Curious about Hide Glue, I did a search with AltaVista and found an 
informative thread on a Harp-makers list. I've snipped the names to 
protect their Constitutional Rights ;-) and done a bit of editing that 
was instrument specific. I think some of the insights regarding various 
glues here may be helpful to some who are working on their Galoot's Glue 
Master's Thesis.

Enjoy or delete...your choice.

Darren C. Addy
Kearney, NE

*********************************************
I think of hide glue as "liquid zipper" - something that makes a 
reversable
bond that is certainly not permanent. I know that it is ideal for
instruments of the violin family, for they have to be disassembled to 
effect
most repairs. They also are so designed that their major joints are 
loaded
either in shear or in compression. They do not have the stress that a 
harp
has, and do not have the extreme peel load that harps have in the
soundboard/soundbox joint.

The harp is a different critter. I have a fine example of a
turn-of-the-century L&H harp that was put together with hide glue and 
pegs
in lieu of screws. The board pulled off very early in the instrument's
life, so the harp was shut away in its case without strings for many, 
many
years. Most of the other instruments that have made their way into my 
shop
(all harps, for I don't do violins) have had their boards held down by 
only
the screws, for the glue had failed over at least part of its length. I 
can
only remember two harps with completely intact glue joints. Both of them
were under 15 years old, and both had been assembled with PVA. Every 
harp I
have worked on over 15 years old has had to have some repair to glue 
lines,
regardless of whether they were made with hide glue or PVA. 

Which leads to the question - is there really a permanent glue? I really
wonder. Hide glue, and its contemporary material, caseine, are both 
subject
to microbial induced deterioration, and both are hygroscopic. I know 
that
some of the problems can be reduced by the judicious addition of 
something
like formaldehyde, but they still fail in time. Neither of them will 
result
in a "structural" joint in hardwoods ( a joint where the glue line is at
least as strong as the wood). 

Resourcinol and Plastic Resin glues are "structural", but both require
extremely good wood/wood fit and high clamping pressures, and both
delaminate under sustained peel loads. Urea/Formaldehyde glue works well 
for
some applications, but is glass hard and does not handle shock well. It
does seem to last, however, if the two or three flyable Mosquito Bombers 
can
be used as examples. They were built with Aerolite, a urea/formaldehyde
material, back in the early 1940's. This is one of the glues I use 
routinely.

PVA (whiteglue, and its cousin, yellow glue, which is a cross-linked 
PVA)
are pretty good if one can stand the creep and knows to use alcohol to
reverse the bond. They are easy to use, and are non-toxic. Very handy, 
but
the one harp I built using it has already started to show glue line 
stress
and it is only seven years old. Still, for decorative moldings and low
stress joints it has its place, unless you are using shellac over the 
joint.

The newer formulas, epoxy and polyurethane, have a lot of promise, but 
have
not stood the test of time. Still, one has to make a choice. I use some 
of
both, but tend to prefer epoxy for most applications. The epoxies I use 
are
approved adhesives for aircraft primary structures. Epoxy has been 
adopted
as the glue of choice by a lot of boatbuilders. The chemistry of the 
stuff
is such that once its cure is complete (about 72 hours) it is chemically
inert, and will not get brittle. Accelerated aging tests have been good
enough that aircraft using the material have not been assigned a useful
lifetime on account of the glue. It will age with exposure to 
ultraviolet
light, but glue lines seldom see the light of day, do they? It fills 
gaps
with strength, especially if the eopxy is filled with glass fibres. For
restoration work, where one is constantly faced with warped or damaged
surfaces it is a harpsaver. 

The big disadvantage to epoxy is that you can't open a cured joint 
without
destroying wood on one side of the joint. I can't assure one that the 
stuff
won't break down in fifty to a hundred years, but I do believe it will 
last
at least as long as anything else I have seen used on harps. It still 
won't
hold down a pedal harp soundboard reliably unless it is backed up by 
screws,
though, for the surface wood fibres will peel away in time.

