Hi E-wan,
While I don't own a Nomad Mk2, nor do I own CSS rims or Magura brakes, this phenomena also happens with other combinations. I've seen and felt it on a number of my own bikes, especially my tandem with Pedersen Self-Energizing front cantis and homemade tubular cro-moly brake booster. The combination generates a lot of clamping force, with Kool-Stop salmon pads, even on wet rims.
The fork is juddering in reaction to the stopping torque produced by the brake pads sliding on a film of water, then grabbing and momentarily increasing the coefficient of friction. It has happened to me when:
a) The rim is no longer evenly coated with water after several turns with or without braking.
...and...
b) the pads grab on the drier, dried, or dry spots of the rim in-between the wet spots.
Ref my post to jags regarding brake squeal after he cleaned his Sherpa rims (
http://www.thorncycles.co.uk/forums/index.php?topic=3976.0 ), I think what you are experiencing is a gross example of what would manifest as brake squeal if the rims were dry. Water is the contributing factor that is changing the coefficient of friction between the pads and rim. I realize your brakes
aren't squealing dry or wet, but the effect is similar in that the pads are grabbing and letting go in a cyclical fashion. Here, instead of producing a squeal, the effect manifests as "brake grab" -- and it is severe enough to momentarily halt the wheel, which is what is causing the fork to bend backwards (if the tire contact patch is stuck to the pavement...and you still have forward momentum....something's gotta give. In this case, it is the fork doing what it is supposed to: acting as a spring and torque reaction arm, at once); it winds up and lets go, causing judder. You mentioned this occurs in "very wet" conditions, and those are the ideal conditions for this kind of fork judder. Remember, unlike the rear triangle, the fork is an unbuttressed wheel support, and again, it acts as both spring and moment arm, which makes it want to bend fore-and-aft if the friction coefficient of the pad-rim interface changes rapidly as the wheel rotates.
This has also happened in the opposite way on my tandem when it is really wet. Because of the long wheelbase and the weight of myself plus the aft weight of my stoker, the bike won't endo (flip onto the handlebars) if the front brake locks up. Instead, what will sometimes happen is the rim will go dry enough for the brakes to grab it firmly to the point of lockup, and the
front tire will go skipping along the pavement, making a dot-dash pattern on the pavement, which also causes the fork to judder back-and-forth as the tire sticks and then releases from the wet pavement. It can't skid cleanly as it would on the rear because the wheel is leading the bike, and the fork takes up the force until it can't anymore and springs forward, the forward momentum continues, and the tire breaks loose to repeat the cycle again. It takes riding in a downpour or through a puddle and then braking heavily to make it happen. When you think about it, forks live a pretty tortured life. It is a wonder more of them don't fail, but most forces are well within their limits of plastic deformation, and the blades are more resistant to the forces involved, thanks to the arrangement of section width in profile (they're generally somewhat oval fore-aft), greater wall thickness, and larger diameter at the root (crown). We wouldn't get away with much if the fork blades were round and the size of seatstays.
I don't have experience with hydraulic brakes, but if you were running v-brakes or cantis or sidepulls of some sort, I would suggest either:
1) A change of pads (you've tried two, and that didn't work, so scratch that suggestion).
...or...
2) Adding more toe-in to accommodate the grab-release cycle as you ride through wet water. Do the Maguras have a method for adjusting toe-in? If so, I would try adding a bit more toe-in and see if that helps (it should, but it is a fine balance between too little/much toe-in and wet/dry results).
That said, I have also seen and experienced this when a hard-anodized rim begins to show some sidewall wear, especially in cases where a light rim has been built with relatively heavy spokes that slightly distort the rim sidewalls, or where the manufacturer's eyelet-setter was adjusted to apply a bit too much force when clinching the eyelets. Where the sidewalls are distorted, the anodizing wears off in a dot-and-dash pattern; when it is wet, the brakes will grab and release in the same manner, causing the fork to judder as well.
By any chance, are your rims worn unevenly in a similar manner? It seems unlikely, as Thorn doesn't match light rims with heavy spokes, nor are Rigida Andra's eyeletted, but trying to cover everything I've seen that causes similar effects. By any chance, are you running Mavic or another brand of ceramic rims with eyelets?
One last shot at an answer: By any chance, are the pads tight in their holders, still properly secured by pins or clips? A loose fit here could account for the effect you're experiencing.
Hope this helps,
Dan.
