I have been reading the suspension postings for awhile now and have only once seen the GNX 3-link mentioned. Can any of our suspension knowledgable guys comment on the pro's and con's of the KIRBAN GNX 3-link versus our 4-link? My MC-SS already has every frame and body brace noted in all the postings I have seen, including the rear brace and GNX #5 body bushings. I have the complete SS suspension, 12" B-body brakes, Hotchkis A-arms, springs, upper and lower trailing arms w/braces, Edlebrock IAS shocks, Baer rear disc's, and 8.5" 10-bolt 373:1 w/Eaton posi-rearend. I am really more interested in road handling than strip performance. I plan a 383 this spring and am looking at the KIRBAN GNX 3-link as an improvement over my current setup. What insight can you suspension guy's give me? I would really appreaciate your candid comments.

The Kirban device appears to be more strip oriented. It's really a traction arm, like the GNX piece that it replicates (Buick called it a torque arm, I think). It isn't a true torque arm, nor is it a real 3rd link either, since it's fixed at the axle (like a torque arm) and pinned at the frame (like a real 3-link; torque arms are only supported at the chassis against rotation, not against fore/aft motion as well).

This some-of-each construction means that you're going to get some binding of the Kirban link, some within itself and some vs the LCA's as the suspension moves. How much this binding amounts to depends on how closely the forward pivot of the Kirban link coincides with the chassis pivots of the LCA's as seen in side view (for bump, affecting ride quality) and with the car's roll axis (for roll, mostly affecting near-the-limit cornering behavior) and on how much compliance exists in the bushings. That the LCA pivot and the roll axis aren't too far apart vertically means that you can come fairly close to satisfying both criteria but you can't eliminate both sources of bind simultaneously. I guess this is the point where you have to define the car's mission. Mostly a street driver or one with a larger dose of extreme cornering (i.e. auto-x or on-track lapping).

If the K-link chassis pivot can be carefully located re the above (disclaimer: not sure how much freedom you have in this respect) and you refrain from stiffening everything back there with plain poly (or worse yet, Del-A-Lum) bushings it should be more predictable than the OE 4-link mess, though with the lower and better-controlled rear roll center height (courtesy of the Panhard bar) you'll probably need to stiffen the rear springs, rear bar, or perhaps a little of both to bring the amount of roll down and the understeer/oversteer balance back.

With the increased anti-squat you also get increased anti-lift under braking, which means there's an increased likelihood of axle hop under hard braking. I think this can be tuned away with brake balance and/or pad selection; some LS1 F-bodies have this problem and there has been some discussion over at http://www.camaroz28.com .

Norm

Greetings,

Those are great points, Norm. I've wondered about this same mod, but I hesitate because of a.) what seems like lots of unsprung mass added to the rear, both from the rear housing-mounted traction arm housing, and the added Panhard rod bracketry, plus roughly one-third of the masses of the Panhard rod itself, and the traction arm; b.) the still rather high (about 12") roll center defined by the Panhard bar pivot point on the axle.

The bind potential is very real, since the traction truss has an effective length of 30"-31" (if I remember correctly,) while the LCA's have a side-view effective length of 18.6." As you say, the system is over-constrained; perhaps allowing the traction bar to function as a true torque-arm is the way to go. (I wonder how difficult the conversion would be?)

For me, the absolute clincher is all the added unsprung mass which, just guessing, could top 40 pounds. For those with the 8.5" GN rear, which outweighs the far more common 7.5" rear by about 25 pounds, the combination would push the sprung/unsprung mass ratio almost down to a positively dismal 3:1. In my opinion, this would be intolerable.

