Watts Link for improved street manners

[Norm Peterson]

Best I can come up with is that those little PHB-related lateral movements of axle and chassis come through on a more or less subliminal level as 'vagueness'. Nothing you can put your finger on, but the car feels more composed when they're gone (or at least reduced).

Bushing compliances are probably involved, as they turn all these precise-looking theoretical geometry calculations into rough approximations. Pivot points move away from where you measured them at, and the actual geometry in operation moves with them. Hopefully both pivots and geometry only move a little, but when was the last time anybody saw an aftermarket Watts link with rubber bushing pivots?

Rear suspension travel in 'bump' is supposedly 3.5". At about 1.5" into bump, the PHB levels out and further bump travel now makes those lateral motions move in the opposite direction. Lowered cars will, of course, be more subject to this effect.

For handling, a Watts link behaves the same way in both directions, while the OE PHB arrangement (passenger side chassis mount) causes the rear geo roll center to rise when turning right and lower when turning left. In addition, axle roll steer also varies very slightly differently, right turns vs lefts. Meaning that the understeer budget isn't the same for turns in both directions. Understeer budget is the summation of all manner of effects that individually have either understeer or oversteer effect. This can be quantified, but what you'd need to know goes beyond RC heights, bar & spring rates and roll couple distribution, and axle rollsteer %.


Norm

 

[Sky Render]

Well the axle can move up and down but it needs to do so in a straight vertical line relative to the chassis, right? So think about it, when one wheel hits a bump either the wheel with traction will need to slide over or the chassis of the car will need to shift over slightly in order to keep the vertical alignment between the chassis and the pumpkin. Either of which would seem to be a helpful source of vertical inertia which might reduce the net unsprung weight to be controlled on the side that hit the bump. Another way to look at it, you can usually feel the improvement if you lighten a wheel by 10 pounds or so. This effect might be the same?

I understand some people say a watts link does nothing more than a PHB in this situation but the anecdotal evidence is pretty substantial. Either a lot of people are wrong in their reviews or there is an engineering explanation for it.

The CENTER of the axle moves up and down relative to the chassis. There is nothing preventing one wheel from moving up and causing the axle to rotate about that center pivot, which is what would happen going over a bump with one wheel.

And there is no effect on un-sprung weight.

 

[stevbd]

The CENTER of the axle moves up and down relative to the chassis. There is nothing preventing one wheel from moving up and causing the axle to rotate about that center pivot, which is what would happen going over a bump with one wheel.

And there is no effect on un-sprung weight.

I hear you but I don't think that's right.

Seems to me on a WL setup what is helping prevent (or a better word might be control, or damp) one wheel from flopping up and down on a one wheel bump is the fact that the other wheel remains in contact with the ground. Thus as one wheel hits a bump, the center pivot point on the WL moves slightly to one side and the chassis of the car needs to move with it. There's no such limitation on a PHB car, where all the vertical movement of the axle needs to be controlled solely with the spring and damper.

It also seems to me that putting the pivot point in the center of the axle like with a WL design, rather than on one end like a PHB, should measurably reduce the unsprung weight the spring and shock needs to control when one wheel hits a bump. Basically, the spring and shock are controlling the weight of half an axle instead of the whole thing.

 

[stevbd]

By the way, I've looked all over for an explanation as to how a WL reduces the skittishness on rough roads of these cars, assuming they really do. It'd be great if someone who designs or sells these things could explain the science or point me to a link. Thanks!

 

[Sky Render]

There is no magical reduction in un-sprung weight. When you hit a one-wheel bump with a PHB or a Watts link, the other wheel stays on the ground (unless you're at Dukes of Hazard speeds), and the other wheel is still going to be affected because it's connected by a dang beam.

There's no such limitation on a PHB car, where all the vertical movement of the axle needs to be controlled solely with the spring and damper.

LOL wut? Control arms control (hence the name) the vertical motions of the axle.

It also seems to me that putting the pivot point in the center of the axle like with a WL design, rather than on one end like a PHB, should measurably reduce the unsprung weight the spring and shock needs to control when one wheel hits a bump.

No, the roll center is still in the center of the axle. The PHB pivots at both ends.

Basically, the spring and shock are controlling the weight of half an axle instead of the whole thing.

BOTH WHEELS ARE CONNECTED BY A BEAM.

The lateral "looseness" that people refer to with PHBs was eliminated when I switched to a rod-end panhard bar.

 

[SoundGuyDave]

The lateral "looseness" that people refer to with PHBs was eliminated when I switched to a rod-end panhard bar.

And there we have it. Several points to consider when looking at "awesome results with a Watts install."

1) Bushing compliance. This is particularly true if you go from a stock PHB to a Watts, which will be equipped with either poly or rod-end bushings. Just losing that compliance will tighten up the rear end "feel" dramatically. Between stock bushing compliance and tire carcass squirm, all the other factors are just lost in the "noise."

2) Roll center location and migration: When installing a Watts, it usually winds up changing the roll center as well, unless the pivot bolt is in EXACTLY the same vertical location as where the PHB crosses the axle vertical centerline. Migration of the roll center is indeed an issue, and is what causes the "different feel left to right" that some report with a PHB.

3) Other mods. Generally speaking, a Watts install isn't done in isolation, but is in conjunction with other things, like dampers/springs, control arms (bushing deflection!!), or even a torque arm. So once you throw somebody's entire catalog at the car, what, exactly, is responsible for the effects you observe?

