Robbie said he was still confused about the efficacy and overall
strength of hide glue versus the carpenter glues.
>[ Well, I'm still confused. Metal fasteners are routinely tested for
>[ strength and elasticity, and so forth. Surely there is a publi-
>[ cation about glues and adhesives that gives the physical properties,
>[ and describes testing methods which are irrespective of the appli-
>[ cation. I would much rather read numbers than a flat "yes" or "no".
I could oblige with some facts and figgers, but they are all mute,
because there are too many variables when using glues to have anything
in "just the facts, Ma'am," when you get through. Let me explain a
bit, and remember too, tests have been run by independent testing labs.
Whether or not their results are usable is still not really useful,
because if I bought the exact same batch of glue and tested it on the
exact same lumber, my test joints would vary in strength. I know --
I've done it.
We know how the electric motor companies rate their motors, and how air
conditioning manufacturers rate their equipment, how herbalists and
pharmaceutical companies alike claim great things. So we go to the
Consumers Reports or some independent lab for our data. Are they
right? Maybe, but is it so in the real world -- what they discovered
about your brand and year of car? Not at all.
So when you have a product that you can actually fudge on tests with
and nobody is going to call you a liar, do we suppose they won't fudge
(wouldn't think of it)? When I was in the electronics industry, even
something as cut and dried as a microcircuit chip spec sheet omitted
crucial data if it wasn't particularly helpful to the sale -- and
that's supposed to be "completely objective stuff." Here are two
similar circuits and one circuit has one set of data, and the other one
omits a few points, while adding another less needed one to make up for
it. Objective? Not really.
In the case of hide glue, the gram strength is everything. New glue of
any viscosity is strongest right after it is made. But the strength of
hide glue is directly related to its viscosity, so here's how to tell
what you have:
As I have said before, hot hide glues vary purposely in strength from
30 to 600 gm. The reason is to give the glue more latitude at the low
gram weights, and high strength at the high end. Glue at about 40-60
gm is what is used to chip glass! When it shrinks, it is stuck with so
much force it spalls out big glass chips -- or little ones, depending
on what you want. That means, it has to stick very tight to glass.
Now the higher strength glues will not even bother with glass because
their molecule is too big. But because it is, it is able to have
tremendous strength when you need it. So it depends on the
Hot glue, when covered with water (that is free of iron!) and left for,
say, 12 hours maximum, has absorbed as much water as it can hold. If
the mass is still very solid and just rolls out of the jar, that is
very strong glue with usually undesirable shop characteristics in
general. Its strength is great, but its adhesion may be very poor.
If the lump is soft but mostly cohesive, and the rest of the water has
darkened, that is about right for general shop work.
If the glue just lays there in the bottom of the jar, and as you slosh
it around it looks like sludge, that will be too weak for general
purpose shop work. But, you can mix the extremes and get somewhere in
the middle! You can add things like urea fertilizer to widen the
acceptance of the glue to more difficult materials, or add KCl or
"potash." And there are other tricks, as well. You can even turn hot
glue into bottled glue that won't set up by gelling, but you also
change its properties somewhat, and its strength is considerably
The strength of glue is due to its adhesion and it own inherent shear
strength. Since this is theoretically greater than the surface of
wood, anyway, what does it matter how much stronger hot hide glue is,
than say TiteBond? And granted, even though adhesion is half the
battle, some woods allow much less adhesion than others. And you can
lay 3 dozen bonds out on a table and wait for them all to set up. When
they are fully hard and dry, you can test these bonds and you will find
that the wood itself breaks at vastly different loads, anyway. In
every bond, some parts of the glue break, and some parts of the wood
break, often (but not always, though).
I have, however, tested TiteBond, Elmer's, and hot hide glue in similar
fashion. I butt-glued after sizing each piece first, I also lap-glued
perhaps a dozen sample 1" wide strips 1" overlaps. What I was
delighted to find was this: After half-a-week had probably passed (10
-15 years ago, now), I removed the clamps from the PVA samples, but I
did not clamp the hot hide glue samples. The hide glue, which at the
time was Hudson Industries' 165 gm chrome glue beat all the PVA
samples, although, if I recall, I had several really good and strong
PVA samples -- equal to the hide glue.
So overall, hide glue won. But I was amazed at the differences between
the joints of the PVA samples. They varied all over the map, even
though I really do know how to glue things well, and did them all the
same exact way. I decided the differences were caused by the wood
itself, and the way that wood soaked up the water, allowing the PVA to
set. None of the PVA butt-glued samples held at all -- even after
pre-sizing them. But the hot hide glue butt joints were just fine!
So in summary I would just say that experience is the only real proof,
and that erudite glue strength tests (which, by the way have been
performed) are worthless in a shop, overall. Don't worry -- when you
try to take something apart and it splits a perfectly good board in
half, then that is strong enough. Glue strength has never been the
issue, anyway. It is the reparability and originality that is the
issue. Hot hide glue is just one of the nice little things I do for
myself -- just in case I'm the one who must disassemble it again --
which is always the case. So, be selfish like me.
[ Thanks, Craig. I'll never have the experience, though, and that's
[ why I'd like to see test procedures and test reports, with numbers.
[ The data must be somewhere, if we only knew where to search!
[ -- Robbie