Effects of thread wrapping, Series 2:

Plan D


So far, in our attempts to restore the bore proportions of Potter's old flute, we've tried several treatments, but only Plan B, heavy humidification, appears to have done much at all.  And what it did do was a little confusing.  We'd hoped that the heavily compressed tenons would magically reconstitute.  They did reconstitute a little, but not enough, but the startling effect was that the whole section gained diameter.  At first we thought this might be due to hysteresis, but a subsequent experiment suggests this isn't the explanation.  That only leaves us with the scary thought that the binding of the tenons at both ends had actually compressed the entirety of the section.  If that were so, then this flute is operating significantly outside of the maker's original intentions.  Clearly, we need to get to the bottom of this.  But how? 

One possible line of investigation is to try a different reconstituting approach and see if that produces the same result, a greater or lesser enlargement, or none.  Plan A, just playing, and Plan C, simulated playing, proved not particularly effective at introducing moisture and motivating change.  I think it's time we got a whole lot more bold. And to do so, we're going to invoke one of the great innovations of Potter's own era.  Steam!

Speaking of steam

We can't slip into this though without a sideways glance toward one of the great names associated with the development of steam power, James Watt.  Watt's dates are similar to Potter's.  Seems he also had an interest in flutes ....

"As a young instrument maker in Glasgow, Watt may have found the going hard. As well as scientific instruments he sold musical instruments like violins, flutes, bagpipes and guitars.  Watt wasn't particularly interested in music but with less demand for scientific instruments than he had anticipated, the sale of musical instruments may have provided an extra source of income.
In the 1760s fashionable young men considered learning the flute as a respectable pastime. From the 1730s Thomas Lot had made high quality flutes for players and aristocratic families across Europe. But while Lot was the Stradivarius of the flute world, many lesser makers sold flutes for only a few shillings each. There is a possibility that young Watt counterfeited copies of the finest flutes available in the seventeenth century.
The evidence is in James Watt's garret workshop. This was presented in its entirety to the Science Museum in 1924 and it remained virtually untouched following Watt's death in 1819.
Curator Michael Wright discovered a number of flute-making tools among the workshop's contents. These include a crude stamp for marking wooden goods with the maker's name. The stamp has four letters: T LOT. To an amateur it presents a 'very passable imitation' of the stamp officially used by Thomas Lot.
What was the stamp doing among Watt's possessions? Being short of money, Watt may have marked his own cheap flutes as being made by Lot, in order to sell them for more profit. A real musician wouldn't mistake a Watt flute for a Lot flute but a gullible amateur might easily do so."

(From "Making the modern world - James Watt, life and inventions".

You can also see some of Watt's flute-making tools.  Follow the link, press Continue, choose Explore the Garret Workshop, and press Flutemaking Tools.)

Steaming flutes?

So, what's the attraction of steam in the context of reconstituting a strangled flute?  Lots of things, we hope.  We can expect faster results in terms of how quickly the moisture penetrates, but also greater results, as the high temperatures should also play a role by making the lignin, the glue that holds wood together, more flexible.  Just how much quicker and greater, we will only learn from the experiment.  That alone makes it worth doing!

The Experiment

So what does Plan D entail?  We've already permanently changed the flute's LH section, so further experiments there might be inconclusive.  But the RH section pretty much escaped alteration, so let's see what steam can do for it.  Good at this time to review its current condition.  It's the section between X=200 and X=330mm in the graph below:

There appear to be two areas of suspicion in the RH section.  The bore just after the socket (X=210 to 220 in the graph above) appears to be narrowed.  This is not surprising, given that the original ivory ring is long gone and was replaced by a crude brass one.  The cylindrical brass ring appears to have been hammered on to the taper at the start of the socket, with the effect that a cylindrical crack has formed in the area where the compression can be seen.  Nasty bit of work, and not something we normally have to deal with.

And there's the compression under the tenon (X= 310 to 330mm) and probably in the lead-up to it (X=290-310mm).  Now of course, we don't know these are not original, but we certainly suspect they aren't.  It would be fabulous if a miraculous recovery proved we were right, but that might be too much to hope. 

And of course there's the new question, prompted by our experience in Plan B.  Will the RH section grow like the LH section did?  And will that tell us anything?

Steaming or Cooking?

About now I had an attack of Gung-Ho, and decided to really push the limits.  It was probably a reaction to the profound under-excitement of Plans A and C.  Instead of judicious use of steam (is there such a thing?), I decided to pressure cook the section.  As you probably know, a pressure cooker heats steam up to about 120˚C by approximately doubling the normal air pressure, taking it some 20˚C above water's normal boiling point at sea level.  This should be interesting....

