MonteCarloSS.com
MonteCarloSS.com

CELEBRATING 20 YEARS!

Previous Thread
Next Thread
Print Thread
Page 1 of 5 1 2 3 4 5
#1068445 - 11/03/20 08:38 PM SS rear suspension: thoughts and musings  
Joined: May 2002
Posts: 4,324
MAP Offline
15+ Year
MAP  Offline
15+ Year
Member

Joined: May 2002
Posts: 4,324
Yuma, AZ
Hi Folks,

I thought I'd try to consolidate various thoughts and observations about the MCSS rear suspension that has been evolving in this forum over the past two decades. Hopefully, this will be of benefit to those reading here who want to improve their car's handling. I must say that the body of information and testing contained in this forum is truly impressive, especially if we explore what was written before roughly 2005. Earlier posts were more theoretical than practical while later posts have been more practical than theoretical. Both aspects are necessary, but if we have good theoretical comprehension to guide the practical, then we can advance the state of the art. If we only have the practical, then the best we can do is copy/paste to maintain the state of the art.

1. To settle the biggest question of the last two decades, the OE converging four-link design ("C4L" henceforth,) most emphatically does not geometrically bind. Roughly 18 years ago I posted a mathematical proof that for relatively small deflections for which certain approximations could be made (e.g., sine (theta) proportional to theta, the first term in the sine Maclaurin series expansion,) there is no tendency to bind. Larger deflections get very messy mathematically. Around 2002 Marcus (Mark Savtiske, author of How to Make Your Muscle Car Handle) posted that his 1:10 RC scale model cars with the C4L rear didn't bind even under extreme roll. In this same pragmatic spirit, I just built and tested a 1:5.76 scale model of the OE C4L rear using parts with, for all practical purposes, infinite longitudinal stiffness and almost zero stiffness/resistance under rotation. I tested a range of motion that is roughly double that possible in an actual MCSS. First, with zero roll, there is no geometrical bind with full up/down motion. No surprise here, of course. Second, I also tried the full range of roll in conjunction with full up/down motion, and once again, zero geometrical bind. Surprise, maybe!

2. What do I mean by geometric bind? I mean a condition where a deflection in the suspension would cause the distance within all four pairs of pivots points embodied by the four CAs to change. This is equivalent to saying that if we had CAs and bushings/bearings with infinite stiffness in length, that is - the condition approximated by the model - then geometric bind would cause the rear to seize-up under deflection.

But having said this, it's quite clear that any deflection of the rear causes the OE cylindrical bushings at all 8 locations to undergo not only internal cylindrical torsional shear but also a misaligning twist between the ID bushing sleeve and the OD bushing sleeve. This misalignment is surprisingly pronounced and pronouncedly non-linear as the rear suspension moves, especially with the rear's short UCAs. No wonder GM used soft rubber bushings! The substitution of TPU cylindrical bushings at these 8 locations was the universal "upgrade" that folks used here years back, but the easily order-of-magnitude higher modulus of TPU versus the OE rubber meant that even small rear suspension motion created extremely high internal bushing stress that, in turn, tended to create very high friction of bushing motion. I'm convinced that this high friction was pronouncedly telegraphed back to the driver, creating the impression, but not strictly speaking the reality, of bushing bind. Stiction was also possible. The result was a twitchy, nervous handling over bumps and on smooth surfaces with sudden steering inputs. No wonder this behavior was thought to be an inherent geometrical bind problem with the OE C4L. But the model disproved that beyond any reasonable doubt...

3. Mass
  • OE C4L rear, with typical wheels (about 50lb each,) OE aluminum drum brakes, and 7-1/2" ring gear, including 1/3 mass of springs, CAs, driveshaft, and shocks: roughly 285lb.
  • As above, but GN 8-1/2" ring gear rear: roughly 310lb.
  • As above, but with typical rear disk brake conversion: roughly 335lb. I'm guessing that most folks who change their rear axles are close to this mass.
  • Sprung mass over rear axle: roughly 1,250lb. 1/2 -full tank of gas, no vehicle occupants. N.B.: rear sprung/unsprung mass ratio: roughly 3.7:1.
  • OE front, with typical wheels (about 50lb each,) larger aftermarket brakes, including 1/3 mass of springs, CAs, and shocks: roughly 220lb (again, both wheels.)
  • Sprung mass over front axle: roughly 1,850lb, +/- a lot, depending on choice of motor and transmission. 1/2 -full tank of gas, no vehicle occupants. N.B.: front sprung/unsprung mass ratio: roughly 8.4:1.
  • Ford 9" unknown. Does anyone have data?




Last edited by MAP; 11/03/20 11:21 PM.
#1068451 - 11/04/20 02:36 PM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: Apr 2011
Posts: 1,193
Hunter79764 Offline
10+ Year
Hunter79764  Offline
10+ Year
Member

Joined: Apr 2011
Posts: 1,193
Grand Prairie, Tx
My stream of consciousness response, which I think is just restating what I'm hearing you say, is as follows.
So in the absence of mechanical bind, what are the factors that cause the rear to "misbehave"? I've understood it to be (and purely theoretical, I'm still running stock rear suspension) bushing compression forces and torsion in the connecting links causing a sort of non-linear anti-roll action that is heavily dependent on axle elevation. The fix for that seems to be either leave things as flimsy as possible (GM's approach originally) or stiffen everything up, but provide for adequate degrees of freedom at each end, a la Roto-Joint etc.
Based on that, it seems that the basic goal of an upgraded suspension would be to have stiff links with free rotating ends, and achieve any roll stiffness via roll bars and any roll resistance via shock absorbers in an effort to get all of the suspension jobs done by discrete components instead of having one component carry over affects to a different parameter. Conversions to 3-link seem to be in line with that, just another way to skin the cat. And any of that is somewhat independent of actual link geometry, which is a different discussion entirely.
I'd have to think about it, but with the factory geometry and free rotating ends, is there a roll stiffness due to the torsion in the axle housing when one side droops and the other compresses? I'd expect it to not be significant in normal driving (as compared to rock crawlers etc.), but that might be another area of non-linearity that the bushings otherwise mask/muddy.

