Asymmetric 3link braking stability

[BillyShope]

As many of you know, I'm a strong advocate of an asymmetric rear suspension for certain applications. Such a suspension, when used with a beam axle rear wheel drive car, is capable of providing, during forward acceleration, equal rear tire loading with no squat or rise. One form of such a suspension, an asymmetric 3link, was used in the Jaguar C-Type, a Le Mans winner (with, of course, a Panhard). What is not often mentioned is that the designer of the car had second thoughts about the design, after the Le Mans win, when further testing revealed certain braking stability problems during "panic" braking. Panic braking involves complete lock-up of the wheels, a condition not encountered in the Le Mans race. It is desirable, then, to provide equal tire loading in BOTH forward and rearward acceleration modes. The normal asmmetric 3link has two symmetrically placed lower links and a third link above and offset to the passenger side (US). Braking symmetry can be achieved with the addition of a fourth link. This can restore link symmetry for equal rear tire loading during braking and yet provide equal tire loading during forward acceleration IF that additional link can only act in compression. The first Ramcharger dragrace car accomplished this with a telescoping link which "bottomed out," but a simpler design would use a front bracket shaped, when viewed from the right side of the car, as a Greek upper case gamma (mirror image of the number seven). This bracket would have two pivot points. One, located at the bottom tip of the vertical section, would attach to the front of the link. The other, at the "corner" of the gamma, would attach to the chassis. During braking, the upper surface of the horizontal portion of the gamma would press against a hard rubber pad situated between the bracket and the chassis. During forward acceleration, the bracket would "swing free," and the link would be unable to carry a tensile load. Thus, the desired situation of equal tire loading, in both braking and forward acceleration, would be achieved.

Another interesting asymmetric arrangement would consist, essentially, of a single "ladder" offset to the passenger side. Again, braking symmetry would require, for instance, a link, mounted offset to the driver side and pointed at the rear tire patch, with a bracket which, in this case, would have the appearance of the number "7" when viewed from the passenger side. This link would be unable to carry compressive loads.

For simplicity, it is assumed that the links exist in planes parallel to the XZ plane.

[mitch_04]

This is interesting, but it is missing something... pictures!

[BillyShope]

Sorry. When I went from Vista to Windows7, I lost my CAD software. Would appreciate it if someone who follows all the verbiage would supply some artwork, but I realize that's a lot to ask.

[rustomatic]

When you can fit a useful, centered upper (single) link under an MG GT, why deal with the potential issues of an offset link? There's also the torque arm...

[BillyShope]

If you're satisfied with the unequal loading of the rear tires during forward acceleration, then there's certainly no reason to go asymmetric. Jaguar, with their C-Type, was not satisfied with a symmetrical 3link; Ford, with their Mustang...recognizing the "potential issue" (singular) of braking stability with full wheel lockup...WAS satisfied. It's simply a matter of priorities. I should explain that, with full wheel lockup, consideration of slip angles and steering input "goes out the window" and everything reduces to sliding friction.

Also, when you consider that the three links are similar to the three legs of a stool, there are an infinite number of combinations which could provide the desired asymmetry. Plus, as you point out (and, as I mentioned in my original post), there are other suspension designs...such as the torque arm and the single ladder bar...which would do the job.

[Ron Sutton]

Billy,

I read something of yours awhile back where you were changing the angle of the lower link & upper links to cancel the torque rotation on launch in drag cars. Is that something you still do & recommend?

.

[BillyShope]

Billy,

I read something of yours awhile back where you were changing the angle of the lower link & upper links to cancel the torque rotation on launch in drag cars. Is that something you still do & recommend?

.

Yes, the cancellation of the driveshaft torque is the only reason for asymmetry. That driveshaft torque causes all kinds of safety and performance problems. Since the link loads are proportional to the driveshaft torque, the proper setup cancels all values of driveshaft torque. This thread addresses an undesirable "side effect" of asymmetry, which can also be totally eliminated (original post).

[Ron Sutton]

Yes, the cancellation of the driveshaft torque is the only reason for asymmetry. That driveshaft torque causes all kinds of safety and performance problems. Since the link loads are proportional to the driveshaft torque, the proper setup cancels all values of driveshaft torque. This thread addresses an undesirable "side effect" of asymmetry, which can also be totally eliminated (original post).