Best,
MAP

First, thankyou Norm and MAP. This is a great beginning for me. The posting I read that leans me toward the KIRBAN package is this one: http://www.ffcobra.com/FAQ/3link.html
It discusses the mustang 4 to 3-link options extensively. The 4-link bind discussions are explained very well. I just cannot fathom it all at once. The data provided and the high number of discussion points require multiple readings. As Norm noted, the forward pivot point of the KIRBAN link is shorter than that on an F-body, but still longer than the LCA's. Being totally raw in this arena, I do not understand properly the unsprung weight issue that MAP brings up. Please correct me when I get this stuff wrong, OK? I believe Norm suggested, as does KIRBAN that the rear springs need to be stronger, and yes the weight of the unit is much heavier. The KIRBAN unit also reccommends a very large rear sway bar. As heavy as the our cars are, with say 450 to 500 ft/lbs of both torque and horse power available, what effect other than ride would be most appearant? Norm suggested the need to bring the amount of roll down. Can you explain this a little better for me? Also, what did you mean MAP when you said the roll center of the panhard rod might be too high at 12"?. The panhard rod in the KIRBAN pictures is parallel to the ground (when stationary) and is attached to the axle on the drivers side and the frame on the passenger side frame with another rod running from the passenger side frame at an upward angle to the drivers side frame. One comment in the link above suggests that a 4-link would be closer to ideal, if the UCA was a "V" ending in a single attachment point centered on the pumkin, which would almost eliminate bind. That not being feasible, the 3-link with pan hard bar would be the next best choice for street and auto-cross. The 4-link for drag. As I will opt for street handling capable of dealing safely with the power above, I keep coming back to the KIRBAN unit. It is not a cheap add on compared to the mustang 3-link options. Does anyone know a contact with a GNX? by the way. Thanks guys. Bob

I've printed out the ffcobra stuff to see what's been said and will post again later (got a granddaughter's soccer game to go watch in 1/2 hour). But briefly, I've met Klayfish (apparently just after he did swap up to the 3-link) and he was highly impressed with the changes. And he'd done a gear swap at the same time, from 2-somethings (I think) to 3.73's. That he didn't get into trouble just from that change says something.

Generally the 3-link offers the smoothest operation over the range of suspension motions, making it more predictable than any arrangement that exhibits bind. It's easier to hold it near the limit.

Norm

After looking underneath a GNX earlier this summer, I think it may
have rear steer of the axle under body roll- the PHB is both non-
level and ~5 deg non-parallel with the axle. The front TA pivot is
about 6" ahead of the rear LCA frame pivots, and at about the same level. Kirban's installation guide says " axle height determines the
location of the crossmember. This is important to do correctly as it
determines the pivot point and radial arc of the ladder bar. Make sure
the 4.2" spacer blocks are snug between the axle and frame of the car."

The distance between the rear axle CL and the crossmember pivot is 27"- very short for a TA. Under powerbraking, the rearend rises 4-5"-
with a front plunge mount, you'd need hold down straps with high torque inputs to avoid pinion angle problems. It would be interesting
to "decouple" this setup by providing a front plunge mount and by
modifying the lower arm mount so that the diff cover would contact it
under acceleration, and fall away under braking. Braking axle roll
would be controlled via a rubber biscuit style 3rd link of the type
used by oval track cars.

The frame/axle mounts of the GNX I looked at were 1/8" steel, and the
PHB & stiffarm tubing was 1 1/4" OD. The most signifcant unsprung weight would be the diff cover- I would guess 20- 25 lbs. The TA/PHB
are both 50% unsprung- ~15 lbs?, and the rest is sprung weight. Compared to the 230 lbs of a Ford 9", it doesn't seem extreme. I
would use a level PHB (like Maximum Motorsports Mustang PHB) to get
the roll center to 7-8", and would mount the chassis bracket on the
driver's side.

I'ved been working on a similar setup for my SS, and have come up with
two different TA designs. If you go past 27" with the TA, you need to
box the C-section- Mark- did you tuck your 1x3 tube inside the frame,
or did you weld it on the bottom? I'll try a level, 5 deg non-parallel
PHB with rod ends- if I still have rear steer, I'll have to go 8" out
from the axle like the circle track guys.

