Blowing machine

[Terry McGee]

Just checking my two new stepped calibrators, I noted that the 5mm section of the 10 @ 5.0mm /20 @ 4mm is a bit oversize. The drills normally fit tightly in Delrin, whereas this one is a loose fit. I think I must have nipped it up a bit funny in the chuck. So safer to regard that one as 10 @ 5.1mm /20 @ 4mm, as corrected below. I can remake and retest if that screws you up.

Code: Select all

10 x 5.1, 20 x 4mm Calibrator				
Mano/2	MM(H20)	√A/P	Flow	Resistance
140	280	16.7	40	0.42
71	142	11.9	30	0.40
33	66	8.1	20	0.41
6	12	3.5	10	0.35
0	0	0.0	0	
				
20 x 5, 10 x 4mm Calibrator				
Mano/2	MM(H20)	√A/P	Flow	Resistance
136	272	16.5	40	0.41
70	140	11.8	30	0.39
32	64	8.0	20	0.40
9	18	4.2	10	0.42
0	0	0.0	0	

[Terry McGee]

I was worried about the 5.1mm @ 10mm, one, so I redid the measurements, adding two more data points. The dramatically different results down low probably due to poor resolution!

I also wondered what would happen if you took the other one, the 10 x 4mm, 20 x 5mm Calibrator, and fed it from the narrow (10mm) end. Interesting!

Code: Select all

10 x 5.1, 20 x 4mm Calibrator Revisited				
Mano/2	MM(H20)	√A/P	Flow	Resistance
132	264	16.2	40	0.41
68	136	11.7	30	0.39
30	60	7.7	20	0.39
16	32	5.7	15	0.38
6	12	3.5	10	0.35
2.5	5	2.2	5	0.45
0	0	0.0	0	
				
10 x 4mm, 20 x 5 Calibrator, small end fed				
Mano/2	MM(H20)	√A/P	Flow	Resistance
91	182	13.5	40	0.34
50	100	10.0	30	0.33
22	44	6.6	20	0.33
4	8	2.8	10	0.28
0	0	0.0	0	

[Terry McGee]

We have a 30mm @ 4mm Calibrator. I wondered if we should also have one of half the length, ie 15mm@4mm to test the effect of windway length, and one at twice the bore, to test the effect of bore size. That would suggest a 30mm @ 5.66mm Calibrator. It might come as somewhat of a shock to more sensitive readers, but my workshop doesn't boast a 5.66mm drill. I could get pretty close at 5.6mm, which would give us 1.96 times. Or I could drill two 4mm holes side by side. Any attractions in these bold doublings?

[Terry McGee]

And, even after cleaning and refilling the manometer with a more diluted coloured water, I still get a sense of stickiness when I change pressures. Someone had previously mentioned adding a small amount of surfactant. A single drop of detergent?

[trill]

. . . the reported pressure is clearly normalized to sea level.

Look what I found: https://www.weather.gov/bou/pressure_definitions

I think what we want is "Station Pressure".

This site offers a page to calc "pressure at altitude", with one of the inputs being "Sea Level Pressure", which I now know, is commonly reported:
https://www.omnicalculator.com/physics/ ... t-altitude

[trill]

Also, on the subject of density, here's a comparison of Nitrogen and Oxygen:

https://www.engineeringtoolbox.com/gas- ... d_158.html

n2 = 1.1651
o2 = 1.331
At NTP - Normal Temperature and Pressure - defined as 20oC, and 1 atm .

The flow meters are based on flow-drag lifting the ball.

Drag force is cdrag*one-half-rho*v^2

That could be a 15% difference right there.

I wonder if the flow indicated should be "adjusted" to reflect a 78% N2 + 22% O2 mix, like that of our atmosphere.

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.

In my spare time.

[stringbed]

And, even after cleaning and refilling the manometer with a more diluted coloured water, I still get a sense of stickiness when I change pressures. Someone had previously mentioned adding a small amount of surfactant. A single drop of detergent?

A single drop is both an obvious and good place to start. The need for more will be apparent from the result. The density of the solution will increase with its concentration so calibrating against an H20-only reference is probably a good idea.

[stringbed]

The term back pressure is a misnomer . . .

Naaahh. It's a classic portmanteau: a language construct intended as a convenient label for something novel or new to the speaker. A whistle player can tell that some whistles take more effort. More "push" by the muscles. The "push" is "against" something. That something is feels like it is "pushing back". It's that simple.

Musicians and instrument makers use terms that might confuse an outsider. My suggestion that we steer clear of "back pressure" was founded in its alternate appearance in the present discussion with the term "resistance." Both labels can reasonably be applied to the source of a sensation that "something feels like it is pushing back." However, back pressure is not the same thing as pressure with the sign reversed. Treating it as such in a discussion such as this one strikes me as a good way to get further into the thicket rather than out into the clearing.

So the question becomes, are these all equally valid and acceptable statements?

