------
Dennis, don't do it. Removing one of the upper links will reduce the bind associated with the splayed upper control arms, but it is NOT the right way to do a three link.

Take a look at Evolution Motorsports, they are doing a four-to-three link conversion for the late model guys. I don't really like the setup, but it is intended to be a bolt in affair. Good guys too, I like them, and they are knowingly making compromises to satisfy a particular market niche.

By removing one of the upper control arms, you lose the bind, but gain some nasties in return. The upper arms are obviously slanted relative to the vehicle centerline. With a three link, the upper link is responsible for reacting to axle torque (acceleration, or braking), "only." These forces are longitudinal, but the remaining link would not be, inducing a big time bending tendancy on the link. Not good, especially for the stamped stuff that GM used, and even worse considering the fact that GM put TWO of those little suckers up there to react to torque, but now the load would be handled by only one: not good.

This does not even come close to addressing the kinematics of the system, most notably the "anti's," as in squat and lift.

All that said, you have an advantage over the leaf spring guys, there is attachement potential on top of the dif' from the upper mounts (again, reference EVM for ideas). With a 12 bolt leaf differential, making a suitable upper mount looks a bit challenging at first glance.

Norm Peterson's late model 'bu uses the same basic rear suspension approach, I would guess that he has thought of this/talked to others that have tried, hopefully he'll drop into this thread. My opinion, nothing good can come of that attempt. (By the way, not sure if you mentioned it, but the upper pair is responsible for lateral location, so you would NEED!!!!! a different means of locating the axle laterally, Panhard or Watt's, for instance).

Mark

ddennis68
Registered User
Posts: 266
(7/25/04 4:29 pm)
Reply Re: No!!!!
--------------------------------------------------------------------------------
I'm sorry, I thought I had stated that I would be running a PHB in place of the UCA's for lateral positioning. Also , the UCA's are not really that canted-I think the UCA mounts where they are in conjunction with a solid ended UCA would provide more than adequite strength for the differential "twist" during throttle/brake. I would NEVER even consider doing this with the stamped factory peices-I would use some heavy tube style arms with heim ends or johnny joints. I have looked at some the factory attempts at a 3-link type suspension and my ideas are much stronger than say the Jeep which doesnt even positively mount the 3rd link to the housing-it has a ball joint type mount running from the center of the housing to both the R & L sides of the rear frame with rubber bushings attaching. The mount is constructed out of 1/4 stamped steel.
Dennis-

check out progress of Bondobucket

o1mrquick
Registered User
Posts: 2036
(7/25/04 8:25 pm)
Reply 3 link revisited?
--------------------------------------------------------------------------------
Hey Dennis, when I was building my GTO I was playin with ideas of installing a bolt on 3 link that uses most of the stock UCA mounting holes. I had a bracket that bolted to the mounts then spanned across the frame that had a kick down for the upper link and dampner mount.The LCA mounts were extended foreward to utilize a 26" johnny joint equipt 1 1/2" DOM LCA. I sent the plans to a guy and he said it would work out nicely.The adjusable PHB mounts at rear with custom bracket that would also mount the LCA which was also adjustable.I later found that the stock 4 link design is more than enough for the way I drive. It might give you some good ideas. C-yaaa Vince

ddennis68
Registered User
Posts: 268
(7/25/04 8:47 pm)
Reply Re: 3 link revisited?
--------------------------------------------------------------------------------
Thanks Vince-still have those plans by chance?

Dennis-

check out progress of Bondobucket

MarkM68
Registered User
Posts: 1617
(7/26/04 6:41 am)
Reply 3 link
--------------------------------------------------------------------------------
I'm really considering doing this on my '66. Since I'm going to mini-tub the car using the stock frame. I'd just remove the uca crossmember all together, and use a more solid c/m.

mean69, do you have a contact email?

katz
Unregistered User
(7/26/04 7:55 am)
Reply re
--------------------------------------------------------------------------------
Here's a pic of revised 3-link/PHB on my Riviera (I'd call it project Banana Boat for now b/c of its ugly yellow paint job). The car originally came with 3-link/PHB from factory, but the geometry is crappy at best especially with 4" lowered ride height. Factory LCA packaging is similar to A/G-body sprayed 4-links.

New geometry is pretty good considering I retained all factory chassis mounting points. I know it is similar to what Mean69 is doing and he's planning to market his as a kit, so I won't go into further details. But I'll tell you it is very stable, and traction during out-of-corner acceleration is phenominal. You wouldn't believe you're driving a 4200-lb full size car. Ride quality is a bit stiffer and rod ends transmits more vibrations, but it's still pretty good overall.

Mean 69
Registered User
Posts: 82
(7/26/04 9:26 am)
Reply Yeah Baby!
--------------------------------------------------------------------------------
Now that's what I'm talking about! Very cool, Katz, we'd expect nothing less from you! One of the things that is pretty common to do with the upper link is to offset it as shown, relative to vehicle centerline. It is pretty well documented that if you place it appropriately, you can compensate for axle torque under acceleration and have a very neutral response under throttle as a result. The downside (reported) is that under braking, the rear will react assymetrically, causing a twitching effect. I have never driven a car in this orientation, and it clearly is used on many other cars (FFR, for instance), so I question how big of a deal this is. Can you feel any funny business, Katz, under braking?

Does anyone have a similar picture for an early A-body? I had one for years, but never did much with the rear suspension (other than a home-built set of traction bars ala Comp. Engineering, very effective for straight line). I can't recall too well how things are configured.

MarkM: I'll drop you an e-mail.

Mark (M also, interestingly...)

o1mrquick
Registered User
Posts: 2038
(7/26/04 10:54 am)
Reply nice riv
--------------------------------------------------------------------------------
Hey Katz, looking good. I plan on doing the same...can't say more either,but did you have to move them rear ward also? and what do you think about moving the front mounts outwards a bit to get the LCA abit more parallel? Just some funny thoughts I had layin on my back one day. now you got me goin. take er easy ayh C-yaaa vince

Norm Peterson
Registered User
Posts: 158
(7/26/04 10:54 am)
Reply Re: No!!!!
--------------------------------------------------------------------------------
The so-called "Poor Man's 3-Link" was tried, and mostly abandoned by the Mustangers for the reasons that Mark mentions. The whole idea of a 3-link is to get each of the various links to locate along only one principal vehicle axis. If you rod-end the still-splayed 3rd link, bending in the UCA won't be an issue (zero end moments and no intermediate lateral loads equals zero bending everywhere). Leave them softish OE rubber and there will be some bending, and it will be about double what it was in either UCA when both UCA's were in place. But the lateral component of the PM3L and the now-necessary PHB or WL will get into some unsymmetrical dispute over the lateral loads. Enough to affect the axle roll steer and pinion angle, I think.

As a side note of sorts, you probably hear more about Fox and later Mustang owners completely revising their rear suspension configurations than 3rd and 4th gen F-body folks in part because the torque arm-equipped F-cars didn't need nearly as much help for extreme cornering applications.

I also know specifically of a G-body 3-link under development by another member of this site.

Dennis - last I knew, new membership at cc had been re-opened and has remained so at least into July. However, I do seem to recall that registration attempts from certain ISP's were/are specifically rejected (aol might be one of them, but see the Wiki to be sure).

Norm
'79 Malibu (too unstock for Street Mod)

o1mrquick
Registered User
Posts: 2039
(7/26/04 10:56 am)
Reply upper link bracket
--------------------------------------------------------------------------------
Hey dennis i'll get those plans out to you asap im not gonna be at the shop till Sunday so i'll draw something up for you.Just to give you a couple ideas. Vince

katz
Unregistered User
(7/26/04 11:44 am)
Reply re
--------------------------------------------------------------------------------
Mark,
Thanks. The UCA is offset from the factory on this car. This being my daily driver, I have to keep the downtime minimum so I decided to re-use factory chassis mount locations. If I had more time and money, I'd probably done 9" with central UCA. The offset has been decreased by 0.53", however, by mounting a narrow-width rod end to the inboard side of factory brackets (the factory bushing measures 3.275").

I haven't experienced any pull or twitchness under braking, but that's probably b/c I can't brake hard enough with factory 4-wheel drums Unfortunately, I won't be able to do disc conversion for a while.

Vince,
I moved back UCA mount on HSG quite a bit. New UCA is 5" longer than stock. I had to put 20-degree bend in it to clear axle housing at full bump, but it should hold up fine with my current setup (about 350hp at flywheel and street radials).

Decreasing the converging angle of LCAs is a good idea. On my setup, the roll axis flips towards oversteer direction on 2" rebound travel when PHB is mounted in bottom two holes (but not by much) due to the large skew. With annual Bonnville trip nearing, I couldn't afford parts to build new LCAs, and as you know, I already spent too much time building the exhaust for this car and I can't move the LCA chassis mounts outboard w/o re-routing the pipes. If I ever decide to convert it to coil overs, then I'll probably make LCAs with rod end on each end and decrease the skew within the existing mounts.

The stock geometry @ 4" lower height was terrible. SVSA length was 40.2" with 14.6% roll understeer. The car used to wander a lot on freeway, which I thought was due to excessive bumpsteer. Now the car tracks pretty straight. It's quite likely that excessive rollsteer and loose PHB bushings were causing the problem.

ddennis68
Registered User
Posts: 270
(7/26/04 12:32 pm)
Reply Norm
--------------------------------------------------------------------------------
So Norm, I am guessing the biggest issue then is the that the 3rd link is not parallel to vehicle centerline. If the upper crossmember were removed and another built that would provide a different angle would the angle of the LCA as they are become an issue? Also, it is my understanding that the longer the better with the LCA. Since it is an El Camino and intrusion into the cabin area is not an issue, should the 3rd link be extended as well, I know that the 3rd link needs to be 65% of the LCA but is there a limit to how long it should be? Are you suggesting NOT to run solid ends on the UCA(3rd link)? Lastly, as I understand it the 3rd link should run from the diff housing DOWN to the crossmember, this in contrast to how the factory built the half-ass 4-link currently in place. I also understand that adjustments to that angle make a significant difference in how the rear suspension responds (often used to tighten/loosen the rear-end), is there a formula for figuring out a starting point for this angle?

Thanks for the enrollement tip, maybe I'll try again. I'm sure the "big meanie heads" would love some fresh meat to tear up.

Edited by: ddennis68 at: 7/26/04 12:37 pm

Norm Peterson
Registered User
Posts: 159
(7/26/04 2:32 pm)
Reply Re: Norm
--------------------------------------------------------------------------------

Quote:
--------------------------------------------------------------------------------
So Norm, I am guessing the biggest issue then is the that the 3rd link is not parallel to vehicle centerline. If the upper crossmember were removed and another built that would provide a different angle would the angle of the LCA as they are become an issue?
--------------------------------------------------------------------------------

LCA skew already is an issue, as the plan view intersection point of the LCA's goes to defining the axle's own roll axis (in conjunction with the PHB/WL/whatever). There's probably some optimum range of skew angle for the LCA's too.

Zero skew (parallel) LCA's have their virtual intersection point located at infinity, hence the axle's own roll axis is always parallel to the LCA's. 3rd & 4th gen F-bodies use this geometry. As the chassis rises into rebound to put the pivot at its end of the LCA above the axle pivot, you move into vehicle oversteer on a one-to-one basis with the side view LCA inclination. So a 2" chassis rise above level for 20" long (plan view parallel) LCA's gives 10% roll oversteer (neglecting any effects due to changes in chassis rake). Here's where the LCA length comes in most strongly - make them 25" long and that roll steer figure drops to 8%.

Most recent OE converging (triangulated) 4-link LCA's are splayed at something like 18* or so. The ones those shown in katz's picture above look to be more like 30* (correction requested). A little converging skew angle in the LCA's toward the chassis ends takes away some of the sensitivity of rollsteer % to ride height, which is arguably a good thing.

To put some numbers on the steer percentages, I have a fairly old paper written by a person then employed at Ford that gets into this a bit and ultimately concludes that 7% roll understeer adversely affects slalom performance. With my G-body being somewhere in the 7% - 8% range empty but with a light trunk load and being a bit clumsy through Solo II slaloms (me in it, otherwise empty) I'm inclined to agree.

Quote:
--------------------------------------------------------------------------------
Are you suggesting NOT to run solid ends on the UCA(3rd link)?
--------------------------------------------------------------------------------

Not at all. That was to illustrate that bending stresses in the UCA can be eliminated even if the UCA was to be left splayed. That part is fine with solid ends. Not so benign is the lateral force component of the UCA splay, as it and the PHB's axial load will then have different ideas concerning axle movement during roll to accommodate the various arcs. The result is bind and/or *unexpected* axle movements, and gets worse as the link end connections are made more rigid along the 3rd link's axis. Make the UCA parallel to centerline with rod ends and it won't be fighting with the PHB. But make the UCA bushings unmodified one-piece poly and you'll introduce bind even through a parallel UCA.

Quote:
--------------------------------------------------------------------------------
Lastly, as I understand it the 3rd link should run from the diff housing DOWN to the crossmember, this in contrast to how the factory built the half-ass 4-link currently in place. I also understand that adjustments to that angle make a significant difference in how the rear suspension responds (often used to tighten/loosen the rear-end), is there a formula for figuring out a starting point for this angle?
--------------------------------------------------------------------------------

With a 3-link, the side view inclination of the 3rd link only affects anti-squat. In the OE 4-link it also affects the axle's own roll axis somewhat.

In the 3-link/PHB setup, the construction line for axle steer runs between the center of the PHB and the intersection point of the LCA's projected into side view. Statically, anyway.

Norm
'79 Malibu (too unstock for Street Mod)

ddennis68
Registered User
Posts: 271
(7/26/04 2:56 pm)
Reply Re: Norm
--------------------------------------------------------------------------------
-hanging my head in shame-
Norm, I am embarrassed to say I barely understand half of what is stated above, altough it sounds good. I am running out right now to buy "Suspension Design for Dummies" or what ever suspension related book I can get my hands on at the local bookstore-then I'll order everything on the net I can find. Scary, I've been a journeyman tech for over a decade and I am considered the suspension/alignment specialist at most of the dealerships I have worked at. I guess I am good at fooling alot of people because apperently I don't know @#%$, which doesn't say much for those I have had to help over the years. Thank you for taking the time to help me sort this out. After I educate myself alittle perhaps I'll get back with everybody on what I decide to do. Thanks again-
Dennis-

check out progress of Bondobucket

o1mrquick
Registered User
Posts: 2041
(7/26/04 4:13 pm)
Reply hey now
--------------------------------------------------------------------------------
Don't feel bad Den, I was in your shoes about a year ago but after lots of reading and advice by several members here I have just now tighten my grasp of this stuff.Several books recommended by Norm,Katz and Jon have opened my eyes and it is actually very interesting.Definately buy the books and your brain will follow. Vince

katz
Unregistered User
(7/26/04 4:28 pm)
Reply re
--------------------------------------------------------------------------------
First of all, correction on what I said:
Decreased converging angle will actually increase roll oversteer % on rebound travel on my setup. If LCAs' sideview angle is relatively steep (I'm talking about 3~4 degrees here) at ride height, then large skew seems to help decrease the effect. But as the skew angle increases, LCAs would be subjected to increased bending force under acceleration and effective length of LCA decreases. So there definetely is a point of diminishing return. RC height is also a factor here. The combination of low RC height and steep LCA angle on my suspension is what's causing the roll axis flip.