I think part of the "Debate" has really been about terms, more often than not. It is a relatively known condition (as you said) that putting stiff bushings and/or boxing the control arms without having roto-joints at enough locations causes "Bind". It has been argued many times that it does not cause "bind", but people remain convinced that it does. Maybe we substitute a different term in place of the loaded "Bind" and say "Axle-No-Move-Smoothy-Smoothy", and most people can all get on board with that. Is that a fair statement?


Shawn

'85 MC with budget 5.3L swap, TH350 with stock 2.14 rear end
It ain't much off the line, but it's nice on the highway
#1068452 - 11/04/20 03:56 PM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: Dec 2007
Posts: 6,504
SSLance Offline
10+ Year
SSLance  Offline
10+ Year
Member

Joined: Dec 2007
Posts: 6,504
Peoria, AZ
I'm with MAP on all of the above. I've seen first hand how delrin lined rod ends completely free up the C4L and it can move in any direction without bind or stiction.

And like Shawn says above, once you free all of that up and control roll with shocks, springs and bars as one should, you can make the C4L work just fine.


Lance
1985 Monte Carlo SS Street Car
#1068456 - 11/05/20 01:03 AM Re: SS rear suspension: thoughts and musings [Re: Hunter79764]  
Joined: May 2002
Posts: 4,324
MAP Offline
15+ Year
MAP  Offline
15+ Year
Member

Joined: May 2002
Posts: 4,324
Yuma, AZ
Hi Shawn,

Quote
My stream of consciousness response, which I think is just restating what I'm hearing you say, is as follows.
So in the absence of mechanical bind, what are the factors that cause the rear to "misbehave"? I've understood it to be (and purely theoretical, I'm still running stock rear suspension) bushing compression forces and torsion in the connecting links causing a sort of non-linear anti-roll action that is heavily dependent on axle elevation. The fix for that seems to be either leave things as flimsy as possible (GM's approach originally) or stiffen everything up, but provide for adequate degrees of freedom at each end, a la Roto-Joint etc.


Exactly. The OE rubber bushings are so soft that they can easily handle purely cylindrical shear (i.e., the kind of deflection you get in the front suspension bushings,) as well as the "weird" twisting action between the ID and OD sleeves that you get with all eight rear bushings with pure up/down motion, as well as pure roll, plus any combination thereof. This weird deflection occurs whether the CAs maintain their low torsional stiffness per their original open-channel design, or whether the CAs are radically stiffened torsionally by boxing-in the arms, the "other" popular mod with the rear suspension from years ago (obviously the bushing distortion is far worse with the boxed arms.) I tried the TPU-bushing/boxed-arm approach years ago and thoroughly regretted it. Another problem with TPU bushings is that all the units I've ever seen have a bit of a compression fit between the bushing material proper and the ID/OD sleeves defining the bearing races, so that even without this "weird" twisting effect, motion tends to be accompanied by lots of friction. This friction, which is kind-of visco-elastic as well as kind-of Coulombic, tends to induce a lot of harshness within the NVH spectrum. Again, it's easy to confound with geometrical bind.

But the trouble with the soft OE bushings is that they allow a lot of transverse compliance in the location of the rear axle. So much so, in fact, that Marcus reported that he could install a Watt's linkage set to a radically lowered rear RCH without inducing bind under roll! On the plus side, however, one-wheel bumps with soft bushings tend to induce a slight deflection steer that helps maintain the car's directional stability.

Summary: no matter the cylindrical bushing material, we get lots of rear handling mushiness and/or twitchy, sticky, nervous handling. Both register most unpleasantly with the driver. However, it's a mistake to attribute this problem to the C4L configuration as many have seemingly done in the past. CAs terminated with pivots that are unconstrained under rotation (Roto-joints, Heim joints, and the like,) are definitely the remedy, like what Lance did.

Shawn, as to everything else you said, I agree. I never measured the torsional stiffness I had in my old rear suspension solely arising from the TPU bushings, but I suspect it was a sizeable fraction of the total roll stiffness. And yes, it's also non-linear; that is, getting stiffer (and stiction-ier) as we increase suspension deflection.

Last edited by MAP; 11/05/20 01:39 AM.
#1068458 - 11/05/20 01:36 AM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: May 2002
Posts: 4,324
MAP Offline
15+ Year
MAP  Offline
15+ Year
Member

Joined: May 2002
Posts: 4,324
Yuma, AZ
Hi Lance,

Quote
I'm with MAP on all of the above. I've seen first hand how Delrin lined rod ends completely free up the C4L and it can move in any direction without bind or stiction.

And like Shawn says above, once you free all of that up and control roll with shocks, springs and bars as one should, you can make the C4L work just fine.


Thanks, Lance. I also agree a C4L can work fine. Now, setting rear RCH and dealing with the low sprung/unsprung rear mass ratio in our MCSSs are other problems, but these aren't the fault of the C4L.

Last edited by MAP; 11/05/20 01:36 AM.
#1068460 - 11/05/20 05:05 AM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: May 2007
Posts: 5,028
Buick Runner Offline
10+ Year
Buick Runner  Offline
10+ Year
Member

Joined: May 2007
Posts: 5,028
Dharma station 1 the Hydra
Are you saying that C4L suspensions require joints that allow for angular movement along a primary axis of movement?