Bob- not many people have tried IAS on G-bodies- what"s your experience been, especially with durability? I talked with Stranoparts, and they told me they had 2 warranties out of the
dozen sets they've sold vs 1 warranty in 10 years with Bilstein. If
you want more GNX info, try: http://www.kirbanperformance.com, http://www.gmhightech
performance.com-March 2002 issue, pp47-50, http://www.buickGNX.com, http://www.corner-carvers.com, go to suspension tech & look for Decoupled
Torque Arm by Richard P (presently on pp 3)- he built one for a road
race Mustang, and has 5 races on it so far. HTH.

Axle steer is mostly a function of the LCA's and their orientation in side view. And some function of the LCA orientation in plan view. As I see it, the effect of adding a slightly angled PHB is a small indirect one, by virtue of whatever influence is added to the LCA bushing deformations. And with a long horizontal PHB at static ride height it should be minor.

BTW, I saw your post at strano - heard anything since? I'm sort of looking at Bilsteins myself, but not sure if I want to go with revalved G-body pieces up front or try off-the-shelf S-10 pieces (near identical lengths and stroke as the G-body, but with a bigger diameter).

Norm

Finding the rear steer is merely a matter of connecting the dots. Find the center of your lateral control (PHB, in this case). Find your instant center for antisquat. Connect these points together and the inclination of this line will tell you the direction and severity of your rear steer.
FWIW, I don't believe a TA/PHB setup can outperform a 4-link once you've eliminated the material execution problems (squidgy bushings and limp-wristed arms.
The popular opinion around here is that the 4-link will bind by design if it's arm lengths are fixed. This is due to some problem the 'stang guys have, and it is assumed that we have it also.
I say that everything makes perfect sense once you've got it figured out properly, and this simply doesn't make sense. Geometrically, there is no source of bind in this system. More importantly, I have a source in local circle track asphalt racing for lots of empirical data. I'm satisfied that the 4-link bind problem does not exist for the GM metric chassis.
Dispensing with the lateral rigidity/ bind problems, what is left in favor of the TA/PHB setup?
Different and adjustable rear steer characteristics. This is nice, and preferable for circle track racing, where every 100th counts, but for us? Rear steer is still adjustable with ride height and pickup point placement. Unfortunately, this also requires changing the antisquat%. Point in favor of the TA/PHB.
The disadvantage is lateral axle movement with ride height, which makes nasty rear bump steer. Is the tradeoff worth it? I don't think so. I'd rather pilot a 4-link over a bumpy course. This is the main reason why the 'stang and Camaro were so maligned in car&driver and other magazines. They incorrectly referred to it as "axle tramp", but it wasn't really.
I believe that the 4-link is actually preferable to the TA/PHB arrangement. Now throw in a decoupled TA, and it's a whole new ballgame...

Just as a little friendly rebuttal, John . . .

The Mustang is a converging 4-link while the late-model F-bodies are TA/PHB. If they both were maligned in C&D etc., then it's not the specific method of axle location that's at fault. More likely it's a complaint being levied at the live axle and its poor unsprung weight issues.

Regarding the C4L and bind. When the car rolls, either the UCA's have to change their effective length or the roll axis has to take a hike upward for geometric compatibility. Actually, you probably get some of each with most bushings (including poly), and a roll center height change only with all rod ends.

Given that you'd be using the longest practical PHB and mounting it so that it was horizontal at static ride height, the lateral axle motion will be pretty small. Smaller than the lateral motion that is permitted by the OE rubber-bushed UCA's, and not much more than with an all rod-ended configuration. Something like 0.01" of lateral axle movement per inch of rise/fall of the chassis. It's there, but it's not a big enough thing to worry about.

The ffcobra site referenced has a couple of good pictures that illustrate the all rod-ended arrangement. It's about 3/4 of the way down (page 23 of 30 on what I printed out). Yes, it shows that the axle can roll with respect to the chassis. But it's much harder to determine from them just what has happened to the roll axis of the car (most of which is not present). Particularly since the two pictures were taken from slightly different points. And it's also worth noting that the LCA's in that model appear to be parallel, not slightly converging as is the actual case.