4) Confirmation bias. Let's face it, this is human nature. If you just dumped a thousand bucks and a bunch of man-hours into a project, it's natural to say something on the order of "OMG!!!!1111 It's incredible!!" on a subjective level. On an objective level, there may have been no change, but the perception is still that there was one.

Pros/cons: A Watts is absolutely superior at acting as a lateral location device for the rear axle. There is (+/- bushing deflection) NO lateral shift whatsoever under bump or droop, the axle just moves vertically. A Panhard does move in an arc, which allows some shift to occur. Again, taking bushing deflection out of the equation, and applying the ole Pythagorean theorum, we find that that lateral movement is something on the order of 0.050" or so, not the "INCHES!!!1111" that are sometimes bandied about. So, why NOT go with a Watts? First, weight. The PHB is lighter, both sprung AND unsprung. Second, complexity. The PHB is simple to set up, and stupid-simple to maintain. The watts has a lot more moving parts. Third, price. You can find a nice rod-ended PHB for around $160, but Watts links can push past a grand.

I'm honestly not trying to push anybody one way or the other, it all comes down to what YOU want to do with YOUR car. Autocross folks that hit slaloms will feel that roll-center migration a lot more than road-course drivers ever will. And not everybody is running wall-to-wall meat in the rear wheelwells where that slight axle shift can cause rubbing... All I'm trying to suggest is that you think it all the way through, understand what the parts are doing, and then make an informed choice. Nothing more.

 

[stevbd]

All I'm trying to suggest is that you think it all the way through, understand what the parts are doing, and then make an informed choice. Nothing more.

And there's the problem, especially the understanding part.

As I said, I've spent good time researching this and the anecdotal evidence seems pretty strong that a WL works substantially better than a PHB at reducing rough road "skittishness." Apples to apples, all else being equal, just focusing on rough road DD. Not talking about track and autocross, migrating roll centers, or any of that stuff. Either all these reports of reduced skittishness are wrong, which seems unlikely but certainly possible, or there is an engineering explanation. But finding that engineering explanation - either why the WL is indeed more settled on rough roads, or why it makes no difference and is just snake oil on this point - I think is very difficult to find.

I don't agree with some of the recent points here. But I don't want to talk out of my depth or accidentally spread misinformation, so I am going to leave it here and get back to my day job - which is not engineering!

Thanks and have a good day.

 

[DevGittinJr]

I read a quote from Rehagen Racing where they claimed a one-second average reduction in lap times from the combination of the watts link and cambered axle they run. They're at the top of the food chain in terms of consistency and well sorted, so I have to believe that at best a watts link could only be worth about .5 seconds.

He also made the comment about the PHR and it's movement through the arc / travel range - which is a very good and legitimate point. The issue is that amount of calculated movement, plus the mentioned deflection, plus the additional fore/aft force that has not been mentioned and then the position of the RC - all culminates.

What people often forget is that the PHR also affects thrust angle. Thrust angle affects steer...and also plays a factor in how the car acts during roll, based on how it is affected by the geometry of the PHR.

Since we are being technical here, this culmination just keeps on getting amplified. You now have lateral axle shift, that affects thrust angle, that affects steer, that affects grip, that affects bushing operation and binding and on and on and on.

The above quote can't be ignored though. I've got heims at every location on the axle, except the shocks, and on both ends of the PHB. While the difference in articulation (and therefore grip) is pretty dramatic compared to poly, throttle application over some rough patches is still a bit precarious.

IMO it's not so much a single bump, but rather a series of bumps (rough patch) that tends to upset the rear end, on-throttle. Now, it's not just that the axle moved .050," in one direction or the other and back, but it's doing so a bunch of times really quickly. I picture the axle kind of sawing back and forth in addition to it's other movements. It'd only make sense that removing this tiny "sawing," and the roll-steer affects of the PHB, would make the car noticeably easier to drive and/or more confidence inspiring, even if only marginally quicker.

 

[SoundGuyDave]

I hear you but I don't think that's right.

Seems to me on a WL setup what is helping prevent (or a better word might be control, or damp) one wheel from flopping up and down on a one wheel bump is the fact that the other wheel remains in contact with the ground. Thus as one wheel hits a bump, the center pivot point on the WL moves slightly to one side and the chassis of the car needs to move with it. There's no such limitation on a PHB car, where all the vertical movement of the axle needs to be controlled solely with the spring and damper.

It also seems to me that putting the pivot point in the center of the axle like with a WL design, rather than on one end like a PHB, should measurably reduce the unsprung weight the spring and shock needs to control when one wheel hits a bump. Basically, the spring and shock are controlling the weight of half an axle instead of the whole thing.

Okay, a little suspension explaination is in order, here, I think...

SPRINGS: The only thing they do, is keep the body up above the axle. Period. Full stop.

DAMPERS: Are there to control the rate of the spring compression/expansion, and to damp oscillations from spring movement.

CONTROL ARMS: Are there to locate the axle fore/aft (lower control arms), and prevent it from rolling in side-view (upper control arm) or yawing in plane view. They do nothing (disregarding bushing compliance) to keep the axle centered under the longitudinal centerline of the car. Without these three links, the axle is free to move on any axle, be it pitch, roll, or yaw.