SBC powered 1987 Regal with TES headers, ZZ4 intake, ZZ4 PROM chip, mini starter, THM2004R, 2500 stall converter, 2040 cam, CCC system, and 3.73 posi rear.

2008 ex NPS P71 Crown Victoria, cop motor, cop shocks, cop brakes, and Jmod.

Never argue with an idiot.
They will just drag you down to their level and beat you with experience.
#1068467 - 11/05/20 08:36 PM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: May 2002
Posts: 4,324
MAP Offline
15+ Year
MAP  Offline
15+ Year
Member

Joined: May 2002
Posts: 4,324
Yuma, AZ
Hi Buick - the joints need to be completely unconstrained under rotation because motion occurs around all possible triads of orthogonal axes. That's why cylindrical bushings like the OEM design are a poor solution, but they worked well enough back in the day because tires and bushings were very soft by today's standards, and thus hid a multitude of suspension sins, which the OEM exploited to the absolute hilt to make all kinds of cost and weight cuts. The very high rear RCH, for example, allowed the factory to eliminate a rear sway bar for most A/G models.

Last edited by MAP; 11/05/20 09:37 PM.
#1068474 - 11/06/20 03:21 AM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: May 2002
Posts: 4,324
MAP Offline
15+ Year
MAP  Offline
15+ Year
Member

Joined: May 2002
Posts: 4,324
Yuma, AZ
A few more observations:

  • The stock rear geometrical understeer is about 10%. That is, for every degree of body roll, you get a tenth of a degree of axle steer toward understeer. Understeer is generally good for directional stability, especially as speed increases. But it isn't the fastest way around a corner.
  • The large difference in side-view length between the UCAs (about 7.95") and the LCAs (about 18.52") means that pinion inclination varies greatly with up/down suspension motion, tending toward nose-down with high deflection up and down, and toward nose-up at neutral ride height. Trying to set pinion angle with great precision tends to be futile since it varies so much with suspension motion.
  • Rolling back the clock to the design priorities of 1975, the length of the UCAs was no doubt minimized to maximize rear-seat leg room. Handling linearity was no doubt accorded minimal priority. Since the A/G body muffler was mounted transversely right behind the rear axle, there was little room to lengthen the UCAs by moving the differential mounting points more rearwardly.
  • On a stock 1978 Malibu I measured, I was surprised to find that the rear anti-squat was 48%. If the rear mounting points of the LCAs are dropped by about 2.4" and the UCAs are left untouched, then rear roll steer can be canceled and anti-squat increased to about 128%. (Assumption: COM is located about 20" above the pavement and about 50" behind the front axle.)


Last edited by MAP; 11/06/20 04:43 AM.
#1068475 - 11/06/20 03:37 AM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: May 2002
Posts: 4,324
MAP Offline
15+ Year
MAP  Offline
15+ Year
Member

Joined: May 2002
Posts: 4,324
Yuma, AZ
Definition of coordinate system: a Cartesian right-handed orthogonal triad consistent with the convention commonly used for cars, aircraft, and ships: X is the car's longitudinal centerline (CL henceforth) with + forward. Y is transverse with + to the left. Z is vertical with + pointing up. The origin of the system is the car's center of mass.

The stock rear sway bar creates stiffness in roll by the torsional twisting in Y of its transverse section, and by bending about Z at the corners where the transverse section turns into the side arms that are bolted to the LCAs. The 1:5.76 model I created surprised me in that even modest roll creates a lot of bending at these corners, suggesting to me:

  • A significant portion of the sway bar stiffness in roll is developed by the deformation of the bar's corners. I have no quantitive data to get a good handle on what friction of total bar stiffness is derived this way, however. But, we can deduce that this fraction should decrease as bar diameter increases, because the corner bending stiffness for a constant bend radius should go nearly as r^3, while bar torsional stiffness will go as r^4.
  • The stress imparted to the sway bar side arms in roll very likely transmits a lot of force to the LCAs, and may become critical if the LCAs are left unboxed and the bar diameter is quite high, like the 1.50"-diameter Herb Adams bar of years past.
  • Bar designs that use pivoting end links are a much better solution despite higher cost and complexity, in that they not only require less bar diameter to achieve a desired roll stiffness (because the bar side swing-arm length can be made much shorter than the OE length, which amounts to the LCA length,) but because we can remove some of the bar's mass from the rear's moving mass. This coupling mass is minimized by mounting the bar transverse section to the frame and the side swing arm through end links to the axle.
  • And speaking of bar mass, using a hollow transverse section reduces mass in relation to bar roll stiffness, since the bar stiffness goes as Rout ^4 - Rin ^4 while bar mass goes as Rout^2 - Rin ^2. The wall thickness can't get so thin as to permit elastic shear buckling, however.


Evolving acronym decoder ring:
COM = center of mass
LCA = lower control arm
UCA = upper control arm
C4L= factory rear suspension configuration - "converging four link"
RCH = roll center height
TPU = thermoplatsic polyurethane
OE = original equipment
OEM - original equipment manufacturer
ID = inner diameter
OD = outer diameter
NVH = noise, vibration, and harshness. The first is auditory; the latter two are tactile.
GN = (Buick) Grand National.

Last edited by MAP; 11/06/20 04:29 AM.
#1068484 - 11/07/20 03:48 AM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: May 2002
Posts: 4,324
MAP Offline
15+ Year
MAP  Offline
15+ Year
Member

Joined: May 2002
Posts: 4,324
Yuma, AZ
The rear OE RCH is about 18", and corresponds closely to the height of the rear pivot points of the UCAs. This is extremely high and has profound, and usually negative, consequences for handling. More on this to follow.

