Project Mustang 5.0 Part 3: Optimizing Weight Transfer and Roll Steer

By Vince Illi

In the last installment, we tightened down Project Mustang 5.0's chassis with Whiteline sway bars and strut bracing, dramatically increasing its tarmac-holding capability and making the chassis much less prone to being "upset" in quick transitions such as slaloms.

Read about Part II here!

Now, we're going to turn to the rear suspension.  All suspension configurations are compromises.  The engineers who design them have various objectives and must make compromises within a Trade Space to satisfy as many of those objectives as possible while working within constraints such as space, comfort, performance, and budget. 

The S197 Mustang has a rear suspension configuration known as a Three-Link with Panhard Bar.  A diagram of this suspension is illustrated below.  Three-link suspensions are one of the most widely used geometries for solid axles with coil springs.

Dang it, Jim!  I'm an Electrical Engineer, not a draftsman!

The three-link suspension uses three Trailing Arms (also known as "control arms").  Two of these trailing arms are attached to the bottom of the axle directly adjacent to the wheels.  We'll call these the Lower Control Arms, or LCAs.  The third link is attached to the center of the axle, directly on top of the differential.  This is the Upper Control Arm, or UCA.  The purpose of the trailing arms is to locate the axle longitudinally (or front-to-back) on the chassis.  As the vehicle traverses bumps and goes around corners (causing weight transfer), the axle pivots on the trailing arms, but it does not move longitudinally relative to the chassis except in the arc defined by the trailing arms.

Although it is called a "three-link," there is a fourth component of this system: the Panhard Rod.  As I said previously, the trailing arms locate the axle longitudinally, but they don't do a very good job of locating it laterally (or left-to-right).  The panhard rod can be thought of as another trailing arm, but one attached in an axis perpendicular to the others.  On the Mustang, the panhard rod attaches to the axle directly adjacent to the driver-side wheel and to the chassis near the passenger-side wheel.  The panhard rod prevents lateral movement as the car traverses bumps and goes around corners.

I'd like to think my drafting skills are improving, but we all know they're not. Note that these diagrams are not to scale.

Looking at the axle from the side and ignoring the panhard rod for a moment, we see that the rear axle rotates around these control arms as the suspension compresses and decompresses.  The axle does not move perfectly up and down.  If we draw an imaginary line from the UCA and LCAs, the point of intersection is the Instant Center.  This is the point about which the axle rotates as the suspension moves.  The instant center changes location depending upon where the suspension is in its travel; it is not constant.

instant center diagram
The Instant Center is defined (from the side) as the intersection of imaginary lines drawn through the upper and lower control arms. (Not to scale.)

You can see that the axle travels in an arc as the suspension moves.  This arc is defined by both trailing arms, but primarily by the longer, lower ones.  Because of this, the angle of the LCAs relative to the chassis dramatically affects the handling characteristics of the vehicle.  This angle affects a vehicle's cornering, braking, and acceleration.

The lower control arm angle affects straight-line performance first and foremost.  This is why drag racers often swap out the stock LCAs as one of their first modifications for making their car "stick" during a hard launch.  To understand why, let's think about what happens when the Mustang accelerates hard from a stop (or coming out of a corner). 

Looking at the diagram above, we see the axle, wheel, and control arms from the driver's side of the car.  When accelerating, the wheel is turning counter-clockwise.  Due to Newton's Third Law, the axle tube is thus trying to turn in the opposite direction, or clockwise. 

The OEM attachment points for the control arms are via bolts and rubber bushings.  Those rubber bushings, while great for a comfortable ride, deform when the forces of hard acceleration are applied to them.  This creates slop, or "binding," which leads to wheel-hop and less-than-ideal acceleration.  Furthermore, the stock control arms are simply made of stamped steel, which is not the most rigid design.  Therefore, the first step to upgrading the control arms is to replace the bushings with stiffer ones (or rod ends if you're hard-core) and replace the arms themselves with a more rigid design.

Ignoring the bushings for a moment, you can see that the axle tube "pushes" the car forward via those lower control arms.  This is where the angle of those LCAs comes into play.  From the factory, the lower control arms are parallel to the ground.  During hard acceleration, that control arm also "pushes" the car forward parallel to the ground.  There is also a tendency for the control arm to rotate upwards around its front mounting point, moving the axle upwards relative to the chassis and contributing to "squatting," or rearward weight transfer.

