It has been proposed that a spring will twist under compression, and that full compression of the long springs in the forks will put enough torque on the forks to take you off of your line or at least twist the handlebars. Some have gone so far as to install needle thrust bearings in their forks to reduce these effects. The tests presented here are intended to give a quick look at the amount of twist produced by compressing a fork spring.
Click on any of the following images for larger versions.
I used a stock KLR fork spring for testing. The first step was to add a reference line to the spring that would visually indicate the presence of any twist. I made a simple striping rig by securing two lengths of 3/4" aluminum angle to a base, with the spring in between. I drove a nail at either end, along the centerline. To make the stripe, I used a used a device like a chalk line, except I used enamel paint instead of chalk. I tied a small loop in one end of a piece of nylon line, and dipped the line in white enamel. I hooked the loop over one nail, pulled the string taut and wrapped it around the other nail. Snapping it like a chalk line left a straight line (of dots) along the coils of the spring. (Apologies for the lack of depth of field in the photos, but I didn't have time to try different photo setups.)
Reference line painting jig.
Resulting reference line on free spring.
The spring will be compressed between two nuts on a length of 5/8" threaded rod. This particular piece of rod is left-hand threaded, which becomes important later. Normally, the spring is contained in the fork tube, which keeps it from buckling laterally when it is compressed. To prevent buckling along the threaded rod, ten teflon spacers were used to constrain the spring internally. The spacers are 1" long, 0.860" OD x 0.625" ID. Total spacer length is then 10", which is roughly the free length of the spring minus the suspension travel. The holes and corners were chamfered to prevent catching on the spring coils or threads. The spring is roughly 0.875" ID, but the spacers were cut slightly smaller to facilitate installation in the spring. Also, aluminum spacers center the ends on the rod and provide a bearing surface for the nuts.
Of course, we do not want friction from the nuts to twist the spring while it is being compressed. Needle thrust bearings were installed on either end with bearing washers and a flat washer under the nut, as shown in the photos.
The spring was compressed, and the reference line photographed at roughly one-half and full compression. The pictures are shown below.
From the photos, we can clearly see that the spring does twist when it is compressed. The twist was not induced by turning the nuts. In fact, since the threads are left-handed, the spring is twisting in the opposite direction of that requred to tighten the nut. So, the spring is really twisting due to compression. Like threads, springs can be wound in either a left- or right-handed direction. The direction of twist will depend on the direction of winding. It is fortunate that the rod happened to have left-hand threads, else we could not have ruled out bearing friction as the cause of the twist. How much did the spring twist? Unfortunately, I forgot to measure the amount of deflection before relaxing the spring, so the actual number will have to wait for another day.
So, should everyone run out and buy needle bearings for their forks? Not just yet... It turns out that it took barely any effort to untwist the spring while it was under compression. In fact, the "self-twisting torque" wasn't even enough to overcome friction of the needle bearings to re-twist itself after I had straightened it out. So, I do not believe that spring twist will put any significant torque loads on the forks. If you really do want needle bearings, don't blow $20 on them; you can get them from McMaster-Carr for around $7.25 plus shipping.