If LCAs are close to level at ride height and/or RC is high, smaller converging angle may be more desirable.

Norm,
The skew angle of my LCAs are about 42.6 degrees (21.3 degrees per side). It was probably done so due to X-shaped or Hour-glass shaped frame. LCAs are 24" long, but the effective length is more like 23" due to large skew.

As Norm said, converging angle of UCAs/LCAs define roll axis & RC height on sprayed 4-links. My guess is GM pointed up the UCAs towards front in order to reduce RC height and roll steer. But this isn't really desirable for traction as it locates IC behind the rear axle, which theoretically will have negative anti-squat (or pro-squat). If you understand this, you'll understand why AME's tri. 4-bar is designed the way it is, unlike typical street rod tri. 4-bars on market. I did the best within parameters given by my bosses.

The beauty of 3-link is that, one thing can be adjusted w/o causing drastic effects on something else. I can change IC and SVSA length w/o changing RC height and roll steer, for example. This simply isn't possible with sprayed 4-links.

Dennis,
As you found out, engineering that goes into seemingly simple link suspension is actually pretty damn complex, and this is only a part of the game. Once you design and build suspension with good geometry, you'll have to go through tuning springs/shocks/anti-roll bars/tires etc. so everything works in harmony. Gather info from whatever sources (credible sources) you can find. CC.com is VERY good. Don't let "beginner" thing bother you and keep you from learning. Pay close attention to posts by ROUSHGTR, Norm, Mark (Mean 69), Preston, and many other very sharp folks. Read everything, then re-read everything again. Repeat if necessary. Then you'll know what to do. Good luck to you.

streetfytr68
Registered User
Posts: 1223
(7/26/04 4:42 pm)
Reply Re: hey now
--------------------------------------------------------------------------------
I'm with you Dennis--but I'm getting there. No that I have settled on a ride height, I started by plotting out my rear axle in relation to my rear frame rails (kick up) to get a sense of where the LCA pickup points might be located (at both ends) From there you can conceptualize whether or not the LCA's will be parallel or angled. (I think?) Get the LCA's figured out first and then you can build the third link to suit. In your case the LCA's are carved in stone (or are they??) So you can measure their length and inclination and then cut a great big hole in the bed of your truck to figure out the 3rd link.

I'm thinking about a Watts link to avoid the exhaust packaging problems associated with a PHB on a 1st gen F-Body (Much easier on that big ol' Riv or an Elco). I have eliminated my back seat so the 3rd link can be located as needed.

Front suspension is done (or close). So once I get out west, I want to try Chicane's leaf springs. Only reason why I'll yank the leaf springs is because I want to cut the crusty NY floors out. At that point I'm sure I'll get carried away and go 3 Link. Only after that upgrade will I cease to consider myself entry-level.
/Steevo

p.s. Vince Asaro has been their done that with his full frame Elco. A de-coupled torque-arm (too clunky) replaced with a 3-link. The car is radically different than stock but, he should be able to provide some valuable driving feedback if you can get a hold of him.

www.lateral-g.net

MarcusUSA1
Registered User
Posts: 141
(7/26/04 6:37 pm)
Reply Re: hey now
--------------------------------------------------------------------------------
The technical aspects have been pretty well covered but I`ll add a little practical info. We`ve (Savitske Classic & Custom/ SC&C) been developing a 3 link package for G body (and later on A body) rear suspensions. We`ve intentionally kept it simple to keep the price down. The LCAs remain where they are,in the factory mounts. The UCAs come off and a mount is bolted in between the factory UCA mounting "ears" on the rear. It`s centered on the diff (see above for reasons). The frame side is mounted to a fabbed mount welded into the stock crossmember,which is reinforced. The mounts for both ends of the 3rd link have several mounting holes to allow a lot of adjustment for antisquat. The weld in PHB bar mounts also have several holes to allow RC height adjustment. These are basically out of the box stock cars parts. It`s all mounted in a caged `87 Cutlass with C-5 based front suspension,full cage,LS1/6 speed and adj. QA1 shocks. We`ve learned some interesting things since installing the new 3 link/PHB out back. The car has a totally different feel now. It used have a vague, twitchy feeling in the rear probably caused by the non linear binding of the factory C4L rear suspension. The rear end now feels planted solid as a rock. It`s very confidence inspiring in the driver`s seat. You can drive the car harder and faster with less butterflys in your stomach. That`s saying a lot because this car had very high cornering limits before. Not very scientific I know but this is what we`re really after isn`t it? It was designed in Suspension Analyzer and on stacks of spread sheets but that`s what it boils down to. The adjustable antisquat has worked out well too. We`ve run as much as 200% antisquat with no rear brake lockup problems (even in the rain). The car hooks up HARD now. Big improvement,it used to have both traction and wheel hop problems. That said the much lower rear RC (9"-12" vs. 18" stock) is causing more understeer at the limit. It wants more rear spring rate and more rear anti roll bar. We`ve been running 167lb/in rears with a 7/8"OD rear bar. We`ve got 250lb/in springs on (back) order and we`re going to a fabbed decoupled (from the LCAs) rear swaybar. The stock G and A style rear bars have a much lower rate than you might think due to the huge length of the arms (LCAs in this case). We`re going tubular with splined arms so it`ll be lighter than stock. The torsion bar is a custom order piece we`re still waiting for but even without it the 3 link/PHB has proven itself enough to keep going with the R&D process. I`ll post an update when we get the complete setup in place and on the road. Mark SC&C

o1mrquick
Registered User
Posts: 2043
(7/26/04 6:42 pm)
Reply 3link
--------------------------------------------------------------------------------
Hey Marcus would you beable to post pictures of your 3 link design, would love to see what you have come up with. Thanks Vince

ddennis68
Registered User
Posts: 272
(7/26/04 6:44 pm)
Reply Re: hey now
--------------------------------------------------------------------------------
Thanks Marcus, kats?, Mark, Steve, Norm, Vince, and anybody else I left off my list! I went by every book store and auto part store in town-no luck so I bit the bullet and ordered:

"Engineer to Win" by Carroll Smith

"Race Car Engineering and Mechanics" by Paul Valkenburg

"Chassis Engineering/Chassis Design, Building & Tuning for High Performance Handling" by Herb Adams

On Amazon-give me a few weeks to receive and then digest the vast information before I come up with anymore ass-clown type ideas-see I do learn things over at c-c.com
Dennis-

check out progress of Bondobucket

Edited by: ddennis68 at: 7/26/04 6:45 pm
08-16-2004
chicane67

Page two

Norm Peterson
Registered User
Posts: 160
(7/26/04 7:02 pm)
Reply No need to hang your head, Dennis . . .
--------------------------------------------------------------------------------
I was generally aware of your occupation, but not the specifics (profiles are there for a reason. so I'm apt to peek). But to reinforce what Vince has said, there's no need to feel bad.

At the service end of the automotive business, I'll bet that you haven't had the opportunity to see even a summary of the engineering tech that went on during the basic vehicle design, let alone any of the details. It hasn't been necessary for service to know that stuff in order to perform a quality alignment, and the effects of moving the pivots a little bit this way and that to physically accomplish the alignment has only minor effects on the basic geometry as long as you stay within recommended ranges. Even the recent technical articles in "Brake & Front End" are barely scratching the surface of what's going on at the design level. As a heads-up, this all may be in for a little shift though, depending on the overall volume of lowered cars, raised trucks, and maybe state inspection attitudes toward such vehicles.

My book recommendation starts with Fred Puhn's "How To Make Your Car Handle". Softcover, and should still be available, with enough math to be useful without being overwhelming. I think the full book list was posted back around March of this year.

FWIW, I spent some time last year and early this year between engineering jobs working in the service department of a local Chevy dealership, so I've seen the industry from the inside a little myself.

Norm
'79 Malibu (too unstock for Street Mod)

Mean 69
Registered User
Posts: 83
(7/26/04 7:21 pm)
Reply Nice!!!!
--------------------------------------------------------------------------------
Hey Marcus,

Sounds like you guys have your hands full, great project! Congrats! I have a question, that if you don't care to answer due to confidentiality, I totally understand, but:

"The adjustable antisquat has worked out well too. We`ve run as much as 200% antisquat with no rear brake lockup problems (even in the rain)."

That is awesome. Knowing basically how things work out with the packaging restraints, this begs the question for me, what is the approximate SVSA in this configuration? For those of you that have seen the infamous TCP thread on CC, you will note that the SVSA is pretty short on their awesome race car, which would imply that there may be issues with brake hop in the rear, but their car didn't experience this in practice. I trust this statement in that I know a guy that actually drove their car in an AI event, cold, just got in and drove, and claimed the car was incredible.

Without having real data and experience, the setup I designed and am prototyping has a ~70 SVSA at 100% anti-squat. I wanted to try to keep it relatively long in hopes of not meeting Mr. Axle Tramp.

Dennis: Good for you. Herb's book is outstanding, and believe me, it is going to take SEVERAL read-throughs before the info really takes set. But if you are interested enough to read and try to understand some of the details in this post, then you are interested enough to understand WHY things are the way they are. And you, my friend, represent the minority in this sport/obsession. That minority is growing though, and it is a really good thing. Keep asking the questions! And another vote goes for Alan Staniforth's "Competition Car Suspension," or something like that. He's a brit, and this book contains the best explanation of weight distribution I have read to date, but bring the calculator, lot's of calc's.

Oh, and by the way, Norm, no one who knows me would ever call me a kiss up (quite the contrary, actually), and that's not the intention here, but you have a great talent for writing, and expressing ideas. I actually drafted a similar type response regarding the roll axis stuff the other day, but re-read it and tossed it in favor of another weekend beer. Nicely done on your account.

So, then, if we've been paying attention, what is the moral of the roll axis riddle for linked suspensions? Hmm? Anyone? Bueller? DEFINE the roll axis by creating TWO lateral restraint points, rather than letting one "float" from parallel (plan view) LCA's. Okay, so technically the roll axis is defined for parallel LCA's, but heck, map things out, you can't get a whole bunch of anti-squat with a moderate-to-long SVSA and still preserve favorable roll steer characterisitics if you keep the lower arms parallel. Ahhhhh, the compromises we must make.

Guys, I can honestly state that I have spent more time here, than I have at cc, in the last several weeks. There is some great tech going down, and some very interested people, nothing but good stuff all around.

Mark

yody
Registered User
Posts: 316
(7/26/04 11:50 pm)
Reply Re: 3-link
--------------------------------------------------------------------------------
you guys should put togehter a class for all this stuff, make a small fee for the class and put it towards the site! i would come if it wasn't too far! I find myself scanning quickly through all the technical stuff, but i bet myself i will be reading everysingle word with a magnifying glass when i come around to actually changing the rear suspension on my bird.
firebird website

jon
Unregistered User
(7/27/04 7:52 am)
Reply x
--------------------------------------------------------------------------------
just to (over?)simplify things a basic 3 link plan is 24" lca's,parallel with each other and the ground,18" upper,centered and angled downhill 5-7 degrees,phb adustable from axle centerline to 5" or so below.am i missing anything?starting with a 9" rear or quick change would help as there are upper link brackets available from many sources.all the trailing arm and phb brackets you could want are also available from many sources.roundy-round cars are commonly built,with 3 links,on garage floors by average joe kinda guys.don't dismiss the speeds those guys do,it LOOKS easy til you get out there.i think the technical term for the speeds they run is'holy crap".about the same as running 100 mph in a mcdonalds parking lot.

Norm Peterson
Registered User
Posts: 162
(7/27/04 8:50 am)
Reply Re: x
--------------------------------------------------------------------------------
The actual elevations of the axle end pivots of the UCA and LCA's need to be known (to determine the SVIC, SVSA length, and anti-squat%), as does the axle centerline height (for determining the PHB elevation, or just provide the PHB elevation).

Meaning no disrespect whatsoever to the circle track racers (they do an amazing job, and I have plans involving some CT-origin parts in the near future), but it certainly has to help the design and fabrication processes to have peeked around the pits and found out what dimensions more or less work.

A street-oriented ride will have to work over a wider range of loading and road conditions than a race car and most folks don't want to have to sort through all the combinations of four of five sets of springs and a couple of bars until satisfied. So in that respect, building a street car can actually be more demanding of thought.

Try to narrow the range down a bit on the PHB first, as that goes toward the selection of springs and bar (and, by implication, possibly shocks).

Norm
'79 Malibu (too unstock for Street Mod)

Edited by: Norm Peterson at: 7/27/04 8:52 am

ddennis68
Registered User
Posts: 277
(7/27/04 10:22 am)
Reply PHBq
--------------------------------------------------------------------------------
I should just wait for library to arrive, but too exited. Does it matter which side of the axle the PHB attaches to, and if so why?

Mean 69
Registered User
Posts: 84
(7/27/04 10:49 am)
Reply Hand Calc's
--------------------------------------------------------------------------------
Jon, I could whip out a really quick and dirty anti-squat and SVSA calcualtion if you'd like for this. I have some notes from my car with me at work today, I don't think it would be terribly difficult to spend ten minutes on getting a very basic first order answer. You'll need to provide the height of the LCA's above ground, the height above ground for the rear link of the UCA too. I'll just assume an 18" CG height, as this will be for an estimate only, but will give you an idea of what you may expect.

It's a first gen Camaro, right? If not, what's the wheelbase?

Dennis, there is a theoretical advantage for having the frame mount on the driver's side of the vehicle (90% sure, but I may have this backwards), due to a reaction of axle torque under acceleration. I know the circle track guys are hot on this, but in truth, I doubt it is really that big of a deal, especially on a road car. The main design considerations for a PHB are to make the bar as long as possible given packaging constraints, and to keep it level at ride height. The brackets need to be very strong, ALL of the lateral forces will be reacted through this device in a parallel lower link configuration. In addition, it is favorable to have the mounts for the bar be symmetric about the centerline of the vehicle, but not completely required. Simple system, they work really well. Ever consider a Watt's instead?