Last edited by MAP; 11/07/20 04:31 AM.
#1068485 - 11/07/20 04:31 AM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: May 2002
Posts: 4,324
MAP Offline
15+ Year
MAP  Offline
15+ Year
Member

Joined: May 2002
Posts: 4,324
Yuma, AZ
Rear unsprung mass consequence for handling:

In point number 3 in the first post, I mentioned that for an MCSS, the ratio of sprung to unsprung mass is about 3.7:1 at the rear, and about 8.4:1 at the front. This means if the rear encounters a dip corresponding to an acceleration of roughly 3.7g or greater, then the rear will temporarily lose contact with the pavement (I'm neglecting tire compliance because it's about an order of magnitude smaller than the suspension compliance.) If the car is turning at the same time, then the car's directional control could be lost. The situation at the front of the car is much better: the dip would have to exceed roughly 8.4g.

So the point here is that for handling, high unsprung mass is bad. To make this obvious, consider the behavior of an unloaded trailer reacting to bumps and dips in a road while being towed. In this case, the unloaded trailer's sprung mass may be comparable to, or even smaller than, the trailer's sprung mass constituted by its axle and wheels. Everyone has seen this at one time or another: the trailer in this state will bounce about with obvious periods where one or both wheels are entirely up in the air. As a result, the trailer may be tossed about and even sway dangerously from side to side.

To a lesser but not negligible extent, we have the same problem at the rear of an MCSS with its high unsprung rear mass. Shock dampening may reduce the probability of losing ground contact, but the frequency bandwidth of the rear's V part of NVH is simply too wide for dampening at resonance to have a broadly dominant effect. The only possible passive cure is to reduce mass significantly, usually by switching to an IRS.

There is an additional mass problem: the high rear RCH causes one-wheel bumps to displace not only the axle, but the body as well. Years ago, Motorweek described this odd coupling motion in an A/G body car they were reviewing as, "A dog walking with a hip out-of-joint." With a high RCH, when one wheel rises over a bump by an amount z, the body is pushed to the side away from the bump by an amount z*(RCH/half-track width). If we call the car's sprung mass M, and the unsprung mass m, then the fraction of M coupling with the wheel is roughly (RCH/half-track width)^2. To put numbers to this, consider:

  • RCH = 18"
  • Half-track widfth = 30"
  • One-wheel rear bump mass = 335lb/2.5 (roughly) = 135lb
  • Rear spung mass = 1,250lb

Get:

Total one-wheel bump mass = 135lb + 1,250lb*(18"/30")^2 = 585lb! This is where lots of tire sidewall compliance, as well as lots of rear suspension bushing compliance, can help mask the problem. But if we use stiff Heim joints in the suspension and performance tires with stiff sidewalls, we can get in trouble on bumpy surfaces.

What can we conclude from all this?

  • On a track with smooth pavement, unsprung mass is immaterial. A live axle can handle as well as an IRS. Turning in this regime is all about lateral load transfer (LLT henceforth) in the front and rear, and the developed tire slip angles as functions of time with a given steering profile. In this regime RCH still figures prominently, however.
  • On a street with bumps, unsprung mass is highly consequential. If we want the rear to stick to the ground well while turning, then unsprung mass must be minimized. IRSs are highly effective in this scenario. If unsprung mass can't be minimized, then using tires with a soft sidewall can help considerably, by helping to decouple the tire's tread area from the entire rest of the mass coupling chain. If we want to reduce the high tire slip angles this engenders, we can make the tire wider.


Back to shock dampening, high unsprung mass continues to be a problem. Assuming purely viscous dampening, which is admittedly a very loose approximation to actual shock behavior, one can show that with the masses we've used so far for the car's rear, the Q of the tire/road resonance will be roughly double that of the car/road resonance (here I'm assuming negligible tire dampening and that the tire's vertical stiffness is about 4.1e-6 m/N.) Consider, for example, shock tuning that yields a Q of 0.5 (critical dampening) for the tire/road resonance. The Q of the car/road resonance will be about 0.25, which is markedly overdamped. If we instead reduce shock dampening so that the car/road resonance is critically dampened, then the tire/road resonance will be about 1, which is underdamped. In this condition, bumps will induce some oscillatory motion at the wheel: i.e., the wheel will bounce about a bit up and down as it goes over bumps. Ideally, both resonances should be critically dampened, but this can only happen if 1. we reduce unsprung mass, and/or, 2. we increase tire compliance.
____________________________

Well - to take a step back from this a second, I hope this conversation is interesting and helpful. As you can see, my knowledge is more theoretical than practical. For example, I can't contribute anything useful to considerations of limited-slip differentials. But hopefully, we can hit all the major points with your input. Please feel free to dive-in!

Last edited by MAP; 11/07/20 05:18 AM.
#1068489 - 11/07/20 09:14 PM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: May 2002
Posts: 4,324
MAP Offline
15+ Year
MAP  Offline
15+ Year
Member

Joined: May 2002
Posts: 4,324
Yuma, AZ
I wish the window to edit these posts here weren't so short. A few sentences into the post above, "the trailer's sprung mass constituted by its axle and wheels", should read, "the trailer's unsprung mass constituted by its axle and wheels."

To amplify a few points:

Why do I speak of two resonances? Because considering just the vertical dimensional aspect of things, each suspension corner is a two-degree-of-freedom system with two masses (unsprung and sprung,) two compliances (tire and suspension spring,) and one dampening element (the shock.) The tire's dampening is very small and can be neglected in comparison to the shock's action. Each mass-spring pair constitutes its own resonant system. Using a vertical tire compliance of 4.1e-6m/N (a value used by my former employer for active suspension development,) and an unsprung mass of 152kg, we get a tire/road resonance of 9Hz (the so-called "wheel hop" frequency,) and something like 1.4Hz for the car/road resonance (assuming 130lb/in stock spring stiffness.) The bandwidth between these two is almost three octaves, so it's reasonable to consider each resonance as if in isolation.