But what happens if that lower control arm angle changes?

Page 1 of 5 Next Page
Bookmark and Share
Thursday, November 29, 2012 6:07 AM
Seems like so far you've been able to stay in STX or ESP class for autocross, but now that you changed the LCAs won't that put you in CP or something? If you're building the car to autocross it doesn't make sense to me to not optimize it for a class...
Dusty Duster
Dusty Dusterlink
Thursday, November 29, 2012 6:09 AM
It does if the clubs you run with only use PAX indexes and don't have classes.
Thursday, November 29, 2012 12:08 PM
It's really nice to read something about solid axles. I recently got an FB RX-7 and vaulted myself into the world of 'old technology' and hotrods. It's nice to know a little more about my rear suspension. Now I have a bunch of studying to do on what the geometry on the FB looks like. Also, MotoIQ should do a some nostalgic car tuning. I would love to see and FB, 510 or 240Z on the site.
Dusty Duster
Dusty Dusterlink
Thursday, November 29, 2012 12:18 PM
The nice thing about solid axles is that they don't have many "degrees of freedom," so it's very easy to picture in your head what they do and how they react to various inputs.

That's a blessing for me, since I'm a non-mechanically-inclined electrical engineer!
Dan DeRosia
Dan DeRosialink
Thursday, November 29, 2012 3:23 PM
You know, I'm actually working on building an FB for roadrace duties. I should make a build log on the forum or something.

Suffice to say that there's a bunch of things wrong with the FB geometry in the rear, but it's all fixable. Unfortunately one of the best sources for parts and advice on how to do so got out of the business this year, but there's others.
Dusty Duster
Dusty Dusterlink
Thursday, November 29, 2012 4:55 PM
Does the FB have a 3 link or 4 link setup?
Thursday, November 29, 2012 5:27 PM
The FB has, I believe, a 4-link with (interestingly) an asymmetrical watts link.
Dan DeRosia
Dan DeRosialink
Thursday, November 29, 2012 5:37 PM
Stock, it has a 4 link with the upper links a lot shorter than the lower ones, and angled inwards, plus a watts link bolted to the diff. The instant center is really short, and lowered the anti-squat goes up to ridiculous extents... but worst of all, because the way the links are setup, the only way the car can roll at all is by bushing compliance. Mazda threw rear sway bars at it to try to mask it, but with sticky rubber on it, the roll stiffness goes up the more body roll there is and it eventually snap-oversteers.

The race solution pioneered by Jim Susko converts it to 3-link plus panhard; because it's bolt in without sheetmetal surgery, it's legal in the various SCCA roadrace classes, and more or less everyone running FBs in roadrace uses either his actual stuff, or other people's copies of it.

Front end is struts - irritatingly though, the spindle is integral with the strut tube. I may actually end up boring the stock tubes out and fabricating new ones to take Koni double inserts.
Dusty Duster
Dusty Dusterlink
Friday, November 30, 2012 4:58 AM
4-link with an asymetrical Watts link? That is kind of bizzare. I've got a buddy who autocrosses an FB; I need to look under his car and see what he runs.
Dan DeRosia
Dan DeRosialink
Friday, November 30, 2012 5:26 AM
Well, the Watts link mount is on the front of the diff housing for packaging reasons... so it has to be asymmetric for obvious reasons. The lever arms on each side of the pivot are different lengths so it works, more or less. At least it's not the worst thing about the rear suspension.
Friday, November 30, 2012 7:13 AM
Wait a minute, "Less squat means less rearward weight transfer, less wheel-hop, and better acceleration."? Am I misunderstanding this? Squat =/= weight transfer. It's caused by it. The only rearward weight transfer that's reduced is that due to longitudinal CG movement, which is negligible. The point of anti-squat is not to decrease rearward weight tranfer (if the rear wheels are driven, you DEFINITELY don't want that), nor is it to combat wheel hop. It is to SPEED UP weight transfer.

When a car squats, there is a delay between when you start accelerating and when the weight transfers to the rear. The traction force at the rear wheels puts a moment about the car's CG, then it takes a finite amount of time for the rear springs to compress and fully react that moment. Meanwhile your tires are spinning because they have not so much weight on them.