One specific question that I have re the circle track guys running the metric chassis or any other similar beast - are they limited to the C4L general arrangement by class rules? If so, then their success is in spite of the design, not because of it.

Norm

Nope, sorry ...
I should have put that differently; The motoring press complained of "axle tramp" from the Camaro, but not from the 'Stang. Take a look at the reviews. As a matter of fact, they noticed (quite correctly) that the 'stang didn't perform as well with the IRS. They chalked it up to Ford's not knowing how to put together an IRS system (hehe).
Yes, the roll center moves vertically by necessity in an incompressible lattice, but it also does so with the TA/PHB. No advantage either way.
AFA the axle displacement with ride height, you've just got to experience it behind the wheel. I know I push a G-body off a lumpy exit much harder than an F-body. Not because the car won't do it, just that it doesn't feel like it'll hold. I know it's minor, but it feels very major. Moreover, it creates rear steer in response to bumps, as opposed to just general squidge. As I said, you've just gotta try it. Catch a ride with an F-body type sometime on a bumpy course. Pay attention to the rear end and how it responds to bumps. Afterwards, ask yourself if that's really what you want, 'cuz any setup with a PHB will do this.
Very specifically, no, the class does not mandate the C4L, He just runs it to prove he can (and wins quite often, btw). He's of the same mind on this one. They used to have a weight break for running it, but it's disappeared bec's the other arrangements are no quicker. I suspect this is due to a very lumpy turn 4. (the "dogbone" decoupled TAs are illegal for the class, which also says something).
The design isn't as bad as it's made out to be, guys. Certainly room for improvement, but I don't believe the TA/PHB is any better. Certainly not better enough to justify the outrageous price tag, unless you're going for sporty grocery getter and not all-out handling.

I always bring this up in these discussions and it`s never been covered well enough to satisfy my thirst for knowledge in this regard. Take the C4L and replace the upper arms with a wishbone and single pivot on top of the pumpkin. No binding I can see,the axle remains centered at all times and the rest of the geometry can remain pretty much the same. Problems with this are? Marcus

It can't roll without binding, unless there's something I'm not picturing...

Norm-

I spoke with Bilstein about spring ratings- they said 20% difference up or down before revalving was needed. The rear PN for the Monte
is also used for the Caprice wagon and Impala- spring ratings for
these can go to 142 (Impala) and 173 (Caprice) lb/in for HD constant
rate springs.

Marcus-

One of my late 80's Circle Track magazines profiled a Camaro with
a triangular LCA mounted to the diff with a 1 1/4" rod end. It was
being raced successfully at a California track, and also had a revised
front end setup. I can send you a copy of the article if you can
give me a snail-mail address (my email is davidjsaunders@hotmail.com).
down was

Thanks, Dave. That means that probably nothing special need be done in the back for my 165-ish 5413's.

Marcus - Dave's CT article was the one that I was trying to think of earlier. From the few pix I've seen of this arrangement, you either end up with something in single shear or a pair of cantilevered brackets being responsible for resisting some sizeable loads. Although it seems that a little thought and the refusal to worry about another couple of pounds of brackets ought to fix that. Local differential case deformations under these loads might be a concern with really big power and good grip combined with some specific axles.

Norm

I got lost, right after "I have been reading the suspension postings for awhile".

Supernova455

Don't feel bad. Suspension stuff is probably the last thing that the majority of MCSS owners investigate in any great detail. And this thread has gone well beyond spring swapping and shock choices.

It's all about swapping the rear suspension out for something entirely different, with the intent of improving handling/cornering. The Kirban pieces replicate the extremely limited production GNX's rear suspension and work quite differently from the 4 skewed links that are found under the back end of all other G-bodies. The other stuff was never under any G-body at all (save, perhaps, for factory development mules) As you have perhaps noted, opinion isn't exactly unanimous on the direction to take once you've decided to do it or even whether to do it at all.

If you're into reading up on things like this, I've got a list of books that deal with this.

