The Paragon Yoke is a two piece design, with a nicely profiled exterior shape and a relieved interior for lighter weight. It forms a male plug end for the chain stays to fit a .875" x .035" tube and has enough forward material to adjust the overall length of the finished chain stay for a variety of wheel sizes. Originally designed as an elegant and stiff piece to allow 29+ 3.00" wide tires plenty of room without crowding the chain rings, I've also found that it is well suited on smaller width tires when mated with a belt drive chain wheel, which typically takes a bit more real estate.
I've had numerous folks already ask if it is possible to braze the two halves together. The pieces form a meticulous fit with zero gap or room for a brazing material to flow, so I do not feel that brazing is a suitable option to join the two halves reliably. Here is the process I used for tig welding the pieces.
I approached the steel and the Ti using the same process, just to insure uniformity in results. That being said, the first thing I did was fixture the individual pieces to the mill table to be able to create channels for gas flow throughout the yoke. Once clamped securely in place, I ran a .125" ball end mill through the three square/tapped junctions that are found on the inside of the pieces, leaving the solid end sections intact. With the inside now channeled for purging, I mated the two individual sections in the vice, and using a small bit, drilled a single hole in each of the bottom bracket and chain stay ends. The yoke was taken out of the vice, blown out with filtered air, and then surface sanded with 120 grit shop roll. I then hit all the weld areas with a fine stainless wire wheel, a grey Scotchbrite wheel, then scrubbed with soap, water, and a rinse.
Once rinsed, the pieces went into the ultrasonic for 20 minutes at 120 degrees, along with a couple pieces of filler rod. The items come out of the ultrasonic, are rinsed in hot distilled water, then dried in the bake box at 250 degrees.
While drying, I took two simple air needles (like the kind you use for filling up balls) and created a Y shaped purge hose that is attached to a Parker micro valve, for precise control.
With the pieces totally dry, they get a quick wipe with Acetone and then are clamped together for tacking.
I made a quick sketch with dimensions showing the location of the thicker sections of the yoke, and used those reference points for the first tacks, to insure that I could put enough heat into them to hold alignment once the welding really got going, reducing the probability of blowing a hole in the thinner wall of a high dollar piece. I ended up with fourteen tacks in all, working alternately back and forth around the piece to keep it all aligned.
To weld the yoke, there are two primary concerns; seeing the mating line, as the machined surface camouflages it well, and insuring that you can create good gas coverage for the longer beads. I accomplished this by taking two parallels wrapped in foil to protect their surface, clamping them to the yoke, and then placing everything in a vice at a slight upward angle. Fortunately for me, the thickness of the yoke matches the diameter of a #12 cup and lens perfectly, so that you simply have to keep the torch at a right angle in the channel and your electrode will be spot on center to the seam. A little foil at the end of the parallels creates a nice dam. As it is positioned in the vice slightly uphill, as you weld up the channel, the gas will cover your active puddle and slowly dam up and keep a pool of argon present to continue to shield the bead as it cools, allowing for long runs. You can see the basic set up below...the bead is pure silver, though it is reflecting a rare bit of Ohio sunshine from the adjacent window :)
The process is repeated for the interior sections, using the parallels as we go.
Filler rod is added to all the externally mated surfaces. The actual male plug sections that fit into the chain stays were fusion welded to the shoulder, to prevent interference with the tube id.
Of note is the inside profile of the bottom bracket end, it's a bugger to get your torch, filler rod, and still create sufficient shielding at a good torch angle. One of the few drawbacks to the design in Ti.
A steel yoke in the fixture, ready for the next step...
As this product has a fixed shoulder at one end, mating it to plate style dropouts is preferred, as it will allow a little wiggle room at the axle end. If you are a masochist or just really like a challenge use a hooded dropout and be prepared to be able to fabricate a mating tube section that is EXACT in bend angles and length, then replicate it for the other side.
Put some serious heat and filler into the tube/yoke joint, as there is sufficient material to take it.
Other interesting notes....
The 4130 yoke, once welded, is super hard. Two points to adhere to for coping...know the point you'll need to push the cutter through and weld up to that point and not beyond. The hole saw will engage one side of the yoke's profile fairly easily, the other is a bit of a tough job. To get through the more difficult arm, I ran the saw backwards and let the teeth cut a small groove in the surface, then ran the saw forwards, allowing the groove to keep the teeth from wanting to skate or flex.
I choose to use a detail belt sander and smooth off the weld bead on the Ti yokes, as I was planning on masking it to leave the inside face where the tire is located bare, with Cerakote on the rest, the customer's name negatively masked on the outside face to show the polished Ti. Just remember to keep the belt dead flat so that you do not get into the base material...not an easy task.
Welding to the bottom bracket, the inside edges of the yoke arms can be tough to sneak into with a good torch angle. A large cup and screen that allows maximum electrode extension is really nice, giving good access and gas coverage...
So that's about it...now if we could just convince Mark to make a nice piece to fit true Fat Bike tires, that would be SWEET!