But with anti-squat, you change the line of action of the traction force so it isn't parallel to the ground. It now has an upward component putting an opposite (and equal if 100% A/S) torque about the CG. Now you have a (mostly) rigid link opposing the squat instead of springs that take time to compress. This makes it so you can hit the tires harder and put power down faster.
Friday, November 30, 2012 4:01 PM
Bowenaero: I was actually thinking about that too as it didn't quite make sense in my head. And you phrased what I was thinking exactly
Dusty Duster
Dusty Dusterlink
Monday, December 03, 2012 4:07 AM
I think you're reading into things overmuch. I never said that squat caused weight transfer; I was merely listing the effects that the anti-squat brackets had.

If the anti-squat effect only affected the center of gravity, then lowering the car (which lowers the Cg) would increase anti-squat, since the rear wheels would put a moment of force around the Cg in a counter-clockwise direction (looking at the car from the driver's side) that would cause the back end to raise up, which was clearly not the case. Remember that this car has a solid rear axle, which reacts somewhat differently than an independent suspension.

And you are simply incorrect in saying that the anti-squat brackets have no effect on combating wheel hop.
Monday, December 03, 2012 7:54 AM
You wrote in the article, "Less squat means less rearward weight transfer," implying that more squat means more rearward weight transfer. And if instead you meant to say that the brackets decreased weight transfer, that also is incorrect. Suspension geometry doesn't change weight transfer. You will have the same amount of weight transfer front-to-rear unless you change CG height or wheelbase (given a certain acceleration).

I don't really understand your middle paragraph, but I agree that lowering CG doesn't increase anti-squat.

I said, "The point of anti-squat is not to...combat wheel hop," in response to where you wrote, "Less squat means...less wheel hop." Whether or not A/S can combat wheel hop, it's not the main point of A/S. The point is to make weight transfer faster/more efficient.

The biggest reason you saw a decrease in wheel hop was that your suspension geometry (once the car was lowered) was providing "squat" geometry (links pointed down toward front of car). Even just bringing them back up parallel to the ground would have reduced the wheel hop significantly. But again, I'm not arguing that anti-squat can help further.
Tuesday, December 04, 2012 3:51 PM
Kenku: I would love to read about a FB project on MotoIQ. I need some ideas.
Tuesday, December 04, 2012 3:58 PM
Also, has anyone on here ever worked on a lotus or Chapman link? Saw it on the RX7 forums and it seems interesting/light but possibly very complex.
Dan DeRosia
Dan DeRosialink
Wednesday, December 05, 2012 9:37 AM
Honestly, in the real world, unless you're trying to put the roll center somewhere weird that requires a Mumford link, I'd just use a Panhard bar. There's theoretical advantages to other stuff sure, like the axle not moving sideways a little under axle movement... but it all comes at the expense of complexity, weight, and often lack of adjustability.
Wednesday, January 16, 2013 8:09 PM
As Bowenaero said, changing the anti-squat does not affect weight transfer but it does affect the amount of load transferred to the rear tires. Transfer enough load off of the front tires and onto the rear, and you'll lift the front tires off of the ground (drag racers).

"Adjustable panhard rods" aren't necessary since there is minimal axle shift in S197 cars even at 1.5" lowered ride heights (which puts the panhard rod ~parallel to the ground). BUT an aftermarket panhard rod is important since the stock ones are weak and can bend from track use.
Dusty Duster
Dusty Dusterlink
Tuesday, February 12, 2013 2:31 AM
The axle can shift as much as an inch or more when lowering the car. That is hardly what I call "minimal." And if you don't believe me, call up anyone who else races Mustangs and ask them if the axle needs to be centered or not.
Michael Lin
Michael Linlink
Monday, April 20, 2015 9:17 PM
Great article! I learned a lot, but I do have a question...

You mentioned the lower control arms are parallel to the ground from the factory. If you set the new control arm to be the same length as the OEM piece and then add anti-squat (thereby tilting the arm down at an angle), won't that actually reduce the wheelbase of the car at rest?

If that's true, then is it be better to adjust the control arm length to match the factory wheelbase, or leave the control arm at OEM length? Thanks!
Post Comment Login or register to post a comment.

MotoIQ Proudly Presents Our Partners:

© 2018 MotoIQ.com