Blowing machine

[Tunborough]

So, my suspicions about the manometer sluggishness seem to be borne out. Thoughts?

I tried plotting the new numbers, the original numbers from that calibrator, and the numbers from the 10 x 5.1 + 20 x 4 calibrator (because it had 6 data points). The new numbers, other than the first data point, have a pretty-much linear relationship between flow and pressure. The other data sets showed the square relationship, but one that would bottom out before it got to (0, 0). For the new data set, you started low and increased. For the other data sets, did you start high (40 L/min) and decrease? If yes, then that does suggest the manometer is lagging. A drop of dish soap is in order. I don't think it would affect the density of the water enough to change the calibration. Do you give the manometer a tap before you read it, in the manner of old sailors reading mercury barometers?

[Terry McGee]

OK, back again after a few good tunes. Ah, maybe that's why we're going to so much trouble!

So, I resolved to get rid of the current blue liquid and replace it with plain water, which I did, after a good deal of flushing to make sure. Same reluctance to start moving, so I gave that away immediately. And introduced a good sized drop of concentrated dishwashing liquid into the open end of the tube. Then pressurised the manometer till they conjoined, then wiggled the flow knob back and forth to mix them. Hmmm, tiny bubbles visible rising to join the miniscus. Poured them off and added some fresh water to make up. And did it again. Looking better. Then took this data series:

Code: Select all

30 x 4mm Calibrator, water + detergent					
Mano/2	MM(H20)	√A/P	Flow	Resistance	Gauge used
0	0	0.0	0	0.00	5L/Min
1	2	1.4	2.5	0.33	
2	4	2.0	4.3	0.47	
2	4	2.0	4.5	0.44	20L/Min
3	6	2.4	6.3	0.39	
4	8	2.8	7.3	0.39	
5	10	3.2	7.9	0.40
6	12	3.5	8.5	0.41	
7	14	3.7	9	0.42	
8	16	4.0	9.6	0.42	
9	18	4.2	10	0.42	
10	20	4.5	10.8	0.41	
20	40	6.3	15.3	0.41	
30	60	7.7	18.9	0.41	
40	80	8.9	22.7	0.39	
50	100	10.0	24.8	0.40	
60	120	11.0	26.7	0.41	
70	140	11.8	28.6	0.41	
80	160	12.6	30.6	0.41	
90	180	13.4	32.6	0.41	
100	200	14.1	34	0.42	
144	288	17.0	40	0.42	

Looks much better to me, but I'll leave the experts to judge. I do note a little discrepancy where I change from the 5L/Min to the 20L/Min gauge but we were aware of that from an earlier experiment.

Are we there yet?

Oh and a note for anyone silly enough to want to make one of these systems. Make sure you put all the meters up at convenient eye height. Mine aren't!

I am experimenting with better lighting, having a magnifying glass standing by, and possibly mounting the rule so it comes halfway across the front of the manometer tube, rather than to the side of it. I'm thinking it might assist in taking measurements. Jury still out on that one.

And have to decide whether to reintroduce the colourant, as it doesn't look like it was the cause of the problem, just the surface tension of the water. If it's enough to let a water spider walk on it, it's enough to cause issues in a narrow tube....

[Terry McGee]

Mine doesn't have the ivory beak or mouthpiece, so here's an image of a similar one that does:

That photo comes from a YouTube demonstration of the depicted instrument by jemtheflute. He plays it as he would a tin whistle for ITM, here. As might be expected, it has a sound of its own, but it certainly suits the musical purpose and Jem doesn’t seem to be having the slightest difficulty getting it to do so.

I understand that you are supposed to have some sea-sponge in the windcap to absorb the moisture rather than it clogging the windway. I don't know if that's just a liner, or a blob through which the wind passes. Possibly more resistance?

The sponge is normally an annular bit held clear of the direct wind path. Nonetheless, the geometry and aerodynamics between the entrance to the mouthpiece of an English flageolet and the windway exit are more than a just tad complex. Streamlining them would surely have been a consideration, alongside reducing production costs, in the derivate keyless flageolet that is the focus of our current attention.

(Anyone interested in the parallel with keyed and keyless flutes who didn’t follow the YouTube link might find it worthwhile to click now. Jem uses the keys on his flageolet to good advantage and their subsequent abandonment arguably entailed the loss of a useful technical resource.)

I'd heartily agree, stringbed. That's a lovely instrument, and Jem's doing a great job with it. The tune is "The Old Pigeon On The Gate" and I can play along on a D whistle. It would be interesting to ask him if he is a) aware of increased resistance, b) feels hampered by it or even, c) likes it!