The Rake of Spuds

While we wait for the pressure cooker to come up, just time for an aside on matters cultural.  Here's the quick'n'tasty way to produce that essential Irish-Australian culinary masterpiece, the Rake of Spuds*.  Wash a dozen or so small new-season potatoes and cook in their jackets for 11 minutes in the pressure cooker (or until just nicely cooked).  Unload into a warmed bowl, smear with a little butter, season with salt and freshly ground pepper, sprinkle with chopped parsley and consume with the fingers.  Useful these pressure cookers, eh, and not just for cooking flutes!

*I think I first heard this name for this dish from "Little Joe" Conlon, who has also been the source for a few good songs.

Back to the flute....

Hoping for quick results, I decided to check after just 10 minutes.  Probably a good thing!  The graph below records what happened.  It's important to note that the horizontal axis is not to scale.  If it were, it would be hard to see what was going on as the cooking intervals would be negligibly narrow blips.  So remember to read the X axis time intervals carefully.

The measurement parameters are:

  • Weight of the section (in navy)

  • the maximum and minimum diameters of the socket (mustard and pink), and

  • the maximum and minimum diameters of the tenon (aqua and purple).

Of these, the tenon diameters are of the most interest, as we hope they will increase substantially to shuck off all those years compression.

You'll see I've normalised all the measurements and expressed them as percentages, so they can be read on the same scale.  As we start, column 1, they all stand at zero.


After only 10 minutes of steaming in the pressure cooker, it's clear that action is indeed rapid.  Incidentally, I'd gone a bit soft and only used the "vegetable" weight, rather than the full "meat" regulator weight on the cooker.  So we've probably only been up around 110˚C.  I may need to use that in my defence!  The weight has only gone up by 2% (not bad in itself) but the dimensions have risen by 4 to 6%!  I decide to give it one more blast.  After a further 10 minutes, I call it quits.  Weight now up 4%, but dimensions from 6 to 8%. 

Interesting to compare that with the Heavy Humidification approach we had applied to the LH section, as shown in the graph replicated below.  It had taken 10 days to achieve dimensional changes up to 8%, and it required a similar increase in moisture levels.  Here we've achieved the same thing in 20 minutes, with only half that moisture intake!  The power of steam!

But it's not without impact on the flute section.  Potter applied a varnish to the outside of his flutes, and this section was now looking considerably darker than the rest of it. Not a big deal in my case, as a lot of the varnish has been scratched off over the years, but probably not good to do to a nicely varnished flute.  And where the varnish is missing, the grain has been raised, so not good to do to nicely polished flutes unless you are prepared to refinish them.  Which is possibly going to cause problems in terms of preserving the maker's mark!  Heh heh, thinking about it, what's the value of a maker's mark when the operating conditions he'd carefully set up are lost?

The high temperature also far exceeds what superglue can live with, so the cracks I had closed up earlier were now open wide.  That probably casts some doubt on the precise validity of the socket measurements from here on, but the tenon measurements are uncompromised, as well as being of more interest to us.

Airing, Drying, Airing

Getting the moisture back out takes a lot longer!  I let it air on the bench for 6 days, before giving up on the natural process.  It was behaving just like the LH piece - refusing to drop to the original weight or size.  I decided to give it a downward nudge and so popped it into the desiccator for three days.  That pulled the weight well below the original value, but didn't have that much effect on the dimensions. 

Airing on the bench over the next two days, the weight returned to original, but the dimensions remained on average 5% larger than where they had started.  Just to be sure, I waited 6 more days, with no further change.  This bird has landed, but its feathers are still up the tree.


So, time now to measure up the RH section carefully.  And what do we see?  Something truly awe-inspiring!  You can see it below, in thick green, at the middle of screen, to be compared with the thick navy line which represents the dimensions as found originally.  Five things are immediately evident:

  • the compression just after the socket is gone.  Indeed there's now a little flare where once there had been compression.  That's possibly due to the opened crack, and can be corrected by closing the crack and installing a replacement ring.

  • the compression at the tenon has disappeared almost entirely.  This is stunning news - the very question that got us into this whole business appears answered!  Can it be this easy?

  • the compression just above the tenon is much ameliorated.  It seems to suggest that it was indeed an artefact of the tenon compression.

  • the once bent bore now reasonably approximates a straight line.  Remember Dr. Bigio's observation that reamers of the period are unlikely to involve compound tapers?

  • the whole section is now larger, by an average 4%.  Remembering that 6% in pitch terms is a semitone might help put it in perspective.