Norm

Blue_,nope no binding at all. I built that very setup (machined toolsteel links with ballends and coilovers) for both axles on one of my welded steel tube frame 1/10 scale RC monster trucks and it freely articulates 45 degree on each end for a total of 90 degrees! It handles much better than it did with the converging 4 link I had on it before too. FWIW Land Cruisers,Isuzu Trooper2s and RangeRovers use this type of rear suspension so if it flexes well enough for offroad use it should roll plenty for a street Monte. As for upper mount strength,a mount could be made using the stock mounting ears on the rear with say a 3/4" or even 1" hardened shaft going through it,spacers on each side and a 3/4" or 1" chrome moly heim joint for the center pivot. That would be much stronger I`d wager than the flimsy stock sheetmetal arms. Stock mounting points could be used on the frame side but new brackets would have to be fabbed so that each leg could pivot on the same axis without binding. No biggie. In side view the suspension remains exactly as it was,antisquat is unchanaged (unless you want to change it). Roll center height would seem pretty good to me and should be very stable laterally. Seems it would be pretty effective to me and it`s so simple. So again,why not use it? Marcus

Marcus,
Sorry, I just can't picture it. Could ya send me a sketch? My e-mail is JKingTech@Netscape.Net

Norm,
Isn't it refreshing posting with a group who disagree about everything, yet get along so famously?

Yup. And sometimes that little bit of disagreement ends up forcing some second (third?) looks at things. And a on me too. At Cheers!.

About that upper wishbone arrangement – the axle end pivot for the upper (now a single wishbone) link is now a ball joint rather than any kind of cylindrical bushing. The only bind that would develop would come from whatever moment resistances exist in the LCA bushings. Rear roll center is fixed at the ball joint center for all combinations of bump/rebound and roll. Really, it's what the OE pair of converging uppers is trying to be, just that it has a real physical pivot instead of a virtual pivot for the pair.

Norm

I see what you're saying...
So the roll center is very high, which suggests a possibility of jacking, but decreased roll moment. An antiroll bar would probably be superfluous.
Also a possibility of inducing roll steer unless the lower rear pickup points are moved down.
I can see how this is doable. Certainly it's more rigid laterally than the stock arrangement.
This would require soft rubber at the frame end (or spherical bearings) to avoid pounding out the holes during vertical travel. unless there's some way to make the frame ends articulate along the same axis.<----On second reading now that I have a clue, I see you've already thought of this!
But with a fixed pivot point, your lowers will need to articulate more. Picture the suspension rolled 45*. Now change the vertical displacement. See how busy the lowers are? In the C4L, this same lateral articulation is needed, but it's divided between 4 arms instead of 2. I don't think this can be done without spherical bearings on the lowers at all 4 corners.

I believe you`ve got it! True the lower arms have to move a bit more when in roll but forget 45 degrees (although 1/10 scale heims handle this on the LCAs of my truck) and try 10 degrees. I think stock "flex-o-twist" LCAs with rubber bushings could handle it. Monoballs would be better for hardcore use of course. Maybe it`s a deadend and a 3 link and PHB would be easier and more adjustable but I`ve always wanted to try it on some thing full scale. Marcus

"I've always wanted to try it on something full scale"
That's all the reason you need. Question is, how do you fab up something modular with the stock pickup points (just in case it doesn't work out)?

Greetings Folks,

Another outstanding thread! (Before I delve into specific comments, first a clarification: I've always used the term "bind" to indicate a condition of complete immobilzation. I've noticed here that people use it rather to indicate any tendency toward immobilization. Sorry if my previous posts at Turbobuick.com have been misleading as a result!)

