what humidity and how do you keep it?

[le_koukou]

Actually I like to think that cork is not like a sponge at all. It is very waterproof and still often used to seal bottles. A cork will protect a good wine for decades.

One should try to measure the amount of humidity a piece of corck would absorb. I would not be surprised if that would be less than the same piece made out of blackwood.

[GaryKelly]

Actually I like to think that cork is not like a sponge at all. It is very waterproof and still often used to seal bottles. A cork will protect a good wine for decades.

One should try to measure the amount of humidity a piece of corck would absorb. I would not be surprised if that would be less than the same piece made out of blackwood.

I'll let Terry take care of that one I think!

[Jayhawk]

To rehydrate cork, you need to use hot, nearly boiling water, so I agree that cork is not sucking in water and causing the swelling.

Besides, my flute is without cork (threaded with nylon), and I have this issue, too (tighter joints with longer periods of playing). The nylon threads are coated with cork grease, and they don't absorb any water. Even more telling is in owning at different times two delrin flutes, one with cork and one with nylon thread, neither ever had this issue.

I really can't see any explanation working except that the wood is swelling with humidity. It happens to doors, floorboards, pieces of plywood left in damp basements, etc. I don't think it's instantaneous, but as Terry said a flute is a thin piece of wood exposed to significant humidity at 98.6 degrees farenheit.

Eric

[GaryKelly]

Actually Terry didn't say that, he said we had to completely ignore the temperature and humidity of human breath because the flutes are lined, and the bores oiled, so the breath can't affect them.

It's odd, I have a flute with an ABS headjoint and blackwood body. The top tenon is lapped with dental floss and greased (I removed the original cork, which was damaged when the parts were separated for the first time in a very long time). I don't experience any problems taking the headjoint off after playing.

But the bottom tenon is cork-lapped, and after playing for an hour or two, I do notice that taking the footjoint off is harder. Weird.

The more information I see shared in this thread, the more these flutes seem like living creatures, all with their own unique personalities.

[Jayhawk]

The more information I see shared in this thread, the more these flutes seem like living creatures, all with their own unique personalities.

I agree with you completely on that! Perhaps the real reason flutes crack is that they're simply mad at you...being the temperamental beasts that they are with limited self expression abilities.

Eric

[dcopley]

With all this experimenting going on, I could not resist getting in on the act. The central point of the discussion seems to be whether cracks are caused by temperature changes or by humidity changes. My opinion at the moment is that for metal-lined head and barrel joints the primary cause is a dry environment, with a major contribution from stresses induced by poor manufacturing practice. I don't think that temperature is an important contributor, but I could be wrong on that.

So a fairly direct experiment would be to make some simulated head-joint sections (wood tubes with a metal lining) and subject some to high temperature and some to low humidity. See which, if any, develop cracks. Here is a preliminary test which I started with available materials.

I took a brass-lined head joint. It already had a 3.8 cm long crack running from the bottom (I'm not willing to sacrifice a good one at this stage). I corked the top end and the embouchure hole to seal them.

Phase I (temperature): I poured boiling water into the open end, until it filled the liner tube. I emptied and re-filled it a couple of times to get the highest temperature I could. I saw no new cracks and no lengthening or widening of the existing crack.

Phase II (humidity): I put the same piece in the oven, at around 110 degrees F and 15% RH. It has been there for just over an hour. No changes seen yet. I'll keep checking it for new cracks and changes to the existing crack. It can stay there for the next few days.

Meanwhile I'd be interested to hear ideas for a larger scale experiment. I am thinking in terms of making a number of short simulated head joint sections (maybe with round "embouchure" holes), and following a similar method of producing temperature and humidity changes. I also plan to do the humidity test on some unlined pieces to measure the deformation when not constrained by a metal tube. I may even be able to provide test specimens to someone who has laboratory facilities and is willing to share the results publicly, provided it does not take up too much of my time and budget.

Dave Copley
Loveland, Ohio

[GaryKelly]

Brilliant Dave! I hope someone with the facilities will snap up your offer!

For what it's worth, I have a flute in my freezer as I type...

Don't panic, it's only one of those awful pakistani jobs, so no great loss if I blow it up!

[Terry McGee]

Phase I (temperature): I poured boiling water into the open end, until it filled the liner tube. I emptied and re-filled it a couple of times to get the highest temperature I could. I saw no new cracks and no lengthening or widening of the existing crack.

Whew, a breathtakingly direct approach! Well done - hard to argue with that. I think we can officially forget temperature, and even temperature shock as an issue, providing of course Phase II produces a result. If it doesn't, we're thrown back onto Divine Intervention.

Phase II (humidity): I put the same piece in the oven, at around 110 degrees F and 15% RH. It has been there for just over an hour. No changes seen yet. I'll keep checking it for new cracks and changes to the existing crack. It can stay there for the next few days.

Sounds good to me. Now, can you tell us a bit about the head and its crack so we know how to interpret whatever results you get. I assume it's one you have made. Did it come to grief while making or in the field? Have you been assuming so far that the crack is humidity induced or do you know that an accident befell it? How tight was the brass to the wood at the time of making - tight, smooth, loose, rattling fit? What were you using to keep it in place - burrs, scratches, tightness, urethane glue, epoxy, etc?