In the last paragraph about shock dampening, in order to get the same Q for each resonance, I said, "...this can only happen if 1. we reduce unsprung mass, and/or, 2. we increase tire compliance." There's a third option to add: we can increase rear spring stiffness. But we'd have to roughly quadruple it to arrive at the same Q for both resonances, which would yield about 520lb/in. Not only would this yield an uncomfortably stiff ride at the rear for most people I'd venture, but then we'd have to boost stiffness at the front commensurately to preserve a halfway reasonable vehicle pitch response. However, if we did that, then the front's set of resonant Qs would be thrown off-kilter. In other words, by playing around with spring stiffness, we can't hit a target of good pitch response and good Qs for both front and rear vertical resonances, so in reality we're still hosed when rear unsprung mass is too high...

What all of this analysis highlights is that for road-racing on smooth surfaces, the high mass of the live rear axle is immaterial, as I wrote before. And if all of the streets of the world were similarly smooth, we'd probably all be driving vehicles with live axles, even at the front. The "gotcha" enters as soon as we want sticky, tenacious handling on bumpy surfaces. That's when unsprung mass needs to be minimized, and that's the real reason why live axles have all but disappeared from performance cars. It's also the reason why we want to reduce the rear RCH from an unreasonable 18" to something like 9" or possibly even lower. Remember that the body's coupling mass goes as the square of RCH, so halving this height divides its coupling mass to the wheel by a factor of four. In the example cited a few posts back, one-wheel bumps would go from an effective moving mass of about 585 lb to about 250 lb!


Last edited by MAP; 11/07/20 09:40 PM.
#1068493 - 11/08/20 02:59 AM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: May 2002
Posts: 4,324
MAP Offline
15+ Year
MAP  Offline
15+ Year
Member

Joined: May 2002
Posts: 4,324
Yuma, AZ
Rear unsprung mass consequence for NVH:

As mentioned above, each suspension corner is a 4th-order, dampened system. Regarding displacement transmitted to the cabin (and ignoring the compliance of the body bushings for the time being):

  • Below the car/road resonance at about 1.4Hz, displacement is stiffness-controlled and constant over frequency.
  • Above the car/road resonance at about 1.4Hz but below the tire/road resonance at about 9Hz, displacement is mass-controlled for the sprung mass and displacement declines at 12dB/octave with increasing frequency. Per Hamilton's principle, the center of mass between the unsprung and sprung masses tends to remain stationary.
  • Above the car/road resonance at about 1.4Hz and above the tire/road resonance at about 9Hz, displacement is mass-controlled for both the sprung and unsprung masses and declines at 24dB/octave with increasing frequency. Per Hamilton's principle once again, the center of mass between the unsprung and sprung masses tends to remain stationary.


Hamilton's principle has an important consequence when unsprung mass is high: since the center of mass of this two-mass system isn't coincident with the sprung mass, vibration is transmitted to the cabin, entirely apart from the frequency-dependency aspect of transmission addressed above. The lower the ratio of sprung to unsprung mass, the higher the transmission of vibration to the cabin. For an extreme case, think of the unloaded trailer once again. Since the rear mass ratio is roughly 3.7:1, about 27% of the vibration profile experienced by the rear axle is transmitted to the body (for the front with 8.4:1, the fraction is about 12%.) High unsprung mass is therefore not good for vibration.

On the other hand, the tire/road resonant frequency is reduced with higher unsprung mass, so we get better frequency attenuation at higher frequencies, although Hamilton's principle is still fighting us. The net result here is that high rear unsprung mass is bad for vibration, but neutral to possibly somewhat good for harshness at higher frequencies. Vibration in the 1.4Hz-9Hz range is usually perceived as "jiggliness" or "looseness," and never thought of as good.

If the rear were converted to an IRS, the front and rear VH response would be much more nearly the same. More than half of the live-axle rear's jiggliness and looseness would disappear.

Last edited by MAP; 11/08/20 03:21 AM.
#1068495 - 11/08/20 04:08 AM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: Jan 2000
Posts: 5,430
mmc427ss Offline
20+ Year
mmc427ss  Offline
20+ Year
Member

Joined: Jan 2000
Posts: 5,430
Pottstown, Pa
Mark don't think your musings are going unnoticed. I'm trying to digest all of it although difficult at times.

Just being a hobbyist at this handling game, to old to get to serious about it anymore. But over the last 20-25 years have put a lot more importance on braking and handling of my toy. Have done a lot of changes in both aspects in the name of improving the 86. Like others I chose the G-body for numerous reason, but mostly because in 1991 full frame cars were disappearing in the mid size range. I wanted that frame to builds on. In 1990 my son totaled my 72 SS Chevelle, I understood
the A body, the G is just a A on a diet. So for near 30 years of owning the 86 can't say there is much I haven't changed for what i hope is the better. So far so good.

Projects over the past couple years are to make significant changes front and rear to get the fat, front end heavyyyy, thing a little more neutral in and through a corner without turning my knuckles white. With as good as the front suspension is chasing the rear becomes the issue.

Not being a physics guy and really not a whole lots of brain cells left, I'm still intrigued with the way this topic is going. Many of the things noted here I've stashed away for when I try to put the big picture together.

Rear roll center is my area of interest currently. Have looked at numerous ways to control roll center and just don't see an easy fix for a C4L, Fays2, hmmmm. Three link, truckarm, several home built three link including Aaron's car. Had the pleasure of riding in his G-Velle at UMI KOTH in 2019. Here's a link to the page of his front and rear suspension in his 65 Chevelle.
https://lateral-g.net/forums/showthread.php?t=41839&page=33


It's a very nicely done three link of his design. It works very well, he's very competitive with that car. His build here from page 1 to the end is worth reading. Page 33 has his front suspension and at the bottom of the page his home built three link design.
He does exceptional quality work.