Mark

ddennis68
Registered User
Posts: 278
(7/27/04 12:30 pm)
Reply Re: Hand Calc's
--------------------------------------------------------------------------------
watts link, hmm....I don't have much knowledge of these (not that I know cr@p anyways, but I really don't know these) but have heard that they are somewhat pricey-I can slap a PHB on for about 100.00 including the scrap metal to fab up some brackets. They look real neatoo though

Maybe when books show up and I have a better understanding of the geometry involved and how each component is responsible for controlling what I will look into that as a possibuility. Thanks for the input.
Dennis-

check out progress of Bondobucket

Davezz28
Registered User
Posts: 8
(7/27/04 1:03 pm)
Reply UCA / LCA length ratio
--------------------------------------------------------------------------------
Hello

Good to see a lot of talk here about 3 link susp. I am in the process of setting up my 69 Camaro with a 3 link and have most of the ideas drawn up. My main question to any of you that may have a idea is: Is there a common excepted ratio between the length of the UCA and LCA.

My current design has a LCA of aprox 24" and an upper of around 12" so the ratio would be 50%. SVSA is around 60" and anti squat is around 40%, if I WAGed the CG height correctly.

It seams to me that a short upper arm will induce radical pinion angle changes as the axle travels vertically.

Dave

Norm Peterson
Registered User
Posts: 164
(7/27/04 1:08 pm)
Reply Re: Hand Calc's
--------------------------------------------------------------------------------
A driver's side PHB chassis attachment will result in the rear RCH dropping in right turns. This has the effect of reducing the total rear lateral load transfer slightly (which includes that due to the pinion trying to climb up the ring gear and also unloading the RR as you accelerate on on corner exit. In left turns, the RC rises, and the greater lateral load transfer helps plant the RR against that same R&P effect. In both cases, the tire loadings are less different than they'd be with a passenger-side pickup, meaning slightly better overall rear grip. Basically, you're using the PHB's asymmetry to work at least partially against the R&P's always-asymmetric effect.

That's more desireable for street, autocross, and road course duty. Seems it was just mentioned on the broadcast for the NASCAR road race a couple of weeks ago that even Cup cars use a driver side PHB pickup on the road-race chassis. But for left-turning only, there may be advantages to a passenger side chassis attachment (depending on other specifics).

Norm
'79 Malibu (too unstock for Street Mod)

Edited by: Norm Peterson at: 7/27/04 1:13 pm

katz
Unregistered User
(7/27/04 1:12 pm)
Reply re
--------------------------------------------------------------------------------
It would also help to know where UCA is located longitudinally as SVSA gets shorter when you move it back (for a given pivot height and angle), though most CT stuff seem to have central UCA with pivot located very close to axle centerline.

Like Mark said, there's an advantage of mounting PHB in driver side frame/passenger side HSG orientation unless you're running a reverse-rotation motor. PHB, LCA sideview angle, and top view skew (if there's any) define RC height and roll axis. PHB brackets usually end up really long as you can see in my pic if you were to locate RC at good height. That intorduces lots of leverage so brackets need to be pretty stout.

Watt's don't have the advantage of PHB's assymetricity for counter-acting driveline torque, but it has its own big advantage (IMO). I would have done Watt's on my car if I could cough up money to buy bell crank bearing and extra two rod ends. If you do Watt's, be sure to mount the bell crank on chassis.

Now that Mark let the cat outta bag, I can tell you how mine is set up. I also stayed somewhat conservative. I'm not quite as comfortable designing rear suspension as SLA IFS, mostly due to lack of experience. SVSA is 69-ish, anti-squat is roughly 85%, and 25% anti-lift (30% rear brake bias). Roll steer is 0.9% understeer at 10" RC height. My guesstimated CG height is probably off slightly. The roof of my car is lower than most 1st gen F-bodies, and I really don't feel much squat. PHB ended up at 36.5". A bit shorter than I wanted, but it's still 4.5" longer than stock PHB.

Like I said, I'm more than happy with the current conservative setup. But after reading Marcus' 200% AS susp. and TCP's racecar, I'm tempted to try 42" SVSA and boost AS to 112%. Once I get better brakes and front suspension, I gotta attend HPDE type event.

streetfytr68
Registered User
Posts: 1235
(7/27/04 3:15 pm)
Reply Re: re
--------------------------------------------------------------------------------
I'm 700 mi away from my Camaro right now--otherwise I'd be posting from under my car with measuring tape in hand. So I've got questions:
1) Are level LCA's feasable on a 1st gen F without modifying the frame rails or hanging too low or sacrificing bracket strength?
2) Rear LCA pickup point should be under the axle at centerline. What if pivot point were slightly forward of axle centerline so that I could raise it some?

www.lateral-g.net

Mean 69
Registered User
Posts: 86
(7/27/04 4:11 pm)
Reply Input
--------------------------------------------------------------------------------
It "feels" to me that a 12" link is too short, or rather, as you stated, would induce some pretty aggressive pinion angle changes in bump etc, as well as IC migration, quick changes in anti-squat, etc. Point of clarification, is this the effective length of the upper link, meaning from pivot point to pivot point, or is this the length of tube (minus linkage ends, for instance)?

Steevo, the answer to both of your questions is yes (most likely), but I don't feel that you'd want to do either. Using the forward spring pocket as a placement seems to be a good idea at first glance, but it is pretty high off of the ground. And by NO means would I use the stock spring mount as a LCA mount. Second, placing the rear mount of the link CLOSER to axle centerline increases the radial loading that the rod end/linkage would have to deal with, for a given torque, scaling linearly with a decrease in offset. It means your parts will wear faster.

If the main question is one of ground clearance, ask yourself this question: If I ran slapper type traction bars on my car with leafs, would I be worried about ground clearance? I'm not, but that's just me.

Katz, those numbers look really good given that there wasn't "too" much done to arrive at them. Nice work! I will agree also, the TCP car specifically had a huge roll understeer axis, super low rear roll center, and a really short SVSA, but still seemed to work really well. Go figure. Still though, that car also had a NASCAR style cage, so if bad stuff happened, at least there would be that to help keep the driver safe. Hmmm....

M

MarkM68
Registered User
Posts: 1622
(7/27/04 4:54 pm)
Reply .
--------------------------------------------------------------------------------
Dennis, here is an good article to get you started.

www.rodandcustommagazine...._0401_ifs/

jon
Unregistered User
(7/27/04 6:59 pm)
Reply x
--------------------------------------------------------------------------------
i think maybe this is being made a little more complex than it needs to be.why not just call coleman,order a housing with trailing arm/coilover mounts,phb mount,and uca mount already installed,and use the same geometry as countless short track cars,which do make good road course cars in general.you can buy blueprints to do you own frame,or order pre made front and rear clips.i built my front clip myself on the shop floor and i think it holds it's own against the kind of cars i tend to play with.i run leafs and feel my car is well developed -enough to stop messing with it too much,anyway.

streetfytr68
Registered User
Posts: 1237
(7/27/04 7:09 pm)
Reply Re: Input
--------------------------------------------------------------------------------

Quote:
--------------------------------------------------------------------------------
Using the forward spring pocket as a placement seems to be a good idea at first glance, but it is pretty high off of the ground. And by NO means would I use the stock spring mount as a LCA mount.
--------------------------------------------------------------------------------

Mark,
I don't want to use the leaf spring pocket. I'm just trying to picture where along the stock frame rail to locate a new mount. I'm picturing that I would hang a mount straight down from the frame rail. Now draw a line straight back to 2" below the axle centerline. It seems that forward pivot would have to hang pretty low in order for that LCA to be level to the ground (and in order to allow for travel w/o the arm being too short). See?
So'z I was trying to raise the rear mount to compensate--but I understand why that wouldn't work based on your reply.

I'm trying to maintain the stock frame rails as well as the present ride height (axle location).

www.lateral-g.net

MarcusUSA1
Registered User
Posts: 142
(7/27/04 7:31 pm)
Reply Re: .
--------------------------------------------------------------------------------
I`m not ready to post pics yet,it`s still in the working prototype stage. The geometry is close to production but the brackets and small details are different (it`s not pretty yet). The SVSA length is 46" at 100% anti-squat. At 150% AS it`s 31". It`s rediculously short at 200% and it shouldn`t work at all. It was an experiment just to see what would happen and it worked out better than it had a right to. That said it was ONLY an experiment and I wouldn`t recomend leaving it like that. Currently it`s got over 1000 (hard!) street miles on it set at 150% AS and it works really well. The rest of the car`s setup is probably a big part of it. Right now the front is much stiffer than the rear. The front end has little dive under braking and the rear is running shorter than stock springs which have a fairly rapid rate drop in rebound. This much AS % and this short of a SVSA may not work well at all in a differently setup chassis. Being a test mule,the car`s gone through several stages and the rear is just starting to catch up to the front. We`ll have a better indication once the whole car is sorted out better. The amazing thing to me is how well the car works even at this point. Mark SC&C

ddennis68
Registered User
Posts: 281
(7/27/04 8:01 pm)
Reply Re: .
--------------------------------------------------------------------------------
Thanks Mark-I'm a little past that, not much though. Got just enough info bouncing around between my ears to be dangerous. Give me 2-3 weeks and "somebody stop me", I'll be outta control.

Damnit Mark (the other Mark), now I'm all over the net looking for watts link info.

Now that you've got me pulling what very little hair I have out, can you direct me to an outfit that will sell a partial "kit". I can't even find a bellcrank-except for racefab inc, they have a kit for 830.00
Dennis-

check out progress of Bondobucket

Edited by: ddennis68 at: 7/27/04 9:29 pm

o1mrquick
Registered User
Posts: 2051
(7/27/04 11:36 pm)
Reply front LCA mount
--------------------------------------------------------------------------------
Hey Steve, I've been planning on using a 2 x 3 .120 wall cross member for the front LCA mount.It mounts flush with the bottom of the existing frame so my LCA's will run alittle higher than the leaf springs in stock location.The axle mount will feature dual mount so I can experiment with different LCA lengths and angles. On my Camaro it will run a 1/2" foreward of the spring eye so I can start with a 24" LCA. On the Chevy II im doing, it will run across at the same legnth as the front spring eye mounting. Just something I wanted to throw out there instead of front perch mounting or a lower one.
Dennis the watts linkage does impose exhaust routing problems as well as being alittle more exspensive but it is easier on tires than a PHB. C-yaaa Vince

streetfytr68
Registered User
Posts: 1240
(7/28/04 3:32 am)
Reply Re: front LCA mount
--------------------------------------------------------------------------------
Vince,
That's what I'm looking for. Will your forward pivot points be located behind the crossmember or below?

www.lateral-g.net
08-16-2004
chicane67

Page three

Davezz28
Registered User
Posts: 9
(7/28/04 6:56 am)
Reply Re: Input Mean69
--------------------------------------------------------------------------------
The lenght of the UCA us the effective length (eye to eye of the rod).

I found a comment by katz in another thread and he suggested that a geometry be used that will maintain a SVSA of at least 40-45" at full bump. I went back to the simulator and ran the numbers with the 12 in UCA and the SVSA is too short.

Went back to the car to look for a cure and I can get an extra 2-3" by moving the mount forward. By using a 13.75" effective length UCA & 18" CG height the SVSA in now 71.11" w/ 66.9% AS. With full bump at 3", the SVSA is 42.1" w.80.9% AS.

Do these # sound more reasonable? or do you think a longer rod will create MUCH better geometry.

For ref. the car will be street and road track driven (I live 2hrs from Buttonwillow). Aprox 500-550 hp.

thank

PS I guess I should start a new thread, I feel like I am hijacking and getting multi conversations in one thread.

ddennis68
Registered User
Posts: 282
(7/28/04 7:24 am)
Reply hijack away
--------------------------------------------------------------------------------
Naw go ahead, it's related. Vince exhaust routing isn't a problem-I always dump before the axle anyways

o1mrquick
Registered User
Posts: 2056
(7/28/04 8:10 am)
Reply hello jack!
--------------------------------------------------------------------------------
Hi Dave, yeah its related.Anyways its nice to have everything in one post so you wont have to search...im lazy
Dave I wanted to ask you how long your UCA ended up to be.I'll be running a 18" UCA and 24" LCA's.I moved my front UCA mount as far as I could go and the rear (axle side) back 3" behind the axle center line. Havent done any testing yet but i'd like to see if someone eles would run the numbers to see if that would work.
Den, side exhaust is a good thing. c-yaa Vince

Mean 69
Registered User
Posts: 87
(7/28/04 2:06 pm)
Reply Hmm...
--------------------------------------------------------------------------------
Hey Dave,

It's is hard to say, and I'll further qualiofy that by saying I have never driven a car with this setup (which is really what's important in the end). That said, it seems reasonable, and it does not appear that things are moving around really fast from the info you have here. I think that is one of the more important things to consider with most aspects of a suspension system, things should migrate nice and smooth, linearly if at all possibe. Quick jumps in the kinematics through motion can result in sudden, erratic behavior to the driver. No one likes surprises!

I would guess that if you went longer with the upper link, that things slow down even a bit more, it may be worth a try to just move things around a little bit and see how sensitive the various aspects are. Based soley upon the results of other cars with geoemtry on "both sides" of yours, though, I'd say you are in a pretty good starting spot. Never hurts to experiment.

Nice job!
Mark

ddennis68
Registered User
Posts: 284
(7/28/04 2:45 pm)
Reply Another thought
--------------------------------------------------------------------------------
Is there a limit on UCA/LCA arm length as long as the UCA isn't longer than the LCA? My thought is by increasing the the CA lengths to like 48" and 38" the amount arm angle change and therefore pinion angle change would be much less.

I'm ready-hit me.

katz
Unregistered User
(7/28/04 3:54 pm)
Reply re
--------------------------------------------------------------------------------

Quote:
--------------------------------------------------------------------------------
Dennis, here is an good article to get you started.

www.rodandcustommagazine...._0401_ifs/
--------------------------------------------------------------------------------

Hey, that article looks familiar!

I must admit that I haven’t fully grasped the concept of UCA/LCA length ratio on rear suspension yet. One one hand, it’s desirable to minimize the IC migration, pinion angle change and fore/aft movement of rear axle. This calls for long UCA. But on the other hand, it also seems to be desirable to build in progressive anti’s, and that requires shorter UCA. Perhaps the latter isn’t so important on road racecars whose wheel rates are generally high. I actually tried to calculate progression (or digression) of AS on my car taking suspension travel and pitch angle into account, but I haven’t had time to measure front pickup points so I don’t know where the Pitch center is. This brings up yet another concern: migration of Pitch Center. Since it’s nearly impossible to make Front SVSA and Rear SVSA symmetrical, it’s probably difficult to control it, but it should have parabolic path just like RC migration…Damn, it is sooo involved!! Or am I just thinking too hard?

I guess the finding a good compromise is the key just like everything else, and perhaps that’s where 65% (generally accepted number) came from. I just made mine as long as packaging allowed, and it ended up at 66.7%. My SVSA shortens up to 41” or so at 3” bump.

