Measuring the bore profile of a conical flute

Irish Flute
1

I have a couple of nice antique flutes that I’d like to measure to better understand their playing characteristics, and I’m wondering how to go about getting an accurate profile of the bore. I’m familiar with the use of telescoping gauges and micrometers to measure the ID at a random point for cylindrical bores, so my concern is not the precision of the ID measurement itself (the Y axis of a bore profile graph), but precision along the bore (X axis). How do you ensure that you are measuring the ID at a precise given point along the bore (say 10 mm, 20 mm, etc.)? Are there telescoping gauges that are long enough for this task? The ones I’ve seen for the diameters of interest seem to be about 10 cm long, which would not even get me halfway of the LH section of the flutes I want to measure.

Any suggestion would be greatly appreciated. :slight_smile:

2

Good on you, Flutern, for delving into the dark chunk of the space-time continuum we call the flute bore. Keep in mind my old quip, flute acoustics isn’t rocket science. It’s much harder…

Not aware of long-handled Telescoping gauges, though they might exist. And you’re right, they’re not much good to us as they stand. Here’s what I do.

Firstly, I put marks every 10mm along the handle of the T-gauge shaft, with the first of them in line with the two sprung jaws. So now I can insert the gauge any desired distance into the bore (until I’m running out of handle), or determine how far into the bore to see where say 16.7mm is, by estimating within the 10mm marks, EG 67mm. If you want more precision, you could measure the gap from the end of the bore to the 70 mark.

But then I run into the problem of not enough length. A Prattens body for example is about 320mm long. Being able to measure say 90 in from each end leaves a lot of unexplored territory! So I took some white Delrin or similar plastic rod, 260mm long and 10mm in diameter, faced off the end, and drilled a hole that would allow it to just accept the knurled knob at the end of the T-gauge. And I drilled the hole just deep enough so that the end of the rod was 100mm away from the line between the jaws. Then I continued adding marks along the white rod to extend my measuring capacity to be able to handle the longest flute sections I was likely to encounter.

I can use this setup in either of two ways:

How far in is 15.6mm?
Set the jaws to be 15.6mm using a micrometer. Insert T-gauge into bore until it stops. Note the depth of penetration. Or…

What is the bore diameter at 150mm penetration?
Lightly tension the gauge so it just springs open. Introduce into bore and slide it down to the 150mm mark. The jaws will keep compressing as you go. When you get there, twist the handle to lock the jaws. Confirm that worked by feeling for where the gauge now stops, it should be the 150mm mark. Withdraw and measure jaw width.

Come back with questions if I haven’t made that clear enough!

3

Dasqua make a set of long reach telescoping gages which go a little over 12 inches and cover the diameter range you will need. Penn Tool Co sells them and there are probably other sources as well. The Delrin extender described by Terry also works well and is what I used before I got the long reach set.

4

I remember visiting Rod Cameron’s shop many years ago and what he told us was fascinating.

He made Baroque flutes, high-end instruments, played by some prominent flutists.

First he explained his philosophy, which I got the impression wasn’t his alone but common in the Baroque flute world.

That is that we can’t hear with 18th century ears. We’re not 18th century players. We can’t know what timbre 18th century audiences expected, or what performance characteristics 18th century flutists demanded.

The closest we can come is to find the best-preserved 18th century flutes and play them. Or, since many are in museums who won’t allow such, make the most painstakingly accurate copies we can and play them. Play them, and let the instruments tell us how they want to sound, how they want to be played.

As opposed to modern makers being “inspired” by original antique instruments but not hesitating to “improve” them to make them sound and perform the way modern flutists want.