Comments in turn:

BobK 87: that Mustang thread was a fascinating one. About 4-links tending to bind - this is real, although the tendency is normally small. If we look at a basic 4-link with angled UCA's that are separated horizontally at their point of attachment to the differential case, and with LCA's that are parallel to each other and with the ground, as well as parallel to the UCA's in side view, then zero bind will occur as the rear rolls, provided that the rear ride height doesn't change in the process. In general, bind will increase as 1.) the UCA's and LCA's depart from nominal parallelism with the ground, 2.) the separation between the UCA anchor points on the differential case widens, 3.) the LCA's depart from parallelism in a top view, and instead converge toward the front of the car as they do in the G-bodies (the included angle of convergence is 30 degrees by design,) and, 4.) as roll angle increases. This isn't an exhaustive list by any means, but it's a start. (In saying this, btw, I'm assuming that the arms are unconstrained under rotation, and thus offer negligible reaction moments to any angular deflections.)

As some have pointed-out here, the compliance of the suspension bushings, plus the less-than-infinite torsional rigidity of the entire rear axle assembly, make this a problem of little practical consequence under typical circumstances.

About a live axle's high unsprung mass: now this *is* a matter of profound practical consequence. A car's wheel is a fourth-order mechanical system: there's the mass of the car at that wheel, sitting on the spring, dampened by the shock absorber. The spring, in turn, sits on the mass of the wheel, which, in turn, sits on the dampened compliance of the tire. So we have two masses, two "springs," and two dampeners. The mass of the car sitting on the car's spring is the "sprung mass." The mass of the wheel is the "unsprung mass." The car's spring and shock absorber constitute the vertical coupling between the car and wheel. The mass of the wheel, in turn, is coupled to the ground by the compliance and dampening of the tire. Since we have little direct control over these last two elements, we usually ignore them.

But for chassis-tuning purposes, the interaction of all six elements is crucial. If we have a high unsprung mass, and a stiff tire, we tend to get a high-Q resonance between the wheel and ground, which is only partially controlled by the car's shock absorber. As a result, as a heavy rear with performance tires travels over a bumpy surface, it will tend to "jiggle." This has two consequences: 1.) it creates a rough ride, and 2.) it causes an unusually wide variation in the force the ground exerts on the tire. *This latter factor is the clincher*, since if one is trying accelerate *while* traveling over a bumpy surface, then one is likelier to lose traction when the contact force between the ground and tire drops low.

For these reasons, one would like to minimize unsprung mass. There's another way to look at the same problem: the maximum downward vertical acceleration that the wheel can experience as a result of the car's sprung mass sitting on top of it, is simply g*(sprung mass/unsprung mass). In the case of the G-body, where the rear sprung/unsprung mass ratio varies (depending on the configuration from the factory) between roughly 4:1 and 3:1, the rear can only be accelerated downward to a maximum of 3 or 4g's. As you can see, if the rear happens to be traveling over an especially bumpy surface, or at a high speed - or, worse yet, both - then the rear can easily lose all contact with the ground over dips. Needless to say, one wouldn't want to be making any fast turns, or stomping on the throttle, at such an inopportune time.

So that's an introduction to why high unsprung mass is bad for acceleration over bumpy surfaces. Over smooth surfaces - no problem! That's why this is rarely a consideration at the drag strip. But if you want drag-strip acceleration on the street, you'll be fighting unsprung mass at every bump and dip.

About why the Kirban rear would need more roll stiffness than the stock 4-link: the imaginary intersection of the UCA's defines the rear's roll center. The design height at rest is 18.0". The Kirban's roll center is defined by the PHB's pivot point where it attaches to the rear axle, which, judging by pictures I've seen of it, seems to be about 12" above the ground. If we assume that the car's center of gravity is fixed in height, and that the roll center height of the front suspension is fixed in height as well, then the Kirban rear will drop the car's roll axis (defined by a line joining the front and rear roll centers) by about 6" at the rear, or roughly 2-1/2" beneath the car's center of gravity, assuming a 58/42% F/R weight distribution. This drop constitutes a sizeable increase in the moment arm exerted by the centrifugal force acting on the car during a turn, and so the car will roll more for a given sideways acceleration. To restore the car's lean in a turn, roll stiffness must be increased by the same factor by which the centrifugal force's moment arm length increases. Assuming, very roughly, 1.) zero front roll center height, 2.) a decrease in rear roll center height from 18" to 12", 3.) a 58/42% F/R weight distribution, and 4.) a center of gravity height (guess only) of 20", we get that the Kirban rear swap will change the centrifugal force moment arm from roughly 12-1/2" to 15". As a result, the car's roll stiffness would need to increase by roughly 20%. One caution, though - this doesn't mean that the increase should be accomplished solely at the rear - otherwise the car may tend toward oversteer. Rather, roll stiffness should be adjusted at both the front and rear, so as to generate the desired handling balance between understeer and oversteer.