Hey Billy,

Thanks for the reply. I'm pretty familiar with all the variations of 3-links. As you know, the term asymmetric can be vague, or interpreted a few ways. In your first post on this thread, it sounds like you're referring to an offset 3-link, with parallel lower links & the top link offset to the passenger side.

Is that correct?

Or are you referring to the asymmetric 3-link you recommend in which you have the links at different angles to each other?

[BillyShope]

Hey Billy,

Thanks for the reply. I'm pretty familiar with all the variations of 3-links. As you know, the term asymmetric can be vague, or interpreted a few ways. In your first post on this thread, it sounds like you're referring to an offset 3-link, with parallel lower links & the top link offset to the passenger side.

Is that correct?

Or are you referring to the asymmetric 3-link you recommend in which you have the links at different angles to each other?


I'm generally concerned with any form of asymmetry which provides the necessary cancellation, but, yes, I was referring to the normal 3link in the first post, as you described it.

[Ron Sutton]

I'm generally concerned with any form of asymmetry which provides the necessary cancellation, but, yes, I was referring to the normal 3link in the first post, as you described it.

Thanks for clarifying which version you're discussing here.

[ace_xp2]

I was going to use a 90/10 shock setup with a short internal bump stop for the braking linkage and then a circle track style pull bar for the accelerative one.
But, I've notice that it seems a lot of modern high performance stuff seems to shy away from anti squat of any amount, though this is just going off of arm angles, and independent suspensions, so it's only of so much use. However, Koenigseggs triplex damper setup is specifically described as letting them do away with anti-squats negatives while resisting squat (it's basically a left to right connected damper so that it only works in bump).
Now I'm thinking more along the lines of a low/no anti-squat arrangement, and some way to either increase spring rate when it's time to get serious or some similar damping arrangement to deal with those forces.

[Ron Sutton]

I think you may have misinterpreted the Koenigseggs set-up. They are most definitely using anti-squat, they've just "dampened" it with the 3rd shock.

[David Pozzi]

Billy, The Hotchkis 3 link for first & second gen Camaros has an offset upper link & it has been working very well. We have not encountered any rear lockup problems, however the upper link offset is not extreme. I think it's around 5" - 6" to the right of center. I suggested Hotchkis try it, I'd been wanting to do it for years. I've seen more offset used in 90's Trans-Am cars. Nearly every TA car used it near the end of the series.

[Ron Sutton]

Hey David,

If you don't mind me asking, how much anti-squat have you been running ?

[David Pozzi]

Around 50% Anti-Squat. Maybe a bit below that like 45%. I'd have to look it up. It is adjustable, both ends of the upper link, & also the lower links have a few holes.

[ace_xp2]

I think you may have misinterpreted the Koenigseggs set-up. They are most definitely using anti-squat, they've just "dampened" it with the 3rd shock.
None may be an overstatement,but it doesn't sound like they're using much, He even makes reference to being able to use less and seeing it as an advantage here:
http://www.youtube.com/watch?v=bbgjRBT4ltM
And here's a pic of the carrera gt rear suspension mounts (This really only lets you see that it doesn't look like a ton, how much is difficult to guess.)
81702
And venom gt (not that this is a design masterpiece, rather that it's an example of a car that sure seems to have no a/s, and manages to put down a lot of power apparently without any obvious handling snafus being brought up):
81704
Ortiz talks about the complications that antis can bring here:
http://www.theoryinpracticeengineering.com/resources/ortiz_newsletter/2004_12.pdf
I bring this up because my calculations put my offset link just about over top the lower link to cancel out torque reaction at 100% anti squat, which is about as far over as it can go without tire contact. Now based on the info I've found I'm thinking of anti squat in the low tens to low twenties, and while there would (I think) still be some torque cancellation at this rate, I question how noticeable it would be at that range.

[Ron Sutton]

Around 50% Anti-Squat. Maybe a bit below that like 45%. I'd have to look it up. It is adjustable, both ends of the upper link, & also the lower links have a few holes.