Since most of the best surviving instruments are in museums, you can’t go around shoving tools up their bores. In any case that won’t tell you anywhere near all that you need to know. As you all may be well aware, the bores of Baroque flutes are quite complex, because after being reamed on the lathe the makers went to work on the bore with specialised little files, making subtle concavities at various points. These are necessary because most of the chromatic notes were achieved through cross-fingering, so most of the finger-holes were responsible for several different tasks, for example Hole 3 would need to make Low Octave A, Low Octave B-flat, 2nd octave A, 2nd octave B-flat, as well as being the vent-hole and/or tone-hole for one or more 3rd octave notes.

How to measure all these bore variations without being allowed to touch the flute? Rod Cameron (who as I recall had an engineering background) created a probe thing that could be passed up the bore, measuring it quite precisely. It was connected to some kind of electronic device that stored the data and allow his lathe to re-create the bore.

Thus each flute he offered in his catalogue was an exact copy of a specific original 18th century flute. Tests performed using original 18th century flutes in the hands of top players showed that indeed his flutes sounded and played like the originals.

5

For most flutes you can get all the readings you need using the following two Starrett Telescoping gages with 12" handles:

579A-12 Telescoping Gage: 5/16 - 1/2"/8-13mm Range, 12"/300mm Handle Length
579B-12 Telescoping Gage: 1/2 - 3/4"/13-19mm Range, 12"/300mm Handle Length

I actually have the next larger one too (579C-12) but I only need it for flutes in lower keys than D.

I take measurements at regular intervals along the bore (either every 10 mm or every 5 mm, depending on how detailed I want to be).

The reference point for my measurements is the start of the bore (head end of the upper body section), and this is where I insert the gage. The shaft of the gage can be marked with the relevant distances you want to insert.
The section of the bore from that point to the embouchure is generally cylindrical (ignoring for now the steps due to tuning slide), so it can be simply measured at each end. Its length varies depending on the tuning slide extension, so I keep that part of the data separate and only really use it when making a new head for a flute, perhaps with some head length correction for playing at modern pitch with less tuning slide extension (and hence less of a tuning slide cavity in the bore).

You can profile each body section and foot separately (inserting the gage from the wide end) and then measure lengths of the sections and do the calculations needed to convert your measurements to distances from the reference point mentioned above. Then you can graph the bore as a whole. But make sure you remember to take into account the socket depths and tenon lengths correctly, and verify that the tenons are the same length as the sockets they fit into! Sometimes they are not!

Most antique wooden flutes have experienced movement due to shrinkage, and this results in the bore being oval in places. You can test for, and account for, this by taking widest and narrowest measurements at each point and by averaging the two. Acoustically, it is the cross sectional area of the bore that is most important. This is one advantage of using these telescoping probes vs pre-machined cylindrical probes. A cylindrical probe can only measure the narrowest dimension. Of course, you can’t quite do this perfectly at the location of the tone holes.

By taking measurements in this way and using them to carefully construct reamers you can make quite precise copies of antique flutes, which play like the originals. Not in their original state, of course, but in their current state, which is sometimes what you want.

Rod Cameron used similar techniques to this, and I learned quite a bit from his posts years ago, but he also made a device for scanning the interior of a bore without the need for metal on wood contact, which in theory risks damaging the flute. In practice, though, I’ve never noticed any damage to the bore from using the kind of gages linked above. They have very smooth, rounded ends, and are spring-loaded with fairly weak springs. You lock them off before removing them, but so long as the bore is not actually chambered (containing a local maximum between two narrow points) removing them to the wide end is simple. If it is chambered (and not just at the point where sections meet) you can’t get accurate measurements of the chamber width using these gages because you lose the measurement in order to extract the gage.

6

Reportedly, Vincent Broderick won the All-Ireland flute competition in 1953 playing a copper-pipe flute that he had made himself. Now I’m wondering how that sounded and if anyone ever measured the bore.

Best wishes.

Steve

7

Thanks everyone, I really appreciate your detailed answers and I now have all the information I was looking for :slight_smile: The Starret gages look really nice but are a bit pricey. The Dasqua long reach set is more affordable and is readily available from amazon.ca, so I’ll probably go for that.

