Since then, I've found that this is the most common place for a Cannondale to crack; on mentioning it to biker friends, they've all asked, unprompted, if it was a downtube crack.
I tore the bike apart, to take the frame in for warranty, and investigated the remaining components pretty carefully, prior to installation on a new frame.
On cleaning off the headset, which is a heavy Sakai model, I found that the fork crown race had beaten the seat of the fork so heavily that it fell off in my hand, rather than requiring pressing off, a halfway expected side-effect of the aluminum fork seat in this intensely high-stress area. The upper, threaded locknut weighed a ton, so I machined one out of magnesium, with a locking nut on it (a la the Gorilla/Mavic headsets) and it was pretty heavily corroded on the threads, where they contacted the replacement Cannondale steel steerer tube. My experience has been that magnesium in intimate contact with steel is a bad idea, especially in environments like this, where there is contact with salt water (sweat) to provide a good electrolyte/chloride concentration.
I was using a Control Tech stem, 1 1/4"/10deg/150 mm, and it exhibited extensive corrosion on the back side of the quill, where it was pressed against the steel steerer.
This is a common occurrence in high-strength, structural aluminums, like the 2000,6000,7000-series aluminums -- under conditions of stress and exposure to corrosives, they tend to fare badly. The stress actually seems to open small channels between adjacent crystals in the metal, and pump the liquid into these channels by peristalsis. The resulting corrosion takes the form of extensive pitting that reaches deep into the aluminum, and is usually treated by careful sanding and buffing to remove the stress risers formed. In doing this, I cleaned off the whole stem and found that it, too, was cracked. That made me feel better about the frame crack -- better that than a stem failure. The stem was cracked where the quill meets the riser, on the BOTTOM side, for a distance of about 14 mm. This point is typcially in compression for 90% of the time, and I'm not sure why it cracked, but my extensive trials riding might have had something to do with it.
My previous tactic was to do the following preflight: check air pressure in the tires, check front, then rear brake to make sure they locked and could hold full brake pressure without slipping, lunge hard once on the handlebars to check for play or cracks, lock the front brake and rock the bike back and forth to check play in the headset or looseness in the stem.
Henceforth, I'm planning on checking for cracks in the following places: just behind the headset, at the top of the fork crown on the back side, and at the seatstay/top tube joining (we had a frame failure there once, a while back) and check the stem, (top and bottom), the handlebars just at the edge of the stem/bar clamp, and possibly the pedal spindles, although I suspect they'd go from no crack to fully sheared in a very short time.
I've seen several stem failures. This is the first non-full failure I've found, and I'm grateful. A friend's Kalloy stem snapped across the quill, just where it entered the frame, and only held together because of the wedge tightening bolt. A Klein bar/stem combo sheared in a race, actually breaking diagonally from the left side of the stem/bar clamp back to the right edge of the wedge tightening bolt hole. A nameless (Suntour?) stem from the '60's broke at the bar clamp bolt hole, crosswise, so the clamp no longer held.
I've broken one seatpost, a Fast Feather which snapped across the seat fixing bolt and the swaged cup that allows for angle adjustment. It was a poor design, as it was serrated for indexing of the adjustment, and the serrations were fairly sharp-edged. Every seatpost failure I've seen has been in roughly the same place; I've seen a number of seatposts BEND at the point where they enter the frame, but I've never seen a shaft break.
In each case I have encountered, the crack was CLEARLY VISIBLE because the paint was raised, flaked, and discolored on either side of the crack. I would not have noticed the frame, stem, or seatpost cracks without the paint indications.
My conclusions are: painted components are Good Things. The paint serves to show the crack as it progresses, and also acts to electrically insulate materials. If my headset or stem had been painted where they contacted the frame, they would have less corrosion. Aircraft routinely plate everything with sprayed pure aluminum to prevent corrosion, and follow that with a coating of zinc chromate, but surface oxidation or anodization would also work quite well, I think. If bare metal is present, consider covering it in lacquer or spray fixative of some type.
This page was created on 1/20/97, last modified 3/5/97