Hi Dave,

I have ran a wide variety of cars anywhere from 30% to 80% anti-squat, but always with the I/C far forward creating a long "swing arm." Within this range, we too have not encountered any uneven braking leading to a yaw condition under brake lock conditions. The cars track true if locked up in a straight line.

I design & build offset 3-links so the amount of top link offset is adjustable ... in the range of 8-12% of trackwidth. (I.E. 5.0-7.5" on a 62.5" track width) I make it adjustable, because it is challenging to accurately predict the frictional drag of various rear ends & gear assemblies. But 8-12% tends to cover the range pretty well.

[David Pozzi]

Here is an old 3 link thread with Trans-Am pics: https://www.pro-touring.com/threads/1109-Three-link-suspension-used-in-TA-race-cars?p=7228#post7228

[feeble]

Would it be correct to say that a car with ABS and an offset 3 link would have less of an issue with braking stability due to the offset link? And as a corollary, cars without ABS would be better off with a centered upper link, at least with regards to braking stability?

At my skill level, (a novice with no experience), I doubt I will be able to notice a difference between centered and offset, so for packaging reasons, I'm centering mine.

[Ron Sutton]

Would it be correct to say that a car with ABS and an offset 3 link would have less of an issue with braking stability due to the offset link? And as a corollary, cars without ABS would be better off with a centered upper link, at least with regards to braking stability?

At my skill level, (a novice with no experience), I doubt I will be able to notice a difference between centered and offset, so for packaging reasons, I'm centering mine.

Shawn,
What Dave & I are saying, is that we have not experienced any braking stability issues with offset 3-links.
Dave said his anti-squat is around 45-50%. I have ran cars from 30%-80%.


Billy,
Do you think anti-squat plays a role in whether or not a track car will experience braking issues?


.

[jerome]

I am considering making an offset upper link as an acceleration link (collapsible, but not extendable) and a centered upper link as a brake link (extendable, but not collapsible) to decouple braking and acceleration geometry. The acceleration link will run downhill from rearend to chassis (high anti-squat under acceleration) and the brake link will be flatter (high SVSA length under braking). The centered upper link takes care of any asymmetry under braking.

A couple questions:

1. Acceleration and brake links will travel along different arcs.
Chassis pickup point heights are different (to create high anti-squat and high SVSA length)
Chassis pickup points are slightly different longitudinally (packaging)
Rearend pickup point heights are slightly different (packaging)
Rearend pickup points are slightly different longitudinally (packaging)

I have not run through a simulation yet to determine whether there will be kinematic bind. Anybody have guesses on whether this arrangement works? The commercially available acceleration and brake links do not seem to have more than 0.5" of travel in the collapse/extension non-load directions.

2. Any other issues with this concept?

[feeble]

Shawn,
What Dave & I are saying, is that we have not experienced any braking stability issues with offset 3-links.
Dave said his anti-squat is around 45-50%. I have ran cars from 30%-80%.
.


Since I don't have an accurate CG for my car yet, I had to make some assumptions and build in some adjustability. If my CG estimate is close, my AS for anything but drag racing is right at 50%. If I go to the drag strip, I can drop the rear of my lower control arms into a lower hole and get a little bit more than 100% AS.

[Ron Sutton]

I am considering making an offset upper link as an acceleration link (collapsible, but not extendable) and a centered upper link as a brake link (extendable, but not collapsible) to decouple braking and acceleration geometry. The acceleration link will run downhill from rearend to chassis (high anti-squat under acceleration) and the brake link will be flatter (high SVSA length under braking). The centered upper link takes care of any asymmetry under braking.
Can we assume you are utilizing poly bushings to dampen the contact or are you planning something different?

A couple questions:

1. Acceleration and brake links will travel along different arcs.
Chassis pickup point heights are different (to create high anti-squat and high SVSA length)
Chassis pickup points are slightly different longitudinally (packaging)
Rearend pickup point heights are slightly different (packaging)
Rearend pickup points are slightly different longitudinally (packaging)

I have not run through a simulation yet to determine whether there will be kinematic bind. Anybody have guesses on whether this arrangement works? The commercially available acceleration and brake links do not seem to have more than 0.5" of travel in the collapse/extension non-load directions.
There's your issue. In race cars, as long as the rear of the car rolls like we want it to ... this set-up works fine. On the street, with very little compression travel, the ride is rough, as any bump transmits energy to the other side, as the rear end rotates versus compresses. The bumps are not absorbed ... they shake & rock the car from the rear.