Incidentally, I realize that I am addressing this issue from a different
perspective than someone who works with folk instruments or very old,
lightly strung pedal harps. The stress that Al or Glenn deal with is 
much
lower than thaat on a modern pedal harp.

*************************************
 

Thanks for your input. I tend to agree with what you are saying about 
the
lack of known permenant glue. Epoxies are the most likely candidates. 
The
Gougeon Brothers have used epoxy (their own formula and brand name) for, 
I
think 30 years in yacht building and I believe that provided that they 
the
work is protected from UVs joints hold (which is the case of harp 
joints).
No one knows what will happen to man made glues over a longer period 
indeed.

I differ a bit about what you said regarding the impossibility of 
separating
epoxy joinery. Not so many people, and I was once one of them, seem to 
know
that some epoxy joinery can be undone... Epoxy too is thermo-plastic,
although it takes a bit more heat than carpenter glue or hide glue. You 
can
slide a hot knife between the edge and the soundboard and melt epoxy 
glue. I
have done it. If you meet screws, heat them up so that you could remove
them first.

Excellent tip! I had forgotten what heat does to epoxy, though I should
know. When I was designing filtration equipment for nuclear plants we 
had a
competitor who used epoxy to assemble their filters. On test, these 
filter
cartridges would fail at about 150C due to the weakening of the epoxy 
materials.

Incidentally, another boatmaker's supplier, System Three, makes resin 
with
three grades of hardener. I find it extremely useful to be able to use 
the
epoxy in my shop in the winter without having to heat everything to 25C;
their fast hardener will work with reasonable cure times at 12C. 
Wonderful
stuff!

***********************************

If anybody in the world is interested in good glue joints, its the folks 
who
fly in wooden airplanes! There was a really nice series of articles in 
the
Experimental Aircraft Association's "Experimenter" magazine a few months 
back
about glues. Of course, they were worried about some things that we are 
not,
like leaving the 'plane parked outside in the sun all the time, but it 
still
had a lot of good information. 

A properly made joint using any kind of glue is stronger than the wood 
around
it, provided it is loaded in such a way that the glue is not taking a 
stress
it is not suited for. Phillipe's experience with the peeling soundboard 
is a
good example of that. The glue line held, it was just that there was not
enough wood to take the load! The trick is that the glue, as a bulk 
material,
is almost never as strong as the wood; what is strong is the bond 
between
glue and wood. Thus you want to minimize the amount of glue in the 
joint, and
maximize the wood to wood contact. 

One part of this is making clean joints. Carleen Hutchins mentioned a 
study
of failures in wooden airplane propellers during WW II, when lots of 
them
were laminated out of ash and mahogany. If the laminations were planed 
to
thickness the props held up well, but not so when they were sanded. 
Sanded
surfaces are much rougher than scraped or planed ones on a microscopic 
scale,
and it is also almost impossible to get out all of the dust. Gluing wood 
to
dust is not productive! Note however that a well-fitted sanded joint is
probably preferable to a poorly-fitted planed one. 

Hide glue is actually one of the strongest, both in terms of the bond 
and as
a material. Don Warnock mentioned once that one of his students left 
some to
dry out in the jar, and it peeled off a layer of glass as it shrunk. It 
is
also not subject to creep in shear, if I remember correctly. The 
problems
with hide glue are that it is not moisture resistant, there are things 
that
eat it, and it will melt when heated. BTW, I have some violin plates in 
my
(very humid) cellar that were glued with hide glue ten years ago or so, 
and
the joints are holding up well while some other stuff that was assembled 
with
Titebond is coming apart. I would say that the *humidity* resistance is
actually pretty good. 

The main problem with white glue is, as I say, the lack of creep 
resistance.
Yellow glue is much better, but still not as good as hide. White and 
yellow
glue joints can be dissolved with acetic acid. Vinegar is OK, but for 
really
tough joints that can't be heated, try the 38% Acetic that you can get 
in
photography stores (Stop Bath). I have also mixed it with Methyl 
Cellulose to
form a gel. The one drawback is that it will react enthusiastically with 
the
iron in a steel knife, and the Ferric Acetate thus formed is a very 
permanent
black! A stainless steel knife solves the problem. 