I thought once we sort out the current issues with the manometer etc, it would be interesting to make up a replacement beak for mine, and test the contributions in resistance terms of the windway, the windcap tubing and the beak.

[Terry McGee]

Now I'm aware I haven't answered some suggestions from others, but I thought the new data from "water + detergent" might need to be assimilated first. Let's see what we think about that.

[Terry McGee]

. . . Oh, yes, flutes, I remember now. . . Whistle makers, I don't know....

Mr. McGee, good sir,

May I make 1 more request before you clean-up/close-up on the great whistle-flow-extravaganza ?

The request: using the pitch-meter, please list out the mmH2o+lpm for the Gen as it hits the notes going up the scales from D5-D7 ? The needle-valve on the flow meter might help.

I'm not worried about the up/down hysteresis. Going up is fine.

I'd be most grateful.

I'll be very happy to do this, trill, but I'd like to clear up the hysteresis/sluggishness/whatever-is-going-on issue first, as I think it is messing with us. But in the meantime, let me make sure I understand correctly - you want me to indicate pressure and flow for when each note in turn hits its correct pitch when set to its correct fingering. And if perchance I can't get a particular note up (or down!) to correct pitch, note how close I can get?

Now I have three Generation models (I think). The one I call Old has the remnants of a ridge down the middle of the head visible front and back. The one I call Newer doesn't have that. They are both red heads. I also have a green one, which appears to be identical to the Newer one. I haven't a clue how old they are! Your preference?

[stringbed]

That photo comes from a YouTube demonstration of the depicted instrument by jemtheflute. He plays it as he would a tin whistle for ITM, here. As might be expected, it has a sound of its own, but it certainly suits the musical purpose and Jem doesn’t seem to be having the slightest difficulty getting it to do so.

I'd heartily agree, stringbed. That's a lovely instrument, and Jem's doing a great job with it. The tune is "The Old Pigeon On The Gate" and I can play along on a D whistle. It would be interesting to ask him if he is a) aware of increased resistance, b) feels hampered by it or even, c) likes it!

Jem may already be aware of our discussion but I’ll send him a PM to make sure.

I assume that the point of comparison for the question of whether “he is aware of increased resistance” would be a tin whistle — but of which manufacture? We’ve been using Generation whistles as a benchmark for the present exercise. But they don’t similarly typify, much less provide a normative point of reference for, the playing characteristics of the range of whistles one sees in the hands of performers.

Overton whistles and their spawn require far more measurable blowing pressure than Generations do, with other commonly encountered varieties forming a broad intermediate spectrum. It also remains to be determined how well players’ subjective assessments of resistance rank against values calculated from volumetric air flow and manometric air pressure.

[Terry McGee]

Indeed. I guess a starting point with Jem would be asking what his favourite whistle is, and how the flageolet compares to it. Or perhaps if he has something "plain vanilla" like a Generation, a Mellow D, a Feadog Mk 1 that we have access to, how it compares with that. (I do have a later Feadog which I could measure too.)

Assuming we can get past this current manometer issue, we can hopefully go on to agree on a Resistance measurement approach that anyone appropriately equipped could use to put a figure on an instrument's resistance. And then build up a list of instruments in Resistance order. I know we've survived in the past without it, but I think the more we know, the better are our purchasing decisions. And with such choice nowadays, and little opportunity to try before you buy, it's all got to help. Which is why I'm experimenting with different bores and windways, looking for what suits me!

[trill]

. . . pressure and flow for when each note in turn hits its correct pitch when set to its correct fingering

Exactly ! Within 20c is plenty good.

. . . Now I have three Generation models (I think). The one I call Old . . . Your preference ?

Same one as used previously.

Bless you + your "first-in-the-world" test-bench !

[trill]

. . . Overton whistles and their spawn require far more measurable blowing pressure . . .

Indeed they do !

That's one reason I love them !

Maybe we could all pitch in and buy one for Terry ! That way he could: a) enjoy a great whistle, and b) send us the data !

[trill]

Also, on the subject of density, here's a comparison of Nitrogen and Oxygen . . . If I'm thinking clearly, I think the lower density of N2 would make the "indicated" flow rate (if calibrated for O2) less than the true flow rate.

I think I have this worked out. It's late. I'm tired. But I think the true flow (with air vs. oxygen) will be higher by 0-2.5 l/min, depending on flow rate.

I assumed a linear indication on the flow-block: 0-20l/min over 15cm.

Terry, could you send a snapshot of the flow-meter you're using ?

My curiosity is driving me nuts . . .

[Terry McGee]

Now, something that has crossed my mind (it doesn't take long!) is wondering what interests trill and david_h have in this question. Are you making, thinking about making, just interested in the physics, or is that TV in your area is really boring?