Davezz28,
Your new geometry looks pretty good to me. I wish I could give you a better advice, but rear suspension kinematics is still a bit of mystery to me. I just haven’t had opportunity to spend LOT of time analyzing them (my job revolves mostly around SLA IFS). Keep experimenting, and I will let you know if I discover something.

Mark (Mean 69) and Mark (Marcus), - we seem to have many Marks lately…
Good point about the safety, considering my driving skills I'm thinking about doing a bolt-on roll bar w/ safety harness, but I highly doubt I'll be doing a full cage on this car. I have three adjustment holes on UCA - each hole decrease SVSA by 14" or so. I'll get used to the current setup, and start experimenting. And there’s no way I would dare try SVSA shorter than 40" other than on drag strips, so no worries there. My car doesn't have much power yet, so I don't need much traction aid from AS and I'd rather not to use up torque on excessive (for my car) AS.

Steevo,
LCA don't necessarily have to be level, though doing so seems to minimize the change in slope of roll axis. Mines are pointing up 3.5 degrees towards front at ride height. This was the only way to get good SVSA length AND good AS. Now I was able to do this while maintaining very slight roll steer because of large skew angle. If you're building 3-link from scratch, finding a good skew angle is the key (IMO).

Here's a pic of my rear susp at ride height (I know LCAs are below the scrub line, but new LCAs will be tucked up higher). You can see LCAs are pointing up quite a bit. I intentionally lifted throttle in the middle of a moderate corner (just enuf to make tires squeal) the other day trying to induce trailing throttle oversteer, which is caused by combination of rear tire unloading and/or excessive roll steer (and excessive positive scrub radius on IRS cars). My car just slowed down w/o any drama.

Davezz28
Registered User
Posts: 10
(7/28/04 4:06 pm)
Reply Another thought
--------------------------------------------------------------------------------
I'm not sure what you would gain by extending your UCA/LCA by the same ammount be sides slowing down the rate your pinion changes through the axle travel. You could have a longer UCA than LCA if your geometry is correct, your pinion angle would just move up instead of down as the axle moves up. You can also have a design with 2 UCAs and one lower.

As far as the system I am thinking of building, I plan on using a Chevelle 12 bolt and use the upper ears as a mouning point for the UCA bracket. The UCA attach point will be aprox 3.75" behind the axle center line and centered above the diff.

I plan on fabbing a crossmember that will be welded between the frame rails and go up over the driveshaft tunnel like __/--\__ which the front bracket will be welded to and extend into the rear seat area aprox 2".

I will have to see how much it interfers with the rear seat since I would like to have the rear seat if just for looks.

I have been using this web page to help with geometry calcs.

www.isd623.org/ben/jp/fourlink/

You can enter your measurements and it will calc many perameters and give you a visual picture of the suspension. There is also an excel version you can down load.

I was set on doing a 4 link suspension like the Mule for a long time and about a month ago I really started thinking three link and really want to make it work if I can.

Davezz28
Registered User
Posts: 11
(7/28/04 4:29 pm)
Reply Another Idea
--------------------------------------------------------------------------------
Since we are talking ideas and brain storming here is a thought of another design I was thinking of with 2 UCAs and one LCA

You would have to build a cross member like Anthony Rose did on is Firebird for the front mounting point for the LCA. The axle mounting point would be a bracket welded to the lower radius of the differential if using a Ford 9", or a bracket welded under axle tube offset to pass side if using a 12 bolt. You would have you LCA running next to the drive shaft and may have to box out the tunnel for clearance.

The UCA would be a copy to the UCAs on the Mule. The attach point would be fabbed and welded to the torque box under the seat.

Your rear axle would pivot around the LCA instead of the upper.

I can't see why this system would not work, but I will be the first to admit I don't know what I talking about, I just started contemplating this stuff.

Dave

ddennis68
Registered User
Posts: 285
(7/28/04 5:12 pm)
Reply Re: Another Idea
--------------------------------------------------------------------------------
Dave, the biggest concern I would have with running the 3rd link in LCA position would be packaging. I would think it very difficult to locate the link centrally without having driveline interference issues.

I was thinking today (no, it wasn’t a first nor did I hurt myself) while I was supposed to be working. I happened to have a PT Cruiser on the hoist and HOLY @#%$ this thing has a watts link out back. Well, not one to turn away somebody else’s R&D, I checked out the parts catalog and the bell crank is only 40 bucks. Point is I’ve been looking high and low for a source and here it is if anybody is interested-part#4656465AF. I haven’t actually seen one out of the vehicle yet so I don’t know how durable it is built as far as bearing sizing but I think I’ll order one for Friday so I can see how it looks.
Dennis-

check out progress of Bondobucket

katz
Unregistered User
(7/28/04 5:47 pm)
Reply re
--------------------------------------------------------------------------------
Hmmm, it seems it's more of a matter of roll axis geometry to me.

In most cases, UCA will have steeper static angle than LCAs for good side view geometry (exception is super high CG applications such as rock buggies). This means you'll have fair amount of roll understeer to start with if two UCA's are parallel in plan view. Unless UCAs' axle side pivot height is very close to the axle center (ie, more stress in UCAs), slope would still be too much even if you converge UCAs towards the rear. If you converge them towards front, you'll probably have roll oversteer assuming PHB/Watt's is located at good height.

And for good Anti progression, UCA really should be shorter than LCAs (if not equal length). Shorter link will always have greater change in angle for a given suspension displacement, so roll axis' slope will change more.

Don't mean to disappoint you, but there are reasons why you don't see 2 UCAs/1 LCA configuration.

o1mrquick
Registered User
Posts: 2063
(7/28/04 6:08 pm)
Reply PT
--------------------------------------------------------------------------------
Hey Den, I would not be lookin to use a PT set up, not very HD if you get my drift plus I've seen a few failiures.There is another source however in the Durangos, they are built alittle bit better but I would not promise durability in race usage. Too bad your not using a Mopar diff, that cover is beefy C-yaaa Vince

Davezz28
Registered User
Posts: 12
(7/28/04 6:10 pm)
Reply re
--------------------------------------------------------------------------------
Yea it was just a thought.

I have a hard time understanding the roll over/understeer thing due to rear geometry. Been trying, but can't seem to visualize it yet.

Thank for your help

MarcusUSA1
Registered User
Posts: 144
(7/28/04 7:04 pm)
Reply Re: re
--------------------------------------------------------------------------------
FYI many Lincoln Town Cars and Mercury Marquis from the mid `90s up also have a watts linkage. Interestingly it lays almost flat and is bolted to the top of the pinion snout with it`s pivot axis just above the centerline of the axle tubes. Nicely packaged and made of cast aluminum. Mark SC&C

Mean 69
Registered User
Posts: 88
(7/28/04 7:14 pm)
Reply Visualize this.
--------------------------------------------------------------------------------
Rather than get into the definition of the roll axis, as it has been covered here already, just think of things this way. Imagine the rear axle has a long shaft welded solid to it, that runs down the chassis centerline. Fix both ends of the shaft vertically, so the axle can rotate about it, but the axle can't move in any other direction, like a big cross. The angle of the shaft represents the "roll axis." If the roll axis/shaft is parallel to the ground, and you push up on one wheel, it will lift, vertically, straight up. The other wheel will drop, vertically, straight down. The influence on steering is neutral because the pointing of the tire centerline has not changed, relative to the vehicle centerline.

Now think of the shaft inclined, from the rear of the car, towards the front (i.e. higher in the front). If you lift one wheel in this case, it will move upward, but instead of moving straight up, it will move back a small amount, towards the rear of the car. The amount depends upon the inclination of the roll axis/shaft. Because the axle is rigid, the other wheel will drop, and move a small amount towards the front of the car. If one wheel moves forward relative to the centerline, the other moves back, and they are connected by a rigid longitudinal axle, the tire directions HAVE to be on an angle relative to the vehicle centerline. This is rollsteer, and in this particular orientation, it is called roll oversteer. Incline the shaft/roll axis the other direction, and things swap around.

How does this feel to the driver? Well, not good. It is very much like the back end of the car coming around unexpectedly ( in the case of roll oversteer), the driver feels helpless in trying to control the car (well, most of us would be, the late Mark Donahue was probably an exception). What makes it perhaps even more uncomfortable is the fact that it is variable in regards to roll angle. For instance, if you are loaded in a steady state turn under high loads, and hit a dip. The roll angle will change from the steady state condition reacting to the spring response. Because the roll steer is a function of this, the rear end of the cara will be squirming all over the place, depending upon how steep the roll axis inclination is, and how the roll angles changes.

So, suppose you have an unfavorable roll axis geometry that you can't do much about. Are you stuck? Well, maybe not. One way to reduce the effects is to reduce the amount roll, by INCREASING the roll resistance. You can do this by several means, but for a given vehicle, changing spring rates to stiffer, or by adding more rear sta-bar (this is a function of a fixed rear roll center, by the way, changing the height of the PHB/Watt's, for instance will change the roll axis, that's a whole different story). 'course, if you stiffen the poop out of the springs, the wheel rate will increase, and the car may tend to scamper a bit, not desireable. The better way is to add more rear sta-bar.

But then, alas, you have to determine what the impact of lateral loading will be, relative to the front for instance, and how this will change things.

It's like poking a balloon, trying to make it smaller. You can jab it, but something else will always poke back out at you! But like most things, in moderation, you can deal with the nasties and still get a great outcome.

Time to hit the garage, hope this helps.
Mark

SSWANAB
Registered User
Posts: 102
(7/28/04 8:25 pm)
Reply Re: 3-link
--------------------------------------------------------------------------------
I just started reading this thread, and WOW!. I am at the point of making brackets for the rear end on my 68 Camaro and I just thought I had my mind made up. I have already built a narrowed rear subframe that follows some of this discussion, and I am using a narrowed Chevelle 12 bolt. Keep up this great info! Here is pics of my rear end. There will be a short piece of 2x3 to tie the rear rail to the modified inner rocker which is where the tabs for the LCA will go. I should have at least a 24" LCA. The upper crossmember is 2x3 3/16" thick, but now I have to decide to stay with stock UCA setup or convert to 3 link. Right now it is easy to go either way with a little guidance. The car will be mostly street driven but will see a track whenever I can get near one. Would you stay with the stock (will be using adjustable tubular arms) upper UCA's, or go 3 link? I am a little concerned with packaging the 3 link. I am not willing to give up fuel capacity as I plan to make long haul trips with this car. I am always willing to listen and learn. Keep up the good tech!
MY CAR

ddennis68
Registered User
Posts: 286
(7/28/04 10:44 pm)
Reply Re: 3-link
--------------------------------------------------------------------------------
Ladies and Gentleman, we have a winner. Thanks Vince, I didn’t think the PT pivot would work, but would have been good for R&D purposes. The Durango unit looks like it will work better, although a little more $$$. The Durango unit goes by part# 52113258AB and lists for 92.60. I’ll get one coming so I can check it out. The housing bracket looks good although I’m going to run frame mounted bell crank and mount the links to the differential housing. My research (thanks to mystery expert) has revealed that doing this keep RC at a constant even during “bump”. I’ll post a pic when it comes in, maybe even a few of it on the ‘rango to show how DCX thought it should work.

BTW, gonna have to run the watts setup. After evaluating the PHB geometry I will run into tire to fender interference on bump. I’m only running about ½” clearance on the fenders, I’d like more but not willing to give up footprint.

Dennis-

check out progress of Bondobucket

o1mrquick
Registered User
Posts: 2066
(7/28/04 11:26 pm)
Reply Steves answer
--------------------------------------------------------------------------------
Man this thread is moving fast, Steve the LCA's will mount to fabricated brackets that will be welded behind the crossmember.I would like to mount the x-member and brackets so the LCA's will be a 1/2" lower than the frame, which is about 2-3" above the stock leaf location(at lowest point). Also making height adjustable in a 5" range.
Den Im gonna start with a PHB to start out with. A watts linkage will go on the next car. I feel you gotta crawl before you can run.plus my brackets are already made! Post up pictures of the bell crank so others can see...you going with chassis mounted crank right? C-yaa Vince

streetfytr68
Registered User
Posts: 1246
(7/29/04 3:45 am)
Reply XXX
--------------------------------------------------------------------------------
Xactly!! Are you'z guys all running Xhaust thru the X-member?
/Steevo

www.lateral-g.net

Norm Peterson
Registered User
Posts: 165
(7/29/04 4:00 am)
Reply Re: re
--------------------------------------------------------------------------------
katz - Billy Shope is one individual that I know of who has at least put some serious thought into a 1 LCA/2 UCA arrangement. But that was specifically to simultaneously equalize rear tire loading under acceleration and obtain 100% anti-squat. The same can be done with 2 lowers/1 upper, but IIRC either the link angles or the 3rd link's offset get rather large. On another forum, he presented some discussion complete with formulas for the link inclinations. As you might expect (given his Ramchargers background), this has a hard-core dragstrip slant. But I suspect that the formulas for the 2 LCA/single upper could be massaged to determine how much tire load equalization occurs with specified link inclinations and 3rd link offset.

OTOMH, a single lower had better be sturdy.

Norm
'79 Malibu (too unstock for Street Mod)

MarkM68
Registered User
Posts: 1625
(7/29/04 8:10 am)
Reply .
--------------------------------------------------------------------------------
Katz,

Is it better to mount the watts link center/pivot to the frame of the car, rather then to the rear end?

I was looking at some pics of Smokey Yunicks '66 Chevelle, and the center of the watts link is mounted that way. With the arms extending out to a brackets mounted below the lower control arms. Seems like it would be plenty strong, and would ad less sprung weight. Correct?
08-16-2004
chicane67

Page four

o1mrquick
Registered User
Posts: 2067
(7/29/04 8:17 am)
Reply xcited?
--------------------------------------------------------------------------------
He he yeah I got lazy towards the end. Im gonna cut 3 1/2" holes for my 2 1/2" exhaust to pass. Also grafting a C Alston 1x2"x 12" drive shaft loop in the center with a removable lower plate.There isn't much room so I will most likely run 2 moroso tube muffers infront of the x member and 2 Dyno max ultra flo rounds after the x member, I want a quiet car this time around...for a change. But I will have Lake pipe style cut outs right after the X cross over going into side exhaust. C-yaaa Vince

katz
Unregistered User
(7/29/04 9:28 am)
Reply re
--------------------------------------------------------------------------------
Norm,
I seem to recall browing that thread at CC.com sometime last year. I'll have to check it out again. You're right, single LCA better be really stout especially on drag application.

What are your thoughts on UCA/LCA length ratio? (please see my post, 6th post on this page).

MarkM68,
Ideally, RC should be affixed to CG of sprung mass. This should answer your question. Properly-oriented Watt's does the best job at it, though RC still moves around due to change of the other lateral restraints (LCAs in the case of 1 UCA/2 LCAs 3-link). And yes, less unsprung weight to boot.

