DB-6 Suspension Modifications Part 1

» Posted by on Feb 20, 2012 in Current projects and Tech Notes, RB6 Formula Ford | 1 comment

DB-6 Suspension Modifications Part 1

Get ready for more “spilling of the beans”! Part 1 of DB-6 Suspension Modifications is about suspension geometry mods. Part 2 will be about hardware modifications.

If you don’t have a copy of Think Fast yet, now you have another reason to pick up a copy for yourself. I laid out everything that I consider to be important about race car suspension geometry in Chapter 14, Geometry Goals. If you read this along with the book, you will see that what I told the world in Think Fast is exactly the same as what I actually did with my own race car.

Step 1 was measuring the suspension geometry. I did that before disassembling the car. In many cases, I had to use my precision “eyecrometer” to estimate the pivot point dimensions to within 0.03″. It had larger Formula Continental wheels and tires on it when I got it, so I deflated the tires enough to match the loaded radii of Formula Ford tires. The suspension geometry measurement consisted of a combination of direct measurements on the assembled car and measurements of individual parts like the uprights, steering arms, etc. I used Alibre to combine the individual part measurements with the direct measurements to make the geometry as accurate as I could. Then I tweaked the suspension alignment in CAD to the usual baseline values before beginning geometry analysis: -0.5° camber all around, 3.5° caster, 0.4° front toe out and 0.4° rear toe in per wheel, with 1″ front and 2″ rear ride heights. Then all of the pivot points went into William C. Mitchell’s suspension geometry analysis software to find out what it all meant.

The deeply cannibalized state of my car resulted in a mishmash of front suspension parts. I have the “A series” front uprights, but the standard steering arms. I analyzed all of the possible combinations and decided that what I have will work just fine.

I’m not surprised in the slightest by this, but as it turns out, Bruns did a magnificent job on the suspension geometry. The roll center locations and stability, virtual swingarm lengths, lack of bump steer, motion ratio linearity, scrub, halfshaft joint angles and plunge etc. are all nearly what I would do given a clean sheet of paper. There was only one result that I had a strong motivation to change: the anti-s.

I basically declared war on handling interactions in Think Fast. In my opinion, anti geometry produces far more harm in the form of handling interactions than any benefits that it might offer in sprung platform motion control.

Here is a graph that you probably don’t see very often:

This graph shows the variation of anti-dive versus suspension travel resulting from the standard DB-6 front suspension geometry. There aren’t very many people anywhere who pay attention to things like this. It’s possible that I’m the only one.

The anti-whatever percentage usually varies with suspension travel, and it usually varies in the wrong direction, like this example. That variation is not much of a problem until the car rolls or encounters a one-wheel bump or dip while accelerating or decelerating and cornering. When those things happen, the anti percentage, and therefore the jacking force, is different on each side. That changes the distribution of vertical loads among the 4 tires, and because that distribution has an extremely strong influence on cornering balance, anti geometry makes cornering balance a stronger function of longitudinal acceleration, roll, and pavement undulations than it would otherwise be. There are far too many handling interactions inherent in race cars without exaggerating them with anti geometry. I’m therefore opposed to anti-anything suspension geometry. I’m anti anti.

If you turn the anti picture 90 degrees from side view to front view, you can see that a high roll center can be called anti-roll geometry. That is yet another source of handling interactions that we can do without.

The only anti geometry that I have seen race cars tolerate in large doses is rear anti-lift, and then only on very high downforce ground effect aero cars. In that case, rear anti-lift in excess of 100% held the rear end down farther into each braking zone, so the downforce didn’t fall away with speed as quickly as it would have without anti-lift.

My plan for modifying the suspension geometry was focused on minimizing the front anti-dive, rear anti-squat, and rear anti-lift. The inboard pivot points of the upper front wishbones shared the same width and height dimensions, so duplicating the shared dimensions on the lower wishbone would zero the anti-dive.

As an initial exploration, and because it would have been easy to do, I analyzed the suspension geometry with the aft leg of the lower front wishbone mounted below its mount clevis. That lowered the pivot point too much, producing pro-dive geometry. So, I did it the right way and created a new pivot point clevis on each side that is directly aft of the forward pivot. It took a fair amount of design work, welding, and battery shopping to make it all work out, but it did. The new suspension mount clevises were the most complex of the frame modifications. This change lowered the front roll center 1/8″, which is not a significant change. Custom lower wishbones were also part of the package. I left the original clevises in place just in case I decide to test the original suspension geometry at some point. They aren’t in the way of anything.

In order to zero the bump steer with the modified lower wishbone pivot, each end of the steering rack had to be shortened by 1.33″ and the toe link lengthened to match. I did that by swapping the big heavy clevises that were screwed into the rack bar with spherical ball pivots like the Champ cars used. It took some machining of the rack bar to fit them. That was the first of many superb machining jobs that Edge Engineering did for me.

Modifying the rear suspension geometry to get rid of anti-squat was far easier. All I had to do was to raise the aft pivot point of the lower wishbone on each side. That was easily accomplished with a simple machined bracket, a spacer, and a longer bolt on each side. That zeroed the anti-squat, but produced 3.5% pro-lift. Because that pro-lift percentage is constant with suspension travel, and it’s a small number anyway, I don’t expect it to affect the behavior of the car much. This change also raised the rear roll center 0.26″, which is probably enough to make a noticeable difference.

Next time, we will cover the hardware modifications I made to the suspension, including extensive changes to the front anti-roll bar. I seem to have become an expert in complex anti-roll bar design, and of course I have good reasons for that complexity.

1 Comment

  1. Great discussion even a country boy like myself can understand. Thanks for sharing.