Panhard bar or Watts link (or DeDion tube, or....) has only one function, and that is to positively locate the axle relative to the centerline of the car (top or rear view). They work on different principles, but the effect is pretty much the same.

PANHARD BAR: On the S197, is a long (~48" from memory) bar connecting a chassis pivot point (passenger side) to an axle pivot point (driver's side). With a two-wheel bump (like a speed bump), the axle will move vertically, which WILL cause the distance between the two pivot points (as measured on the ground) to change, forcing the axle centerline to move relative to the chassis centerline. Depending on the angle of the bar when the suspension is static, it may move towards the left or the right. To figure out the net change in centerline position, let's make some assumptions. First, that when static, the axle centerline and chassis centerline are in vertical alignment. Second, that we're discounting bushing compliance completely, to keep the math simple. So, we're going to look at some triangles. Assume that static, the chassis side is higher than the axle side pivot point. And let's say that it's by 2". So, we can define the three sides of a right-triangle: (1) the length of the bar itself, which will remain constant, connecting the two pivot points (“C”). (2) The vertical height difference between the pivot points (“A”), and (3) the "unknown" side, defined by the lateral distance between the two pivot points (“B”). So, let's plug in some numbers. (1) is 48", (2) is 2", and (3) is "X" that we need to solve for. Pythagorean Theorum says that A squared plus B squared equals C squared. Cool! That means that B squared is equal to C squared minus A squared. Plugging in the numbers, B squared equals 2304 (C squared) minus 4 (A squared). Or, B squred is 2300 even. The square root of that is 47.958”.

Now, what happens with our two-wheel, 1" bump? (1) remains at 48", since the length of the tubing hasn't changed, (2) changes to 1 instead of 2, but what happens with (3)?? Well, 48" squared is still 2304. 1 squared is simply 1. That makes B squared 2303 instead of 2304. The square root of 2303 is 47.990. Thus, comparing the two, for a 1" change in bump, the axle will shift 0.032" to the driver's side (47.958-47.990= -0.032).

How about for a 1" droop, like dropping the tires in a big wide pothole? Again, 48 squared is 2304. -2 squared changes to -3 squared, which is -9. Sum is 2295. Square root is now 47.906. Difference of 47.906 and 47.958? 0.052", this time shifted to the passenger side.

Now, for a one-wheel bump, the math becomes a lot more complex, but it should still go to show that we're not talking about very huge lateral movements here. If you think about it, even if you were to DOUBLE the lateral motion from 0.052" to 0.104", it's still spit in a bucket compared to tread squirm or carcass shift in the tires during lateral loading. Even with Hoosier R6, you can see a full INCH of carcass shift in hard cornering... That's a full order of magnitude higher than the axle centerline shift.

Now, if you want to compare that to a Watts link, though, you've got a whole different mechanism going on. There are essentially two styles of Watts available for our cars; differential mounted or chassis mounted, and they each have their own sets of pluses and minuses. With either style, you now have a propeller bolt fixing a vertically oriented "football" in place. This football is now connected to the chassis (for a diff-mount Watts) or the axle (for a chassis-mount Watts) by two links, attaching to the top and bottom "points" of the football. Let's assume that you have the chassis-mount style, which has a nice crossmember spanning the frame rails that the prop bolt and football attach to, and has the two arms attaching to brackets mounted on the axle tubes. You can think of this a TWO Panhard bars, one on the driver's side attaching to a mount connected to but above the axle, and the other on the passenger's side attaching to a mount connected to but below the axle. Instead of one triangle (same legs as above: Fixed-length arm between the football and the axle, height difference, and height difference), you now have two, and they are mirror images both vertically and horizontally, offset by the height of the football.

As you hit a bump, the axle housing itself moves upwards, relative to the propeller bolt on the crossmember. This obviously changes the height differences of the pivot points, causing the "C" value to change as well. But, since the one point is moving away from the prop, and the other is moving towards it, the prop simply spins in place, and as a result the axle goes straight up and down. Continuing the double-PHB analogy, the two arcs described cancel out, and there is no lateral shift.

Now, roll centers... On a PHB, the roll center is described by where the PHB intersects the vertical centerline of the axle. Under bump, it shifts up and down, but only about half of the travel distance of the chassis. It also shifts laterally, slightly, with the mid-point of the PHB describing an arc. Again, very small movements. But, in roll, things change. since the body pivots around the (duh!) roll center, it can do all kinds of "weird" things with the roll center placement in right and left turns, and whether you're trail-braking or on-throttle. With a Watts, the lateral roll-center is constrained, which is good. However, the vertical roll-center is actually freer to move than with a Panhard. With a PHB, under squat in the rear, the roll center drops, but only by about half the distance that the chassis squats over the axle. We’ve already ascertained that the lateral motion (with proper bushings) is actually quite minimal, so again the RC migration, while there, isn’t all that huge. With a chassis-mounted Watts, that roll center is fixed relative to the chassis, and thus drops at the same rate as chassis squat. With a diff-mounted Watts, the roll center is fixed vertically relative to the ground, but is now variable to all the chassis points, and at full rate relative to the motion of the chassis.
What does this all mean? They do the same job, lateral axle location, but they do it in different manners. Each has advantages, and each has disadvantages. With proper bushings, though, neither will allow that “floaty, disconnected” feeling. Which one to choose really comes down to the design goals for the car, and other contraints. If you’re looking for an ultra-leightweight build, PHB. If you’re stuffing all the rubber Akron ever made under the wheelwells, then Watts. Autocross? Watts. Road-course? Well, either, but I would go with the PHB for lack of complexity.
[FONT=&quot]From a subjective standpoint, and this is ONLY my opinion, the Watts just doesn’t bring enough to the table over a rod-ended Panhard bar to make the conversion worthwhile. I’ve driven (and raced) a LOT of S197 Mustangs, with all kinds of power levels, and all kinds of suspension packages under them. When I drove a Watts-equipped (well built) S197, I was frankly underwhelmed. I had honestly expected some SERIOUS handling differences out back compared to mine, and they just weren’t there. For the kilobuck and the extra pounds, I’d stick with a rod-ended PHB. That being said, I don’t autocross, so my experience was all open-track[/FONT]