Incidentally, concerning the biggest advantage that the IRS has over the live rear, namely - unsprung mass, check out this thread at Turbobuick.com, from several months ago:

http://www.turbobuick.com/forums/showthread.php?threadid=41235

MaineSS: If the TA is 30-31" long, which is consistent with your statement that the forward TA pivot point is some 27" forward of the axle CL, and the effective swing length in side-view of the LCA's is 18.6", then the two pivots points must be separated by more than 6" in side view - can you confirm? About the short TA swing-arm length and the high anti-squat: Herb Adams strongly recommended against such a short combination. As has been mentioned here, what you gain in traction under anti-squat, you lose in traction in "anti-anti-dive," since the wheels are unloaded significantly during any hard braking. Some racers at the TB.com site warned about weird and potentially dangerous acceleration/braking transition behavior at the big end of the track as a result. However, as you said - some means of changing this dynamic balance as a function of forward accleration/deceleration, should mitigate the problem. Herb Adams is a big proponent of this system as well.

About the weights: I've got 225lb without brakes for the 8.5" GN rear, which probably makes it quite comparable to the 230lb you quote for the Ford 9" (with or without brakes?) About coupling masses: for objects undergoing linear strain (such as springs) or for objects which rotate about fixed chassis points (such as control arms,) and assuming uniform mass distribution along the object in question, the dynamic coupling mass is 1/3 of the total mass, measured at the "wheel" end. The derivation isn't too hard, but I'd admit that the result isn't intuitive.

About the side frame reinforecments I used: I bolted them along the bottom of the existing frame rails, such that roughly half of the 3" width of the 1"x3" box tubing I used for the reinforcements, projected inside the innermost part of the frame rails. This makes a very handy "shelf" for anchoring things like custom transmission cross members (as I did,) or any other cross member, for that matter.

About "IAS" on G-bodies - what is this? And Stranoparts - what's their URL? Thanks!

Blue Hog: I agree that fundamentally, a 4-link isn't inferior to a 3-link or TA with a PHB. In fact, "flip" the stock 4-link rear that GM designed upside-down, and you've got Herb Adam's beloved Satchell link! But practically speaking, the problem with the GM 4-link is the absurdly high location (18.0") of the rear roll center, created by having to mount the UCA's rear pivot points, near the top of the differential case. As I wrote in my "Anyone swap an IRS into a TR?" post over at Turbobuick.com, the rear moves as a dog "with a hip out of joint," as was critiqued in the TV series "Motorweek" many years ago. It's not good for handling by any means: when the car goes over a bump while in a turn, the car's center of gravity resists lateral jerks. This creates a lateral reaction jerk at the rear axle's tire contact area with the ground, which thus reduces traction. And so we have a double-whammy for the heavy GM live rear with a high roll center: the tires definitely don't want to stay planted on the ground when negotiating a turn over a bumpy surface! It's even a *triple* whammy if we add stiff, short-sidewall tires to the mix, since these are far less forgiving of lateral jerk inputs.

Now add any simultaneous throttle/brake inputs, and we have a recipe for, well, at the very least, "spirited sliding."

The live rear desperately needs two things: 1.) to go on a diet (one approach would be to mount the sway bar on the frame, and couple to the rear via links, such as the Wolfe bar,) and 2.) get the rear roll center height down! To reduce the roll center height much below the existing 18.0", one must abandon the existing 4-link design. The only other feasible 4-link method would be to convert it to its "flip" cousin, the Satchell link - but how and where do you anchor the LCA's at the rear? And do you cut the floor to make the space to anchor the forward end of the UCA's?

