 The Clarinet BBoard
|
Author: Gordon (NZ)
Date: 2002-10-30 11:27
A couple of pages down there was a thread "Help! My barrel won't come off! " started by 'Nick'.
There was discussion on whether a jamming tenon joint was caused by the tenon timber swelling or the socket timber shrinking as a result of moisture in the bore.
On 22 Oct I started an experiment. I removed the reinforcing ring from one end of a barrel, measured the bore at this end, corked the other end, and left the barrel full of water. Today, 9 days later, I took measurements of the bore again.
Ken asked me to report in a new thread:
Original diameter = 14.49
The bore is now somewhat oval
New minimum diameter = 14.50
New maximum diameter = 14.57
This suggests to me that when a grenadilla cylinder gets damp inside, it behaves like a metal cylinder that is heated, i.e. the inside diameter increases. This suggests to me that the socket bore diameter must increase if anything, and that it is expansion of the tenon that causes jamming. However the result raises more questions:
1. Does the bore diameter reduce if the outside diameter is constricted by a tenon ring. I think this most unlikely, but who knows!
2. Is it possible that although the tenon bore diameter increases when wet, the outside diameter remains essentially the same. This too, seems most unlikely to me.
The next step of the scientific process would be for somebody to duplicate the experiment, and also to test the above two possibilities with alternative experiments.
|
|
Reply To Message
|
|
Author: Henry
Date: 2002-10-30 15:48
Bravo, Gordon (NZ), this is the kind of work we need to throw some light on the issue of dimensional stability of, and the possible creation of moisture-induced internal stresses in, grenadilla wood! The fact that the bore becomes oval is disturbing, although perhaps not too surprising, considereing the fact that almost any piece of wood will be somewhat anisotropic (i.e., with properties that depend on the direction). I do believe that the tenon rings are there to reduce the swelling and the effect of the anisotropy. Have you measured the bore diameter at the other end of the barrel (the one where the ring was still present)? Omar, I know from your previous postings that you have worked with samples of grenadilla wood. Did you make any measurements of the dimensional stability upon immersion in water or various oils?
|
|
Reply To Message
|
|
Author: Henry
Date: 2002-10-30 16:07
Gordon (NZ), just one more question. Was the barrel recently oiled, or was it totally "dried out"? In the latter case, you may wish to repeat the experiment after a good oiling (with the Doctor's oil, of course!).
|
|
Reply To Message
|
|
Author: Don Berger
Date: 2002-10-30 17:30
Gordon NZ - May I applaud your "scientific method" approach to this frequent problem, and your request for independent verification. My working career was in chem eng. research, oil and petrochems, so I recognize a "fellow" worker. Yes, the wood is capable of greater domensional change with ambient conditions than the nickel-silver rings, so "its a battle" between very [too] tight and loose, which most of us experience. I'll try to respond to your questions before long. Many TKS, Don
|
|
Reply To Message
|
|
Author: L. Omar Henderson
Date: 2002-10-30 17:36
Bravo Gordon !! Every piece of wood however is it's own experiment since the growth patterns, porosity, internal stress patterns, etc. vary from piece to piece. Since Grenadilla wood is so dense, the porosity is diminised in comparison to other wood types, dimensional deformation parameters are attenuated to some degree. The addition of oil to the wood, and the water holding properties of plant derived oils can further mollify the absorption of water and subsequent swelling.
As detailed in previous posts I have been experimenting with oil mixtures that attempt to stabilize or bring back dimensional characteristics of older horns that have been allowed to "dry out". With the help of some engineering friends at GA Tech. and their modern micro-measurement techniques (including miniature strain guages, laser measurement devices and micro-temperature and RH sensors) we have been able to document changes in proportions of wood infused with various oil mixtures and kept in elevated humidity conditions. The effects - changes seem to be more related to the individual piece of wood in terms of the 3D directions of swelling but there is a general increase in most parameters. Wood infused with plant derived oils shows the greatest overall lessening of deformation.
I have conducted some experiments on different segments of the same billet of native Grenadilla wood which have been infused with various types of oils (animal derived, plant derived, and petroleum), untreated, and then exposed to varying temperature and RH conditions for different lengths of time. As indicated before, the wood infused with plant derived oils shows less dimensional deformation than with the other oil types used and untreated wood shows the most change. The caveat of course is the statistical significance (power) to detect changes with a limited sample size, with the number of variables measured, and the number of different samples tested (both time and money are my limiting factors). I have also measured changes in parameters (length, width, and circumference) in Grenadilla tubes where only the inside is exposed to elevated RH and different oiling conditions. These latter experiments indicate swelling - change in circumference of the bore with increasing exposure to moisture while the outer circumference changes little (is not statistically significant).