Mean 69
Registered User
Posts: 89
(7/29/04 9:31 am)
Reply Roll Center
--------------------------------------------------------------------------------
The difference with running a Watts frame mounted versus differential mounted is the question of roll center definition. With a frame mounted Watt's bellcrank, the roll center is fixed in bump, etc. With the bellcrank mounted to the differential, the rear roll center moves vertically the differntial moves in bump. Either will work really well, but the "preferred" method is to have the Watt's bellcrank fixed in position relative to the sprung mass of the car.

By the way, the roll center for a PHB car also moves, and it kind of splits the difference between the two above. The RRCH for (symmetrically applied) PHB car moves one-half the distance that the differential moves in bump.

Mark

katz
Unregistered User
(7/29/04 9:37 am)
Reply re
--------------------------------------------------------------------------------

Quote:
--------------------------------------------------------------------------------
...(please see my post, 6th post on this page)...
--------------------------------------------------------------------------------

I meant page 3. Sorry!

ddennis68
Registered User
Posts: 287
(7/29/04 10:07 am)
Reply Watts pics
--------------------------------------------------------------------------------
OK, I just ordered both the PT and Durango bellcranks for examination (God I love working at a dealer-sometimes). They will be in tomorrow so this weekend I'll post some pics. I also snapped some photos today of both applications as on their respective chassis'. It seems so simple I wonder why all models are not running this setup.

streetfytr68
Registered User
Posts: 1249
(7/29/04 11:18 am)
Reply Re: Watts pics
--------------------------------------------------------------------------------
Okay now!! You guys are pissing me off!! I get rollsteer with respect to control arm location in plan view. (Mean 69 post) But I don't see how it is affected with respect to control arm inclination from side view? (Katz post??) In fact I'm not even sure if it's affected or if I just misread something.

Was in meetings all morning and was completely distracted. I was answering the wrong questions and everything. So the burden is on you guys to fix my brain.
/S

www.lateral-g.net

katz
Unregistered User
(7/29/04 12:09 pm)
Reply re
--------------------------------------------------------------------------------
Steevo,
Have a donut and relax. I had two this morning - old fashioned and plain, though my favorite is custard filled and crullers (tractor tire looking thing). Mmmmm...

I don't know if I understood what you were really asking, but...

Both plan view and side view angle of LCAs affect roll steer. The axle rolls around the roll axis - the long welded shaft example Mark used. On real 3-link susp, this 'welded shaft' is just a invisible virtual line that is defined by LCAs' top view and side view angle, and PHB/Watt's location.

When LCAs are parallel in plan view, roll axis is parallel to LCA's side view angle. In this case if LCAs are level, roll axis is also level (therefore zero roll steer). However, as soon as susp. moves, LCA sideview angle changes, and you'll get roll steer. Longer the LCA, less angle change for a given wheel displacement and therefore less change in slope of roll axis. I don't know what happens when one LCA goes uphill and the other side goes downhill during a pure roll. I'd suspect that there'll be slight lateral displacement, which forms topview skew angle and instanteneous roll axis is defined by this.

When LCAs are skewed in plan view, it naturally forms intersection point. Below this topview drawing, you have a sideview drawing of equal scale. Let's say LCAs are level in sideview to make things simple. So you have this LCA centerline in sideview, which is level in this case. Now you project a line from the topview converging point of LCAs to the sideview drawing below. The line will intersect with the sideview LCA centerline. This is one of two lateral restrain points, let's call it point A. Now draw a line from this point A to PHB centerline (or Watt's bellcrank centerline), which is the other lateral restrain point B. This A-B line is your roll axis. As you can see, roll axis will have some angle even though LCAs are level unless PHB is exactly at the same height as LCA centerline.

PHB is generally higher than LCAs, so to achieve zero roll steer in this case, you'll have to point up the LCAs towards front so that point A will be at the same height as point B. How much to point up the LCAs depends on height of point B and the topview angle of LCAs. This is why my setup have almost-zero roll steer despite 3.5-degree LCA uphill angle.

Does it make sense? If it did, good. If I confused you more, you should have another donut. Donuts are primary fuel for my brain.

streetfytr68
Registered User
Posts: 1251
(7/29/04 1:27 pm)
Reply Re: re: re:
--------------------------------------------------------------------------------
Okey Dokey--Thanks Katz.
Looks like I got hung up on UCA centering. So just to summarize (in english) roll steer is induced by slope of the centerline (front to rearslope when viewed from side). Centering (from above) of upper control arm eliminates left/right variations in degree of roll steer. We'll assume that LCA's are always parallel.

If that's right, then I'm at least in the boat and rowing. Now let's see if I can figure out where the hell we're going.

I want level and parallel LCA's. If I can't have level LCA's I can compensate for roll steer with the vertical orientation of my lateral link (PHB or Watts). I want all of my forward links to be as long as possible with some ratio of upper length to lower length (which has yet to be clarified, I think). My UCA (torque control link) will be angled down (against the direction of axle rotation under acceleration). But not so much as to induce wheel hop under braking.

What say we row the boat towards better explaining that UCA angle.
Once we get there, I guess we need to look at how all of these attaching points relate to the front suspension. I'd hate to have to vary rear ride height in order to better match my already-complete front suspension.

Great! My car is just a big a$$ Rubik's cube.

www.lateral-g.net

Edited by: streetfytr68 at: 7/29/04 1:31 pm

ddennis68
Registered User
Posts: 290
(7/29/04 6:04 pm)
Reply Re: re: re:
--------------------------------------------------------------------------------
Put part of a page together with some different stuff. One thing I noticed on all the vehicles I looked with a factory watts style suspension was that none of the links were parallel and ride height. In every case they were angled up toward the wheel.
http://www.geocities.com/[email protected]

Different aspect, I was under the Elco today looking at possibilities. I would like to run 48-54" LCA but will have trouble running them parallel to the vehicle centerline without moving them inboard several inches on the differential housing. What would be better, move the LCA toward the center section or run the links at an angle from the differential toward the center?
Dennis-

check out progress of Bondobucket

Edited by: ddennis68 at: 7/29/04 6:48 pm

baz67
Registered User
Posts: 189
(7/30/04 8:12 am)
Reply UCA angle
--------------------------------------------------------------------------------
Is it time now to wrap my head with duct tape to contain the mess when it explodes. Nice way to summerize it Steve and I agree in steering this towards the idea of the UCA angles. This thread has been one of the best I have read on PT.
Brian

Norm Peterson
Registered User
Posts: 166
(7/30/04 8:48 am)
Reply Re: UCA/LCA ratios
--------------------------------------------------------------------------------
I wasn't stalling. Really.

This seems to be another one of those things that depend on intended usage. Those more serious about high speed cornering seem to be using a relatively long UCA, 75% or even more of the LCA length. That relegates anti-squat considerations to secondary consideration status at best, and even at that, A/S is more of a by-product of brake hop vs anti-lift while still trying to maintain good rear wheel braking %. Something about losing the brake hop being far more useful than A/S% in 3rd gear or higher and not screwing up the axle roll steer or RCH with sharply inclined lowers. Somewhere I have a video of a road-race Corvette where the MPH was much higher than the time would ever suggest, and it simply had no bite off the line (on the gas - go sideways, back off to straighten, repeat several times).

One other thing I looked at via a spreadsheet was the shape of the anti-squat curve as a function of ride height. And some UCA length ratios and inclinations give some really odd plots. Perhaps for primarily street use the shape of the curve is important, and if the A/S is relatively low at static ride height you at least don't want it to drop as the rear of the car squats. [edit] And if the UCA is only around 45% of the LCA length A/S rises with squat/drops as the tail lifts. [/edit] If I'm remembering all those sims I ran, the freaky plots seemed to fall in the 45% - 50% range of UCA to LCA, and much past 50% tended to end up with decreasing A/S with squat.

BTW, when discussing 3-link lengths, it's probably more useful to compare the UCA length against the side view projected length of the LCA's. For whatever it's worth, the side view projected length ratios (of uppers and lowers) for the G-body and the Fox Mustang are in the 43% - 45% range.

Norm
'79 Malibu (too unstock for Street Mod)

Edited by: Norm Peterson at: 7/30/04 9:57 am

Mean 69
Registered User
Posts: 90
(7/30/04 9:49 am)
Reply Getting there, but...
--------------------------------------------------------------------------------
Hey Steevo/guys,

You are getting close, but there are a couple of things that I think still might not be clear. And believe me, don't get too uptight about this, it IS really complicated, your car IS very much like a Rubic's cube. As much as anything, when I attempt to explain this stuff, I am also trying to point out that properly designing a system is not as simple as cutting a bunch of metal and bolting stuff on.

The UCA has no effect on ROLL STEER. Because it is a longitudinal link, there is no way for it to resist lateral movement of the rear axle, hence it does not help define the "lateral restraint points" necessary to define the roll axis. The subtlety here is the offset relative to the vehicle centerline as in Katz' application. There is a difference, when we were talking about the potential negative effects of this type of (off center) location, it was in regards to TORQUE reactions induced by the axle housing (i.e. forward acceleration, and braking torque). I have not done the analysis personally on this, but it seems to me that the negative reaction would be a "bobble," most likely trying to force one side of the axle up and the other down, and it would attempt to unsettle the car. The point being, this is different than roll characteristics.

Your statement of wanting to keep the LCA's parallel in plan and side views is a choice of packaging. IF you want to do this for packaging reasons, then you need to be firmly aware of the roll steer issues associated with this arrangement. If you keep the LCA's in a three link (or Torque arm system, for that matter) parallel in plan view, then the roll axis inclination is DEFINED by the angle of the LCA's in side view. The kinematic reason behind this special case is that the LCA's offer no lateral resistance to axle movement. Think of it this way, if you removed the PHB/Watts, you'd be able to move the rear of the car with no resistance, laterally, because there are no side resistance forces. So, in this case, the roll axis is defined as parallel to the LCA inclination, and runs through the RRCH, which is defined by the pivot of the Watt's, or the intersection of the roll axis at the axle centerline of the PHB height (tough to visualize). If you move the pivot/PHB height, the angle of the roll axis does NOT change, only its vertical orientation.

Now, think about having the LCA's converge, in plan view, such as is the case with Katz' cool ride. In this case, the LCA's DO offer some lateral resistance, due to their orientation. It's not enough to control the axle location on their own, but it does offer some resistance. As such, this case does have an influence on the roll axis definition. See Katz' description of how to find this in a previous reply.

From my research and modelling, nothing has jumped out as a hard, fast rule for the ratio of upper to lower length of the links. I really think that it is terribly dependent upon on the intent of the car, how much of an issue packaging is, etc. For me, high speed stability is of primarary concern, and should be for all of you guys that want to track the car, especially if you are a novice driver such as myself. My starting point is an "effective" ratio of nearly .7, but I have not tried this combination on the road/track yet. I want things to migrate nice and linearly in bump and roll, so using long links, a long SVSA, roll steer considerations, etc, are very important. The downside (for some, maybe, not me) is that packaging requires cutting the car, but no different than would likely be the case for any effective three link application. The only thing that really concerns me, to be honest, is how bad the car will likely push as a result of a well thought-out and performing rear setup! That's okay though, jus another thing to think about, another challenge to overcome, and more lessons to be learned.

Anyway, did this help?
Mark

ddennis68
Registered User
Posts: 293
(7/30/04 10:13 am)
Reply Sooooo
--------------------------------------------------------------------------------
Mark, if I understand correctly it is actually ideal for the LCAs to converge in plan veiw? I was under the impression that to eliminate all bind under bump the arms should parallel to the centerline. This is a good thing, for me anyways. It will packaging much easier. Is there an optimum angle they should converge at? I would imagine too much would be worse than not enough. I think I speak for most of us here when I say a big THANK YOU for all of your time.

Norm Peterson
Registered User
Posts: 167
(7/30/04 10:15 am)
Reply Re: Getting there, but...
--------------------------------------------------------------------------------
We'll get there eventually. I'd hoped that my mention of roll steer wasn't going to be misinterpreted. So I should probably clarify it a bit.

As you say, the 3rd link has no effect on roll steer. Sort of (so maybe call it no direct effect). But if you end up playing around with the LCA's to end up at a particular A/S% while holding the SVSA length relatively constant, it will have indirectly affected the roll steer.

I also think that some smallish LCA convergence is desireable, as it makes the roll steer effect a little less sensitive to ride height changes.

Norm
'79 Malibu (too unstock for Street Mod)

Edited by: Norm Peterson at: 7/30/04 10:21 am

katz
Unregistered User
(7/30/04 10:36 am)
Reply re
--------------------------------------------------------------------------------
Norm,

You're absolutely about the length of LCA. Projected sideview length should be used for ratio calculation as opposed to true length. I noticed my error when I was thinking about this over a glass of bourbon last night. The true ratio on my setup is more like 71.6%.

I realize that ratio depends on applications. I was wondering if there is an 'optimum' ratio for certain vertical span of UCA/LCA axle pivots, just like spindles on IFS. But it sounds like the primary concern is stabilizing the motion of the rear axle itself for road race applications, in which case, UCA/LCA length and angle should be oriented to minimize the change of roll steer while maintaining good anti-lift and no brake hop. On street car that has lower cornering speed and higher CG, it maybe beneficial to take AS and its curve into considerations as you mentioned.

Pitch center and its progression are rarely discussed and none of the books I have cover the subject. Perhaps this is not really that important since wheels roll in the direction of acceleration, but it really stirred my curiosity and I tried to analyze the progresson of AS on my setup yesterday, using IFS drawing I had on hand so I can find where the pitch center is. In this case, susp. travel is just a by-product of pitch angle around the pitch center.

Then I realized that it probably doesn't work this way due to wheel rate difference between front and rear. More I think, more I realize how complex vehicle dynamics really is.

Thanks for your input. It's always greately appreciated!

Tig Man
Posts: 7
(7/30/04 11:54 am)
Reply 3 link
--------------------------------------------------------------------------------
Having been a fabricator in winston cup and trans-am racing for the last 18 years I thought I'd ad my 2 cts worth.I've welded on and cut off a train load of parts designed by some of the top engineers in the business. And have come to this conclusion. It's all trial and error, the key factor being the loose nut behind the wheel and the TIRES! Everyone is different and wants something that they don't have. I wonder what the thinking is sometimes these guys building the water heaters with power to rival all out race cars for the street. The tires, no matter what kind is the limiting factor. What good is it to have all that power when there's now way your going to hook it up. I've seen the guys build a killer motor on the engine and chassis dyno, and the driver actually slowing down because it screws up their rythem.We've won races with the worse motor they took to the track, because they broke the good one and rented out another.There is no magic formula to make up for raw driving talent and communication. It's real easy to over engineer suspension geometry. The best formula I've seen is KISS. Keep It Simple Stupid!!!!! Ask anyone in pro motorsports.

ddennis68
Registered User
Posts: 294
(7/30/04 12:13 pm)
Reply Another Mark
--------------------------------------------------------------------------------
What is that 3 now in the same thread? Thanks for the input Mark, it's always nice to hear from the "been there/done that" crowd. I don't think anybody is looking for a "do it just like this answer". I know me in particular, just need a general idea of what is going on before I crank up the MIG. I hate having to cut cr@p off, especially if half the world already knows it won't work to begin with. I wonder with all your Nascar style expereince if you have any idea of a truck-arm/3link argument. Which is better in and why? Both have their good points, I can't think of a bad point for the 3link right now but I'll come up with one. If the 3link is a better system why doesn't Nascar use that system instead of the truck-arm?

katz
Unregistered User
(7/30/04 4:06 pm)
Reply re
--------------------------------------------------------------------------------
It's weired. When I posted reply this morning, my post was right below Norm's post from this morning. Now there are three posts in between. Maybe it's an unregistered user thing? I probably should register when new board is up and running.

