I am re-posting this here from a build diary thread. Take it for what it's worth, I expect a great deal of disagreement.
...Our team did a lot of analysis on stability and control because of the Sonic Arrow 1996 roll. The most applicable data came from motorcycle stability and design studies. These studies included tests that isolated rake, trail, wheelbase, axis of rotation (flexability in torsion and bending is all axis), etc. This isn't theoretical, they actually built bikes with obscene ratios of rake and trail that we would never consider. The data surprised me at first, but explained everything I had ever felt in 5 years of motorcycle road racing.
The most important factor in steering stability is trail. It is TOTALLY independent of rake. The studies included a bike with 15 degrees of rake and trail between 2 and 20% of wheel base. Anything over 5% led to almost no wobble; trail greater than 10% of wheelbase made steering force very high. The tendency of wobble to couple into weave (tank slapper) was directly proportional to rake due to the long moment arm of the forks associated with high rake angles. Weave was related to how flexible the frame was laterally (left to right bending). This is important: NOT torsion, NOT vertical bending;
Lateral stiffness dominates weave.
It looks like your existing bike has ~10" of trail on a ~75" wheel base. This is a LOT of trail, which seems fine as long as you can still steer. It's also a LOT of rake, which only matters since the lateral flexion of the fork adds to the frame. The frame is of excellent construction, and needs more lateral stiffness to avoid weave. It's a double-down-tube, single -spine design instead of a perimeter or split-spine as most modern race bikes are. I raced superbikes during this changeover from 1980 through 85, and these reports explained every wobble and weave I ever had. Any steering instability will likely couple completely to the longitudinal bending and create weave. Steering dampers take wobble and feed it into the frame to create weave. They only work if the frame is laterally stiff.
I know you're very far along in this and have had lots of success. As we found out, the time to see if the vehicle wobbles or weaves is BEFORE an upset, not after as Craig did at 675 mph. The car was fine to 636 without a crosswind. So we have added as much lateral stiffness as we can and will test wobble and weave at every speed increment in our test program. Everyone needs to do that same. From my own experience on superbikes, we could go very fast with bad steering instability as long as we only went straight and didn't hit any bumps. Once upset, the bike could tank slap fatally.
This suggestion is only for future bikes:
Trail: 5 to 10% of wheelbase;
Rake: as little as possible;
Lateral stiffness: as much as you can get;
Longitudinal stiffness: enough to support your weight.
Good luck, and all the best.
Eric Ahlstrom
Program Manager
Fossett LSR