Meanwhile I'd be interested to hear ideas for a larger scale experiment. I am thinking in terms of making a number of short simulated head joint sections (maybe with round "embouchure" holes), and following a similar method of producing temperature and humidity changes. I also plan to do the humidity test on some unlined pieces to measure the deformation when not constrained by a metal tube. I may even be able to provide test specimens to someone who has laboratory facilities and is willing to share the results publicly, provided it does not take up too much of my time and budget.

Dave Copley
Loveland, Ohio

Excellent, Dave. I think the first step in coming up with new strategies to minimise cracking is to get reliable at producing cracking!

Keep in mind with the unlined sample shrinkage test that it will shrink ovoid, and so measuring maxima and minima is important. Could be good to measure inside and outside diameters. Inclusion of an unlined sample is also good to test that the rate of dehumidifying isn't just too much for the wood anyway - if it also cracks we know that was the reason. I wouldn't expect it to unless the wood was too damp to start with. You should expect the unlined sample to dry at a faster rate as air has access all round it.

It would be interesting to try out some different fixing strategies to see if some do indeed promote cracking and if some retard it. I'm thinking burrs or scratches, urethane glue, epoxy, shellac. You'd probably need to repeat such tests a few times to be absolutely conclusive, but once might be enough to point in general directions.

Terry

[dcopley]

Now, can you tell us a bit about the head and its crack so we know how to interpret whatever results you get. I assume it's one you have made. Did it come to grief while making or in the field? Have you been assuming so far that the crack is humidity induced or do you know that an accident befell it? How tight was the brass to the wood at the time of making - tight, smooth, loose, rattling fit? What were you using to keep it in place - burrs, scratches, tightness, urethane glue, epoxy, etc?

Terry

Here is where I should sound a note of caution in reading too much into what I am doing here. It is definitely a "one-off" and mainly of value to get some ideas of how to design a more comprehensive test.

The head joint was an early version of what has evolved into my current design. There is 43 mm of brass tubing projecting beyond the bottom ring. Going upwards from the bottom ring, there is first a socket, extending 20mm, where the wood has been bored out to a larger diameter to accept the part of the barrel liner which projects from the wood of the barrel. For the next 6 mm, the wood and the brass are in contact. Proceeding up the head joint, the brass has been machined back by 0.2 mm so it is not in contact with the wood. Then for a length of 20 mm centered on the embouchure hole, the brass is in contact with the wood. Then another region of machined back brass, and back in contact again for the top 5 mm. A picture would definitely be worth a thousand words here.

The brass was a smooth fit in the wood. Assembly was with Gorilla glue, a polyurethane glue which foams and expands to fill space. So the deliberate gaps between the brass and the wood are filled with urethane foam.

The crack happened in service, when the flute was fairly new (I'd have to dig through notebooks to get the exact age). It appeared to have started right under the bottom ring and progressed up toward the embouchure hole. It is widest at the bottom under the ring. If this is the case, then the head liner would not have been in contact with the wood at the location the crack started. I did notice that when I closed up the tuning slide, the barrel liner was a very tight fit inside the wood socket, so my hypothesis (rather a weak one) is that stress from the barrel liner caused the crack.

So for further experiments I plan to include some more conventional assemblies where the wood and the metal liner are in close contact, and some where the assembly is a tight or force fit.

Dave Copley
Loveland, Ohio

[Cathy Wilde]

I recall a tech sheet from Taylor guitars giving the specs of their workshop humidity for the same reason - to tell the owners how humid to keep the instrument and prevent the wood cracking. They build the guitars in a humidity controlled environment. Is this relevant?

Relevant to Taylor guitars I'd say. Terry will doubtless say it's equally relevant to flutes, but I would disagree. A guitar soundboard is extremely thin and very long and wide (compared to a flute, for example), so I can understand why they would be sensitive to drying out and warping over time, far more so than a short fat thick-walled headjoint.

I do have to chime in here, and say that our local woodwind repairer recently told me he's talked to the fellows who import LeBlanc (? I'm not sure if that's the brand) clarinets -- they place them in a humidity-controlled room as soon as they arrive, and leave them there for weeks until they "settle" post-trip. Don't know what that's about, but it's another instance of people thinking there's a connection ...

Anyway, thank you, thank you, thank you to everyone, especially the heroic experiments and debating skills of Terry and David and Gary. I'm so grateful y'all are on the case, and I think only good can come of it.



cat.

[Terry McGee]

Thanks Dave, very helpful in visualising the situation.

I don't think your hypothesis about the barrel sleeve cracking the head wood is all that weak. You no doubt had adequate clearance when you made it, or you would have noticed the drag before you dispatched it. But if the wood shrunk a bit it could have taken up that clearance. I've certainly found 19th century flutes where that has happened.

So, we have an already established crack at the bottom, but not at the top, and areas of minimum contact surrounding an area of full contact at the embouchure. That's fortuitous - it's like running several experiments at once! You should push on beyond the first event (need for cakes perhaps intervening!).