If there is a permanent glue, it is epoxy. My dad did a lot of the usual
kinds of accelerated testing on epoxies back in the sixties in 
connection
with making epoxy-potted transistors, and the only MILSPEC test it 
failed was
for heat-resistance. At the time, transistors were purely military
technology, and the company decided not to use epoxy. They make 
lightbulbs
now... 

Gougeon Bros. uses a test for creep resistance that is fairly easy to do 
in a
home shop. It was outlined in a general way in the "Experimenter" 
article I
mentioned, and I will try to get some more details. Their epoxy goes 
under
the name of "West System" and is available through some of the luthier
suppliers. They also mention that most epoxies will soften above 150 
degrees
F., but regain their full strength on cooling provided they were not 
held at
the high temperature for a long time. Higher heats will destroy the 
bond,
with the exact temperature depending on the formulation used. 
Quicker-cure
epoxies are more effected by heat and moisture in general. Moral: don't 
store
your harp on top of the stove!

One issue with epoxies is shelf life. We used to use some stuff in the 
Navy
that had a two-year rated shelf life. Beyond that it would sometimes 
fail to
'kick', or else kick over so fast that it would foam up. The heat 
produced in
the reaction would make it hot enough to vaporize the water that is a
by-product of polymerization. In fact, the excess heat can be an issue
whenever you mix a big batch, and will drastically shorten the working 
time.
One way to lengthen it is to mix in a small amount of alcohol; about 25% 
by
volume makes 5-min. epoxy into fifteen. The cure time is lengthened
proportionally, and you do incur some shrinkage, but it is a useful 
trick,
and doesn't seem to weaken the final bond. 

I have not tried the polyurethane glues, but some of the aircraft people
have, so they are probably reasonably permanent. 

*******************************************

I have heard that Mark had a thing about wanting to see glue squeeze out 
of
a joint, so he clamped with plenty of pressuure. Since the epoxies I am
familiar with all require at least a .003 in. thick glueline to achieve
maximum strength, it may be that he had starved joints. 

I would also like to know what epoxy he used. The normal hardware store
stuff has a lot of fillers, and is too viscous for wood. As noted 
earlier,
the amateur built aircraft hobby is the most critical application for 
glues,
and they have been using epoxies in primary structures for years, so the
materials they use and which are FAA approved, are the ones to go for.

I know that the first part of this reply sounds like an excuse for poor
workmanship. It isn't. I get as much satisfaction as anyone who 
considers
himself a craftsman from achieving a nicely fitting joint that you can't 
get
your smallest feeler gauge into. When using epoxy, however, one must 
allow
for a glue line. Most of these glues need from .003 in to .010 in for 
best
results. 
I generally plan for the low end of this range, but plan I must. For
soundboard/soundbox joints I fit them well, using graphite paper to make
certain the contours match, apply glue, wait 15 minutes, apply more glue
with glass microfibers mixed in, and assemble the parts. The screws are 
put
into predrilled countersunk holes with just enough pressure to bring the
parts together, but I use a .003in. feeler gauge to check for glueline 
in
several places.

*************************************

Since I started (I think) this epoxy+screw story, I might as well carry
on... There are various possible explanations for Mark's pillar joinery
failure. Starving the joint would indeed be a candidate. Epoxies are not
supposed to be squeezed out the way "one part glues" are. You need a 
certain
"body" of glue for strength. The other factors are: "Two part epoxy 
resins"
formulated for woodworking can be made more or less "flexible" by 
varying
the amount of hardener you use. More hardener (maximum 1:1) increases 
the
flexibility of the product; more resin (max 2:1) increases the tensile
strength of the glue. Now a common mistake is to be a little too relax 
about
the measurement of your mixture. The result could be a glue joint which
would end up to brittle or too soft...I know, I have done it!!!! The 
last
point, I think is "shelf life". Not all manufacturers will tell you, but
their product ages and will end up not forming the right bond.

***************************************************