8

Those Dasqua gages look like a clone of the Starrett gages and probably do exactly the same job. You also need a good set of digital calipers to take the measurements from the gages once removed, of course.

Which flutes are you measuring?

9

I do have a good set of dial calipers already :slight_smile:.

Which flutes are you measuring?

The flutes I’m most interested in measuring are two Rudall & Rose (#39XX and #49XX) with medium and large holes, respectively. (The RH section of the later one is a replacement, most likely from Rudall, Carte & Co based on what’s left of the stamp.) I also have a very nice, early flute by Joseph Wallis, with small holes and a fairly narrow bore. It’s not really suited for ITM but it has a beautiful voice and plays really well.

10

I wonder why the makers of these gauges don’t put some calibration on the handles? Or are we the only people who want to know where the bore is 15.3mm, or what is the bore 150mm in?

You can get round this by putting a mark on the shaft and then measuring with calipers or rule between the mark and the line between the jaws. But it’s another step.

11

The hardest part is not the measurement, it’s interpreting the data. What of these measurements were intended by the maker, and what are the subject of damage over the nearly 200 years since?

Now, do be careful, Flutern. Very careful. Delving into the dark innards of a wooden flute has often led to thoughts crossing the mind such as “could I make one?”. And then that leads to thoughts such as “could I make one better?” When you find yourself starting to flick through lathe catalogs, it’s probably too late…

Heh heh, 50 years ago I visited England, Ireland and Scotland, to research what was going on in traditional music in those countries. A few days before we left to come back to Australia, we went once more to Portobello Rd markets, and I found myself paying a lot of attention to a lathe for sale. 6 months or so later, I was making flutes.

Be warned…

12

Yeah I wondered about that. But for now I’d just like catalog my small (but growing :smiley: ) collection of antique flutes. We’ll see how far curiosity will take me.

Now, do be careful, Flutern. Very careful. Delving into the dark innards of a wooden flute has often led to thoughts crossing the mind such as “could I make one?”. And then that leads to thoughts such as “could I make one better?” When you find yourself starting to flick through lathe catalogs, it’s probably too late…

I can’t say I haven’t been warned :laughing:

13

How robust are these flutes? You can use silicone rubber to make a mould of the entire bore of the flute and then make copies from that mould and take as many measurements as you like from it with ease. If you just cover the holes and fill the flute with silicone rubber solution, you’ll need to apply high forces to extract it, and that could rip chunks of wood out or even snap it in half if it’s fragile, but the simple trip is to coat a rod in plastic film (I use a strip of film wound round it in a helical way with sellotape sealing the joins) and then insert that into the flute so that only a relatively thin layer of silicone rubber needs to be poured in between rod and bore. You then pull the rod out (it will slide out easily leaving the foil behind - the layer of plastic foil is essential as the friction of the silicone rubber won’t allow the rod be pulled out otherwise). Once the rod’s out, the foil can be pulled out with ease due to the space vacated by the rod. You can then work carefully to detach the silicon rubber mould from the bore of the flute by gently prising it off with something like a knitting needle inserted between mould and bore, though you’ll likely need something double the length. So, instead of trying to pull the mould out, you push each part of it into the space where the rod had been, and you can use silicone spray to help the mould stay detached and free to slide when you gently pull the thing out after it’s fully detached from the bore surface. Most of the work in detaching the mould is done simply by rotating the “knitting needle”, only gradually pushing it further in as it goes round, so it’s the rotation that does the detaching. The bore only needs to be sufficiently robust not to be damaged by the “knitting needle” being pressed against it and rotating with very low friction, further reduced by the silicone spray - done the right way it’s a very gentle process. The flute would have to be in very poor condition for that to do any damage to it. You can try this method out on modern flutes first to get a feel for how it works. Silicone rubber solution for making moulds comes in various grades depending on stiffness, so you need to use one of the more flexible ones. I use this type: https://www.amazon.co.uk/dp/B0CWV152N4/?coliid=I3KJDE209YP28R&colid=1LLVJM66J1P5&psc=1&ref_=list_c_wl_lv_ov_lig_dp_it. Once you’ve extracted the mould, put the rod back into it before trying to make measurements from it.