2. Any other issues with this concept?
If you're running poly bushings, they have to be replaced on a regular basis, or they crack, break & come out. This varies with durometer rating, load, & number of cycles. When they break & fall out, the suspension links have monster excessive travel, which makes the car unstable on braking & acceleration.

[David Pozzi]

Bad Penny & the One Lap Camaro have the Lateral Dynamics 3 link. I've driven them both & never experienced brake hop, even when rear brake bias was excessive. I always read warnings about it, but I think if the links are reasonably long, it isn't going to be a problem.

[Ron Sutton]

Hi Shawn,


Since I don't have an accurate CG for my car yet, I had to make some assumptions and build in some adjustability.
You can guestimate your CG around 18" height (as most are 16" to 20") and get an idea of where you'll be. You should build in adjustability anyway.

If my CG estimate is close, my AS for anything but drag racing is right at 50%. If I go to the drag strip, I can drop the rear of my lower control arms into a lower hole and get a little bit more than 100% AS.
Again, you should make your rear suspension adjustable, so you can tune it for optimum performance. You may find 50% AS makes the car loose on corner entry under hard braking ... depending on the rest of your suspension set-up. You may not. It simply needs adjustment range to be dialed in for your car, your set-up & your driving activities.

Make sense?

[feeble]

Ron: Thanks for your input. I assumed a CG height of 20" and have built in some adjustability. All of it bolts together as well, so I can redo some or all the parts if the optimum falls outside of the adjustment ranges. I am well versed in redoing things several times until I get them right. ;-)

[Ron Sutton]

Ron: Thanks for your input. I assumed a CG height of 20" and have built in some adjustability. All of it bolts together as well, so I can redo some or all the parts if the optimum falls outside of the adjustment ranges. I am well versed in redoing things several times until I get them right. ;-)

LOL. Take care !

[jerome]

If you're running poly bushings, they have to be replaced on a regular basis, or they crack, break & come out. This varies with durometer rating, load, & number of cycles. When they break & fall out, the suspension links have monster excessive travel, which makes the car unstable on braking & acceleration.




I have figured out a way to get enough travel in these collapsible/extendable decoupled braking and acceleration links to allow the suspension to cycle through full travel. This will allow me to have a link active only during braking and a link active only during acceleration.

Under acceleration:
Acceleration link is offset to passenger side and will help offset driveshaft torque, in addition to providing anti-squat that can be adjusted to any value without affecting any braking characteristics.

Under braking:
Braking link is centered laterally on the rearend and will not have any assymetrical effects. Under braking, rearend will try to forward (pinion downwards) and so the brake link will be loaded in compression. I am trying to figure out what the brake link angle should be when optimizing only braking characteristics (as it does not affect anti-squat).

Uphill towards chassis: pushes up on chassis and down on the rearend during braking. Weight transfer forwards acts through the third link instead of just unloading the springs, making the rearend push the chassis upwards, which makes the rear of the car rise more.
Flat: no upward or downward forces acting during braking.
Downhill towards chassis: pushes down on the chassis and up on the rearend during braking. Weight transfer forwards acts through the third link instead of just unloading the springs, making the rearend push itself up into the chassis, which makes the rear of the car rise less.


I think theoretically #3 would be best because it keeps the CG height down for less weight transfer to the front wheels. Would you generally want to design with #3 in mind and keep increasing the downhill angle towards the chassis (reducing SVSA length) and stop when you experience brake hop?

Jerome

[feeble]

Just throwing this out there without a lot of thought, but wouldn't you want AS to be as close to zero as possible on the braking link?

[Ron Sutton]

So, what are you using to take up the slack & compress when needed? Poly bushing? Spring? Something else?