 

[SoundGuyDave]

And there's the problem, especially the understanding part.

As I said, I've spent good time researching this and the anecdotal evidence seems pretty strong that a WL works substantially better than a PHB at reducing rough road "skittishness." Apples to apples, all else being equal, just focusing on rough road DD. Not talking about track and autocross, migrating roll centers, or any of that stuff. Either all these reports of reduced skittishness are wrong, which seems unlikely but certainly possible, or there is an engineering explanation. But finding that engineering explanation - either why the WL is indeed more settled on rough roads, or why it makes no difference and is just snake oil on this point - I think is very difficult to find.

I don't agree with some of the recent points here. But I don't want to talk out of my depth or accidentally spread misinformation, so I am going to leave it here and get back to my day job - which is not engineering!

Thanks and have a good day.

Okay, sorry for the long post previously, that was in the works before you said you didn't want all that....

Addressing strictly the above, and assuming you're 100% bone stock in the rear suspension, you're fighting two, very well known issues. First, the uber-soft bushings in the stock Panhard. That thin is slopping all over the place laterally even before you get to the PHB arc thing. Second, your rear shocks are severely under-damped. Soccer moms love that "smooth Cadillac ride" that you get from nerf-soft bushings and soft shocks. That is, assuming your factory rear dampers aren't blown by now. Assuming you're stock, none of the other factors are coming into play.

As for all the anecdotal evidence, and being unable to unearth an engineering explanation, I would call that confirmation bias.

If you want to do an experiment, buy a $10 take-off or junkyard stock PHB, and replace the bushings with some poly from Freedom Racing. That's another $10 or so. Swap it out and do nothing else. That will show you what bushing compliance is all about, and for less than $25 out of pocket.

 

[SoundGuyDave]

I read a quote from Rehagen Racing where they claimed a one-second average reduction in lap times from the combination of the watts link and cambered axle they run. They're at the top of the food chain in terms of consistency and well sorted, so I have to believe that at best a watts link could only be worth about .5 seconds.

Interesting. I know Dean as a driver and am inclined to listen very carefully when he speaks. That said, we have the classic two-variable equation here. How much is the axle worth by itself? How much is the Watts worth by itself? My gut is telling me the cambered axle is worth a HELL of a lot more than 50% of that time difference. Think about how much you gain from a simple camber plate up front. In back that lets you get on the power a LOT harder and earlier, which translates to higher exit speed, and that's carried all the way down the straight to the next corner. The Watts can only affect the transitional phases. In other words, it can help you get to max lateral grip in a corner quicker and easier, but it can't add grip in and of itself. Once the car is set, all roll centers are static, geometry is fixed, etc. I could see some very minor gains in that area (assuming a driver like Dean), but you're talking about shaving thousandths between initial turn-in and when the car sets, and it takes an awful lot of those to create a noticeably quicker lap.

He also made the comment about the PHR and it's movement through the arc / travel range - which is a very good and legitimate point. The issue is that amount of calculated movement, plus the mentioned deflection, plus the additional fore/aft force that has not been mentioned and then the position of the RC - all culminates.

What people often forget is that the PHR also affects thrust angle. Thrust angle affects steer...and also plays a factor in how the car acts during roll, based on how it is affected by the geometry of the PHR.

Since we are being technical here, this culmination just keeps on getting amplified. You now have lateral axle shift, that affects thrust angle, that affects steer, that affects grip, that affects bushing operation and binding and on and on and on.

The above quote can't be ignored though. I've got heims at every location on the axle, except the shocks, and on both ends of the PHB. While the difference in articulation (and therefore grip) is pretty dramatic compared to poly, throttle application over some rough patches is still a bit precarious.

IMO it's not so much a single bump, but rather a series of bumps (rough patch) that tends to upset the rear end, on-throttle. Now, it's not just that the axle moved .050," in one direction or the other and back, but it's doing so a bunch of times really quickly. I picture the axle kind of sawing back and forth in addition to it's other movements. It'd only make sense that removing this tiny "sawing," and the roll-steer affects of the PHB, would make the car noticeably easier to drive and/or more confidence inspiring, even if only marginally quicker.

I'll still disagree (politely) with Kelly about the thrust angle. Thrust angle is the difference (in degrees) between the axle centerline and the chassis centerline. Assuming pure bump, both of the lower arms are staying in-plane, and thus there is no change in axle centerline and no change in thrust angle. Now, if you add roll to the equation, yes, the thrust angle changes, but that's called roll-steer, and has nothing to do with the PHB or Watts, as it is a function of the lower control arm pivot points moving differently longitudinally. NOTE: You can have the axle centerline offset with the vehicle centerline (but parallel to it: dog-tracking) and still have zero thrust angle. With roll-steer, though, the more anti-squat you dial into the car, the more pronounced the effect.

