In the past 18 months there have been numerous discussions on the
relative merits of various sealing compounds, most of which were
concerned with pouch subtleness/stiffness and long-term life.  Although
never mentioned, it was assumed that all of the compounds discussed were
more or less equal in sealing ability.

Sparked by these discussions, I thought it might be interesting to
actually seal a number of new pouches with different sealants, then
measure the subsequent leaks over time.

A summary of my preliminary findings are listed below; all measurements
are in cubic inches per minute.
•
   Days
   after  Neatsfoot Silicone  Mink  Rubber
 Sealing     Oil      Oil     Oil  Solution

    0         52       57      59     53    (unsealed)

    1         16        9      11      8
    2         20       10      14     13
    3         21       10      15     17
    4         22       10      16     21
    5         23       11      17     24
   10         24       11      19     32
   15         25       11      20     42
   20         26       12      30     44
   25         28       15      40     45
   30         29       15      41     46
•
All measurements were taken using a pouch vacuum of 6 inches water gage.
For comparative purposes a #71 bleed cup has a leakage rate of 70 cubic
inches per minute under the same conditions.

I don't wish to draw too many conclusions yet, (30 days out of 20 years!),
but the following does seem apparent:

- of the materials tested, a silicone-based compound appears to seal
best;

- a rubber solution (at least with only a single application) appears to
be almost pointless;

- whichever sealant is used, it seems necessary to let the sealed pouches
"age" for several weeks before any bleed optimization is carried out.

Although only one pouch per sealant type was measured, I have no reason
to believe that the results are abnormal.  If these results generate any
interest, further work could be done, e.g. application techniques,
multiple applications, proprietary brands.

While on this subject, there is a related topic which I'd like to raise:
Should pouches be sealed?  A good theoretical argument can be made for
not sealing pouches.

In a pouch/bleed situation, it is usually assumed that when a pouch and a
bleed pass an equivalent amount of air at some stack vacuum then these
two components can be considered to be identical at any other stack
vacuum.  This is far the case.  The leakage flow in a pouch is laminar in
nature -- the flow varies directly with the pressure drop.  The flow
through an orifice, by comparison, is turbulent in nature -- the flow
varies by the square root of the pressure drop.

An example may help to get my point across.  Compare Case A, where you
have a "leaky" pouch and no bleed, with Case B, where you have a leak-
tight pouch and a large bleed.  At 6" stack vacuum the flows are set to
be identical.  Look what happens to the flows as the vacuum level
increases (all flows in cubic inches per minute).
•
  Stack Vacuum     Case A          Case B
  inches water     Leaky pouch/    Leak-tight pouch/
                   No bleed        Large Bleed

          6             50             50
         12            100             70
         24            200            100
         30            250            112
•
Thus the simplistic approach of adjusting for a leaky or a leak tight
pouch, by adjusting the bleed size, can lead to all sort of complica-
tions.  From a theoretical point of view at least, having no bleed but
rather a very leaky pouch might have advantages in forte repetition.
I think I remember reading somewhere that one piano manufacturer actually
did this.

I hope the above is of interest.

Phil Dayson

 [ The Stephenson paper mentions capillary bleeds of about 1.0 mm
 [ diameter.  Has anyone tried this method? -- Robbie