This is a cracking result!  The test tenon in the first series had proven that thread could crush the tenons, but also gave us the hope that enough residual memory might remain to auto-reform, if the binding were removed and the conditions for sufficient wood mobility were met.  The heavy humidification process in Plan B met some of those conditions, but not enough to reconstitute fully.  But just twenty minutes in the pressure cooker reversed 200 years of oppression!

And does this invite us to propose a yet more radical scenario, "Join the ends", as illustrated by the thin blue trace in the graph above?  Fits nicely with the newly reconstituted RH.  Aligns well with the largest part of the reconstituted LH.  But only touches the so-far-untouched foot at its distal end.

Looking back though, the "just flatten the tenons" scenario suddenly looks very meek!

A step too far?

So, a 4% increase in diameter?  Is that something we want, or something we can live with?  Has pressure cooking expanded the section back to where Mr Potter intended, or beyond that into uncharted waters?  Could this even happen? And how would we know anyway?

One disturbing indicator is the socket on the foot.  It can no longer accept the enlarged tenon at the end of the RH section.  But is that a argument against the enlarged tenon or against the shrunken socket?  This socket too had lost its ivory ring sometime back, to have it replaced by one crudely fashioned in brass.  Like the RH socket, this cylindrical ring had also been hammered on over a tapered section of wood, compressing it.  Not much evidence left there, Inspector.

A nervous speculation

Imagine this.  A timber-getter cuts down some boxwood in the forest.  The trunk is cut up into lengths and riven lengthwise to prevent splitting.  The timber is stored under shelter to season.

Water evaporates from the surface, causing it to shrink with respect to the wetter timber inside.  Pressures build up.  If the evaporation rate is too high, the timber will split.  If the people running the seasoning have managed things well, it won't, just.  Slowly, slowly, moisture migrates from the damp interior through the drier exterior and to atmosphere.  The dry exterior band gets thicker and thicker, leaving a diminishing core of damper wood, under increasing pressure.  Slowly slowly, the damper core dries out until all is close to moisture equilibrium.

Mr Potter buys the wood and bores and reams out its interior.  The outer timber is still exerting pressure inwards.  Then he turns off the outside, leaving walls just 4mm thick.  The thin remaining wood is neither inner or outer wood, but from somewhere in between.  Will the locked-in pressure be inwards or outwards, or nicely balanced?  Maybe it doesn't matter to Potter - either way, the timber is dry now and so has little mobility. 

Supposing Mr Potter had popped the freshly turned flute section into an early pressure cooker right then and there, then let it air dry again?   Would the diameter have shrunk or grown, or stayed where it was?  Is it possible that we've just released pressures that originate right back when the timber was first seasoned, as well as releasing those pressures added later by overtight ring and tenon threading?

Ask the family!

Ah, but we're forgetting something aren't we!  We've forgotten to Ask the Family whether this new enlarged RH section, and indeed, the slightly less enlarged LH section, still fit within family expectations.  Let's add these new dimensions to the graph we developed back in Musings.

Ah, now suddenly that doesn't look so outlandish, does it.  The pressure-cookered RH (thick brown) extends nicely from the largest of the three corps de rechange from the Bate flute and is close to if just a little larger than the Bate flute's RH section.  The humidified LH (thick green) approximates the Bate flute at the middle but is still very compressed at and near the tenons at both ends.  But, if no longer outlandish, it's still pretty staggering!

The power to crush

Back in Series 1, we went to some pains to illustrate the kind of forces that thread wrapping could set up.  And, it has to be said, many readers struggled to believe.  I'm hoping that this experiment will put that behind us forever.  Water dripping on a stone may not produce much effect over the period of a minute, an hour, or a day.  But, 10 or 200 years later ....

Where to now?

So have we found the way of the future?  Just gather up all the old flutes and slap them in a big pressure cooker for 20 minutes to enable them to recover their lost youth?  Or can we come up with a more gentle regime that will achieve the same thing, or at least achieve enough that we can honestly say this is the flute the maker intended and not some mangled travesty of it?

We certainly need to do something still with the tenons on the LH section, and the foot of the flute isn't much good to us if it won't go on the end!  Can we use them to find and prove that more gentle regime?


Well, wasn't that stunning!  We appear to have actually achieved what we set out to do - enable at least one part of the strangled flute to auto-restore to its original shape, and assert on the way what its original shape was.  But, if that is its original shape, the rest of the flute has still some way to go to match it.

And if that is its original state, it was a lot more strangled than I had thought.  Remember I'd chosen it and characterised it as "moderately strangled"?  So just how strangled were those severely strangled flutes that got us in to this whole investigation?

But, this is no time for hubris.  We're not out of the woods yet!

On to A Test Tube, or

Back to McGee-flutes Index page...

  Created 22 June 2011