If you want to use the mould to make another flute from, there are a couple of ways to do that (you can make an outer mould and pour craft resin (epoxy) into the space between the two moulds, or you can rotate the horizontal bore mould with a motor (8 revs per minute) and gradually apply the resin to it in layers to build up the thickness over a number of days. Either way, you end up with a very robust copy of the flute made entirely out of resin. Again you need a layer of film between the rod and bore mould so that the rod can be extracted - the friction is extreme otherwise and you’d have to destroy the flute and likely damage the mould in the process before you can get the rod out.

Anyway, if the antique flute is still playable, it should also be sufficiently robust to make a mould from in this way.

14

David, what would you say is the level of accuracy using your silicone rubber mould approach? First you’d have to take into account the thickness of the film and tape layer, which presumably isn’t uniform. This may not present too much difficulty, but you also have the flexibility of the silicone to account for, including how much it compresses or expands when making the initial mould and how much it compresses when you try to measure it using calipers, say. Would you use calipers to measure it afterwards, and if so how accurate is that?

The ultimate test is whether the dimensions of a copy bore made using this method match the original, and how closely, but how would you take measurements to answer that question? It is tricky given that in this game even very small dimensional changes, of the order of 0.1 mm are quite significant.

15

If you need that degree of precision, the film and tape could become a problem as it’ll vary a little - it’s usually damaged when extracting it, but even if it can be reused you’d be hard pushed to line it up again perfectly to use it again.

So, it might be better to use multiple narrower rods so that you can use them without the film and still extract them by removing the fat middle one first, then each of the rest can be pushed into the middle to extract them. When making the silicone mould you would have to seal the outer rods where they contact each other in a way that stops the silicone rubber getting past them into the zone where the central rod has to remain free to slide, but without any of the sealant pushing the rods any further apart. When using the mould afterwards it could then simply have all the rods put back inside again without any film in there and without having to seal the outer rods - just spray some silicone in to help them slide so that the mould is free to move and take up its natural shape. The mould would then have perfect shape every time and its dimensions should be sufficiently precise that you wouldn’t detect any differences between original flute and copies.

(It might be possible to do this with seven rods of the same diameter - once the central rod is removed there’s likely enough give in the silicon rubber to enable the first of the outer six to be prised past its neighbours into the middle, and the rest would all fall out easily after that, but that would need to be tested first to make sure it works. If the outer rods need to be narrower, you’d likely have to make a custom central rod to the exact diameter that enables all the rods to touch it and each other with no gaps. I think it should work with seven rods of the same diameter though, because as soon as you start to move one inwards into the central space, you’re creating more room for the rubber to bulge into as the neighbouring rods compress it. Come to think about it a bit more though, if one of the outer rods is made of two halves split lengthways, you could with correct alignment extract half a rod without any compression of the rubber at all, then the other half and all the other rods would pop out with ease, so I’d look at doing it that way from the start. For sealing the rods, you could simply run candle wax down the inside, then any that gets in the way of the central rod when that’s inserted will just grind off.)

The silicone rubber is free to move while it gradually solidifies, so it ends up with the right dimensions and without any stresses built up in it. When making measurements, so long as you don’t press it hard with the calipers, those measurements should be as good as making them directly from the inside of the original flute, only much easier to access the places you want to measure.