I think Mark and Norm made it pretty clear about how roll steer characteristic is affected by link angles. Now try going back to page 1 and re-read every post, and you'll be able to find more usuful info.

I too feel some LCA skew is desirable. The amount depends on LCA's sideview angle and RCH. Although my Riv drives pretty good, I probably wouldn't use that much skew if I were to design something from scratch.

Tig Man,
I understand what you're saying. Driver preference is a HUGE factor on setup. But, most of us here don't have much experience driving on track. For this reason, I feel it's better to start out with a car with properly designed or at least half-way decent suspension geometry. For example, if a car spins out in a fast turn as a result of whacked geometry, many of us (at least I) won't be able to save it.

And lot of us are building just one car, and don't have financial freedom to build another one. Some will choose to sell their cars when they are complete to start another project, and others are planning to keep their cars and enjoy driving. I'm one of the latter. So I might as well do it right (or as good as my ability allows) the first time.

Dennis,
If I'm not mistaken, NASCAR's chassis and susp are restricted by rules quite a bit. I mean, look at those 15" rims with negative offset up front. Those cars aren't the most stable things at high speed compared to other modern racecars. No offense to NASCAR fans, but I'd look into other racecars if I want to incorporate some racecar chassis technology on my ride.

That said, truck arms work pretty good for what it is IMO. And if some average Billy Backyard was to build a system from scratch, there is far less chance of him screwing it up compared to 3-link b/c of its much simpler kinematics. SVSA can be made pretty long on long WB cars, and roll axis change is relatively small due to long arms and planview skew. It's low profile, so you can keep the backseats (doesn't matter on your Elco). The down sides are inherent roll bind (it relies on arm's twist), low AS as a result of long SVSA/low SVIC, difficult exhaust routing, lack of adjustability, etc.

Other than rear seat invasion of the third link, 3-link's most apparent (and not so apparent to average Joe) down side is difficulty of design. It's so easy to screw up and end up with a dangerous suspension. The fact that 3-link is sooo extensively discussed here and CC.com speaks for its advanteges for road race applications.

streetfytr68
Registered User
Posts: 1253
(7/31/04 4:33 am)
Reply Re: re
--------------------------------------------------------------------------------
One thing I'm not is "uptight" Nor am I afraid to ask questions that expose my fundamental lack of brain cells. I would like to further these discussions for my benefit as well as the benefit of the 100's of people viewing this thread (Thanks Dennis).
That means that I will on occasion--attempt to summarize (dumb it down) your wisdom and the wisdom of guys like Norm, Katz, Chicane and others. I do this for my benefit as well as the probably hundreds of people viewing this thread.
As is, this discussion is omitting some important elements such as instant center, Side View Swing Arm, etc. (turns out SVSA does not stand for the Society of Vehicularly Stupid Americans--go figure) But it's better to dumb it down for now.
At this point, here's what I have: I can go out to the shop, put my car on jackstands and take some basic measurements starting with a virtual 3D rectanglular box limited by:
-The upper and lower limits of my suspension travel.
-Rearmost location of a Panhard Bar or Watt's link.
-Forward location of a lower control arm crossmember (variable for now)
-Max Width of the box would be limited by frame rail location near axle (Lateral location device such as PHB and rear LCA pickup points would be closest to this max width)

That's the big box. All other physical (I said physical) points would fall within the max width. i.e. forward LCA pickup points would fall between max width--In my case defined by frame rails. Axle centerline falls within max depth (suspension travel)
Then there are the imaginary points that fall outside the big box such as instant center (right?) and SVSA (right?) Other theoretical points fit within the box i.e. rear roll center height (right?). Still others are calculated from all of the above such as anti-squat/anti-lift

SOAPBOX:
So without spending a penny, any hot rodder worth his measuring tape can develop a basic foundation that can be used to practically apply the complex and valuable knowledge of rear suspension design.
Figure out the physical limitations of your car > build a (bad) model > hit the books > get a clue> revise model >repeat.

MORE SOAPBOX:
The end result is that you can exceed the basic hot rod (and hot rod mentality). Your car will hopefully run circles around basic "bolt-on" hot rods AND > Instead of just crawling under a hot rod and saying "Hey cool! a Watts Link! Wow! Gosh!", you can declare: "Huh? Why the hell did he mount that there?"

I for one will get started on "The Box" as soon as I get home. Will post when I have some basic dimensions.

www.lateral-g.net

Mean 69
Registered User
Posts: 91
(7/31/04 8:38 am)
Reply Yep!
--------------------------------------------------------------------------------
"Mark, if I understand correctly it is actually ideal for the LCAs to converge in plan veiw?"

I do. Emphatically, for the reasons discussed here. There really is no optimal angle, just like all of the other things in the overall system. The trick is to find a way that allows suitable packaging, and most importantly, defines the forward intersection point (a.k.a. lateral restraint point) at a suitable place. Clearly this is vehicle dependent, and there really is no right answer.

The drawback for angling the arms relative to a parallel setup is that now you are introducing a bending force in the LCA's. A parallel setup induces primarily a tension/compression force set. As such, the LCA's need to spec'd appropriately to deal with this factor. Not a terribly big deal though.

Steevo is on the right track by first defining the available envelope. The other thing that needs to be determined, up front, is what aspects of the performance you are trying to optimize, 'cause you're going to have to make a compromise somewhere. As I stated, the approach I took was one to attempt to maximize high speed stability will maintaining a very high anti-squat figure. I took a couple demo laps at California Speedway's road course, as well as a few laps around LVMS' oval, and I'll tell you, seeing dark skid marks trails up to the wall while you're cooking around the track will make you think twice about what your car can handle with you behind the wheel. In my case, packaging was secondary to "function," which is obviously yet to be determined, but with several iterations under the car, behind the drawing pad, I figured out a way to make it all fit while not having to alter the frame rails. It has been really educational, and very fun.

Mark

And by the way, I don't think you are dumbing anything down, stating things in a different manner can usually turn the light bulb on.
08-16-2004
chicane67

Page five

maineSS
Registered User
Posts: 1
(8/4/04 9:54 am)
Reply "Decoupled" 3-link
--------------------------------------------------------------------------------
You could get around the compromises inherent in one upper link by using a centrally- mounted compression link for braking loads, and an offset link on the pass side of the diff for accel/antisquat & torque reaction cancellation. This would also eliminate the "push-pull" fatigue loading of the chassis mount, which seems to be a problem with sheet metal mounts.

ddennis68
Registered User
Posts: 314
(8/4/04 10:17 am)
Reply WHAT???
--------------------------------------------------------------------------------
Maine. please elaborate. Centrally-mounted compression link? How will that work?

Mark- I got some reference material in the mail yesterday, I was up reading until about 1:00. Now that I am totally confused, let me ask a question. I now have basic understanding of roll center/roll axis, we had previously discussed mounting the bellcrank to the frame instead of the axle. As I understand it, the roll center on a 3-link/watts setup is determined by the mounting location of the pivot. Since the pivot in this case is the bellcrank and the frame is sprung weight and subject to height change during droop/bump, wouldn't the RRC change as the suspension moved? It looked like mounting to the axle would eliminate RC change during bump, am I not understanding this yet?

Anyone, is there a prefered RC height for the front-end? Just guessing without any measurements yet it looks mine is going to be about somewhere in hell at this point. Does FRC have any affect on anti-dive, if not is there a prefered angle the UCA should be mouted at? I will obviously have to take some acturate measurements before starting but it looks like the UCA mouts are going to have to moved around a little bit.

Norm Peterson
Registered User
Posts: 169
(8/4/04 11:48 am)
Reply Re: WHAT???
--------------------------------------------------------------------------------
A chassis-mounted WL bellcrank seems to be the better arrangement, and I think it's got something to do with the car's roll axis (as defined by the front and rear RC's) more closely following the attitude of the chassis in side view as you accelerate, brake, and as the car "heaves" over gentle rises/low spots in the road. That keeps the roll moment more constant.

There's lots of opinions on RC heights, but the consensus for stick axle cars over 3000# and >100" wheelbase seems to be 2" - 4" above ground level up front and something like 6" - 8" higher than that in the rear. If you have a "clean sheet of paper", anyway.

Norm
'79 Malibu (too unstock for Street Mod)

Mean 69
Registered User
Posts: 95
(8/4/04 2:10 pm)
Reply Correct, but...
--------------------------------------------------------------------------------
"As I understand it, the roll center on a 3-link/watts setup is determined by the mounting location of the pivot. Since the pivot in this case is the bellcrank and the frame is sprung weight and subject to height change during droop/bump, wouldn't the RRC change as the suspension moved? It looked like mounting to the axle would eliminate RC change during bump, am I not understanding this yet?"

Excellent question, but the tricky thing is what the roll center is moving "relative" to. If you can imagine, the roll center of a vehicle (front or rear) is an imaginary point that you could "push" on, and the result would be no roll movement of the sprung mass, it is sort the fulcrum, if you will. A car rolls because lateral forces due to acceleration are "felt" through the CG of the sprung mass (simplified, there is more to it, but not needed for this discussion), and the sprung mass rotates about the roll center. So far, so good.

Now what is important to realize is that the difference in height of the CG and the roll center is a thing called the roll moment. The taller the roll moment, the more a car will roll given a given lateral force, think of it in terms of torque. If the roll moment is "fixed," then the force to roll response is predictable, and linear (for the most part).

For your specific answer, you are correct, the RRCH will change as a function of bump, roll etc, relative to the ground, for a fixed-to-frame bellcrank Watt's link. However, the roll MOMENT will NOT change in bump/roll, because the CG of the sprung mass does not change relative to the RRCH, it's fixed by definition.

If you mounted the bellcrank on the axle, the RRCH would not change relative to the ground in bump/roll, but because the CG does move relative to the ground, the roll MOMENT will change. If the roll moment changes dynamically, then the roll angle of the car will also change dynamically because the roll resistance from the bars/springs is (relatively) fixed, and something's got to give.

In practice, my guess is that there isn't a huge difference between the two, but if there were, it would likely result in a squirming effect of the rear of the car dynamically. That said, why fight a demon that might be an issue, easier to apply engineering fundamentals up front in the design, in my opinion.

Make sense?
Mark

ddennis68
Registered User
Posts: 315
(8/4/04 2:50 pm)
Reply I see...
--------------------------------------------------------------------------------
Agreed, just wanted to validate what I am learning. I understand now, next thought, if WL pivot detremines RRCH then pivot placement is very important to set RRCH. If I understood Norm in the above post, the WL pivot point should be approx. 8-12" or close anyways. Short of cutting off the pivot mount and relocating it how would one suggest leaving some adjustment of that mount for tuning later? All the mounts I have looked at are fixed.

ddennis68
Registered User
Posts: 316
(8/4/04 5:35 pm)
Reply ohoh
--------------------------------------------------------------------------------
Just Herb Adams' "Chassis Engineering". Big mistake to read that one, please explain why no one is running a Satchell suspension? According to Adams, the Satchell design suspension is all-around the best design with the converging LCA and parellel UCA. This would be a very simple system to design, especially on an A body that already has a half-ass 4-link. I know I'm getting a little off topic but since I started it -I can

With the Satchell system, does the converging LCA provide enough lateral support-it doesn't seem like it would under severe cornering loads? Is the RRCH determined the same way as if it were a conventional 4-link system?Why is the sky blue?Why is the grass green?
Dennis-

check out progress of Bondobucket

maineSS
Registered User
Posts: 2
(8/4/04 6:23 pm)
Reply "decoupled" 3-link
--------------------------------------------------------------------------------
Basically you have two upper links- one mounted over the center of the diff to evenly divide the braking load. This link telescopes in compression against a rubber bushing (a standard stock car item), but does nothing in tension. The second upper link would be mounted towards the pass side- the exact distance requires knowledge of engine torque and rear end ratio, since you're counteracting the engine's torque reaction force that unloads the pass side tire. If you get it right, you wouldn't need a locking diff, since each tire is equally loaded under accel. The length and angle of each link can be optimized solely for it's intended purpose, instead of being partially right for either. Not having a constant push-pull fatigue load on one point also is attractive- the chassis point sometimes fails repeatedly on tube frame race car chassis that aren't correctly designed with this loading in mind.

SSWANAB
Registered User
Posts: 109
(8/4/04 7:03 pm)
Reply Re: "decoupled" 3-link
--------------------------------------------------------------------------------
Isn't the rear suspension under a Chevelle basically a Satchell system just reversed from what Adam's says in his book? I am using a Chevelle 12 bolt rear under my 68 Camaro attempting this type of setup now.

Mike
MY CAR

ddennis68
Registered User
Posts: 317
(8/4/04 7:13 pm)
Reply Satchell link
--------------------------------------------------------------------------------
Yes, it is a "junk" GM 4-link inverted, however the LCA are at a much greater convergence angle and the arms are much stronger and free of bind. I have been doing "research" over at c-c.com among other places. I'm getting excited, I really like this design (assuming it works OK). There is a lot of "theory" talk and "on paper" discussion, I am really looking for anybody who has actually run this or knows somebody who has for feedback. 6 months ago I knew "for sure" I was running a truck-arm design, now I wouldn't run that if it were given to me. It is amazing how much info is out there if you take the time to look and ask questions.
Dennis-

check out progress of Bondobucket

Mean 69
Registered User
Posts: 96
(8/4/04 7:38 pm)
Reply Yeah baby!!!
--------------------------------------------------------------------------------
"Basically you have two upper links- one mounted over the center of the diff to evenly divide the braking load. This link telescopes in compression against a rubber bushing (a standard stock car item), but does nothing in tension. The second upper link would be mounted towards the pass side- the exact distance requires knowledge of engine torque and rear end ratio, since you're counteracting the engine's torque reaction force that unloads the pass side tire. If you get it right, you wouldn't need a locking diff, since each tire is equally loaded under accel. The length and angle of each link can be optimized solely for it's intended purpose, instead of being partially right for either. Not having a constant push-pull fatigue load on one point also is attractive- the chassis point sometimes fails repeatedly on tube frame race car chassis that aren't correctly designed with this loading in mind."