So Mark I'll keep reading and try to figure out how this will help me polish my turd a little more.
Bob

#1068496 - 11/08/20 05:05 AM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: May 2002
Posts: 4,324
MAP Offline
15+ Year
MAP  Offline
15+ Year
Member

Joined: May 2002
Posts: 4,324
Yuma, AZ
Hi Bob - I very much hope this turns out to be useful to you. (Just writing about it sure has been useful to me!)

That link is interesting. The thread seems to be very long, but he's got his rear RCH at about 9", which I think is a great compromise for various reasons which I have yet to talk about. The entire rear frame and the rear itself look very stout, which is good. But it also looks heavy, even with that aluminum center section. That's bad. But on a smooth track, it doesn't matter. I need to read more of that thread...

The factory C4L is very limited in that if you use CAs without compliant bushings, you're more-or-less stuck with an RCH of 18". Or, as I mentioned in another thread, you can flip the C4L geometry upside-down and get the RCH down to about 8'-9" (the "Satchell link" that Herb Adams wrote about years ago.) This is very good, but it's awful to package, which probably accounts for the reason why I've never seen it implemented. Basically, the new LCAs need to run through where the mufflers go, and that's a big problem.

To give a foretaste of the punchline of where I think this whole thread is going: C4Ls don't deserve the bad name they've gotten. The geometry doesn't bind, but cylindrical bushings in a C4L can come close, especially if they're made of TPU. If we can free-up the CAs, reduce RCH, and do what we can to minimize unsprung mass with the live axle, we'll do great on the track, if not so great on the street. If we want great performance on the street too, we need an IRS. If that sounds surprising, would it also be surprising if I said we need an independent suspension in the front? Of course not.

Thanks,
MAP

Last edited by MAP; 11/08/20 07:18 AM.
#1068498 - 11/08/20 02:49 PM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: Dec 2007
Posts: 6,504
SSLance Offline
10+ Year
SSLance  Offline
10+ Year
Member

Joined: Dec 2007
Posts: 6,504
Peoria, AZ
A friend recently built a 95 F150 to do the same stuff I do with my car. He went all out, completely custom front suspension and a Mustang IRS for the rear, big power, stiff chassis etc. Hes raced with us a couple of times now and one big takeaway is while the IRS does have some benefits in spots on course, it also has draw backs. Im sure with more fine tuning he'll get it handling better, but in the meantime I do enjoy beating his high dollar build by several seconds with my C4L setup.

We are at similar power levels, similar weight, pretty close front to rear weight bias and according to him similar driving ability. It'll be fun to see just how close he can get to my old OEM chassis style. I do know this, my ability to plant my power on takeoff and corner exit is leaps and bounds better than his with his IRS and that is a common theme. Seems when you fix the forward bite with IRS, you hamper the cornering ability...


Lance
1985 Monte Carlo SS Street Car
#1068499 - 11/08/20 03:29 PM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: Jan 2000
Posts: 5,430
mmc427ss Offline
20+ Year
mmc427ss  Offline
20+ Year
Member

Joined: Jan 2000
Posts: 5,430
Pottstown, Pa
Here's a link to Sean's 95 F150 truck build.
https://www.pro-touring.com/threads/129476-95-F-150-track-ready-street-beast

Another build worth reading from beginning to end. Just the fact that it was done in a carport along side his house is remarkable, but his trials and tribulations of his build are impressive.
Bob

#1068500 - 11/08/20 04:44 PM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: Dec 2007
Posts: 6,504
SSLance Offline
10+ Year
SSLance  Offline
10+ Year
Member

Joined: Dec 2007
Posts: 6,504
Peoria, AZ
This just dropped today as well...long watch but really shows all he put into it.

https://youtu.be/EQmJx2ohduk


Lance
1985 Monte Carlo SS Street Car
#1068502 - 11/08/20 09:39 PM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: May 2002
Posts: 4,324
MAP Offline
15+ Year
MAP  Offline
15+ Year
Member

Joined: May 2002
Posts: 4,324
Yuma, AZ
Hi Lance,

That was a bit ahead of where I wanted to go at this point, but the disadvantages with an IRS are limited anti-squat (about 25% according to Herb Adams,) and limited RCH. The former limits how much traction you can get from the rear, and the latter, after mass effects have been considered, influences steering transient response (not discussed yet.) The traction is best addressed by redistributing mass in the MCSS to be much more rearwardly-concentrated. Hardly an easy fix! Given the poor weight distribution in an F150, I absolutely wouldn’t use an IRS on the track. IRSs work best in rear-heavy vehicles.

Even so, on bumpy streets, the great low unsprung mass advantage of an IRS is something a live axle can’t match. Crudely put, if you want to pirouette like a ballerina, you don't want to weigh 300lb. According to Greg Locock, suspension designer for GM and Ford, "For a track car [an IRS's] other virtues may dominate, but for traction in corners on rough roads even the crudest IRS will be superior to any beam axle."

Bob and Lance, I promise to get to those videos.

Thanks,
MAP

Last edited by MAP; 11/08/20 11:12 PM.
#1068503 - 11/08/20 10:11 PM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: May 2002
Posts: 4,324
MAP Offline
15+ Year
MAP  Offline
15+ Year
Member

Joined: May 2002
Posts: 4,324
Yuma, AZ
Speaking of steering transient response and RCH:

At the front, when a steering input initiates, the wheels are turned and front tire slip angles develop almost immediately. As roll grows within a few tenths of a second, the difference in inner/outer tire loading increases and the difference in tire slip angle between the two front tires increases. (It's this difference in slip angle in relation to the yaw center that Ackermann is intended to address, in order to minimize scrub.)