That's why one is virtually forced to convert to a 3-link or TA and PHB.

Jumping down to your next post, you mention that you felt you could "push a G-body off a lumpy exit much harder than an F-body." This is very interesting. Why do you feel that the F-body rear just "won't hold?" Is it the rear-steer characteristics of an F-body rear when the PHB isn't parallel with the ground, or just mushy suspension bushings? Unsprung mass should be comparable. BTW, the G-body rears rear-steer as well: they were designed with 10% rear roll understeer from the factory (i.e, for every 1 degree of body roll, the rear rotates 0.1 degree when viewed from the top, in the direction toward understeer.)

Marcus: about converging the 4-link rear UCA pivots to a single pivot - yes, this should eliminate all bind (assuming control arms which are completely free to rotate,) since three non-colinear points define a plane.

Maine SS: Bilstein uses the same rear PN for the Caprice wagon and Impala? What an enormous potential range of weight, especially if the wagon is loaded! Someone here (was it Norm?) mentioned that the S-10 Bilstein shock was more resistive than the standard G-body shock - this should be good for setups with unusually heavy rears, such as the 8.5" GN rear with a thick sway bar, for instance; the reason is the high-Q vertical resonance I mentioned before, for heavy rears with stiff tires.

Norm: you mentioned that you've got a list of books about suspension design - could you post this? Thanks.

Best regards,
MAP

Re. MAP,
"Jumping down to your next post, you mention that you felt you could "push a G-body off a lumpy exit much harder than an F-body." This is very interesting. Why do you feel that the F-body rear just "won't hold?"
It's not a matter of roll steer, that's predictable. It's the bumpsteer. In a hard left turn, the TA/PHB tends to "jerk" the rear end away from the turn, whereas the C4L doesn't. This is directly attributable to the arc that results from a finite length PHB. The C4L does not suffer this same drawback, as it's vertical travel is linear. Understand, this probably does not affect the handling of the system, but it does affect driver confidence. This is often overlooked, but nonetheless important. When you're coming off a turn at 50 and there's a wall, your instinct is to back off when the rear end jumps out, regardless of the reason.
I agree that the RCH is too high for this system, but lowering the RCH will not materially affect the inherent drawbacks of the live axle. It can tolerate alot of RCH anyway because it's roll IC equals the track. There is no cure for high unsprung weight, other than a diet, as you said.
You end up with a tradeoff; more antisquat or less unsprung weight. The determining factor is the track itself. Smooth tracks prefer live axles, bumpy tracks like IRS (but it's gotta be reeeal bumpy).
Since the TA/PHB has such minor advantages over the existing C4L,and some disadvantages, How then to justify the expense of a conversion? It's like trading in your Civic for a Hyundai.
I maintain that it's not worth doing unless it's done right. Bring in a decoupled fully adjustable TA setup with a Watt's link, and you'll have something worth the effort. Otherwise, the money's better spent making the C4L perform like it was designed.
How much $$ to convert? How well will it run compared to the original C4L with the same money spent to make it work better? See what I mean?
BTW, this is a really engrossing thread!

I took a ride out to Marcus's shop a couple of days ago and this very thing came up in conversation. On possible direction to go might be to run something between the existing ears for the UCA's to mount the pivot on. I've seen pictures of an installation where the passenger side ear on a Ford 8.8" rear was used to stabilize an offset 3rd link axle bracket that was otherwise just clamped around the axle tube. You'd also need to pick up a couple of the diff cover bolts to eliminate rotation of this added bracketry as seen in side view.

I'm thinking that the difference between the rear roll center always hiking up by a pretty constant amount regardless of which way you're turning (C4L) vs moving in either direction as a function of the roll angle (TA/PHB) is part of the issue with "feel" here.

With that in mind, perhaps a real 3-link is the best way to go. Even one of the TA/PHB's strongest supporters on a number of other forums concedes that the 3-link has potential advantages over the TA.

Norm