Soundboard tension & compression

[Delwin D Fandrich]

This is a multi-part message in MIME format.

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

  ----- Original Message -----=20
  From: Richard Brekne=20
  To: Pianotech=20
  Sent: November 24, 2002 2:36 PM
  Subject: Re: S&S D Duplex


  Hi Del...=20
  I think I understand what you mean here. Basically you are saying that =
if you use flat ribs, whatever crown you end up with is more or less a =
measure of how much internal compression the panel is under. Otherwise =
there would have to be a difference between flat ribbed / flat cauled =
and flat ribbed / curved cauled.... or what ?=20

  So please indulge me for a bit so'st I can perhaps get this right.. =
all this goes to the flat rib / dished caul variant of Compression =
crowning. Follow along and tell me where / if I hop off the wagon.=20
   =20

    --If you start off with just a panel in a dished caul and push it =
down... then you have a panel thats tensioned on its outside and =
compressed on its inside.
  In theory, yes. there will be some slight amount of tension on the =
outside of the curve. Keep in mind, though, that in a practical =
soundboard the radius of this curve is quite large (relative to the =
thickness of the soundboard panel) and the amount of tension will be =
very slight.

    --Push in a few ribs and the sides of the ribs that meet the panel =
are tensioned.
  Yes. The top (or front) of the rib will be under some tension and the =
bottom (or back) of the rib will be under some compression. How much =
tension and compression will depend on the height of the rib and the =
radius of the curve.

    --If panel and ribs are glued thus and allowed to dry, all stays =
unchanged as long as the thing is in the caul. But when you take it out =
then the whole thing wants to try and return to its flat state but cant. =
The compressed underside of the panel and the tensed sides of the ribs =
work against each other being glued tight as it were.
  So far, so good. A stress interface develops between the soundboard =
panel and the ribs at the glueline.

    --What does happen tho is that it returns about half way and you =
have some crown already before taking on any humidity, and the inside of =
the panel is less compressed then it was in the caul, the outside less =
tense, and the ribs less less tense.

  And now we're on the edge of the wagon getting ready to hop. Actually, =
we're probably already on the bunny trail. I would question that the =
outside of the soundboard panel would be just 'less tense.' Even at this =
point it is almost certainly under some amount of compression.=20

    Let the panel take on humidity..=20
    --and the compression on the underside of the panel increases, the =
tension on the ribs increases...=20
    --and the top side of the panel... hmmm.. its still under some =
degree of tension if I am correct up to this point. So at some point =
enough humdity will put the top side under compression.
  Ah, now we're definitely hopping on down the bunny trail. And we're =
not alone out here. For some reason this seems to be a point of =
confusion with many.=20

  As the panel becomes compressed to the point it is able to develop a =
glueline stress interface sufficient to form or even maintain crown the =
whole panel is going to be under compression. And, as it takes on =
moisture sufficient to increase that crown the amount of compression is =
going to increase. And this increase is evident through the entire =
panel, not just the surface nearest the ribs.

  Keep in mind that amount of wood cell compression required to generate =
the amount of force -- the stress interface -- necessary to bend a set =
of flat ribs into a crown radius of 18 m (approx. 60') is considerable. =
There are a lot of variables but it's going to be something over 1% to =
2%.=20

  Now it doesn't matter to the wood if this compression comes as a =
result of being placed in a vice and squeezed or if it comes as a result =
of developed internal compression through the mechanism of being dried =
(shrunk), restrained (ribbed and brought back to some higher MC =
(expanded). Hence if a soundboard panel starting out at, say, 1,000 mm =
across grain, is put in a vice and squeezed -- compressed -- by 1% it is =
going to end up being only 990 mm wide. If by 2% it will be 980 mm wide. =


  Or if a panel is stabilized at 4% MC and cut to a width of 1,000 mm =
and then taken back up to 12% it is going to expand to somewhere around =
1,015 mm to 1,020 mm (yes, I've done the experiment). That's an =
expansion of about 1.5% to 2.0%. Now, if you dry this panel back down to =
4% (shrink it), put it into a fixture that will not allow it to =
physically expand, and then take it back up to 12% MC, it will still be =
1,000 mm wide (the fixture sees to that) but it will now have a =
considerable amount of internal compression. Something on the order of =
1.5% to 2.0%. At least it will for a while. Since wood cells were only =
designed to tolerate compression levels up to about 1.0% they will =
quickly begin to fail.

  So, what happens if this panel is dried to a MC of 4% and bent in to a =
set of realistic curved cauls (let's stick with our 18 m radius). Yes, =
the outside surface will stretch by some small amount and place this =
surface in tension. Probably by some fraction of 1% -- I'm guessing =
here, but (assuming a soundboard of 8 mm thickness) I'd be surprised if =
the stretch (and tension) was much more than 0.01% . And the bottom =
surface will compress by about the same small amount. In other words, =
the top surface might want to stretch by about 0.1 mm and the bottom =
might want to compress by about 0.1 mm.=20

  Now with a set of flat ribs glued to the back of this panel =
restricting its expansion as the panel takes on moisture all of this =
would-be expansion turns into compression. And here is where we hopped =
out of the wagon: The amount of compression developed will be =
substantially more than is needed to neutralize that slight amount of =
tension and turn it into compression. What you will end up with is a =
panel with, for example (and using my admittedly hypothetical numbers), =
1.49% compression on its top (outside) surface and 1.51% compression on =
its inside surface. But the whole panel will definitely be under =
compression. And when the compression ridges develop they will develop =
all the way through the panel, top to bottom and bottom to top.

  Can we get back in the wagon now?

  Del



---------------------- multipart/alternative attachment
An HTML attachment was scrubbed...
URL: https://www.moypiano.com/ptg/pianotech.php/attachments/88/da/dc/54/attachment.htm

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