Question for you: When you're talking about being "on-throttle over a rough patch" are you in a straight line or cornering situation? If the axle is moving vertically a full inch of bump that's a lot more than a "rough patch." I'm thinking about things like expansion joints, or those patches of pavement that are cut to make audible warning strips when approaching a toll booth or something like that. More on the order of 1/4" in height. In a straight line, with that small a vertical motion, even one as rapid as the rumble strips, the PHB arc-induced lateral motion is so vanishingly tiny that it's overwhelmed by carcass flex, among many many other things. If you're skittish there, you're either over-sprung (moving the axle vertically lifts the rear of the car, unloading the tire) or you're over-damped (dampers won't track the bump, are effectively solid rods, and again unload the tire). Remember that traction is all about the contact patch. With the proper spring rates, the chassis of the car won't really move too much even when the axle oscillates up and down. With proper damping (matching the spring rates), the axle tracks the surface, and doesn't unload that contact patch. As an example of being overdamped (but in rebound, rather than in compression) google "jacking down" with respect to damper tuning. The rebound bleed is too slow to let the spring extend back up before the next bump comes and compresses it even more.

All that being said, I'm not discounting the final sentence in your post about perhaps being more confidence-inspiring. So much of racing is a mental game, and that comfort could be an edge.

 

[stevbd]

Okay, sorry for the long post previously, that was in the works before you said you didn't want all that....

Addressing strictly the above, and assuming you're 100% bone stock in the rear suspension, you're fighting two, very well known issues. First, the uber-soft bushings in the stock Panhard. That thin is slopping all over the place laterally even before you get to the PHB arc thing. Second, your rear shocks are severely under-damped. Soccer moms love that "smooth Cadillac ride" that you get from nerf-soft bushings and soft shocks. That is, assuming your factory rear dampers aren't blown by now. Assuming you're stock, none of the other factors are coming into play.

As for all the anecdotal evidence, and being unable to unearth an engineering explanation, I would call that confirmation bias.

If you want to do an experiment, buy a $10 take-off or junkyard stock PHB, and replace the bushings with some poly from Freedom Racing. That's another $10 or so. Swap it out and do nothing else. That will show you what bushing compliance is all about, and for less than $25 out of pocket.

Don't apologize I love your detailed and thoughtful posts.

For context, last summer I replaced my stock suspension with P springs and Bilsteins, and after that a Whiteline adjustable PHB. Consistent with what you're saying, the springs and shocks made a massive difference in rough road daily driving stability. The adjustable PHB with poly-whatever bushings - not really so much.

Since then I've spent the winter driving some of the worst roads in New England and have found that the car is still just a little unstable on the worst patches of road, although it's world's better than before. Echoing DevGitten, it seems to be not single big bumps that get it, but a series of closely spaced bumps spaced closely together. Not a high load corner like on a track or autox, just a horrible road. In that situation it will still skip a little bit left and right, although again it's like 85 percent better than before. This last 15-20 percent is what would spend some money fixing with a WL, if it works. But obviously I don't want to just throw away money.

My issue is not on the track. There I feel the car is very easy, stable, and fun to drive quickly with the current PHB setup, at least for a comfy daily driver. I am loving my car on the track.

Thanks again.

 

[SoundGuyDave]

Don't apologize I love your detailed and thoughtful posts.

For context, last summer I replaced my stock suspension with P springs and Bilsteins, and after that a Whiteline adjustable PHB. Consistent with what you're saying, the springs and shocks made a massive difference in rough road daily driving stability. The adjustable PHB with poly-whatever bushings - not really so much.

Since then I've spent the winter driving some of the worst roads in New England and have found that the car is still just a little unstable on the worst patches of road, although it's world's better than before. Echoing DevGitten, it seems to be not single big bumps that get it, but a series of closely spaced bumps spaced closely together. Not a high load corner like on a track or autox, just a horrible road. In that situation it will still skip a little bit left and right, although again it's like 85 percent better than before. This last 15-20 percent is what would spend some money fixing with a WL, if it works. But obviously I don't want to just throw away money.

My issue is not on the track. There I feel the car is very easy, stable, and fun to drive quickly with the current PHB setup, at least for a comfy daily driver. I am loving my car on the track.

Thanks again.

Okay, so not stock... Cool. You've eliminated the two big evils back there, and I don't remember the P-springs (or the Bilsteins) being overly aggressive, so you're probably not radically off in rate or damping curve.

Assuming you're looking to tame the beast, then I would ask a favor. Swap to a Watts (you seem to be leaning that way, looking for some confirmation that it will solve your ills), but do NOTHING else when you swap, and then assess what just that one change did. Before you swap, mark the diff cover (at static ride height) where the PHB crosses the center of the cover, and install your pivot bolt at the same height to eliminate the change in RC. That will give you a true Watts vs. PHB comparison. After you assess, feel free to move the RC as you see fit, obviously!

I, for one, am genuinely interested in the results. Yes, it will be purely subjective, as there is no objective frame to measure against like lap time. Who knows, you might be the one that finally puts this Watts vs. PHB shitstorm to bed once and for all!

