Hi All,  The importance of sizing wood prior to making a bellows is
a topic that seems to have been overlooked in the MMDigest.  And,
since I could not find any postings in the MMD Archives on the topic,
I thought I'd write something.

In Webster's dictionary, 'size' is defined as follows:

  size2 (sïz) n.  Any thin, pasty or gluey substance used as a glaze
  or filler on porous materials, as on plaster, paper, or cloth. \
  vt. sized, siz-ing to apply size to; fill, stiffen, or glaze with size.

Wood, and most especially new wood, is typically quite porous.  The end
grain of new wood is extremely porous, and this presents a real problem
when making a good air-tight bellows.

Sizing wood before covering it with cloth is an important process that
is easily overlooked by those with little or no experience in making
a bellows.  But the question of how to tell if the wood is adequately
sized before applying the cloth is not easily explained in words.

Basically, you have to keep applying a relatively thin coat of hot hide
glue to the wood until it stops soaking into the grain.  And, since hot
hide glue contains water which will raise the grain, the wood must be
sanded smooth after the sizing has dried in order to get a flat surface
for gluing the cloth to the wood.

The amount of sizing that's required before a filled surface is obtained
varies depending on the porosity of the wood.  Generally speaking, you
can see how much the glue soaks into the wood when the sizing coat is
applied.  Logically, you have to keep applying coats of sizing until the
glue no longer soaks into the wood.  Generally speaking, you can see
when this happens because the glue will stay shiny, almost like glass,
after it has dried.

Realistically speaking, it's almost impossible to put too much sizing
on the edges of a bellows.  If you do, it will simply run, or drip over
the edge and make a mess... ;-)

The acid test of whether or not the wood is adequately sized will be
discovered when the cloth is initially glued to the wood and then
quickly removed.  As it is explained in the web page listed below,
the seal between the wood and the cloth must be examined to see if
there are any voids in the seal.  Should voids exist, another thin coat
of glue should be applied and then the seal checked again.  If the wood
is perfectly sized, this process will only need to be done once.  See
http://www.player-care.com/recovering_bellows.html

In closing, some people advocate using a clothes iron to reheat the
cloth and glue after the bellows has been recovered.  As it has been
explained in articles I've read, the heat (and weight) is applied
just long enough to soften the glue to the point where it begins to
squeeze out along the edge.  Fundamentally, this insures a good seal.

However, it cannot be overlooked that if glue is squeezing out on the
outside edges of the bellows, it is also squeezing out on the inside,
and you can't see how much glue is being squeezed out on the inside.
That can be a real problem in the area by the hinge, and it can cause
the bellows to become 'hinge-bound'.  On the other hand, if the wood is
properly sized and the seal is tested as explained above, you will know
that you have a good seal, and the need for using an iron can be avoided.

Note: On some exhauster bellows, like those in a rotary pump, the use
of an iron to flatten, or tighten the corners of the bellows is widely
accepted.  This is because the cloth is quite thick and the angles at
the corners are relatively steep, which precludes the cloth from lying
flat against the wood.  And even though it's highly doubtful that the
original manufacturers used an iron to flatten the cloth, they
routinely used tacks to hold the corners down to insure an air-tight
seal.  (Personally, I think they used some sort of a clamp, in
conjunction with the tacks, that applied pressure to both 'long' sides
of the bellows immediately after the cloth was glued to the wood.)

Musically Opinionated,
John A Tuttle
Player-Care.com
Brick, New Jersey, USA