Blowing machine

[Terry McGee]

Oooh, you're right, thanks stringbed. It's 10:30pm, so a little late to be running compressors, so I'll see what we can make of it tomorrow.

[Tunborough]

I was troubled by the linear section of the latest sample, around 7 to 11 L/min. I was afraid it might be a problem with the instrumentation. Then I remembered the Hagen-Poiseuille equation, which is what happens to Darcy-Weisbach when the flow is low enough that there is no turbulence, what's called laminar flow. Hagen-Poiseuille does have linear relationship between pressure and flow, so this linear section isn't a bug, it is a feature. This linear section may help with the general pressure-flow model, but even before that, it clearly shows where the flow through the windway is laminar (not turbulent).

Speaking yet again from the perspective of an erstwhile day-job voicer of duct flutes, and applauder of the present effort at plugging numbers into the study and application of that process —

The Rottenburgh recorders were delivered to the voicers with radially arched, axially flat windways. Among the routine subsequent actions was adding a varying degree of concavity along the length of the windway’s ceiling. It began at its midpoint and was worked toward the inner edge of the chamfer, the latter first having been cut to its full extent. The axial arching had a few purposes and one of its consequences was a reduction in the instrument's resistance — making it "easier blowing" in the workshop vernacular.

Is there any chance that laminar flow was guiding some of your work? I wouldn't be at all surprised if an experienced voicer could tell the difference between laminar and turbulent flow from the sound, and if some of their work was bringing laminar flow to notes higher up on the scale.

[david_h]

An interesting observation. What's then happening below 7 l/min? I'll give you that a straight line there is a very slightly better fit than a straight line of Flow against P to the 0.5 (i.e. square root). But the latter does almost go through zero. A power relationship fit gives a power of 0.44, which is nearer to 0.5 than to 1.

It looks a bit like a linear relationship from 100 to 180 as well.

Need better data to be convincing.

[Terry McGee]

OK, quick run through comparing the digital and analog manometers after learning how to zero the digital one!

Code: Select all

Mano/2	Analog	Digital	Dig Zeroed	Ratio
0	0	-6.5	0	0.00
10	20	14	20.5	1.03
20	40	36	41.5	1.04
30	60	56	62.5	1.04
40	80	77	83	1.04
50	100	97	104	1.04
60	120	117.5	124	1.03
70	140	138	144	1.03
80	160	159.5	166	1.04
90	180	180	188	1.04
100	200	200	208	1.04
110	220	221	229	1.04
120	240	242	248	1.03
130	260	263	269	1.03
140	280	282	289	1.03
150	300	301	309	1.03
160	320	321		0.00
170	340	341		0.00
180	360	360		0.00
				
			Average	1.03576776926777

New data in the Dig Zeroed column. It only reads down to 1mm, so when it vacillates between say 159 and 160, I've called that 159.5.

Ratio of Dig over Analog in final column. Average 3.6% over-reading. Or is the Analog one under-reading? How can we tell?

Interestingly, my meter doesn't identify itself in the way that the image I posted does. Which is why I didn't know where to look for the manual!

[trill]

OK, quick run through . . . How can we tell?

For those of us who like graphs . . .



Now, what puzzles me are:

1. The variation in disparity between the zeroed and non-zeroed. I was expecting a fixed offset. Turn-on-bias is a common sensor "feature".

2. The close agreement between dig + analog prior to zeroing, for readings 8 and up.

[Terry McGee]

Time to bring into play my special skill? To enter a conversation wracked with confusions, and make it more complicated?

I wondered if we take the flows into account, would that reveal anything?

Code: Select all

30 x 4mm Calibrator							
	Analog Manometer		Digital Manometer		
Flow	Mano/2	MM(H20)	√A/P	Res.	MM(H20)	√A/P	Res
0	0	0	0.0	0.00	0	0	0
10	9	18	4.2	0.42	18	4.2	0.42
20	37	74	8.6	0.43	75	8.7	0.43
30	75.5	151	12.3	0.41	155	12.4	0.41
40	140.5	281	16.8	0.42	289.5	17.0	0.43

Note that the mm(H2O) columns differ noticeably and increasingly, but by the time you take the Square Root of them, or calculate the Resistance (by my so far unapproved simplistic approach!), we're down to very little difference. Disturbing though, I'll admit. Any suggestions for other tests to crack this one?

Does picking either one help answer some of your issues, Tunborough?

[Terry McGee]

Hmmm, and recalling our earlier conversations that concluded we didn't really need to measure Flow and Pressure, apart from once to calibrate their relationship, are we allowing ourselves to be bogged down?

Now trill, you wanted me to do a run on the Old Gen? Do you think we're ready for that or would you rather wait to see if we can come to a conclusion on the choice and management of manometers?

And can you advise:
Include C naturals (C6 oxx ooo and C7 oxoxxx?)
C sharps (both ooo ooo or?)
D6 (oxx xxx), D7 (oxx ooo) or ?

[trill]

. . . Now trill, you wanted me to do a run on the Old Gen? Do you think we're ready for that . . .

Yes, I think the agreement between the manometers is good enough for my purposes. Use whatever is easier for you.