If I want to push a flow of 10L/Min through my Mellow D whistle, I will need to apply a pressure equal to 25mm of water.
If I push a flow of 10L/Min through my Mellow D whistle, I will be able to measure a pressure equal to 25mm of water at the inlet to the beak.
If I blow 10L/Min through my Mellow D whistle, I will feel a backpressure equal to 25mm of water.
If I apply a pressure equal to 25mm of water to my Mellow D whistle, I will need to keep up a flow of 10L/Min.

If we ignore the difference between quantifiable physical quantities and their perceptive correlates, "pressure" and "back pressure" designate the same thing here. The third point is the only one where the latter is noted. But how does a whistle player even begin to gauge the volume of air being expended or the manometric equivalent of the pressure/backpressure?

There are differences between the vocabulary used in even mildly rigorous discourse about the physical behavior of duct flutes and the everyday jargon of whistle players. We’d probably benefit from not conflating the two In the present context.

[trill]

. . . perceptive correlates . . .

Very nice term !

[trill]

. . . But how does a whistle player even begin to gauge the volume of air being expended or the manometric equivalent of the pressure/backpressure?

If I hand you a 1kg weight, and then hand you a 2kg weight, you'll know the difference. Everyone will. It feels heavier. Your neuromuscular anatomy tells you right away.

If you pedal in 10th gear, it's harder than 1st gear. Everyone can tell.

Gauge ? I might not get the actual ratio, but I will know its more.

Among different whistles, the feeling in my breathing muscles is very plain.

In reading whistle player comments, I recall some other terms like: "easy blower", "hard blower", "gallons of air", "medium blower".

It's true, different groups+tribes have their own vocabulary. The numerical metric ? I'm too tired to sort that out right now.

With the work-in-progress of this thread, though, it could be quantified.

[david_h]

... (I tried several vernier calipers - they read between 3.94 and 3.97, but I'd trust the one that read 3.95 more than the others.)

I think that highlights an issue (or lack of it) over the instrumentation. Like the calipers the manometer and flowmeters will be 'fit for purpose' for one thing but not necessarily for others. They don't sound like the sort of thing that comes with a calibration certificate or have a sticker with the last calibration date.

For most readers of this forum (if not this discussion) it looks to me as if Terry is well on the way of being more quantitative about the comparative 'backpressures' of whistles, which is something that comes up in almost every review. For a static pressure (easier said than done) manometers are a simple and reliable way of measuring the difference from atmospheric pressure. For a U-tube type one so long as we have a good ruler and know the density of the liquid** we are good to go. The reservoir ones, if one is not reading the reservoir level each time, have potential complications.

For Tunborough's modelling it looks like Terry is not there yet. There are all sorts of questions, in addition to instrumentation, that need addressing or demonstrating to be insignificant - e.g how close to atmospheric pressure is the windway exit? As air is flowing away from it into the room it must be higher. How does it vary with head geometry?

** if a single drop of added material or some die is the level of concern you have then you need to be measuring the water temperature at the time of each measurement.

[david_h]

My perception of how hard I am blowing comes from my ears (tympanic membrane maybe) and somewhere in my head (blood pressure probably).

[Tunborough]

We have a 30mm @ 4mm Calibrator. I wondered if we should also have one of half the length, ie 15mm@4mm to test the effect of windway length, and one at twice the bore, to test the effect of bore size. That would suggest a 30mm @ 5.66mm Calibrator. It might come as somewhat of a shock to more sensitive readers, but my workshop doesn't boast a 5.66mm drill. I could get pretty close at 5.6mm, which would give us 1.96 times. Or I could drill two 4mm holes side by side. Any attractions in these bold doublings?

I think I've got a handle on the effect of length, although I wouldn't object to a shorter 4 mm dia calibrator, maybe even 10 mm long just to be sure.

To better match our whistle windways, I'd like to go to smaller diameters rather than larger, and more extremes of taper. For example, 10 mm of 3 mm dia + 10 mm of 4 mm dia + 10 mm of 5 mm dia. Do you have any drill bits smaller than 3 mm that
could safely handle the delrin? (For that matter, the calibrator could be something softer than delrin for this purpose.) It is looking like the difference between inlet size and outlet size is a significant variable.

[Tunborough]

For Tunborough's modelling it looks like Terry is not there yet.

I'd put it differently ... For Terry's measurements, it looks like Tunborough's modelling is not there yet.

There is sufficient consistency in Terry's measurements that I'm not ready to challenge the instrumentation. I still have some work to do.

[trill]

So, I went back to Terry's geometry for the Feadog Mk-1:

Feadog Mk1, flat
L D W.
24.81 1.5 8.23

And the flow test:

Feadog Mk 1
L/Min P(meas) mm SQ RT A/P P (calc) mm
40 252 15.9 252
30 136 11.7 142
20 64 8.0 63
15 33 5.7 35
10 14 3.7 16
0 0 0.0 0

And used the derived Cd to get this:



Some notes:
1) the derived Cd values range from .84 - .9
2) the derived Cd values descend with flow rate, as would be expected, except for the 20l/m data point
3) using the mean Cd value (0.87), the predicted flow at 40l/m is within 2% of measured.