I'm assuming that my and Tunborough's interests are well enough known, and that stringbed have made his background and interests also clear, but let's not rely on those assumptions. Ask or tell!

And if there are any lurkers out there watching quietly, let us know your interests, so we can do our best to fulfill them. This is an unusual journey by any standard, and we want everyone to enjoy and benefit from it.

[Terry McGee]

Also, on the subject of density, here's a comparison of Nitrogen and Oxygen . . . If I'm thinking clearly, I think the lower density of N2 would make the "indicated" flow rate (if calibrated for O2) less than the true flow rate.

I think I have this worked out. It's late. I'm tired. But I think the true flow (with air vs. oxygen) will be higher by 0-2.5 l/min, depending on flow rate.

I assumed a linear indication on the flow-block: 0-20l/min over 15cm.

Terry, could you send a snapshot of the flow-meter you're using ?

My curiosity is driving me nuts . . .

Here's a nice image of a 10L/Min one. Mine look the same but are 20L/Min...

[stringbed]

I hadn’t anticipated cantankerous measuring devices becoming quite the issue they have. The upbeat to this discussion was a desire to refine the calculation of the velocity of air flow at a whistle’s edge from the harvested data. My own greater interest is finding a way to quantify resistance, so I’ve said little about the direct measurement of that velocity beyond noting that it would likely prove inevitable. Now may be as good a time as any to delve further into it. I can already hear a collective groan at the mere thought of a third point of measurement but that’s not what I’m suggesting.

The question is whether volumetric tracking can beneficially be replaced by the direct measurement of air flow. Positioning a miniature hot-wire anemometer at the edge of a whistle strapped into the Test-a-Tron Mk III ought not to be a prohibitively daunting challenge (to which I would consider contributing financially) and it would be interesting to see if it can ease us past the present hurdles. There’s a good intro to that type of measurement here. Several other manufacturers produce appropriate transducers, with a detailed description of a useful range here.

[david_h]

Are you making, thinking about making, just interested in the physics, or is that TV in your area is really boring?

Short answer - it's interesting and I like a puzzle.

Longer answer. I am a flute player thinking of getting back into whistle. Learning flute was greatly helped by reading about the physics (Robin Jakeways, UNSW etc). I find whistle much harder to play than flute because of the need to hit the pressure just right for the pitch of each note, and the 'just right' varying so much between notes and whistles. Better quantitative description would help. I might be in the market for a whistle but I may stick with Feadog Mk1 head on a Generation tube which - from the state of the lacquer on the tube - was what i spent most time on a long ago, or the Silkstone.

As student I spent an afternoon a week for two years doing university physics department practical exercises. Often with the sort of basic instrumentation Terry is using. Checking for banana skins thrown up by the instrumentation, getting a handle on precision and accuracy, and if they were good enough for the task was part of that. As was recognising blunders in record keeping before they waste time. I am not a physicist but the training has led me to working with assorted instrumentation relevant to the task here. That included a precision aneroid that could measure the atmospheric pressure different between the floor and bench top to within about 20% and micromanometer to log the 'breathing' of a disused mine.

I am fairly confident that, if I decide to try any of this whistle stuff at home, that my recreational grade (but bought for work) aneroid altimeter is precise enough with appropriate back calculation to measure the sort of pressures I want to blow to the sort of accuracy I need to satisfy my interest. However, by the time Terry has got his setup finalised I think we will have a design for an appropriate manometer. In that context I'm assuming that the tubing has large enough diameter for the meniscus not to be causing any capillary action, or that any would balance between the two sides. Would a different tube material help I wonder? For the low value measurements the physics allows the section of pipe above zero level to be at a shallow angle as in the one Tunborough illustrated, then going vertical later. It's just the height difference between the two meniscuses(?) that matters

[Terry McGee]

Oh dear, deja vu attack. Back in the late sixties, I was a Trainee Technical Officer at the Research School of Physical Sciences at ANU. For a while I learned my trade from the older blokes. And I was so lucky - they were the best in the business, drawn from all over the world.

My first formal challenge was to design and build a thermocouple vacuum meter to replace older ones that needed very special meters. I was gobsmacked when my design worked, and indeed, became the standard for our School. Arrghhh, I can do this stuff! Which was lucky, as it seemed every department had need of vacuums, and therefore vacuum measurement.

Edited to add that this is the same underlying principle as the hot-wire anemometer stringbed advocated. Essentially you heat a wire, then determine the rate at which that heat is dissipated. In vacuo, that depends on how much air is left. In a wind flow, it depends on the speed of the wind.

But here's hoping we can avoid the need to go this far, he said, blithely tossing the question back to our theorists.....