Nice explanation! The decoupled stuff is highly popular amongst the circle track guys, and from what I gather, the main benefit is trying to not "unsettle" the car during transients, of which the short track guys have hundreds per race. My strong suspicion is that 'Maine has a circle track background, which would be a terrific addition to the tech here. Keep it coming! The more info, the better.

The four link setups have some "inherent" drawbacks, but as Norm will quickly (and appropriately) point out, they can still be made to work, really well. The Satchell is a really interesting approach, it offers the potential to have some terrific kinematic behavior, in a simple package. One of the main drawbacks in commercial four link setups, is bind, or rather, immediate bind when the suspension bushings (required for the convergent type setups to work at all), lose their liberty to move due to compression: snap time. Not so sure that means all the different approaches can't be made to work, time to look further. Anyway, Terry Satchell came up with this design, and it offers the potential of tunable roll steer in the direction we want, good anti-squat, low roll center height, and simple packaging.

So, Dennis, if you don't want the truck arm anymore, just what DO you want to try on your cool project?

M

ddennis68
Registered User
Posts: 318
(8/4/04 8:54 pm)
Reply Re: Yeah baby!!!
--------------------------------------------------------------------------------
Hey Mark-I think I am going to run the Satchell. I have been going over and over it, I have books, graphs, and notes spread out all over the house (wife is not real happy). I just can't find a good reason not top run this setup. The packing is too easy, low RRCH, simple to build, just as easy to fix. For a daily driven (driven hard) street car that might see a couple dozen track days a year this is the ultimate setup. It may not be the best rear suspension out there, but for the amount of work and investment to build it will work the best for me.

Now, about building my own links. Which would be stronger, welding nuts onto tubes for rod ends to thread into or tapping the tubes themselves? Anybody happen to have a good source for suspension grade tubing, maybe like 1.25”X.095” or so? Overkill, yes but like I said-daily driver.
Dennis-

check out progress of Bondobucket

airrj1
Registered User
Posts: 76
(8/4/04 10:21 pm)
Reply So...
--------------------------------------------------------------------------------
I am trying to grasp some of this topic. A couple of questions specific to my vehicle. I have an A-body that is a daily driver first, Auto crosser second, and hopefully an open track day some time.

It sounds like a higher AS % would be more important for my application. Since high speed stabilty would be less of a priority than traction for acceleration. Am I thinking correctly?

Second, to build a 3-link in an A-body would the stock 22" LCA be an acceptable length to work with? It seems like it may be fairly short because it would have more RC movement than longer arms. And it would make for a short UCA. It appears that there is room to make up to a 30" LCA without introuding into the floor. Would the additional 8" significantly reduce the roll steer?

Lastly, it sounds like the decoupled 3-link would be a good choice for an Auto cross application for the same reasons that it works for the circle track cars. Lots of on/off throttle corners.

Dennis, thanks for starting this thread. I am a year away from building a rear from my Chevelle, but this is an excellent education.

R.J.

jon
Unregistered User
(8/5/04 8:54 am)
Reply x
--------------------------------------------------------------------------------
i'm with tig man.pick up any number of race parts catalogs and you'll see that you can buy,off the shelf,trailing arms and brackets,upper 3rd link mounts,phb's, watts link,and the mounts for them.even complete rear clips.there are a lot of cars out there running the geometry i'm suggesting.we can't all be wrong.and yes i've put many together.with off the shelf parts and standard geometry.don't dismiss stock car racing.they are very close to what most of us build.at least check out some of the steve smith autosports books such as 'building the pro stock late model sportsman'.theory is great,but well tested and prover design that can be built with inexpensive off the shelf parts are better.and using standard geometry and parts offers a couple of advantages.if something breaks you can just order a replacment.and a wealth of tuning information is available,including road course setups.what is the reason not to use standard parts?i'm not trying to be a jerk,i just don't understand.seems about the same as building your own wheels.i like being as involved with my car as the next guy,but i'm not going to start a project by mining some iron ore.

ddennis68
Registered User
Posts: 322
(8/5/04 10:13 am)
Reply roundy-round
--------------------------------------------------------------------------------
I am actually running quite a bit of roundy-round stuff. Problem is the suspension system are not optimal in all cases as they have rules to deal with, I do not. Some peices are available off the shelf and I will use those parts, but some peices will require custom links be built. Another drawback is most roundy-round parts suffer some degree of quality control issues, I know I have some peices with pinholes in the welds.Thanks for the suggestion though.

Norm Peterson
Registered User
Posts: 170
(8/5/04 11:17 am)
Reply Re: So...
--------------------------------------------------------------------------------
With a 3-link, the roll center height is primarily controlled by your PHB or Watts link. The point of convergence of the LCA's is much further away from the axle, so it has much less effect. By itself, increasing the LCA length won't guarantee less roll steer, or even a slower rate of change in roll steer as the suspension rises and falls (see note below). Their plan view convergence angle matters, as that's what establishes the LCA convergence point.

As regards longer LCA's and roll steer, it's probably easier to use the OE LCA's. Use some target A/S% to define the UCA length and inclination such that the A/S at least doesn't fall off much with whatever squat you do get. Then set the PHB or WL height to provide some desired rollsteer (perhaps providing for adjustment a la NASCAR/circle track track bar height adjustment). You'll want more rear spring and/or sta-bar anyway (since the RCH is dropping from its OE location and because the 3-link inherently develops less roll resistance than the OE converging 4-link), and the more-spring route will also tend to minimize the variation in A/S for any given maneuver.

Side note about that rate of change parameter: it's entirely possible to have, say, 8% roll understeer at design ride height and 13% roll understeer at a 1" lower ride height (that could occur for a variety of reasons). It's probably a conservative approach to have the roll steer % increase slightly in the more understeer direction as ride height drops, at least if the anticipated passenger and cargo load for your car is expected to vary widely.

Norm
'79 Malibu (too unstock for Street Mod)

Edited by: Norm Peterson at: 8/5/04 11:24 am

Norm Peterson
Registered User
Posts: 171
(8/5/04 11:44 am)
Reply Re: Yeah baby!!!
--------------------------------------------------------------------------------

Quote:
--------------------------------------------------------------------------------
Now, about building my own links. Which would be stronger, welding nuts onto tubes for rod ends to thread into or tapping the tubes themselves?
--------------------------------------------------------------------------------

I think that dedicated fittings exist for this purpose that you insert into the tubes and then weld. It avoids the alignment issues associated with butt-welding the nuts and the machining operation or laborious hand-threading of the tube-tapping approach. I know I've seen this discussed over on CC.com (a ROUSHGTR reply IIRC) and that Jon A did the threading-by-hand method with some aluminum tubing for his 4th gen Camaro (last I knew, a description of the effort still existed on his web page).

Norm
'79 Malibu (too unstock for Street Mod)

Edited by: Norm Peterson at: 8/5/04 11:49 am

maineSS
Registered User
Posts: 3
(8/5/04 5:59 pm)
Reply Circle track parts
--------------------------------------------------------------------------------
One thing you want to check out on race-oriented parts like Watts links is the amount of bump/rebound travel they're designed for- for street use you want at least 4" either way. Also check your state's vehicle laws if you have inspections- many specify that exhaust has to follow OEM routing, etc- it would be a bummer not to get a sticker over that sort of detail. Maine, for example,
prohibits "ball joint kits"- which leaves parts like the Howe
interchangeable stud/take apart BJ in sort of a grey area. I'm presently using monoball K6145 front LBJ's that are 1/2" taller in the front of my Monte SS (VERY perceptible change for the better), and plan to follow up with Pole Position adj uppers and 1/2" taller UBJ's. However, I go to a mechanic that will wink as long as it looks durable.

ddennis68
Registered User
Posts: 323
(8/5/04 11:40 pm)
Reply Re: Circle track parts
--------------------------------------------------------------------------------
I guess we are lucky out here in CA. No inspections like that and definetly no laws pertaining to exhaust OEM mounting or suspension design changes.

Back to suspension design, I was playing with a program tonight and did a mock up thast looks pretty good and I could package OK. I got anti-squat to 210%, but could not get the program to figure RRCH at all. This was using a Satchell design system with 20" UCA and 27" LCA converging at about 30 degrees. Does anybody know how to figure RCH on a Satchell system, using the 4 link method does work even close?
Dennis-

check out progress of Bondobucket

Norm Peterson
Registered User
Posts: 172
(8/6/04 3:56 am)
Reply Re: Circle track parts
--------------------------------------------------------------------------------
A Satchell link is really nothing more than a specific 4-link arrangement, so the general 4-link constructions remain valid.

What may be happening is that your program may be getting indigestion over trying to divide by zero, probably due to inputting UCA's that are perfectly parallel in plan view. Try splaying the UCA coding ever so slightly (in my 4-link spreadsheet, 0.01" difference in the Y-axis (lateral direction) coordinates is sufficient to chase away all of the #DIV/0! rollsteer cell displays while the effect to the A/S result is no larger than 0.0x %).

Edited to add clarification.

Norm
'79 Malibu (too unstock for Street Mod)

Edited by: Norm Peterson at: 8/6/04 5:06 am

jon
Unregistered User
(8/6/04 7:54 am)
Reply rules?
--------------------------------------------------------------------------------
these are the rules for lmsc rear suspension'rear suspension may use leaf or coil spring.coil springs must be minimum of 5" in diameter.3 point,truck arm and tourque arm suspesions are permitted.no coil overs are permitted.no spring rod trailing arms allowed.coil over units permitted on end of tourqe arm only.'as you can see nothing is said about geomerty at all.standard geometry is used because it works-or there are a lot of fools out there.and i'm sure most parts suppliers would disagree that they sell marginal stuff with bad welds.and i buy enough stuff from coleman that rory is starting to memorize my amex number.i don't get bad parts,nobody i know gets bad parts.it seems that a three link isn't cool enough-it needs to be a custom designed,one off,untested design to gain admiration.i still ask what is the flaw in the standard 3 link as used by so many winning racers.side note-i use tap tubing for traing arms and tie rods.but i enjoy working metal and really like running a tap in.i'm not sure why anyone would find it time consuming or labor intensive.wrong size tubing maybe?dull tap or not starting it straight?no cutting oil?should be easy.and while i'm ranting-pro-street cars are 'back halfed' all the time.putting in a 3 link is no different.just use a howe rear clip instead of a cris alston one.why is that a bad idea?maybe by making pro touring more accessable more cars will be built.i live in cali and have yet to chance across one just driving down the road.how many people got partway through this thread and gave up because it was too complicated?reminds me of a 5 page discussion on c-c,with never once any real data(i.e.'i use temp paint and my rotors get up to x degrees').
08-16-2004
chicane67

Page six

Davezz28
Registered User
Posts: 15
(8/6/04 8:10 am)
Reply Suspension Bind
--------------------------------------------------------------------------------
Does anybody have a program or a formula for calculating when a suspension system will bind?

I would like to determine when a 4-link will bind and how, by converging the LCAs (Satchell), will keep the suspension from binding. I just don’t see how using converging LCAs keep a 4-link based suspension from still having some bind, unless it us due to the elimination of the location link.

On a Satchell link suspension system, how much of an angle do the LCAs have to be to eliminate the location system PHB/Watts? It seams to me that the angle of the LCAs play a large part of the suspension design. At a small angle, close to parallel in plan view, the LCAs will not control the rear housing movement, but at 45* the housing movement will be controlled. At what point or what angle will the lateral movement of the housing be controlled with out a location device?

On a Satchell link design, do the mounting location of the links at the axle housing play a part in reducing bind? In Herb Adams book the illustration shows the attachment point for the UCA and LCA at the same point in plan view. The pictures of the chassis over a c-c.com also have the mounting points at the same location on the housing in plan view. If a Satchell link was designed with the LCAs attached at the axle closer to the wheel than the UCAs, will the system be free from bind?

katz
Unregistered User
(8/6/04 8:24 am)
Reply re
--------------------------------------------------------------------------------
Obviously, standard CT 3-link works just fine as many CT racers run the system successfully.

But theoretically(this is the key word), there are rooms for improvement. Engineer-type people get enjoyment out of re-inventing the wheel. Sometimes we don't know whether this re-invention works or not, but we just have to try and see what happens.

And this tremendously helped push motorsports technology to almost insane level as you see today. If everyone had 'why mess with success?' attitude from the beginning, Formula 1 cars will still be running skinny tires with no aero, and Trans Am cars will still be running leaf springs.

If you don't like us designing and building our own suspension, that's fine. You just have to understand we're just doing what we enjoy, just like you do something you enjoy.

I appreciate you chimed in and let people know that there is a proven alternative for those who don't want to design his own suspension.

jon
Unregistered User
(8/6/04 9:10 am)
Reply x
--------------------------------------------------------------------------------
of course i approve of people designing and building new design.i was just looking for the simple disclaimer of 'there is an easier way,no need to get this involved'.i'd never argue with someone doing something for fun,as long as it is presented that way.i just hate seeing people scared off.innovation becomes standard equipment in short order in racing.if someone was winning with a new 3 link design it would take about a year for everyone to copy.surely an alternative opinion is nothing but welcome around here.and no one seemed to argue with tig man.of course i live way up in the hills where cars as still built based on how much 'shine they can hide in the roll cage tubes.

Mean 69
Registered User
Posts: 98
(8/6/04 10:00 am)
Reply Perf. Trends Analyzer
--------------------------------------------------------------------------------
I haven't used the function, and I am not 100% certain that it works as I think, but there is a function in Performance Trends full car Suspension Analyzer that I believe will calculate the effective length increase for a setup to prevent "bind." My demo copy expired, I need to buy the stuff, but I don't need it today 'cause the setup I am building is already designed. It's funny though, I have been quick to bash the A-body/late model Mustang convergent arm system due to bind, but it was because of the bushings that are used. I have not thought too much about the true kinematics, for instance if you put rod ends everywhere, and frankly, the result is not terribly obvious. The Satchell setup caught my eye due to some nice static numbers, and a really simple design.

I would think that a link that needs to control lateral movement in such a setup would "want" to by skewed at least 45 degrees to the vehicle centerline. That would be the point where I would personally start, and model things from there. I am on vacation this weekend, but I bet one of my mech. engineers at the office could whip something up in an hour in Solidworks/COSMOS to show how it would work and where the strain points would be.

That said, you can't argue with the fact against this type of arrangment, there is a huge Mustang industry that eliminates the upper arms, and replaces with either parallel (plan view) arms, or a Torque arm, etc. Hmmm....

And I agree with Katz, this is a much a hobby and mental exercise as it is trying to build a better setup. My backgound is in physics, and I personnally love this type of problem solving. Doing something yourself, good or bad, is the best way to learn. Beyond that, no one offers the type of solution that I am after, although I am buying alot of the stuff from Coleman, SCP, etc. Of the available setups, I can find something in every one of them that I don't like.