. . . . . . And can you advise . . .

1. No Cnats needed
2. D6 (xxx,xxx) please
3. D7 whatever works. (I don't usually play up that high !)

As a side note: the flowmeter I ordered arrived today (yipee!). Now I just have to make time to fiddle with it, comparing air vs. oxygen.

[Terry McGee]

OK, I'll have a go at it. I think I'll give the Digital Manometer a try this time to see how much easier (?!) using that is. I've tidied the analog manometer away and tidied up the workspace, and the greater availability of free space alone is encouraging! We can always use the comparison data to reconvert if we later find that the analog was right!

[stringbed]

The Rottenburgh recorders were delivered to the voicers with radially arched, axially flat windways. Among the routine subsequent actions was adding a varying degree of concavity along the length of the windway’s ceiling. It began at its midpoint and was worked toward the inner edge of the chamfer, the latter first having been cut to its full extent. The axial arching had a few purposes and one of its consequences was a reduction in the instrument's resistance — making it "easier blowing" in the workshop vernacular.

Is there any chance that laminar flow was guiding some of your work?

To the extent that it is a relevant factor in nudging an instrument toward the desired sound and behavior, laminar flow guides such work pretty much by definition. If you are asking whether voicers ascribe any direct significance to it during that process, I never have, nor I have encountered it in any dialog among craft practitioners.

I wouldn't be at all surprised if an experienced voicer could tell the difference between laminar and turbulent flow from the sound, and if some of their work was bringing laminar flow to notes higher up on the scale.

If laminar and turbulent flow produce different sounds, it is reasonable to expect a reasonably experienced voicer to be able to discriminate between them. But where has it been established that the two types of flow result in audible differences that are relevant to the voicing process?

[stringbed]

[This post was a mis-formatted duplicate of the preceding one. I can’t delete it but have stripped its contents.]

[stringbed]

[This post was a mis-formatted duplicate of the one before the preceding one. I have no idea how it came into being but can’t delete it and have stripped its contents.]

[david_h]

Average 3.6% over-reading. Or is the Analog one under-reading? How can we tell?

Fitting a line through the points gives Digital = Analog * 1.033 + 0.319. So 3.3% is better number to use than 3.6%, but in the same ball park. I think we can ignore the 0.319.

The earlier data (without the zero set) over the same range but not using the zero point gives Digital= Analog *1.029 - 5.812. So 2.9% over-reading and as the constant is close to the unused zero point it looks like the zero setting has shunted all the numbers up by a constant. So it was a zero set rather than just clearing the last reading, which is one way the manual could be read.

It's not temperature. Millimetres of water are for water at 4 degrees C (I had been assuming STP) but unless Terry is filling the manometer with freshly brewed tea expansion won't give 3% difference. And Terry would have noticed if his ruler was 3.3% out. Even though the manual for the digital manometer gives accuracy better than 1% I would be inclined to use the manometer reading as a calibration but, just in case, not forget that a 3.3% adjustment has been made.

The scatter around the fitted line is within 2mm. I wonder if that is within what Terry can measure bending down with the ruler. I think the fit between the two measures of pressure means that any odd behaviour in the Flow - Pressure relationship is either from the flow meter or some fluid dynamics to be understood, rather than from the pressure measurement.

Hey, you may be half way there on the instrumentation! Manometer connected to both sides of the sample with the flow metre before the final exit to atmosphere is my suggestion. May be hard for a real whistle head but it would give another view on the wanderings of the flow-pressure curve.

[Tunborough]

I think the fit between the two measures of pressure means that any odd behaviour in the Flow - Pressure relationship is either from the flow meter or some fluid dynamics to be understood, rather than from the pressure measurement.

I would agree. I'm no closer to understanding that relationship over the full range of samples, but a consistent 3.3% error isn't going to change that one way or another. I'm content to take what the digital meter provides; it's easier to read, and more precise (measuring to the nearest 0.5 mm, rather than the nearest 2 mm after doubling). If it matters at the end of the day, I can subtract the 3.3% later.

It's not temperature. Millimetres of water are for water at 4 degrees C (I had been assuming STP) but unless Terry is filling the manometer with freshly brewed tea expansion won't give 3% difference. And Terry would have noticed if his ruler was 3.3% out. Even though the manual for the digital manometer gives accuracy better than 1% I would be inclined to use the manometer reading as a calibration but, just in case, not forget that a 3.3% adjustment has been made.

Definitely not temperature. To get readings consistently 3.3% low, Terry would have to be charging his manometer with extra-dense liquid, not less-dense. If it came down to it, I'd trust the manometer's accuracy over the digital meter.

[Tunborough]

I wouldn't be at all surprised if an experienced voicer could tell the difference between laminar and turbulent flow from the sound, and if some of their work was bringing laminar flow to notes higher up on the scale.

If laminar and turbulent flow produce different sounds, it is reasonable to expect a reasonably experienced voicer to be able to discriminate between them. But where has it been established that the two types of flow result in audible differences that are relevant to the voicing process?

It hasn't been established ... That was idle speculation on my part. It doesn't sound relevant, so I won't attach any more significance to it.