 

[DevGittinJr]

Interesting. I know Dean as a driver and am inclined to listen very carefully when he speaks. That said, we have the classic two-variable equation here. How much is the axle worth by itself? How much is the Watts worth by itself? My gut is telling me the cambered axle is worth a HELL of a lot more than 50% of that time difference. Think about how much you gain from a simple camber plate up front. In back that lets you get on the power a LOT harder and earlier, which translates to higher exit speed, and that's carried all the way down the straight to the next corner. The Watts can only affect the transitional phases. In other words, it can help you get to max lateral grip in a corner quicker and easier, but it can't add grip in and of itself. Once the car is set, all roll centers are static, geometry is fixed, etc. I could see some very minor gains in that area (assuming a driver like Dean), but you're talking about shaving thousandths between initial turn-in and when the car sets, and it takes an awful lot of those to create a noticeably quicker lap.

I agree that the cambered axle is most likely the majority of the gains. "Standard" is -1.5 degrees, and they can go up to -2.5. Also, in any/all of the pics I've seen of their cars (Roush, Stacy, Martin), they're running the pivot in the second hole from the bottom, which puts the RCH about 1-2" lower than it'd be with a PHB, at the same ride height. So the lack of lateral axle movement is probably worth very little in terms of lap times, which is what I was implying above but should've been more clear.

I'll still disagree (politely) with Kelly about the thrust angle. Thrust angle is the difference (in degrees) between the axle centerline and the chassis centerline. Assuming pure bump, both of the lower arms are staying in-plane, and thus there is no change in axle centerline and no change in thrust angle. Now, if you add roll to the equation, yes, the thrust angle changes, but that's called roll-steer, and has nothing to do with the PHB or Watts, as it is a function of the lower control arm pivot points moving differently longitudinally. NOTE: You can have the axle centerline offset with the vehicle centerline (but parallel to it: dog-tracking) and still have zero thrust angle. With roll-steer, though, the more anti-squat you dial into the car, the more pronounced the effect.

What stood out to me most about Kelly's quote was the cycle. How each thing seems to induce or intensify the affects of the other. If you completely remove lateral movement from the cycle wouldn't you gain some stability?

Question for you: When you're talking about being "on-throttle over a rough patch" are you in a straight line or cornering situation? If the axle is moving vertically a full inch of bump that's a lot more than a "rough patch." I'm thinking about things like expansion joints, or those patches of pavement that are cut to make audible warning strips when approaching a toll booth or something like that. More on the order of 1/4" in height. In a straight line, with that small a vertical motion, even one as rapid as the rumble strips, the PHB arc-induced lateral motion is so vanishingly tiny that it's overwhelmed by carcass flex, among many many other things. If you're skittish there, you're either over-sprung (moving the axle vertically lifts the rear of the car, unloading the tire) or you're over-damped (dampers won't track the bump, are effectively solid rods, and again unload the tire). Remember that traction is all about the contact patch. With the proper spring rates, the chassis of the car won't really move too much even when the axle oscillates up and down. With proper damping (matching the spring rates), the axle tracks the surface, and doesn't unload that contact patch. As an example of being overdamped (but in rebound, rather than in compression) google "jacking down" with respect to damper tuning. The rebound bleed is too slow to let the spring extend back up before the next bump comes and compresses it even more.

I'm referring to getting back on the throttle on exit - not driving like a dork, but just kind of hustling it back up to the speed limit around corners. You want to roll into the throttle quickly but have feed it slow until you're past the rough patch. The body's still in slight roll and the bumps are to both sides, randomly. Like stevbd said, it's mostly better, especially with heim-joints and the longer '11+ UCA. I'm not on the watts bandwagon per say, but I've been kicking around the idea lately from an all things considered (RCH adj, large tire fitment, and axle stability) perspective.

 

[SoundGuyDave]

I agree that the cambered axle is most likely the majority of the gains. "Standard" is -1.5 degrees, and they can go up to -2.5. Also, in any/all of the pics I've seen of their cars (Roush, Stacy, Martin), they're running the pivot in the second hole from the bottom, which puts the RCH about 1-2" lower than it'd be with a PHB, at the same ride height. So the lack of lateral axle movement is probably worth very little in terms of lap times, which is what I was implying above but should've been more clear.

Agreed. FWIW, there are aftermarket PHB setups that do allow for RCH tuning, but I don't think they have as wide a range as some of the Watts links on the market.

What stood out to me most about Kelly's quote was the cycle. How each thing seems to induce or intensify the affects of the other. If you completely remove lateral movement from the cycle wouldn't you gain some stability?

To me, though, the heart of his argument is disprovable. He posits that "You now have lateral axle shift, that affects thrust angle, that affects steer, that affects grip, that affects bushing operation and binding and on and on and on." Bolding is mine. I believe that to be incorrect, by definition. The two lower control arms lock the axle into alignment with the chassis centerline, at least under pure bump/rebound conditions. Roll, on the other hand, does introduce a steering effect (roll steer), but that effect is present whether using a Watts or PHB, and is a function of the %AS dialed in through the LCAs. So the lateral axle shift does NOT induce a steering effect in and of itself, which then discounts the loss of grip, etc that he mentions.

