In a recent posting I mentioned in passing the use of silicone rubber
in place of shellac as a sealer for metal valve plates.  Actually,
here is a complex issue that demands stand-alone treatment.

Joining metal to wood in this instance presents a unique situation,
wherein screws are used to hold the joint together, with some sort of
sealing compound to make it airtight.  The challenge is that metal
expands and contracts with changes in temperature, while with humidity
changes, wood does the same thing principally in two dimensions
only, although lengthwise being actually more stable than metal in
the normal atmospheric temperature and humidity ranges to which the
instrument is exposed.  Ideally, this requires a flexible adhesive type
of sealer that never completely hardens.

Adhesiveness is less of a requirement where full access to the moving
valve parts must be easily gained should servicing ever be needed, as
(for example) in the Standard Player action.  Standard apparently
started out using rubber cement as a sealer -- less than ideal because
it contains a volatile solvent which allows considerable shrinkage as
the compound dries.  So they soon switched to a type of shellac having
more "body" and formulated, I believe, to never completely harden (but
of course it eventually did).  Later on, a purplish substance was
introduced which can hardly be called shellac, but which may have been
more resistant to age hardening (for the immediate future anyway).

Way back in the pre-silicone days I used a product called "Sealer 800",
made for boat builders, which gave excellent results in the restoration
of pneumatic player actions.  Then, along came silicone rubber bathtub
sealer, followed by a vast array of silicone rubber products for a
variety of purposes.  Today, there is a critical need for scientific
experimentation with these, to determine which is the ideal one for
player piano restoration work.

In the old player actions in the Aeolian Pianola and Duo-Art, both
upper and lower valve plates are employed.  The upper plates are sealed
with a blotter paper gasket, allowing easy access to the valve chamber.
This allows for sealing the lower valve plates with a more permanent
type of material.

In the older foot-pumped player pianos still having dried-out shellac
as the valve plate sealer, there is nearly always a noticeable amount
air seepage at this point -- and when eliminated, easier pumping is the
result.

I would discourage the use of PVC-E glue as a valve plate sealer; it
seems to have a mildly corrosive effect after many years in contact
with plated brass.  But the real problem is getting the valves apart.
Prying will bend the metal, with a little splintering of the wood here
and there unless it is really hard wood.

Jeffrey Wood