Can I suggest that anytime an event occurs, you pull the piece from the oven, pop it into a plastic bag with the hygrometer, expel most of the air, seal it and see where the hygrometer settles (could take a little while). Although we know what RH your piece is headed for (15%), it will take a few days to get there, especially in the middle region. It would be nice to know at approximately what levels of RH any events occured.

Most interesting.

Terry

[dcopley]

Can I suggest that anytime an event occurs, you pull the piece from the oven, pop it into a plastic bag with the hygrometer, expel most of the air, seal it and see where the hygrometer settles (could take a little while). Although we know what RH your piece is headed for (15%), it will take a few days to get there, especially in the middle region. It would be nice to know at approximately what levels of RH any events occured.

Terry

I can do as you suggest, though I don't plan to be getting up in the middle of the night to check it. This is starting to sound more like a project for someone with a new baby and who probably doesn't get much sleep anyway. Hmmm ... perhaps I should seal the piece in a few ziploc bags and ship it off to Oz.

Dave Copley
Loveland, Ohio

[Terry McGee]

Relevant to Taylor guitars I'd say. Terry will doubtless say it's equally relevant to flutes, but I would disagree. A guitar soundboard is extremely thin and very long and wide (compared to a flute, for example), so I can understand why they would be sensitive to drying out and warping over time, far more so than a short fat thick-walled headjoint.

You're certainly right, in that I would say it is as relevant. It's all part of the same struggle - how to come up with a structure that will permit natural movement of the wood without messing up the operational and acoustic requirements of the instrument. While guitar soundboards are wider than a flute and will therefore move much more, the ribbing is not as immoveable as is our liners, er, your liners.

If you think a guitar is bad, try the harpsichord. I've made a few of these over the years. The soundboard is about 4ft wide and less than 1/8" thick. The difference in width on a dry and a wet day is about 1/4". Deadly serious, not kidding!

When you install a harpsichord soundboard, you wait for a dry spell, or you heat the board over radiators to artificially dry it. You monitor the width to make sure it is at the minimum before glueing it in. Of course the glue is water based, so you have to work very quickly indeed to get it all clamped down before the water in the glue swells it. You work from the worst place - the belly rail at the front of the soundboard. That's where the soundboard is at its greatest width, and where it also tends to be thinnest in the treble region. If you get it right, the soundboard will swell up and rise in the middle during wet periods. If you get it wrong, it will crack in the treble in dry spells.

By comparison, the furniture people have it easy. They use frame & panel construction on cabinets and doors, and sliding tops on solid wood tables. We use frame & panel construction on our flute cases, so they won't crack either. The top and bottom panels are set into slots all round the sides, and only glued in two spots, at the centre of the ends.

Terry

[GaryKelly]

I've been thinking a lot about Dave's boiling water experiment. What panache and derring-do! (Dave you're a nutter!).

I don't think we can totally rule out temperature based on the boiling water test. Remember, we've already established that the amount of increase by temperature of the circumference of the liner is small, certainly too small for it to have opened up the crack visibly to the naked eye alone in Dave's test.

So I still think temperature is a major contributor...remember, in an intact flute, the liner is constrained from expanding by the flute walls until or unless the flute splits.

What we need is a nutter prepared to repeat Dave's experiment with, say, a shiny new Olwell... I'd use my pakistani flute but that would skew the results of my freeze/play cycle testing!

And I have thought muchly about all of this, perhaps too much. I still refuse to believe that a wooden flute reacts near-instantaneously to humidity changes, such that its internal moisture content varies immediately to remain in equilibrium with its surroundings. The wood's simply too dense. I go now in search of answers!

(Oh and Terry, that Clinton flute is fascinating! Are there many still around?)

[Terry McGee]

I still refuse to believe that a wooden flute reacts near-instantaneously to humidity changes, such that its internal moisture content varies immediately to remain in equilibrium with its surroundings. The wood's simply too dense. I go now in search of answers!

Correct - the flute can't react near-instantaneously to humidity changes - remember my unlined, un-oiled head took 3 to 4 days to re-equilibrate from around 45% to around 20% RH. I have to say I still think that was pretty quick, but we have to keep in mind that the furtherest a molecule of water has to travel is about 2mm. So I'd expect Dave's lined head will take longer to react - a molecule of moisture near the liner will have to work its way through 4mm of wood, and then the surface finish.

The ends can react faster, as the moisture can sneak out "along the straws" about 10 times quicker than making their way through them. Dave takes advantage of this in fixing his rings.

So a short burst of dry weather isn't going to bother the flute - it will be integrated out. But in parts of the US and Canada, they have three months of snow, so that's plenty of time for the flute moisture to drop, drop, drop. The ticking of the time bomb becomes palpably loud ....

(Oh and Terry, that Clinton flute is fascinating! Are there many still around?)

No - I can think of one in the DCM collection at Washington, and two in private collections in the US. I suspect that not many were made, as all three have very close serial numbers. Perhaps ebonite had filled the bill. I think it was a lot cheaper.

Terry