Befuddlement??? -Reply

Delwin D Fandrich pianobuilders@olynet.com
Sun, 14 Sep 1997 10:54:02 -0700


Stephen Birkett wrote:
> 
> Del wrote:
> > Wood does not creep? Of course it does. Check the WOOD HANDBOOK, Wood 
> > as an Engineering Material; US Department of Agriculture Handbook #72,
> > Chapter 4; The Mechanical Properties of Wood, Pages 4-4 and 4-36 through
> > 4-41. 
> > Also UNDERSTANDING WOOD by R. Bruce Hoadly; Chapter 6, Strength
> > of Wood. 
> > Also just about any text on the structure of wood or wood
> > technology that you can find. Wood does indeed creep under load.
> >
> Steve wrote:
> We are no doubt talking about different things here...I lent my copy of
> Hoadley so I can't check what you are referring to by this. Obviously wood
> does experience an effect from sustained stress, but I wouldn't call that
> creep, at least not in the same sense creep is used with respect to
> aliphatic glue joints... I've seen harpsichords (less than a tonne of string
> tension) literally fallen apart over time due to aliphatic creep (modern
> ones of course), wrestplanks moved forward into the gap etc....to be fair,
> these were the older aliphatics, and Titebond is clearly much better in
> this respect. But I've not seen any evidence that Titebond creep is ZERO,
> as is creep for hide glue. I would have no confidence using Titebond on a
> big romantic piano, that sustains up to 6 tonnes string tension, held
> together entirely by the glue. I know that hide glue works in this
> application.

Creep is creep. When any structural material is first loaded—whether it
be a wood beam or a glue joint—there will be some, albeit in the case of
animal hide glue, very slight, elastic deformation. If the load is
maintained for a period of time there will be some additional
time-dependent deformation. That is creep. And it doesn’t matter if it
occurs is a wood beam, a soundboard panel or a glue joint. 

Actions, whether ancient or modern, are subjected to stress loads of
such short duration that they are essentially unaffected by creep. But
the evidences of creep have been visible to some extent in every wood
framed fortepiano or harpsichord I have ever seen. Creep in the wood
structure is what has caused (perhaps “allowed” is a better term) them
to buckle—however much or little that may have been—over time regardless
of the adhesive used. Creep, or its cousin, compression set, is also the
primary culprit in probably 99.9% of all soundboard failures that cannot
be attributed to glue failure—at least this is true in compression
crowned soundboards. Again, this is independent of the adhesive used.


> > Del wrote,
> > .... Having worked in a
> > piano factory that used both hide glue and aliphatics I can assure you
> > that there were far more joint failures in the joints bonded with hide
> > glue than in those made with aliphatics.
> >
> Steve wrote,
> Two possible conclusions here. Either the hide glue was used incorrectly
> or hide glue was not appropriate for the modern manufacturing
> techniques/processes used in the factory. 

One of my objections to the use of hide glue today is that, contrary to
your insistence, it is a bit tricky to use correctly. It requires
training and experience to achieve consistently good results. Neither of
which can be assured in either modern factories or in the vast majority
of piano shops of today. Young people grow up today being educated and
trained to do many things, but the use and application of hide glue is
generally not one of them. If they have been given any woodworking
training or experience at all it will have been with the use of either
Titebond or Titebond II (or their competitive equivalents). Since there
is little need in today’s world for this reality to change, it probably
won’t.

But in the joints I was referring to, hide glue was simply an
inappropriate adhesive for the application. It was being used only out
of a sense of tradition. “But, we’ve always done it like that...”


Steve wrote,
> With traditional woodworking
> craftsmanship there are hundreds of years of "proof" of the effectiveness
> of hide glue. In the extreme, there are numerous examples of romantic
> pianos, with no iron braces whatever, and string tensions upward of 5-6
> tonnes sustained continuously over 150 years, entirely relying on the
> properties of hide glue to keep them together (no iron braces
> whatsoever)...and they still are. Aliphatics may be able to match this
> performance, but we will not be able to confirm in our lifetimes whether
> their long term reliability is as good as hide glue.

Our shop specializes in the replacement of piano soundboards. The
majority of pianos coming to us have soundboards that were made, ribbed
and assembled to the rim using animal hide glue. From our perspective,
this is probably a good thing. We know from experience that soundboards
glued to the piano rim will usually pop free fairly easily with the
deteriorated animal hide glue joint breaking cleanly most of the time.
Assuming, of course, that the soundboard-to-rim glue joint has not
already failed—many have around at least around some part of the rim or
belly rail. We also know that it will be difficult to get the old
soundboard assembly out of the rim in one piece because, more often than
not, many of the glue joints holding the soundboard panel together and
the ribs to that panel will have already failed. So much for long term
reliability.

