» Posted by on Apr 28, 2012 in Current projects and Tech Notes, RB6 Formula Ford | Comments Off on FARBalls


Implementing the innovations that I have developed over the years into the front anti-roll bar on my DB-6 turned it into a bizarre looking science project. This is the story of why it looks like that and why it does the things that it does.

An anti-roll bar adds roll stiffness to the end of the car that it is attached to. The reason we care so much about bars is that changing the relative roll stiffnesses of the front and rear suspensions is the most effective way to change the cornering balance of the car. The ideal cornering balance is neutral, with the slip angles of the front and rear tires the same while cornering. A neutral balance is faster, easier to control, and easier on the tires than understeer or oversteer.

Anti-Roll Bar Basics

Here is an illustration of a typical race car anti-roll bar. It consists of 3 springs in series: two blade-like lever arms and a torsion bar. The end of each blade is usually connected to one side of the suspension through a link with a spherical bearing on each end. When the suspension rolls, the blades are loaded in bending and the torsion bar is loaded in torsion. The spring rates of the three springs determine the stiffness of the assembly.

Racing anti-roll bars are usually made so that the blades can be rotated easily, so that the stiffness of the assembly can be changed quickly. This is even easier when the blade angles are connected through a push-pull cable to a lever in the cockpit, so that the driver can change the blade angles while underway. Standard DB-6es have a cockpit adjustable rear bar. I made my front bar also cockpit adjustable so that I will hopefully have enough adjustability to get the car balanced and keep it that way as the tires change. The illustration above shows the blade angles at the full stiff position. Rotating the blades 90° would put them in the full soft position.

Here are illustrations of one blade by itself, showing the two extremes of blade angles and blade stiffness. Ross Stringham optimized the blade geometry for me through the course of several FEA iterations. Thanks again Ross! Edge Engineeering made the blades for me, and did a superb job as usual.

Blade at Full Soft Angle

Rotating the blade so that the link loads it in the thin direction makes the blade deflect a lot for a given load, so the anti-roll bar assembly stiffness is low.

Blade at Full Stiff Angle

Likewise, rotating the blade so that the link loads it in the thick direction makes the blade, and therefore the anti-roll bar assembly, much stiffer. Intermediate angles produce intermediate stiffnesses, but not in a way that is as useful as we would like. The blade acts somewhat like an on-off switch as the blade angle is varied. If the cockpit adjuster has 5 positions, the first 3 give us almost the same stiffness, then the 4th and 5th make the bar much stiffer, but there isn’t much stiffness difference between those two positions. Here is a graph of stiffness versus blade angle for the 4 possible combinations of bar and blade stiffness.

Anti-Roll Bar Stiffness Graph

One way to make a bar act less like an on-off switch, and provide something closer to uniform stiffness increments between adjustment angles, is to tie the two ends of the blades together with another link. That link makes the shape of the curve in the graph above closer to a straight line than a sine wave. David Bruns did that on most of the DB-series cars. The tip link forces each blade tip to deflect about an arc, which reduces the blade deflection in its soft direction. The blade tip link is shown as a green line in this illustration, and the tip deflection arcs are shown in black.

Link Arc with Conventional Adjuster Linkage

However, unless the blades are full soft or full stiff, there is a subtle flaw in that plan: The tip link makes the bar assembly stiffness asymmetric. It’s stiffer when turning one direction than the other. That makes the cornering balance of the car different when turning left compared to turning right. The stiffness asymmetry is because the bar assembly is asymmetric.

Where did that asymmetry come from? It’s because the usual blade angle adjustment linkage rotates the blades in the same direction, as shown above. When cornering one direction, the blades are deflected toward their stiff angle, and when cornering the other direction, the blades are deflected toward their soft angle. I realized that and pointed it out to Bruns after it was too late to revise the bar design on the Swift 007.i Champ car. On that car, the link between the blade tips was a robust I-beam with a bearing at its center that was the mount point for the 3rd spring.

OK, what can we do about that? It’s easy to design a blade angle adjustment linkage that rotates the blades in opposite directions. That makes the anti-roll bar assembly symmetric. Problem solved, right?

Link Arc with Symmetric Adjuster Linkage

Wrong. With that configuration, the link between the blades wouldn’t do anything to reduce blade deflection toward its soft direction. The link would just move laterally with the blade deflection, and it would move a long way. On the 007.i, there wasn’t room to add anything that would restrain the blade tip link from moving laterally. So, the engineers at Newman Haas used that bar stiffness asymmetry to their advantage, with track-specific blade orientations that produced a cornering balance that was closer to neutral through the critical turn in each direction. Clever!

However, I don’t intend to be that clever. On my DB-6, I added a triangular support that prevents the blade tip link from moving laterally. Here is an illustration of the complete assembly, except that the fasteners are not shown. The blade tip link is shown in dark red.

FARBalls Assembly

Like the 007.i, my blade tip link is a robust part because it presses on a polyurethane bump stop mounted in the middle of the frame. That gives me a front bump stop that distributes the excess tire load due to bump stop contact equally between the two front tires. So, it won’t have as much influence on cornering balance as conventional bump stops mounted on the damper shafts. This is yet another salvo in my war against handling interactions.

An additional feature of my blade tip link is a set of 3 holes on each side for the drop link fasteners. Mounting the links in the outboard pair of holes makes the bar assembly a bit softer than standard. Moving them to the middle set adds 25% to the stiffness of the assembly. Moving them to the inboard pair of holes adds another 25% to the stiffness of the assembly. So, I don’t need a big inventory of bars and blades to cover a large range of bar assembly stiffness, and I can change the bar stiffness easily between track sessions.

I would have done all of this to the rear bar as well, but packaging constraints confounded all of my attempts. So, the only change I made to the rear bar is the optimized blades.