Mark

ddennis68
Registered User
Posts: 327
(8/7/04 1:26 pm)
Reply OK OK OK
--------------------------------------------------------------------------------
Last time, I promise. Going over the numbers I got form the 4-link calculator I started dropping some OE “A” body numbers in, the factory arms/locations come up pretty good (with a huge suspension drop). I got 132% AS and a RRCH of 17.7", not too bad but if you move the UCA's in 11" with no other changes AS goes to 174% and RRCH drops to 8.5". Now assuming the front end is set up correctly this would be an outstanding setup I think.

Looking at the model with the UCA's moved in I don't think I would run into bind but I would have to use a PHB or WL to center the axle as the links effectively become parallel, only they are all converging. They converge at about the same angle, with Heim joints or Johnny joints I don’t think there would be a bind issue, would there??? Obviously I would have to run a PHB or WL for lateral stability. Any comments?
Dennis-

check out progress of Bondobucket

Davezz28
Registered User
Posts: 16
(8/7/04 7:03 pm)
Reply Dennis
--------------------------------------------------------------------------------
What length of SVSA are you getting with those AS measurements. If it is too short you may get some wheel hop upon braking. I have read that the SVSA should be not be less than 40-60". I believe Milliken suggests 60" in his book and I think Adams suggests a minimum of 40" in his book.

Just one more thing to think of when you think you have all the geometry nailed down.

ddennis68
Registered User
Posts: 330
(8/7/04 7:42 pm)
Reply Re: Dennis
--------------------------------------------------------------------------------
SVSA is at 39", a little short but still kicking around ideas. If I change the UCA angle to get more SVSA I lose about 110% of the AS and RRCH jumps to 13".

Had another thought-I realize this is all theory but if something has to compromise which should go RRCH or SVSA? I can live with a little bit of AS but I would really like to nail down RRCH as it is the most difficult to change once the pickup points are welded in place. I plugged in some more numbers and got 101% AS, 12.4 RRCH, 51.6 SVSA.

edit for second thought-
Dennis-

check out progress of Bondobucket

Edited by: ddennis68 at: 8/7/04 8:13 pm

Mean 69
Registered User
Posts: 101
(8/8/04 10:27 am)
Reply Huh?
--------------------------------------------------------------------------------
Dennis, I was a little confused by the info you provided two posts ago, not sure what you meant.

In regards to SVSA/AS/RRCH, I'd probably put them in order of importance as SVSA, then AS, then RRCH. As long as the numbers don't get too crazy, I think this is a good way to prioritize.

I'd recommend putting additional pickup points on the control arm brackets to make small changes after you get some seat time in the car. Getting things close with the kinematic numbers is a great start, but it is hard to say how things are going to work out once you get it all together. Especially with a car like yours, which I believe is a bit light in the tail.

A 13" RRCH is not that bad, really. This is a pretty typical height for modified cars using PHB, I wouldn't be too concerned about this height. Again, depending upon how you want to set things up, you can make the RRCH adjustable too, with a Watts or PHB, for instance.

Mark

ddennis68
Registered User
Posts: 334
(8/8/04 11:32 am)
Reply Re: Huh?
--------------------------------------------------------------------------------
Mark, I had originally posted results of the factory 4-link in the car as it sits now; 132%AS, RRCH 17.7”. Now, if I move the forward UCA mounts inward toward the CL of the vehicle, AS goes 174% and RRCH drops to 8.5”. My concern is that in doing this it creates a parallel 4-link, difference being instead of all 4 links being parallel to the CL they are parallel to each but converging at about 15% toward CL. A PHB would have to be added for lateral stability and wanted see if anybody thought this would create a bind condition (4-converging links and a PHB).
I think I found a compromise though, it means relocating all the mounts on the frame side but the numbers look good, on this program anyways. I get 101%AS, 4.2” RRCH, and 51” SVSA. I know RRCH is a little low but I can create some adjustment as you suggested to tune once it’s all together.
When running a converging 4-link with PHB, does the PHB determine RRCH or is it still the angle/height of the links?
Dennis-

check out progress of Bondobucket

Mean 69
Registered User
Posts: 104
(8/8/04 8:11 pm)
Reply I think I get it
--------------------------------------------------------------------------------
Okay, I think I can picture what you are doing. Let me break it down into where I thin you're going:

1. You are dropping the car's ride height, which further slants the UCA's downward, from rear to front. This is creating the additional AS that you describe. How much? The stock ride height of the A bodies as I recall gave a poor AS number, which is why the drag race crowd uses the UCA's relocation brackets, or traction bars (I made my own for my 72, a knock off of the Comp Eng. Design, worked great!). Keep in mind, this also creates a "negative" angle for the LCA's?
2. With the new ride height, you are moving the UCA's more towards vehicle centerline, keeping the effective length the same (i.e shortening the links), yes?
3. Yes, with the UCA modification you suggest, I'd recommend a lateral locating device, such as a PHB.
4. No, the RRCH won't be determined by the convergent point of the UCA's, nor by the PHB height, but rather by some combination of them. This is the ole' roll axis thing we have been talking about. The RRCH is the point at which the roll axis (vertically) intersects the rear axle assembly, along the axle centerline (was that clear?). Because the convergence of the UCA's creates a "lateral restraint point," the picture becomes a bit more complex in the RRCH determination.
5. A 5" ish RRCH is really low, at least it seems so. I'd be concerned about this. Actually, I am having a hard time visualizing this, it's not obvious how this computes, but again, it's hard for me to picture the rear of an A-body, it's been years.

Help?
Mark

ddennis68
Registered User
Posts: 337
(8/8/04 9:09 pm)
Reply Re: I think I get it
--------------------------------------------------------------------------------
Mark, it is lowered about 4.5" in the rear (real low). The LCA's actually sit perfectly level (8" from the ground at the front mount and diff mount), the UCA's sit 17" at the diff and 15" at the frame. The UCA distance from CL at the diff is 4" and at the frame is 14", the LCA distance from CL at the diff is 22" and at the frame is 16". I am proposing to move the UCA's frame mounts in to 3.5" from CL and move the LCA mounts at the frame out to 21" from CL. I figure with these numbers you can plug them in to some virtual software and "see" what is going on.

If RRCH is determined by some combo of the convergent point and PHB height then RRCH will actually come up as the PHB mount will be at about 11", so RRCH should be somewhere in the neighborhood of 6-8" maybe. Please feel free to criticize my plans before I cut my car up.
Dennis-

check out progress of Bondobucket

Davezz28
Registered User
Posts: 17
(8/9/04 8:40 am)
Reply Your CL Measurements
--------------------------------------------------------------------------------
Dennis

Are you CL measurements in side View or Plan (top) View? or a combination of both?

I am guessing you mean top view and in that case you are moving both UCA/LCA to a position where they are close to being parallel to each other. Your two UCA will be about 8" from each other, why don't you incorporate them into one control arm and have a 3- link instead of a 4-link system with arms that are very close together?

I just got done building a 69 Chevelle for a friend and we did not do much with the rear susp. I am very interested in what you decide as well will revisit the suspension on the car after some miles to sort other things out. If I knew how to post pictures I would.

ddennis68
Registered User
Posts: 340
(8/9/04 10:09 am)
Reply Dave
--------------------------------------------------------------------------------
All measurements are in side veiw. I was originally going to do a 3-link but wanted to do a watts link in that case. After much deliberation, I decided on converging 4-link with a PHB. The mounts are already on the diff, I can use off the shelf replacment arms (with rod ends or johnny joints), and don't have to fab a mount for the watts link. I am refering to top veiw in moving the links in. I thought about a wishbone style UCA for awhile-maybe I'll play with that some more. With the wishbone track locater I wouldn't need a PHB or Watts link--hmmm. Packaging is what its all about, With a ride height of 6" there isn't much room back there.

Norm Peterson
Registered User
Posts: 173
(8/9/04 11:56 am)
Reply Re: I think I get it
--------------------------------------------------------------------------------

Quote:
--------------------------------------------------------------------------------
I am proposing to move the UCA's frame mounts in to 3.5" from CL and move the LCA mounts at the frame out to 21" from CL.
--------------------------------------------------------------------------------

With such shallow angles of convergence (plan view) the effective lateral restraint points of either your upper or lower arms will be located far enough in front of the axle that the RCH will be defined almost entirely by the PHB. With a horizontal PHB 11" above the ground at static ride height, you're about where the 3rd gen and 4th gen F's are.

I think I'd make the uppers parallel to centerline rather than converging them inward like the LCA's, though, as your roll steer is also defined by the PHB and any arm convergences. Making the uppers "converge" at infinity would take them out of the rollsteer picture entirely, and you're then left with a rather small amount of roll understeer under most conditions since the LCA's converge lower than and forward of the PHB at car centerline.

Norm
'79 Malibu (too unstock for Street Mod)

Edited by: Norm Peterson at: 8/9/04 12:01 pm

ddennis68
Registered User
Posts: 341
(8/9/04 2:39 pm)
Reply Norm P.
--------------------------------------------------------------------------------

Quote:
--------------------------------------------------------------------------------
I think I'd make the uppers parallel to centerline rather than converging them inward like the LCA's, though, as your roll steer is also defined by the PHB and any arm convergences. Making the uppers "converge" at infinity would take them out of the rollsteer picture entirely, and you're then left with a rather small amount of roll understeer under most conditions since the LCA's converge lower than and forward of the PHB at car centerline.
--------------------------------------------------------------------------------

So, I take you don't like my thought about converging the 2 UCA into 1 creating a wishbone style track locator?

Dennis-

check out progress of Bondobucket

Mean 69
Registered User
Posts: 105
(8/9/04 3:06 pm)
Reply You can, but...
--------------------------------------------------------------------------------
I am not sure that I see any benefit to making a wishbone from the two uppers, as you suggested. Doing this will require a new mount both at the differential, and to be honest, as I write this, I think you would introduce another source of bind. The only way I would see this working is through the use of a large spherical bearing for the rear mount of the upper link, this would allow the rear end to roll, but this arrangement would further add lateral restraint, so as Norm suggested, it could make things more complex with the other ideas you proposed.

If you were to do this, why not just make it simple, do a three link! Seriously though, making an a-arm for either the upper or lowers requires a very stout connection, it is probably not worth the trouble, independent of the other issues.

Man, Dennis, I am proud of you! You are not afraid of this stuff, and have learned a tremendous amount in a short amount of time. Good for you!

Mark

MarcusUSA1
Registered User
Posts: 150
(8/9/04 6:11 pm)
Reply Re: You can, but...
--------------------------------------------------------------------------------
I`m not a huge fan of converging 4 link systems. They introduce a lot of bind no matter what type of ends they use. It`s a geometry issue. I know some will argue with me but I`ve mocked up a complete actual G body rear suspension welded to the top of the workbench and there`s a great deal of non linear binding going on,period. I`ve tried it with heim ends too,same result. The issue isn`t the angle of the links per say,it`s the angle of the pivot axis on each end. Make the front and rear pivot axis parallel (which is an option if you`re fabbing your own) or at the very least parallel enough for heims to take up the difference and the bind goes away. Using converging 4 link with a PHB,sounds like a bad idea unless the links converge so little that they offer no lateral location to the rear. Otherwise they`ll want the rear located in a given location and the PHB,swinging in it`s arc will want it in another which gives us bind which is almost always bad. I`d much sooner see a 3 link with PHB. A wishbone is a possibility too but probably not for your application. A wishbone UCA would offer excellent lateral location and no binding but the PRCH will end up around 17"-18" or about where it is stock. That`s too high to be anywhere near ideal but it would let you run spring and swaybars intended for your chassis,straight off the shelf. The more you drop the RRCH the more spring and sway bar rate you`ll need to make the car corner flat and keep understeer at bay. That means you can probably forget any rear springs or bars intended for your chassis. Ask me how I know,out 3 link/PHB is going through those growing pains right now. I`m not saying that a wishbone UCA would be a GREAT solution but it`s something I`ve been toying with for awhile as a stop gap improvement. FWIW they`ve been using that setup on Land Rovers for years. That said IMO 3 link with a frame mounted watts linkage would be my idea of a near perfect setup. 3 link/PHB is almost as good and much easier to package. Mark SC&C

Norm Peterson
Registered User
Posts: 175
(8/9/04 6:33 pm)
Reply Re: Norm P.
--------------------------------------------------------------------------------

Quote:
--------------------------------------------------------------------------------
So, I take you don't like my thought about converging the 2 UCA into 1 creating a wishbone style track locator?
--------------------------------------------------------------------------------

I was simply tackling one thought at a time. This stuff can get confusing enough as it is.

Wishbone lateral locating mechanisms have been used before, both *sort of* commercially (an early Lotus sports car) and on at least one ponycar or muscle car turned track car that I know of. But in those cases the wishbone replaced the lowers in order to get the rear roll center down below the axle center height rather than putting it several inches above it. A single large pivot was located off the bottom of the pumpkin for the wishbone, and the uppers were located outboard. Reportedly, the track car worked well enough (maybe I can find my xerox copy of a magazine article about this one), and Lotus sportscars as a group are still administratively barred from competing in certain auto-x classes.

In fairness, the Alfa Giulia and its successors used a T-shaped link that functioned as an upper wishbone (but it appears that some effort was expended to minimize its axle pivot height).

It's certainly do-able, just that there's mostly less attention paid to it (with perhaps one other current exception). Hence the details are less familiar (things like all the lower loading concentrated on a single large pumpkin-mounted monoball or finding the room upstairs without adding RCH beyond that of the converging 4-link).

Norm
'79 Malibu (too unstock for Street Mod)

Edited by: Norm Peterson at: 8/9/04 6:46 pm

ddennis68
Registered User
Posts: 342
(8/9/04 6:44 pm)
Reply OK, OK
--------------------------------------------------------------------------------
I give up-you guys win. I'll stick to something basic for now and get my feet wet before experimenting too much. I figure using the stock axle housing mounts I can run a pair of LCA parallel to CL and run 1 UCA from the OE L/S axle mount to a fabbed mount on the frame mounted 1-2" lower than the axle mount for AS and running parallel to CL as well. Sorry Mark, got to go with the PHB, just too damn easy. On paper I get real ambitious but in practical application I know I'm going to take the easy out. I really appreciate everyone help on this and hope many will learn, or want to learn more as a result.

P-T. com rules, Larry where's my d@mn sticker !!
Dennis-

check out progress of Bondobucket

Norm Peterson
Registered User
Posts: 176
(8/9/04 6:48 pm)
Reply Re: OK, OK
--------------------------------------------------------------------------------
Just so you don't miss my last edit - it got time-stamped a couple of minutes after your last post. The last paragraph in particular.

Norm
'79 Malibu (too unstock for Street Mod)