[sccacuda]

Have you looked at circle track pull bars? Lots of travel. I modified one to use as my upper link used under braking (torque arm for accel). You need to preload each with some type of bushing as Ron stated.


https://static1.pt-content.com/images/pt/2013/02/file_zps38e5bba1-1.jpg (http://s1060.photobucket.com/user/340cuda/media/file_zps38e5bba1.jpg.html)

This is my modified pull bar that I now refer to as a "push bar". As you can see, the bushing "cage" can be extended to allow more or less travel.

[jerome]

So, what are you using to take up the slack & compress when needed? Poly bushing? Spring? Something else?

lower link from the below picture for the braking link:
https://static1.pt-content.com/images/pt/2013/09/22751color-1.jpg

acceleration link (run only one of the biscuits):
https://static1.pt-content.com/images/pt/2013/09/ex_q100_w500_h500_images_ePIM_original_1-1.jpg

[Ron Sutton]

The red Coleman unit uses rubber bushings. The unit with the 3 yellow bushings, uses poly bushings.

Both of these work fine, but with use (cycles of compression & release) crack & split. You'll want to replace them annually. Think of them as a wear item like brake pads.

.

[Ron Sutton]

Jerome,

Are you clear that this type of suspension linkage has to compress the lower bushing to compress the rear suspension? The rubber bushing is hard & doesn't compress easily, so it acts like an additional stiff spring in the rear suspension.

These work well in racing applications. And it will do what you're looking to achieve performance wise. But, it will make the rear suspension ride rougher. Because if you hit a bump or dip with both rear tires, your spring rate is substantially stiffer than just the two rear springs. Plus the rubber bushing is a progressive spring, meaning its spring rate gets stiffer as it compresses. So it will not compress far, it will just bounce the rear end up.

When you hit bumps with either individual tire, the suspension on that side doesn't simply compress like in a normal suspension. It rocks the rear suspension and rear of the car's body ... compressing one side & extending the other side. So it "shakes" the car from the rear.

That's when the links are "in line" with each other. If you mount them apart from each other, you will get an interesting "diagonal rocking" situation over bumps. It will rock & shake the rear of the car more over bumps one direction versus the other.

I'm not trying to talk you out of this type of suspension. I have a LOT of experience with this type of suspension linkage and other variations of this in racing applications. If tuned correctly, they add performance for track days. But there are cons for street applications. I just want you to hear it from someone with experience, so you know what you're getting into.

.

[jerome]

Jerome,
Are you clear that this type of suspension linkage has to compress the lower bushing to compress the rear suspension? The rubber bushing is hard & doesn't compress easily, so it acts like an additional stiff spring in the rear suspension.
.

I spent some time drawing and thinking it through - you're right. I was thinking about it too simply and did not realize that in bump, the acceleration link would need to get longer (which it can't) or the deceleration link would need to get shorter (which it can't). In rebound, there would be no bind, the acceleration link could get shorter and the brake link could get longer. The rearend would be free to flop between pulling on the acceleration link and pulling on the brake link depending on what the tires are doing.

[Ron Sutton]

Like I said, I have ran this type and other designs quite a bit. They work well in racing applications. But they do make the rear end spring rate much harder and more likely to rock than to compress.

Spreading the links apart would create a unique, amplified rocking situation.

[sccacuda]

Like I said, I have ran this type and other designs quite a bit. They work well in racing applications. But they do make the rear end spring rate much harder and more likely to rock than to compress.

Spreading the links apart would create a unique, amplified rocking situation.

Is this going to be present in any decoupled 3 link type suspension? I think I see what your saying, but not sure why the spring rate would increase, but do see how it would "rock" between bushings on the upper and lower links (accel/decel links). Wouldn't a bump just cause the rear to rock to the accel stop bushing and then the rear springs start controlling compression?

[jerome]

Is this going to be present in any decoupled 3 link type suspension? I think I see what your saying, but not sure why the spring rate would increase, but do see how it would "rock" between bushings on the upper and lower links (accel/decel links). Wouldn't a bump just cause the rear to rock to the accel stop bushing and then the rear springs start controlling compression?