16

Hi all
You can get a fairly accurate picture of the shape of the inside bore of a flute by measuring the inside diameter at the centre of each of the tone holes. You will need a Vernier callipers (with depth gauge). I’m assuming we have a three-piece system flute. The tone holes are in the main (middle) piece. I’m also assuming that the inside bore of the flute runs central to the outside surface, that it is not off-centre.
Anyway, try this …
Measure and note the inside and outside diameter at both ends of the piece, and the length of the piece. Draw it.
Measure the distance between the wider end of the taper and each of your tone holes. Add this to your drawing.
Measure the outside diameter at each hole (OD), then measure the distance from the outside surface of that tone hole to the opposite inside surface of the bore (ID+WT) using the depth gauge on your callipers. (ID stands for inside diameter. WT is wall thickness.)
Use the following formula to determine the thickness of the wall (WT) at each tone hole ( ’ - ’ stands for minus): OD - (ID + WT) = WT.
Use the following formula to determine the diameter of the inside bore (ID) at each of the tone holes: OD - (2 x WT) = ID. Add this to your drawing.
With the measurements of the the inside bore at each tone hole, you can draw the taper and therefore create an accurate picture of the the bore profile. I suggest you draw the lengths at scale 1:1, and diameters at 2:1, or more. This will make it easier to see how the inside bore is formed. Is it straight and regular or irregular etc. I use three different tapers in the inside bore of my flutes. It has let me bring the tone holes closer together.
Tim Adams
Irish Flute Maker

17

I think I remember reading something about this method before here on C&F, but wouldn’t the ideal method be an X-Ray image of the flute bore? Non-damaging for valuable antique flutes, and the dimensions would be easily calibrated knowing the length of the flute in the image.

The problem of course being access to the kind of hardware that could image something that large, not your local dentist’s machine. But isn’t this technology used in industrial processes for quality control? Maybe you just need the right connections?

18

Unfortunately, only have the Patent head of this Rudall - not even a conical section.

(Not helpful, I just like looking at the image.)

19

One of our Australian flute/pipe makers, Ian Mackenzie, was a radiographer by the day and did an image of one of the old flutes. Turned out to be of no help in measuring diameters. It was hard to discern where the edges of the bore really were, and, as the image was about lifesize, your precision was about that of measuring off a drawing. While our precision with Telescoping gauges is about 0.01mm. Which then brings the risk of hyperventilating over minute details!

Other people have made up probes that involve having washers, or coins or something similar mounted on the end of a rod. These have the disadvantage in not being able to measure ovality, whereas with the Telescoping gauge, you can explore maximum and minimum axes. Did I mention obsessing over minute details earlier?

And I’ve been trying to claw back out of Long Term Memory (a dark and dangerous place at my age) what Rod Cameron’s electronic bore scanner used as a diameter sensor. The only clue so far come to mind was that it employed an LVDT (Google it). Which suggests that there was still bore contact, though that could be very light. Somewhere here I have copy of one of his scans. It looks like the output of a chart recorder, which would make sense. You set the machine in motion and it starts to move down the bore. At the same time, the chart recorder starts to issue graph paper, and its pen moves sideways, tracing the bore diameter. Think earthquake printouts.

Telescoping gauges that cover the smallest bore up to the biggest socket plus Tim Adams’ suggestion of measuring through the finger holes would go a long way for little expense. That would give plenty of detail around the tenons, where we have to try to work out what is original and what is damage or intentional change, and a few points along the middle to see what is the general trend.

20

Building on Terry’s post: The problem with x-rays is two fold. The flute is made from wood, which is mostly carbon. The vast majority of x-ray facilities are sensitive to heavier (higher atomic-number) elements. The pic of the Rudall-Rose patent head that Kevin posted actually contains the wooden part of the head, but you can only see that in the right-hand part of the image, away from the metal that dominates.

The second is that x-ray instruments just make shadow graphs, so the resolution is of order 1 mm, and you need about a tenth that for a good starting point to make a flute.

If someone has a few hundred k lying around, you could go to a synchrotron-radiation facility and use an x-ray imaging beamline that’s more sensitive to carbon. Getting a picture of a flute bore would probably only take 4-8 weeks. (I spent a week at such a facility and we got an image of about 0.1 cubic mm of a computer chip, but that was 100 nm resolution.)