Refer to folder in files, "proa rudders".

These are cambered symmetrical sections, linked rudders.
They kick back and are balanced.
Limit the sweep (rake) with weak string or releasing cleats.

At the end of the post is a detailed description in case the
drawing is not clear.

Regarding the section, I think that one of Tom Speer's
sections is a good bet. It should probably not have more than
1% camber, otherwise it will be outside its groove at zero or
slightly negative lift.
It could have a good thickness, 12-15%, for stall resistance
and high lift, or it could be thinner, 8-10%, for ventilation
resistance. So, I am guessing P30112 is good.
It needs to have a good size. It should have a high 2D L/D at
low lift, which means it can be made big and burly with little
profile drag.

I think these sections might be more demanding on accuracy and
finish than the blunt NACA 0012 for instance. But on the other
hand they should not ventilate as easily if kept inside the
groove. The sweep that is necessary for the balance, should
also delay ventilation.

Description of the drawings.
The top attachment is like a windsurf mastfoot joint. They are
made of rubber, with a wire or rope through them.
Attached on one side of the wire is a disc with a bolt on it
that screws into a nut embedded in the top hull-mounted brace.
On the other side of the wire is a disc with a bolt that runs
through the rudder, and then has a groove in it. The groove is
for locking, by the blue thing. The blue thing is held by bungy.

And fair everything much better than the drawing!

Questions and comments welcome.