Because of the way the rear follows the front as we start a turn, rear tire slip angles develop more gradually. But likewise, as roll initiates within a few tenths of a second, the difference in inner/outer tire loading increases due to the rear's roll stiffness and the difference in tire slip angle between the two increases. (Interesting that the need for Ackermann at the rear is rarely mentioned, but it surely exists. But Ackermann is a fixed correction, while intra-axle scrub is dynamic.)

Perhaps the best way to view this is to look at the vehicle's yaw center as the vehicle starts to turn. If we consider the effective turning center as the imaginary point where the normal to each tire's slip path intersects, and then project a normal from that point back to the car's x axis, the yaw center starts at the rear axle in a turn, and then migrates forward within a few tenths of a second to be roughly coincident with the car's COM. (In reality, the normals to the tires' slip paths rarely intersect at exactly a single point, but we can at least view the average intersection as an estimation of yaw center.)

So the point is: turning starts at the front, with the yaw center at, or near, the rear axle. Front tire slip angles develop almost immediately. As the turn evolves, slip develops later at the rear, and the yaw center moves forward in the vehicle, coming close to the vehicle's COM.

Now, as to RCH: the suspension's lateral force to accelerate the car has an effective height at the RCH. The higher the RCH, the greater the difference in tire loading this creates between inner and outer tires. But this lateral force develops instantaneously as lateral acceleration grows, in contrast with the slower tire loading difference created by the axle's roll stiffness in response to body roll.

Putting this together, we get:

  • A high RCH "hits" the tires almost immediately with a moment that causes a difference in loading between inner and outer tires. Roll stiffness "hits" more slowly. Since the rear lags the front in turning, a high RCH at the rear can help sharpen the vehicle's steering transient response by reducing the rear's steering response lag. Or, going back to the yaw center analogy, a high RCH allows the yaw center to move forward toward the car's COM more quickly than if we just relied on roll stiffness to get the job done.
  • Since RCH transmits a moment that causes a difference in loading between inner and outer tires, it has the same steady-state effect as roll stiffness. So the higher the RCH, the higher the steady-state equivalence of roll stiffness. As mentioned several posts back, GM relied on the rear's very high RCH to avoid using a rear sway bar for most A/G models.
  • One can safely conclude that the more the RCH is reduced, the more we need to compensate with added roll stiffness.


So, a high RCH is not all bad. But it's still mostly bad - the effective moving mass detriment is still a very big one. But on a smooth track, moving mass is unimportant. It's on the street, as usual, where the situation gets much more complex...

If we don't want a high RCH, but still want a fast steering response, what can be done? Increase the car's roll stiffness. (I keep on getting a nagging feeling that this conclusion needs to be qualified. I'll continue to chew on this.)

Last edited by MAP; 11/08/20 11:21 PM.
#1068504 - 11/09/20 12:08 AM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: Dec 2007
Posts: 6,504
SSLance Offline
10+ Year
SSLance  Offline
10+ Year
Member

Joined: Dec 2007
Posts: 6,504
Peoria, AZ
I have found that my favorite tuning tool at the track is tuning how quickly the inside rear tire releases it's contact. If it stays "in the track" the steering response is slow leading to a front end push. If the inside rear tire releases quicker, the turn in is sharp and crisp...everything else remaining the same. I use rear shock rebound adjustment and\or rear sway bar adjustments on my car. I'm also currently learning just how much the differential driving the rear tires affects turn in which again relates to what the inside rear tire is doing.

I've know others that have the ability use rear roll center height changes as a quick and easy track day tuning tool to do the same...affect the timing of the release of the inside rear tire. By changing the front to rear diagonal roll angle you change how both the outside front and inside rear tires react to lateral forces inducing roll.

This response isn't as "engineering speak" as what you are getting to MAP, but it shows in real time what changes to rear RCH do on the fly to dial in a car at the track. It's literally all about getting that inside rear tire to release to let the front turn.


Lance
1985 Monte Carlo SS Street Car
#1068508 - 11/09/20 04:09 AM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: May 2002
Posts: 4,324
MAP Offline
15+ Year
MAP  Offline
15+ Year
Member

Joined: May 2002
Posts: 4,324
Yuma, AZ
Hi Lance,

I think you're saying that the rate of change of inner rear tire loading upon initiating a turn is a favorite tuning trick of yours. If I got that right, then we're saying the same thing: a high rear RCH causes the inner rear to unload faster, so you get a quicker steering response.

I remember when you let me try a sudden step steering input with your car in April, 2019, that I judged the response to be very quick and very predictable and stable. A high rear RCH helps that. Higher roll stiffness can approximate, but never quite equal, the same thing, but too much can be bad for the street (that was the "nagging feeling" I had two posts back. When suspension stiffness gets too high, whether it's in the suspension springs (common and difference-mode bumps) or just in the sway bars (difference-mode only,) the car can get skittish over bumps and lose control in a turn.) Shock tuning definitely figures prominently too in rate of change of tire loading as the car develops roll in a turn, so no surprise to hear this is a valuable tuning tool as well. The only potential trouble is that it affects the Q of resonant response to road bumps as well, so one response can't be adjusted without affecting the other.

Thanks,
MAP

Last edited by MAP; 11/09/20 04:31 AM.
#1068513 - 11/09/20 02:36 PM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: Dec 2007
Posts: 6,504
SSLance Offline
10+ Year
SSLance  Offline
10+ Year
Member

Joined: Dec 2007
Posts: 6,504
Peoria, AZ
Yes, we are on the same path here, I'm just trying to speak in real hands on past experience laymen style ways instead of engineer speak. laugh

Here's another way to visualize how RCH differences affect handling. Think of how high the center of gravity of your car is then imagine how high your rear roll center is. With a high rear roll center closer to center of gravity, the arm or lever that the body has trying to roll the car over in a turn is shorter thereby taking more lateral force to roll the body over.