Consider that the rapid, repeated bump/rebound cycles have to be small. After all, there's no way you're going to find traction driving over a road surface that approximates a series of 2x4 boards stretched across the road. How big a vertical surface change are we really talking about? 1/4"? 1/2" at the most? With something that "small," tire deflection becomes a very large component of how the car handles that evolution. A small and short object (like a pencil, for example) will have the tire actually wrap around it, more than lift the end of the axle. Remember that the tire, loaded with nice compressible air, is a spring; and one that has no damper. VIDEO At around he :42 mark, there's a nice slo-mo fly-by of a F1 car running across some FIA curbing, which is scalloped. You can see the front wing end bouncing, so we know that there is force going into the chassis, but what's instructive is to look at the tire contact patch, AND the wheel hub center. The contact patch flows over and around the scallops on the curbing, but the hub appears to move at around 50% of the expected rate. The balance is being absorbed by the tire. This is with a racing slick. On a street tire, the tread blocks themselves (particularly on an M+S design) will also add squirm in every direction as the block itself distorts under the load. With a 1/4" vertical deflection (and discounting bushings), the axle shift is 0.010" laterally. Those ten-thousandths-of-an-inch simply disappear in tread squirm and carcass deflection. There's a slow-mo video of one of the NASA nationals that I just can't seem to find, where they show about a half-dozen cars dropping wheels onto some very aggressive "gator strips" at track-out that really shows this clearly. If anybody can find a link, I think it would be instructive!

I'm referring to getting back on the throttle on exit - not driving like a dork, but just kind of hustling it back up to the speed limit around corners. You want to roll into the throttle quickly but have feed it slow until you're past the rough patch. The body's still in slight roll and the bumps are to both sides, randomly. Like stevbd said, it's mostly better, especially with heim-joints and the longer '11+ UCA. I'm not on the watts bandwagon per say, but I've been kicking around the idea lately from an all things considered (RCH adj, large tire fitment, and axle stability) perspective.

I'll ask the same favor as I did stevbd. If you do pull the trigger, please do the install with no other changes, and set the RCH as close to the stock PHB height as possible to minimize variables and let us know what you think.

 

[DevGittinJr]

Agreed. FWIW, there are aftermarket PHB setups that do allow for RCH tuning, but I don't think they have as wide a range as some of the Watts links on the market.

That's what I've noticed as well. I only know of the Kenny Brown PHB (only 3 RCH positions) and the Steeda one (weld-in), and they both have their drawbacks.

A small and short object (like a pencil, for example) will have the tire actually wrap around it, more than lift the end of the axle. Remember that the tire, loaded with nice compressible air, is a spring; and one that has no damper. VIDEO At around he :42 mark, there's a nice slo-mo fly-by of a F1 car running across some FIA curbing, which is scalloped. You can see the front wing end bouncing, so we know that there is force going into the chassis, but what's instructive is to look at the tire contact patch, AND the wheel hub center. The contact patch flows over and around the scallops on the curbing, but the hub appears to move at around 50% of the expected rate. The balance is being absorbed by the tire.

I can see what you mean, but Kelly's point made sense at the time too. Now I'm questioning my own judgement.

I'll ask the same favor as I did stevbd. If you do pull the trigger, please do the install with no other changes, and set the RCH as close to the stock PHB height as possible to minimize variables and let us know what you think.

If I go with a WL, one day (the next debate is chassis or diff mount), I'll raise you one and put the PHB back on after a few weeks or so. I find that sometimes switching back is where I notice more of a difference, assuming there is one.

I can't find the above mentioned quote from Martin about the Rehagen cars, but I read this post by VooDooBOSS on TMO, a while back as well:

"I was skeptical about installing the watts link on my car but I talked to Dean Martin about it at the PWC race at Sonoma in 2014 and he convinced me to try it. I was not disappointed. Just like FD stated you can hit the curbing on track and the car no longer gets unsettled. I was a noticeable difference. I also thought it might be BS about noticing the difference from left to right transitions but I noticed it right away. No longer did the car jump to the left when hitting hard bumps either. Instead it just sucked them up and went straight. Filip at CorteX will tell you that the first suspension mod you should make is installing the watts link and I'm a believer in that approach. I'm sure the KB panhard bar relocation kit works fine and I spent an hour talking to both Kenny and Chuck at MM about the merits of their kits. But the watts link made more sense to me and I'm glad I went in that direction. Good luck on your decision."​

I posted this for what Martin said about trying a WL. It's hard to ignore, given his history with this chassis. I know Filip sells WL's, but he also sells other components that are less expensive and are far less controversial and/or the affects are much more widely quantifiable. If he was advising people just for sales, it'd make more sense to break people in with a cheaper first mod. Meaning, (without saying whether or not he's right) I think he's probably giving what he considers to be sound advice based on his knowledge.

I think this is a fascinating debate, FWIW, and it even wages on outside the s197 community. Some really smart people on both sides, but I'm not knowledgeable enough to argue either way.

 

[SoundGuyDave]

I can see what you mean, but Kelly's point made sense at the time too. Now I'm questioning my own judgement.

Ah, don't read too much into it. Kelly's argument really boils down to whether the lateral axle motion induces a change in thrust angle or not, since all the following effects he listed are dependent on that condition. I disagree with his primary assertion about the change in thrust angle. The rest (stacking of effects) is actually quite correct, IF you can get the train started in the first place. Unfortunately for our understanding of the pros/cons of Watts and PHB, the roll-steer effect occurs with both, and has nothing to do with the design of the lateral locating device.