At ride height, imagine the two links in side view where they originate from the same point on the rearend. The brake link is parallel to the ground and the acceleration link points downwards towards the chassis. You now have a triangle where the top link is free to extend and the acceleration link is free to collapse. The problem in bump is that if the brake link stays at its shortest and pivots upwards to follow the rearend, the acceleration link needs to get longer (which it cannot). If the acceleration link stays at its longest, the brake link would need to compress (which it cannot). The stop bushings end up adding spring rate until they are no longer elastic and then you have a bind situation.

[Ron Sutton]

Is this going to be present in any decoupled 3 link type suspension? I think I see what your saying, but not sure why the spring rate would increase, but do see how it would "rock" between bushings on the upper and lower links (accel/decel links). Wouldn't a bump just cause the rear to rock to the accel stop bushing and then the rear springs start controlling compression?

Good question Craig,

Your question implies there is a gap before the bushing engages. For these double-link top link systems to work properly, there can be no gap.

When the car is at rest, at ride height, both the bushing on the decel link and the bushing on the accel link have some load on them. I've had cars with springs in front of the axle, and that set-up causes more load on accel link bushing. I've had cars with the springs behind the axle, and that set-up causes more load on decel link bushing.

If you were to run a double-link top link system with gap in the bushings, the rear end would "flop" back& forth under slight acceleration or deceleration. Plus gap defeats the purpose and function of a double-link top link system. Frankly, when you really dial in a double-link top link system for optimum performance, you will end up with several rounds of preload on the bushings in most cases.

You know that energy takes the path of least resistance. So when you have a double-link top link system ... with the bushings touching or preloaded to a degree ... and you hit a bump on the right rear tire ... the energy goes three places. Some of it compresses the right rear spring, some it compresses the link bushing and some of it rocks the body on the car by extending the left rear spring ... lifting the body on that side.

This is easiest to see with GoPro digital video cameras under the car pointed at the suspension. We typically run two. But I have ran three, so I can watch the affects on pinion angle during all the different motions of double-link top link system.

[ace_xp2]

Bringing this one back up to add, maybe you could birdcage the rear brakes? A birdcage is basically a bearing that goes over the axle tube, so that the rear brake would attach to a big sleeve over the axle, which would then have a seperate link to the suspension. That way you could add all the anti-squat you'd like, and have the brake links set to a more preferable angle.

[BillyShope]

Bringing this one back up to add, maybe you could birdcage the rear brakes? A birdcage is basically a bearing that goes over the axle tube, so that the rear brake would attach to a big sleeve over the axle, which would then have a seperate link to the suspension. That way you could add all the anti-squat you'd like, and have the brake links set to a more preferable angle.

It's not a matter of anti-squat amount. You're going to want something close to 100% anti-squat in any event.

And, with regard to static loading on an asymmetric 3link with a 4th link to achieve equal tire loading on braking, I would remind you that calculations for the 3link totally disregard static loadings. Yes, of course the loadings will always be there. The important matter to consider, though, is with regard to the additional loadings which occur on launch and during braking.

[ace_xp2]

I was thinking in lieu of the decoupled system that was discussed a little later on. It really would be interesting to see the effects of 100% AS on something a little nearer a road course vehicle too, I really haven't come across vehicles built that way recently, which seems to indicate that there's not much benefit in being up in that range during corner exit, though it does seem to often bring benefits during a standing start.

[BillyShope]

I was thinking in lieu of the decoupled system that was discussed a little later on. It really would be interesting to see the effects of 100% AS on something a little nearer a road course vehicle too, I really haven't come across vehicles built that way recently, which seems to indicate that there's not much benefit in being up in that range during corner exit, though it does seem to often bring benefits during a standing start.
Many (most?) sports/racing cars today have IRS. While it is possible to obtain 100% anti-squat with IRS, it's somewhat difficult and usually results in a harsh ride because of the extensive geometry change from stock. When A-arms are involved, it's similar to the design choice regarding braking anti-dive: You either scrape the front bumper on braking or you have a harsher ride than the competition. But, then, ride isn't all that important in a sports/racing car.

Can't really understand why squat or rise would be intentionally designed into a car. Generally speaking, the less jounce and rebound the better. Again, any added stiffness should be acceptable in a sports/racing car. Of course, if you're a very good driver, you might notice slightly increased looseness after the apex.