The Mustang Forum for Track & Racing Enthusiasts

Taking your Mustang to an open track/HPDE event for the first time? Do you race competitively? This forum is for you! Log in to remove most ads.

  • Welcome to the Ford Mustang forum built for owners of the Mustang GT350, BOSS 302, GT500, and all other S550, S197, SN95, Fox Body and older Mustangs set up for open track days, road racing, and/or autocross. Join our forum, interact with others, share your build, and help us strengthen this community!

Panhard bar relocation discussion, round 2

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Norm Peterson

Corner Barstool Sitter
939
712
Exp. Type
HPDE
Exp. Level
5-10 Years
a few miles east of Philly
Grant - I think I found the disconnect that caused a bit of commotion in the other thread. I couldn't reply in the other thread because it had already been closed, so I just forgot about it. But I stumbled back on this thread while looking for a different topic.

I think you were looking for my opinion in part on this:
But mechanically, if the CG height, spring rate and sway bar rate stays the same, there shouldn’t be more roll.
No flame intended, but it looks like you're confusing lateral load transfer with roll, like X amount of LLT can only cause Y amount of roll. It doesn't quite work that way.

Let me get something out of the way first. Aside from within this paragraph, I'm going to be using terms like 'lateral load transfer' and 'total lateral load transfer' and their acronyms exclusively. I know that most of the car enthusiast world likes to use "weight transfer", but the only weight that actually transfers (i.e. physically moves around in the car) is that due to fluid slosh and poorly restrained cargo and passengers. Maybe a couple of lbs from engines and transmissions rocking in their mounts. We're going to ignore those effects, just like anybody who talks in terms of "weight transfer" does as well.

Lateral load transfer can be broken down into three distinct components. LLT through the geometric roll centers (where you get LLT but no roll), LLT through the elastic elements of the suspension (which is where most of your visible roll comes from), and LLT coming from the unsprung masses (wheels/tires/brakes/etc.).

So, if you leave the CG height the same, the total lateral load transfer remains the same. Note that this says nothing whatsoever as to how it gets distributed among the above components of LLT. It could be all going through the roll centers, where you'd get no roll in the suspension (though you would still see a bit due to vertical tire stiffnesses being what they are and the LLT obviously changing the inside and outside tire loadings). Or it could all be going through the elastic suspension elements (the geo-RC would be at grade), where you'd certainly get roll in the suspension. Or in any other combination. But the TLLT would stay the same.

That means if you were to hold the CG height constant, and vary a roll center height, as you get less LLT through a roll center because you lowered the RC, you have to be getting more LLT through the suspension. And that means more roll. This is really just a slightly different way of looking at it than watching the vertical distance between the CG and the RCs increase and computing a larger roll moment. Whatever works, just that sometimes looking at things from a different direction helps.


But as I’m suggesting, the transfer will happen faster. And again, that’s just for the panhard setup.
Actually, lowering the RC height tends to slow down the load transfer. That's because LLT through the roll centers happens almost instantaneously, while LLT through the suspension has to wait for the roll to finish developing (spring and bar forces reflected out to the tires) or at least develop a velocity (damper forces). Even a stiff coilover suspension is relatively soft in comparison to the stiffness of the geometric load path (through the geo RCs), assuming that the suspension linkage and the associated pivots and bracketry are up to snuff.

The effect on handling is more involved, as the transitions from geometric LLT to combined geo+elastic LLTs are no longer the same as before (have to think in terms of force variations with respect to time here). And axle roll steer varies with PHB height as well as with rear LCA plan and side view inclinations.


(truth be told, I was looking to see if Billy Johnson had posted any PHB vs Watts link thoughts).

Norm
 
Last edited:

Grant 302

basic and well known psychic
I think I found the disconnect that caused a bit of commotion in the other thread.

🤔

Pretty sure I found it. ;) 😁

...

Anyway, I do think we went over it pretty well in the other thread that you started, and that we discussed - without interruption.

If I’m looking at the whole subject a little more philosophically, the main difference in wether there is more or less roll as the RC moves is more dependent on the relative rates of the double-sprung model.

At the extremes of such a model would be a road course car with hot tires and OEM springs vs. a drag car on slicks that is sprung past the tire rates.

Kenny comes close to the point where the effect might flip its reaction...when he starts talking about rates in the 600-800#/in range. And FWIW, I suspect it actually would if including the damping force(s), and any time those two exceed the tire spring rate.
 

Norm Peterson

Corner Barstool Sitter
939
712
Exp. Type
HPDE
Exp. Level
5-10 Years
a few miles east of Philly
If I’m looking at the whole subject a little more philosophically, the main difference in wether there is more or less roll as the RC moves is more dependent on the relative rates of the double-sprung model.

At the extremes of such a model would be a road course car with hot tires and OEM springs vs. a drag car on slicks that is sprung past the tire rates.
Yes . . . it is (partly) a springs-in-series thing. In the case of a PHB-located axle, you've got roll in the suspension, which is really just the body roll relative to the axle's own roll (which is due to tire stiffness effects). But when you look at the car in action, you mostly just see the total unless you're specifically looking at tire distortions.


Kenny comes close to the point where the effect might flip its reaction...when he starts talking about rates in the 600-800#/in range. And FWIW, I suspect it actually would if including the damping force(s), and any time those two exceed the tire spring rate.
It's certainly possible to compensate or even "overcompensate" for a lowered RC with suspension stiffening. In general terms (using made-up numbers for illustration purposes), increasing the roll moment by, say, 50%, but doubling the suspension' roll stiffness ought to do the trick . . .

I think one of the things Kenny is doing is reducing the amount of geometric LLT, which ought to improve mechanical grip a bit by reducing the size of the 'instantaneous' geo-RC term and effectively slowing the rates of change in tire loading down. It should also make for less 'progression' from an oversteerish RC-dominated initial condition to the final understeerish (or at least less-oversteerish) condition as the understeerish suspension-stiffness term "catches up" and ultimately dominates. IOW, the car should feel like its behavior stays closer to constant during the transient portion. I'm looking at the car in an overall sense here, not just one end in isolation, and it's a time history kind of problem.

Kenny is probably intentionally doing something with roll steer, but the available pictures I've seen don't really show what I'd like to see. Perhaps by intention.


There's a slightly different way of looking at this that does not use the concept of geometric roll centers (it uses some sort of "anti-roll" geometric approach). I haven't been able to figure out how to model that one, but it might be able to tell us things we can't get out of the conventional TLLTD model.


Norm
 
Last edited:

Grant 302

basic and well known psychic

TMO Supporting Vendors

Top