soundboard stresses

[Delwin D Fandrich]

This is a multi-part message in MIME format.

---------------------- multipart/alternative attachment

We have just seen a bunch of numbers carried out to several decimal =
points of accuracy being tossed about to prove one thing or another.

May I remind the gentle reader that any numbers given for such =
parameters as compression strength perpendicular to grain and shrinkage =
or expansion relative to changes in moisture content are averages. They =
are derived through the careful testing of a variety of wood samples =
within a given species but taken from a variety of locations and growing =
conditions. Some samples will have physical characteristics considerably =
higher and some considerably lower than these averages. These numbers =
are not, sadly, The Gospel Truth!

To complicate things even more, they are based on tests taking a =
relatively short period of time -- minutes rather than hours or days. =
Hoadley, in his book Understanding Wood, suggests that we must degrade =
these figures by approximately 40% if the time under load is going to be =
appreciably longer than the time taken for the specific test in =
question. This number also appears to be an average as I've had other =
wood technologists quote degrade figures of 30% to 50% for time under =
load periods of five to ten years. It depends, apparently, on the =
particular characteristic and the particular species of wood under =
consideration. For a characteristic such as tensile strength parallel to =
grain the degrade figure can be considerably lower. For a characteristic =
such as compression strength perpendicular to grain it will be some =
higher.

To use the published test figures in an attempt to show that a =
soundboard made to a given set of specifications can safely be taken =
right up to the point of short term damage without sustaining either =
short-term or long-term degradation overlooks the cold, hard reality of =
the natural variableness of wood in several ways.=20

First, it overlooks the problem that, out of any given batch of wood, =
some samples are going to measure high and some are going to measure =
low. If we could depend on every single piece of wood that goes into a =
given soundboard panel testing on the high side, or at least on its =
being average, we'd be in great shape. But in the real world we can't =
count on this. Inevitably, some samples are going to test low. In fact, =
about half of them are going to test on the low side of average.

Second, it overlooks the problem that no lumber grader, regardless of =
training and experience, can accurately determine any of these physical =
characteristics without getting into some form of destructive testing =
which, needless to say, is not being done by any piano manufacturer. Nor =
has it ever been that I am aware of. At least not in production. We =
can't get by with random testing, you see, we would have to test every =
single piece of wood.

And, finally, it overlooks the degrade factor that must be applied if =
the structure is going to be expected to support a load over some =
extended period of time. Even if we accept Hoadley's figure of 40% -- =
rather than the more probable 50% for a parameter such as compression =
strength perpendicular to grain for a soft wood like spruce (it's all =
that soft earlywood that does it) -- this still means that we have to =
work from a number closer to 350 lbs/in2 than to the 580 lbs/in2 quoted =
in the USDA's Wood Handbook. And all the time we have to keep in mind =
that about half of the lumber we use is going to test lower even than =
that. Just how much lower, of course, we have no way of knowing because =
there is no practical way of either determining or predicting the =
specific and exact mechanical characteristics of any given sample of a =
natural material like wood.

Del
---------------------- multipart/alternative attachment
An HTML attachment was scrubbed...
URL: https://www.moypiano.com/ptg/pianotech.php/attachments/97/41/90/85/attachment.htm

---------------------- multipart/alternative attachment--