Hi All!
I now have an update on the Sherpa-shimmy problem, and unfortunately, all is not well. The executive summary of continued testing follows:
1) The bike still shimmies with any reasonable rear load, and it shimmies with a total load at the levels referenced by Andy in the Summer 2011 brochure used when I ordered my bike last August.
2) Water placement and mass has essentially no negative effect on the shimmy; if anything, the steerer-mounted bottles damp the shimmy. This was a surprise to me.
3) Unlike a more conventional shimmy, it does not resolve with changes in rider position, letting go of the handlebars, or standing. The result is pretty scary.
4) I can *just* drop the bike into a consistent no-shimmy state with a full load *if*...
a) I fit a Tubus Logo Evo rear rack and fit the panniers on the lower rail,
*and*
b) Fit the 1.5in road slicks borrowed from my tandem, keeping the inflation at maximum pressures described by Andy in the Thorn brochures.
The rack change helped a bit on its own, but not enough to fix the shimmy outright. The larger positive change came with the change to narrower/lower-profile tires. I believe it is because they also had the effect of changing the effective trail, dropping the geometry into neutral-trail territory from the high-trail that occurs when 2.0 tires are used. Switching back to the 2.0 tires showed over-lively low-speed handling and a marked increase in wheel-flop at rest, as expected of the higher trail.
With this thought in mind, I sent for a replacement fork with 59mm offset, versus my standard 52; when fitted with 2.0 tires, the maths show it would have essentially the same trail as when the stock fork is fitted with 1.5 tires. Unfortunately, I have been unable to fit the replacement fork due to problems in shipping and the product itself. I still hold hope it might be the answer that would allow me to tour with at least the recommended load and wider/taller tires offered as an approved option.
Still, if such heroic measures are required, I am concerned it signals a problem with my Sherpa example. All testing results point to a persistent problem when weighting the rear, yet a careful examination shows no outward signs of damage or breakage to the frame. Remember, it rides fine when unladen or with a small load (25lb/11kg ) distributed between the handlebar bag, bottle cages, and rack-top pack.
Continued testing did reveal some additional surprises for me...
5) Thinking my long-steerer might be a problem, but it was not. I juggled the spacers, dropping the stem to very near the Thorn Accessory T-bar, which sits just above the headset race. I inverted the stem and I tried moving it up ad down a spacer at a time. No change. I found this surprising until I recalled that a change from hoods to drops or even sitting upright or letting go of the handlebars also made no discernible change in the onset or continuation of shimmy.
6) Thinking rider mass might be a factor, I added weight to various parts of my body using a fanny pack and the lower portion of a small backpack. No change. I am 5'11"/180cm in height and weigh 172lbs/78kg, right on average for my age cohort. Still, I have a number of friends who are far heavier...I purchased my tandem used from a family whose members all weighed in excess of 375lb/170kg, and it had no problem at all handling that combined mass plus a day-riding load. Perhaps I'm making a leap in logic, but it seems reasonable that if I am of average weight for my age/height cohort, then I would also be of average weight for a properly-sized bicycle like this one. Getting heavier didn't make a difference, and I can't get lighter!
That said, mass on the bike -- and apparently at the rear -- makes the deciding difference in handling and shimmy for my Sherpa. Unladen except for about 35lb/16kg in rear panniers, the bike's handling is really unpleasant and squirrely...moreso than when a similar load is placed only on the rear rack of my other touring bikes with conventionally-sized road bike tubing. In contrast, with only a front or front and mid-load (three 1.5l bottles in the main triangle of the frame), Sherpa's handling remains sweet as can be. The same goes when either a rack-top load is added or rear panniers are added. Things go bad when an upper and lower rear load are carried regardless of whether a front or front and mid-load are carried. This has been a consistent factor throughout all my extensive testing, and is repeatable.
By the way, I have been able to stand with a touring load when pedaling up to speed but standing does not stop the shimmy once it occurs.
Though I can't make it do so consistently, I get the feeling the bike would like to pull left against the road crown when riding with no-hands and a touring load. I have not noticed this when unladen, but I rarely ride no-hands. Just a data point to keep in mind. Yes, both wheels are properly tensioned and true, and centered in the same plane (between the dropouts). This feeling of pulling left was more pronounced with the 2.0 tires (wired-on and folding) than with the tandem's 1.5in road slicks, but remember...the narrower tires also have a lower profile and do significantly affect trail.
7) A definite surprise for me was the involvement of the Schmidt/SON dynohub bolt-on skewer. Though it has standard threads, it lacks nylon inserts to prevent loosening and because it screws to tighten, it is impractical to fix in place with LocTite. Schmidt's package insert for the hub and a reference on their website both clearly indicate a maximum tightening torque of only 6-8Nm (4.42-5.9 ft-lbs). By the way, the SON's bolt-on skewer was tightened to spec with a torque wrench each time, and before the appearance of the original shimmy. Nevertheless, I often found it had loosened after my test runs, and soon removed it and replaced it with a conventional Shimano quick-release. Thanks to its over-center action, the q/r has never loosened in subsequent runs and has not proven to be a factor in shimmy. I am not entirely sure what caused the loosening, whether it was the low recommended tightening torque, the lack of a nylon insert for the nut, the violent action of the shimmy, the weight carried, or some interaction of all the above. Given my experience with the SON bolt-on skewer, it might be a Good Idea for others using bolt-on skewers (i.e. Pitlocks) to give them the occasional check just to be sure they're still tight. Unlike with a q/r, there is no visual indication of tightness with a bolt-on skewer. I've also heard no reports of Pitlocks loosening, so my SON problem may be isolated.
8 ) The front fork blades on the original fork are about 2.5mm wider than the SON dynohub's 100mm, so any wheel inserted into the forks draws them closed when tightened. I don't know if this is a factor or not. The dropout faces appear parallel when the wheel is secure. I think both fork legs are in the same plane (one blade isn't ahead of the other), but this is hard to determine on the bicycle. I had hoped to determine this when I changed forks, but that has to wait a bit. The replacement fork is also spaced a bit wide.
- - - - - - - -
I've certainly learned a lot about shimmy and collected a great deal of data on it in my tests. Still, I do continue to have a problem with it and would dearly love to get my beloved little Sherpa to carry at least a reasonable and recommended load without shimmy. If it would also carry the same weight as my old Miyata 1000LT with conventionally-sized tubing, that would be ideal.
I have waited so long to contact Thorn about it because I wanted to make absolutely certain the shimmy was not the result of some inadvertent action on my part, or due to a mistake in load placement or mass distribution. I had such high hopes that might be the case. By bringing the shimmy issue to the Forum, I hoped members might come up with a solution to the problem. Thanks to extensive testing, I'm now beginning to conclude there is something wrong with my sample and I will need Thorn's help to resolve it.
Best,
Dan.