When you lower the rear RCH, yes you put the pivot point closer to the ground which helps reduce roll, but you also increase the length of that arm or lever pulling on the roll center which will reduce the amount of lateral force need to make to body roll. This increased length of that lever typically needs to be counteracted by some other type of roll resistance to keep actual roll at the same amount. Stiffer rear springs or stiffer sway bar are the go to here...but they have other detrimental characteristics that may hurt your car worse than the high roll center does.

Stiff springs are awesome for lateral grip, they keep outside from compressing without pulling up on inside wheel. But forward bite grip is hurt because they also don't let weight transfer to the rear like we want.

Stiff sway bars on a stick axle are worse as when they reach their resistant force, their only option is to pick up on the inside rear tire which actually causes a loss of grip on both inside AND outside rear tires.

I have personally felt the results of all of the above on my C4L car on track. I have felt how sometimes just slight adjustments one way or the other which each scenario can dial in the feel I prefer. The trick is to have adjustment ability on each setup built in and then decide which strategy works best for you.


Lance
1985 Monte Carlo SS Street Car
#1068514 - 11/09/20 09:15 PM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: May 2002
Posts: 4,324
MAP Offline
15+ Year
MAP  Offline
15+ Year
Member

Joined: May 2002
Posts: 4,324
Yuma, AZ
Hi Lance,

Absolutely I'm all for language that gets the idea across. But I'm also for rigor in physics. Sometimes those two realms don't connect together very easily. When the two are in conflict, I always favor the latter.

Moment is r x F. When the two are perpendicular, as we'll take them to be here, then its magnitude is the simple product of force with moment arm.

What you said was mostly but not completely correct. The moment arm for the car in a turn is the perpendicular distance from the car's roll axis to its COM. If we push the RCH higher, whether at the front or rear or both, we raise the car's roll axis and thus decrease the moment arm. The (fictitious) moment in this case is inertial, and has a magnitude equal to the centripetal force with the moment arm. The car's roll is thus this moment divided by the suspension's stiffness in roll. That may be a bit unclear but its implications are simple.

As we:

  • Turn harder
  • Lower the car's roll axis
  • Raise the car's COM
  • Decrease the car's roll stiffness


The car's roll angle will increase.

If we make everything stiffer - let's say, for example, we double spring and sway bar stiffnesses - we change the forces exerted on each tire's contact patch exactly not at ll, discounting the effect of a slight shift in the car's COM relative to those patches. If we get better grip, it's probably due to the fact that each wheel remains more vertical relative to the pavement. With a stick axle, this will reduce understeer because we improve front local camber; rear local camber remains unchanged.

But high stiffness is generally not good on rough surfaces. Compressed wheels on bumps see too much force, and relaxed wheels on dips see too little and can completely lose contact with the pavement. For the street, soft is better provided we can keep local camber nearly vertical.

To understand the following partially involves understanding that tires do not exactly conform to Coulombic friction. If they did, then maximum frictional force would go in direct proportion to normal loading. It does not. For example, if we double the loading, we would expect to get double the force. But with most tires, and especially when we start to approach the traction circle limit, we might only see an 80% or 90% increase. This fact helps explain:
Quote
Stiff sway bars on a stick axle are worse as when they reach their resistant force, their only option is to pick up on the inside rear tire which actually causes a loss of grip on both inside AND outside rear tires.


When the car "picks-up" the inner rear tire, there is too much LLT on that axle. LLT comes from the lateral force transmitted through the RCH, and the roll stiffness in the suspension. If the inner tire lifts, then one or the other needs to be reduced. If the car still understeers as a result, then other means need to be exploited to reduce it. Too much LLT reduces maximum lateral grip on that axle exactly due to the non-Coulombic friction tendencies of the tire as cited above. When the inner rear tire lifts, we tend to get a jump toward oversteer, because the lower slip angle of the inner rear tire is removed from influencing the rear's direction of travel.

The extreme scenario described above is likelier to arise when the car's COM is too far forward.

Last edited by MAP; 11/09/20 10:59 PM.
#1068515 - 11/09/20 09:59 PM Re: SS rear suspension: thoughts and musings [Re: MAP]  
Joined: Dec 2007
Posts: 6,504
SSLance Offline
10+ Year
SSLance  Offline
10+ Year
Member

Joined: Dec 2007
Posts: 6,504
Peoria, AZ
This is why I try very hard to never recommend a spring rate to anyone without specific knowledge of their complete setup or strategy. The correct answer is almost always "it depends". smile There is so much more to it than that as you show.

What I've been taught is to get the car as neutral as you can in the shop and fine tune the inside rear at the track. Hopefully you are near the center of your adjustments and have enough room to dial the car in to that day's course.


Lance
1985 Monte Carlo SS Street Car
Page 1 of 5 1 2 3 4 5

Random Images
1139/thumbs/photo_1.JPG
by O.G.
875/thumbs/ss_delete_it_i_dare_you.jpg
by BADRACR1
500/thumbs/image32.jpg
by Blown Monte
500/thumbs/P1010096-small.JPG
by ProTourAero
500/thumbs/monte_carlo_ss1.jpg
by shockerman80
Help MonteCarloSS.com


Recent Contributors
86BlackSuperSport
dns87ss
Authorized Vendors
Tell them you saw it
on MonteCarloSS.com!


CustomMonteSSParts.com
Dixie Monte Carlo Depot
GSI Interiors
HRpartsNstuff
Mikes Montes
Savitske Classic & Custom
Powered by UBB.threads™ PHP Forum Software 7.6.0