Originally Posted by WizardOfBoz

I called West Systems (epoxy maker) tech service. Talked to Don. I asked

1) What West Systems product is recommended for carbon fiber? Answer: They recommended their 105 series resin and their 200 series hardener and pointed out that these products are used by high-end bike makers. Don suggested that one use a scarf joint or otherwise increase surface area, but this doesn't apply to the OP's problem.

2) Can you clamp too hard, such that you squeeze out the glue? Answer: The first word out of the guys mouth was "Definitely". Don went on to say that too high a clamping pressure can produce a glue joint that's glue-starved. Andrew's note had an implied problem: naively clamping the joint too hard initially caused poor adhesion, likely due to glue starvation. Andrew had better results clamping lightly initially. This lines up with Don's next suggestion. He proposed coating both surfaces to be glued with a very thin layer of epoxy and letting this cure partially. Then recoat one surface with fresh epoxy and clamp as hard as you like. The initial precoat won't squeeze out so you get a thin but high-fidelity joint. Don mentioned that this is especially useful in wood joints where one wants a super-thin and nearly invisible seam, but it also applies to CF.

Clamping too hard causing glue starvation was documented by the woodworking and guitar-making examples above, and apparently is a well-known problem to one of the top manufacturer's. This is not theory: it's documented practice. In fact, I'd suggest that one possible reason that the OP's "washer" broke off in the first place may have been too high an initial clamping pressure producing a glue-starved joint. Andrew's approach is one-step and more efficient but perhaps requires more "touch" as to how hard to clamp. If I were fixing it, I'd try the "West Systems" approach I think.