The caveat to all of these latter experiments is that they are on tubes without holes (as in the musical instrument) and keywork attached. It is possible to have selective deformations in sections of wood depending on the dimensions of the hole and the surrounding structure of the wood and the imposed structure added by keywork interactions. Hydrostatic pressures in the wood can be tremendous and cause stresses which can be absorbed within the structure of the wood itself by swelling or by deformation of the wood in various planes. There is also differential absorption of water by end grain - e.g. tenon ends and surface absorption. If the pores and tubes of the wood are not plugged then end grain absorption of water is greater than flat surface absorption (completely parallel to tubule orientation) in most cases.
So, the experimentation continues - while time and money permit.
The Doctor
|
|
Reply To Message
|
|
Author: Ken Shaw
Date: 2002-10-30 20:15
Bravo Gordon. It's about time we got real experimental data.
Thinking about, it, I'm not surprised the warping was slightly oval. Very few billets will be made from the exact center of the tree, so the annual layers aren't wrapped conically around the bore. Instead, they're curved. The summer and winter growth will be different densities and so absorb water at different rates. Boxwood is (in)famous for oval warping, which is why museum instruments are so hard to measure.
Best regards.
Ken Shaw
|
|
Reply To Message
|
|
Author: JMcAulay
Date: 2002-10-30 22:27
Gordon, let me add my applause to that of others. What a great post.
And Dr. Henderson, thanks for another fine tutorial.
Regards,
John
|
|
Reply To Message
|
|
Author: diz
Date: 2002-10-30 23:15
Gordon, you're a neverending source of inspiration ... mental note, visit NZ soon!
|
|
Reply To Message
|
|
Author: Gordon (NZ)
Date: 2002-10-30 23:37
In answer to a couple of the specific questions:
1. I thought about the ring-causing-the-timber-to-expand-into-the-bore possibility (definitely not probability!) too late, so I did not take measurements of the end of the barrel that was still retained by the ring.
I doubt that the ring would have had much effect on the dimensional changes of the bore proper (as distinct from the tenon socket), because there is quite a lot of very thin timber between the ring and the bore proper. Any forces exerted by a ring would apply only to the timber in the immediate vicinity of the open end of the socket. Note also that the timber in the region of the metal is not exposed to moisture in the bore of a clarinet, so it could be expected to have good dimensional stability..... which adds fuel to my contention that it is the swelling TENON that causes jamming - jamming almost always includes the fit between the tenon and the ring end of the socket.
2. The barrel had not been used or 'oiled' for at least 10 years, so it was initially as dry as timber gets in this relatively high humidity city of Auckland.
I note also, that if any experiments are done to determine acoustic effects of ovaling of the bore from the effects of moisture, then the another parameter must be eliminated; as a result of all that water exposure, the bore surface is now considerably rougher than it was before the experiment.
|
|
Reply To Message
|
|
Author: Henry
Date: 2002-10-31 00:29
Gordon: An additional few questions.
1. Is it too late now to measure the bore diameter at the other end of the barrel, as I suggested earlier? I still believe that that would be useful info to have. Or could you simply repeat the experiment and then make that measurement after all?
2. Are you considering repeating the experiment after judicious oiling of the barrel?
3. At the end where you removed the metal ring, you could have measured the inner and outer diameters of the terminal section of the wood, unconstrained by the ring. That would almost be like a (short) tenon. Such measurements could have validated your strong feeling that it is swelling of the tenon that is responsible for a stuck barrel. Wouldn't you agree?
Finally, your observation that the bore was much rougher after the experiment is important as well. This could have a significant effect on the tone. Oiling might then have the extra benefit of preventing (or reducing) this roughening. Yet another reason for repeating the experiment after oiling!
Again, Gordon, thanks for taking the time to do this kind of quantitative stuff!
|
|
Reply To Message
|
|
Author: Henry
Date: 2002-10-31 01:00
Sorry for keeping the questions going! My LL Leblanc has a barrel with two sockets, as usual (I assume). The upper joint has two tenons. However, the upper one of these (the one that goes into the barrel) is itself protected by a metal ring at its end. The lower tenon of the upper joint is not so protected. The lower joint has a socket to receive the upper joint's tenon, and a lower "unprotected" tenon that goes into the bell. Is this pretty much standard? If so, then the metal ring on the upper tenon of the upper joint does not appear to prevent the apparently common problem of a stuck barrel. Or does the LL happen to be unique in this regard and might therefore not suffer from this problem? Any thoughts from you, Gordon, or anyone else?
|
|
Reply To Message
|
|
Author: Gordon (NZ)
Date: 2002-10-31 03:32
Henry: There's not much point in measuring the other end because I don't have its original size recorded. The mouthpiece is drying out again now. That is the end of my experimenting for the immediate future. Other commitments call.
"At the end where you removed the metal ring, you could have measured the inner and outer diameters of the terminal section of the wood, unconstrained by the ring... Such measurements could have validated your strong feeling that it is swelling of the tenon that is responsible for a stuck barrel."
In my view, my experiment demonstrated that unrestricted timber swells in a similar fashion to heated metal, i.e. circumerentially, and not INTO the bore as seemed to have been claimed. I have no doubt at all that a wooden cylinder's outer diameter increases when the timber swells from moisture absorption. I don't know how this can be disputed.
|
|
Reply To Message
|
|
Author: Gordon (NZ)
Date: 2002-10-31 03:44
Henry: It is quite common for clarinets to have tenon rings in random places. I've seen it with new Selmers. Perhaps towards the upper end of the instrument because there is more moisture here, creating dimensional instability.