If I go with a WL, one day (the next debate is chassis or diff mount), I'll raise you one and put the PHB back on after a few weeks or so. I find that sometimes switching back is where I notice more of a difference, assuming there is one.

Perfect. The more "data" the better!

I can't find the above mentioned quote from Martin about the Rehagen cars, but I read this post by VooDooBOSS on TMO, a while back as well:

"I was skeptical about installing the watts link on my car but I talked to Dean Martin about it at the PWC race at Sonoma in 2014 and he convinced me to try it. I was not disappointed. Just like FD stated you can hit the curbing on track and the car no longer gets unsettled. I was a noticeable difference. I also thought it might be BS about noticing the difference from left to right transitions but I noticed it right away. No longer did the car jump to the left when hitting hard bumps either. Instead it just sucked them up and went straight. Filip at CorteX will tell you that the first suspension mod you should make is installing the watts link and I'm a believer in that approach. I'm sure the KB panhard bar relocation kit works fine and I spent an hour talking to both Kenny and Chuck at MM about the merits of their kits. But the watts link made more sense to me and I'm glad I went in that direction. Good luck on your decision."​

I posted this for what Martin said about trying a WL. It's hard to ignore, given his history with this chassis. I know Filip sells WL's, but he also sells other components that are less expensive and are far less controversial and/or the affects are much more widely quantifiable. If he was advising people just for sales, it'd make more sense to break people in with a cheaper first mod. Meaning, (without saying whether or not he's right) I think he's probably giving what he considers to be sound advice based on his knowledge.

Yes, Dean knows this chassis very well, and I am not about to argue with his experiences and recommendations. And Filip is a fantastic engineer, but I don't know about "first mod to make." For me, that would be dampers and springs.

I think this is a fascinating debate, FWIW, and it even wages on outside the s197 community. Some really smart people on both sides, but I'm not knowledgeable enough to argue either way.

I just get a knee-jerk reaction to ANY "must have" suspension piece that simply doesn't have the supporting evidence behind it to warrant that amount of hysteria. I run a PHB. I could easily be convinced to run a Watts, as well. I frankly don't really care one way or the other, my only dog in this hunt is to help people understand what they're considering doing to their car, that's all.

One final comment to make on using Dean's car as a "perfect" model... He's in a race series, and is thus constrained by what the rule book allows. However, you need to think like a racer. If the rule book allows a choice of two pieces, and one gives a microscopic edge over the other, you go with it. Think about the guys that are gun-drilling all the bolts on their car to shave weight, because they're restricted to OEM body panels. With racers, they're not chasing the low-hanging fruit, they're after that last fractional-percentage gain, and generally budget is not an issue ($225K Mustang? Hello?). For the rest of us, trying to chase that last 0.02%, when you're on stock bushings and shocks is just pointless, yet the interwebz say that if you don't put a Watts on the car it'll explode the first time you try to take a corner. That's what I'm really trying to get into the spotlight.

I freely acknowledge that the Watts does a BETTER job of lateral location of the axle than a PHB does, but it's just not THAT MUCH better to justify the hysteria that we see in these threads. If you steal Dean Martin's Watts link, and install a rod-end PHB he's still going to drive rings around any of us on here, right? It's just not the low-hanging fruit (high-percentage reward) that people seem to think it is.

 

[Norm Peterson]

To me, though, the heart of his argument is disprovable. He posits that "You now have lateral axle shift, that affects thrust angle

In the strictest theoretical sense, lateral axle shift can affect thrust angle. All that's required is for the LCAs to be angled in plan view with respect to the car centerline. One LCA will push its end of the axle a really tiny amount rearward, the other will pull its axle end a similar (but not equal) amount forward.

This situation actually exists in the S197, but the plan view angles so small and the lateral shifts also so small that the amount of thrust angle change - calculated considering geometry alone - is essentially nonexistent. 0.05" of geometric lateral axle shift works out to something like 0.0001° of thrust angle change.

This effect would have been much greater in the Fox/SN95 chassis (or in the GM Intermediate that I used to have), both of which orient their rear LCAs at significantly greater plan view angles relative to vehicle centerline and have (as OE) less rigid control over lateral axle shift. Perhaps this explains why it occurs to Kelly to mention the effect at all.


Norm

 

[SoundGuyDave]

Norm, why do you have to keep bringing reality into things??? ;-)

I forgot about the minor skew in the LCAs, which I believe most of the aftermarket adjustables (rod-end) cancel out with offset bushings.

I also take your point on the "essentialy nonexistant" change that imparts. I'm going to stick with my statement, where any effect from this is going to be lost in the "noise" of other things like tread squirm or carcass deformation.

 

[stevbd]

I think Cortex's recommendation to do the WL first is because they don't recommend installing their coilovers without an axle relocation solution. I don't read them to be saying the WL is the biggest bang for the buck.

So much of this discussion is focused on the fact that the WL more closely approximates a straight vertical line instead of the arc with the PHB. But what also seems important - although I can't say precisely why - is that the WL attaches to and locates the CENTER of the axle, rather than one end. It seems like this might somehow better balance the substantial weight of the axle, allow the L and R springs and shocks to work better together, and thus reduce the forces that each individual spring and damper needs to control. Sort of like the difference between picking up one end of a heavy log versus having that log balanced in the middle like on a see-saw. Yes, I understand the WL pivot point is not fixed like on a see-saw.