Again, allow me to repeat my standard caution of using the
exceptions—i.e., the survivors—to attempt to prove the rule. I have also
found many very old hide glue joints to be in good condition. But I’ve
found enough failures to know that I am unwilling to trust any critical
structural joint to the uncertainties of animal hide glue. It is
personal experience with the stuff that has taught me to question—not
respect—the long-term reliability of hide glue.


> > Del wrote,
> > I have no problem with that. I just object to the implication that hide
> > glue is somehow superior to some, if not all, of the more modern
> > adhesives. For most applications, it is not. For some few it is as good.
> > And, for an even fewer number of applications, it may indeed still be
> > superior.
> >
> Steve wrote,
> I agree with you Del, up to a point. For modern cabinetmaking/factory
> style construction, hide glue may not be suitable. Apart from such
> industrial applications, I see no reason to use aliphatics in the small
> shop. Del...for what applications do you think hide glue is less suitable
> than aliphatics?
> 
> As I've said many times, hide glue is so much easier to use than any
> modern adhesive, why use these messy aliphatics?
> 
> Stephen
> Stephen Birkett Fortepianos
> Authentic Reproductions of 18th and 19th Century Pianos


All adhesives can be used inappropriately. Including hide glue. Granted
hide glue has nominal creep, but for that very reason it is a poor
choice to use for a lap joint. It is also particularly weak in tension—a
fact clearly understood by early piano builders who provided a method of
mechanically clamping the ends of the ribs to the rim. They notched the
rim and fitted the rib to the notch such that there was actually
physical contact between the bottom of the rib and the rim. (This
practice later became corrupted so that the notches were arbitrarily cut
deep and the excess simply filled with glue.) Hide glue has very poor
gap filling qualities. It becomes quite brittle with age, especially
when applied in a thick glue line. With age, the brittle glue simply
crumbles and falls apart. Hide glue is highly susceptible to moisture
damage, i.e., it is not even moisture resistant, let alone waterproof.
(While I freely acknowledge that we rarely give our pianos baths, it is
not unheard of for them to be exposed to high levels of heat and
humidity.) Animal hide glue is at least as messy to work with as any
other adhesive—more so than most—and is much more tricky to work with
than any other readily available woodworking adhesive. 

Early aliphatic glues did exhibit a fair amount of creep and thoughtful
wood workers understood this and used them accordingly. Modern
aliphatics such as Titebond II, MPA II, etc. also exhibit some nominal
amount of creep. It is very slight, however, and they have been used
successfully for a long enough period of time in bent-laminate
structures so as to prove that this is no longer a consideration in any
practical application. As I said earlier, any remaining tendency to
creep that may remain is insignificant compared to the natural creep
found in wood, especially spruce, sugar pine, etc. We use MPA II for
nearly everything in our work. Contrary to your assertion, it is easy to
use, it is not particularly messy (certainly less so than hide glue), it
is utterly forgiving, it is much better at gap filling than is hide
glue, it requires virtually no training in its proper use, and it forms
consistently stronger bonds than are possible with hide glue. It is also
reasonably water and moisture resistant.

If the possibility of creep is a factor in determining the type of
adhesive to use in a structural wood joint—it rarely is, the joint
should be designed to be structurally supportive if possible—we will
generally choose a urea-formaldehyde based adhesive (sometimes called
plastic resin glue). This is an excellent structural adhesive with many
times the bond strength of animal hide glue and one which also forms an
extremely rigid joint—it is at least as rigid as with hide glue—but with
the additional advantages of being very easy to use, again utterly
reliable. It is even more water and moisture resistant than Titebond II
and MPA II. We use urea-formaldehyde for many belly work applications. 

I don’t consider absolute water resistance to be terribly important for
piano construction, but if this were to be a property I considered
critical for some application—if you intend to store your piano from
time to time in a vat of  boiling water, for instance—I’d use a
resorcinol resin adhesive. This is probably the strongest and most
reliable adhesive extant and it is completely waterproof. It will even
pass the “boil test”—but, therein lies a tale for another day...

My apologies for allowing this to get so lengthy. With this I’ll end my
participation in the great glue debate of ‘97 and allow Steve to have
the last word if he so chooses.
—ddf



This PTG archive page provided courtesy of Moy Piano Service, LLC