It is entirely likely that they are used to repair problems/faults in manufacture, e.g. a tenon turned to too-small a diameter, or a split or chip in the end of the tenon. A joint is hardly likely to be discarded after manufacture when a ring can solve one of these relatively minor problems, both functionally and cosmetically.
I don't recall ever encountering a stuck tenon where a ring is involved, although I have encountered such a tenon where the jamming is at the OTHER end of the tenon where there is no ring.
|
|
Reply To Message
|
|
Author: nzdonald
Date: 2002-10-31 08:11
good one Gordon...
just a quick aside- re "relatively high humidity in Auckland"... well, it's not really very humid at all here (although we do have lots of damp buildings both old and new) but more importantly, we don't have a great variation of humidity throughout the seasons....
Chicago and Cincinnati, for example, are many times more humid than Auckland in the summer but also very dry in the cold of winter. This wide variation is probably quite stressful on wooden instruments- made worse by the drying effect of air conditioning (meaning that in summer the clarinet may be in VERY HUMID and VERY DRY conditions within a short period of time). In Auckland the humidity doesn't vary a great deal from winter to summer, comparably at least.
donald
|
|
Reply To Message
|
|
Author: TDC
Date: 2002-10-31 13:23
Tubes, schmubes. The reason the wooden tube becomes "oval" shaped is because when wood swells, it only swells along its width, and never its length.(a log gets fatter, never longer.) That is why wooden rules (or "rulers" as they're called incorrectly) remain accurate.
If you look at the directions of grain in a clarinet joint, the direction of growth of the tree it came from is parallel with the long axis of the joint. It won't swell that way (at least not so's you'd notice). But it does swell in width. Picture that clarinet joint as a flitch - a 2 X 2 X 14" blank of grenadilla. If you throw it in water, how does it swell? Only across its grain: one direction only. It'll now be 2 X 2 1/4 X 14" (to exaggerate for clarity).
Isn't it natural that a tube made out of it will become ovoid?
What's needed in clarinet technology is to get rid of the dusty, retrogressive "tradition is everything" attitude. Not since Boehm has a new idea caught on. Every damn innovation that would make the instrument more reliable and easier to play has been rejected hands down. Now we're rejecting Buffet's new material as well. We need some Physics!
TDC
|
|
Reply To Message
|
|
Author: Bob
Date: 2002-10-31 14:05
All very interesting but didn't Larry Naylor tell us some time ago that we must differentiate among the different locations of the log from which the part came...as just one variable? I certainly am a believer in "the scientific method" but experiments like this and the glovebox one leave me uninspired. Sorry to disagree with such talent.
|
|
Reply To Message
|
|
Author: GBK
Date: 2002-10-31 17:41
"...That is why wooden rules (or "rulers" as they're called incorrectly) remain accurate..."
-and we all know what a British ruler looks like:
Queen Elizabeth II
(insert rim shot)
Thank you, thank you, I'll be here all day...GBK
|
|
Reply To Message
|
|
Author: diz
Date: 2002-10-31 21:34
Thank god - someone else who knows the term "ruler" for a piece of wood/metal for measuring distance is the wrong terminology!!
|
|
Reply To Message
|
|
Author: Gordon (NZ)
Date: 2002-10-31 22:53
Language evolves. 'Rule', to me, is the verb - "rule a line" - so the derived noun is "ruler" - "rule a line with a ruler". Same as "scribe a line with a scriber". Sensible. Let common sense rule.
Do I recall Mark having language pedanticities as a no-no in this forum?
|
|
Reply To Message
|
|
Author: Henry
Date: 2002-11-01 00:15
Gordon, it seems we have exhausted the subject that you started on so promisingly. Too bad! Perhaps someone else will pursue the issue further. Still, it was fun to look at just three numbers and speculate for a while!!! Thanks again.
|
|
Reply To Message
|
|
Author: Gordon (NZ)
Date: 2002-11-01 09:15
It could make a fairly straight-forward scientuific analysis project for a student at some level. Old mouthpieces can't be too difficult to optain, and equipment required is minimal. Measuring devices could be borrowed from industry or college workshop.
|
|
Reply To Message
|
|
Author: jim lande
Date: 2002-11-02 04:45
radial shrinkage is not zero, but rather about half of tangential shrinkage. In Ebony, however, the difference is not so great. lengthwise shrinkage does tend to be very small in most species. See
http://www.fpl.fs.fed.us/EPW/shrinkage_of_woodFAQ.html#Hardwoods
|
|
Reply To Message
|
|
The Clarinet Pages
|
 |
.jpg){kind=small}
.jpg){kind=small}
.jpg){kind=small}
.jpg){kind=small}
.jpg){kind=small}
.jpg){kind=small}
