I'm going to try that next weekend.
I think I got a pretty good measurement for the ball joint. I took a
cutaway drawing of the ball joint (from the manual) and enlarged it on a
Xerox machine until it matched the outer dimensions of the ball joint
exactly. That gave me the ball center relative to the casting of the real
part. Then I scribed lines on a spare balljoint to establish the center
line. All the Y dimensions were measured with a laser level with a transit
base. The X dimensions I measure from part to part wherever I can reach
them. I made a giant set of calipers using a long aluminum faming square
with a moveable slide. I think I'm accurate to 1/8". I take half the
measurement to get the centerline to pivot center distance. I've done
every measurement three times and I'm within 1/8" for all three sets. I've
managed to drop plumb lines to the rear lower a arm pivots and got pretty
good figures for those. It would all be a lot easier with the engine out,
but so it goes.
The SusProg 3D software actually has the capability of accommodating
trunnions. I found that out last night while perusing the help files. Once
again I learn the hard way to RTFM. They actually show a triumph Spitfire
trunnion in the example. The Front suspension Geometry Pro software does
not accommodate this kind of upright, though you can fake it with two data
files--one with trunnion pivot as the lower ball joint to calculate camber
gain, the other with the trunnion centerline to calculate bump steer.
Your right about the sensitivity. Doesn't take much to make the bump steer
go wild.
-----Original Message-----
From: Larry Young [mailto:cartravel@pobox.com]
Sent: Monday, February 24, 2003 7:08 PM
To: Bill Babcock
Cc: fot@autox.team.net
Subject: Re: Suspension analysis challenges
Bill,
In a previous discussion, I mentioned that I don't believe you can measure
the steering ball joints and pivots accurately enough to predict bump
steer with software. 1/8 inch can make a huge difference in bump steer.
It is very difficult to measure the position of a ball joint that
accurately. After several trips to the garage and measurements from the
scaled drawings, I did exactly what you are suggesting. All the hard work
was in the garage doing the measurements. The camber curve was easy to
match. It took 15 minutes of number massaging at the computer to get a
perfect match of the measured bump curve. Only a few more minutes were
required to predict how to fix it.
What type of camber curve/camber gain do you get with you're
modifications? Although lowering helps a little, the stock suspension
doesn't have enough camber gain. This is probably why TRs like a very
stiff frontend to minimize body roll. Larry Young
Bill Babcock wrote:
> I'm going to try reverse engineering the proper measurement point,
> though I have some doubts about my ability to do so adequately. I'm
> going to graph actual bump steer (as best I can) and camber gain, and
> compare them to the calculated values. Then I'll try different values
> for the lower ball joint location and see what gets me closest. Lots
> of work, ya think?!?
|