Speed of sound in a tube

Exactly. And your pipe based resonance device is going to have free ends at both input and output so the values you get by direct measurement of an open pipe should be useful and meaningful.

The end effect correction is usually taken to be some fraction of the diameter of the pipe used ISTR about 3/4 or so. You could use a bunch of different diameter pipes at constant length if you are really determined to find the answer to this one empirically.

Like Phil I suspect the changes in sound velocity in pipe are tiny but non-zero, and not enough to upset the proposed design.

--
Regards,
Martin Brown

n a

Fun topic, In long pipes you definitely get some dispersion. Did you ever walk in long waste water conduits underneath highways and such... Or just clap in front of a long tube. There's a distinctive higher frequency 'twang' that results.

But perhaps this is from the transverse modes, of which you bespeak.

George H.

u

George,

"Did you ever walk in long waste water conduits underneath highways and such.."

Uh, ...exactly WHAT industry are you in? ;)

you

Grin... memories from childhood. A quick splash underneath the highway got us to a nice woodsy area... saved a few miles of bike riding. And of course once inside the conduit, the throwing rocks was required of all young boys.

George H.

Some of these metal structures are corrugated for rigidity. They definitely produce a twang because of evenly-spaced multiple reflections.

By pure coincidence, yesterday afternoon I was at the mouth of a disused canal tunnel, clapping and listening to the return echo. There was no noticeable twang, but there were unevennesses in the decay curve as the sound wave met different reflecting surfaces.

My position was close to one of the walls of the tunnel (where someone had obligingly left some stepping stones in the shallow water), so this may have affected the result. Next time I go there, I will take some wellington boots so that I can wade across the width of the tunnel mouth and investigate whether there are any position-dependent changes in the echo.

--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

e e

.uk

d d l

I enjoyed your data. +/- 3% range per tube.

I didn't understand your comment about geometric progression, so....

simple assumption resonance of a closed end tube is VERY approx, L =3D

0.25*v/f where L is length of tube in feet, v is veolicty in fps [1100], f is frequency in Hertz

for an audio range of 100 to 3kHz, the tubes should go from 3 ft down to 1 inch

[note to self: I'll bet diameter of the tube 'adjusts' the bndpass. where small diameter relative to length has a narrow bandpass and a wider diameter tube has a much wider bandpass.] *if* each tube has a bandpass of +/- 3% the tubes that would require approx 59 tubes to cover the spectrum, with each tube covering a 6% bandpass: 100, 106, 109, 112, 119, 126, 134, 142, 150, 159, 169, ...2770, 2940

then each length of tube would be, in feet:

2.75, 2.59, 2.45, 2.31, 2.18, ...feet ending for the highest frequencies, in inches 1.42, 1.34, 1.26, 1.19, 1.12, 1.06 inches

those step sizes don't look uniform.

calc step size in inches are

1.9, 1.8, 1.7, 1.5 inches and at the highest frequencies, in mils: 85.1, 80.3, 75.8, 71.5, 67.4, 63.6 mils

those incremental steps do not look like anywhere close to uniform steps, nor a geometric progression..

Back to that note regarding that the bandpass of each shorter tube being slightly wider [as its diameter to length ratio gets larger], allows for each jump in frequency as we're going up in frequency to be further apart and therefore each 'length' adjustment is more than expected and therefore each length cut can be uniform!

Fascinating thread

nd in a

a

tion

t . t

ry

o

ng a

ase

k_z

oup

s

n.

n

ee

ome

at

ur

as

you

Hide quoted text -

Hi Adrian, I've also noticed this with long tubes with maybe

6-8" (15-20cm) diameter. If you clap in front of one you get a similar 'strange' sounding echo.

This may have nothing to do with the speed of sound in the pipe though, maybe it's just that the higher longitudinal modes have a higher Q and echo for a longer time?

I'm sure it's a 'well known' effect. I just don't know what it's called.

George H.

ound in a

n a

ration

ent

t).

hat

n.

dary

so

he

ting a

phase

e k_z

h

group

ess

ion.

.

ion

o

free

some

that

iour

ay

ns

e as

d you

. d o h

k-Hide quoted text -

Googleing "Echo in a long tube" found this,

But no physics explanation.... sigh.

George H.

in a

operation

reply)

quoted text -

When I was a kid there actually already existed Disneyland... believe it or not ;-) In 1956 they had (IIRC) Tomorrowland (?) with various science demos... one of which was a telephone with variable delay that you could adjust until you went into perpetual stuttering mode... sort of like our resident PITA's :-) ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
I love to cook with wine.     Sometimes I even put it in the food.

No. +/- 1.5% per tube, 3% in total between 3dB points.

If the tube responses overlap at the 3dB points, the phase of one will be 90 degrees advanced and the other 90 degrees retarded, so they need to be combined in antiphase. This is clearly not practical to arrange if the tube responses are added acoustically (although it would be easier to arrange if each tube had its own microphone).

It was this which made me realise that there was a big difference between one-mic-per-tube and the Mason & Marshall / Olsen / W.E. design; their designs couldn't have had alternate signal inversion and yet they didn't show a regular series of dips in the frequency response, so something else had to be happening. That 'something' was energy exchange between two tuned pipes, which gave a flatter response.

If I had used one mic per tube, with 3% frequency increment there would have had to be 117 pipes just to go from 100 c/s to 3Kc/s; and the difference in pipe lengths would vary considerably. The W.E. construction had 50% frequency difference between the two shortest tubes (in a very audible part of the spectrum) yet nobody appeared to notice any huge gaps in the frequency response - so that again knocks the high 'Q' resonance theory on the head.

--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

om

tube

same

s.co.uk

ld

ut

3% t

ld

ng

nd

e a

t

be

il

al

.uk

Did you measure the phase shift at the 3dB points? might not have been +/-90. plus adding and subtracting sound energy is like herding cats.

I do have an idea for making a "sound curtain" like you could drape around an airport to mitigate aircraft noise, a little.

I once put a 3 inch diameter tubular path between two chambers in a 6 foot high [huge] very long snow drift. the idea was to be able to talk from one chamber in the snow drift to another chamber in the snow drift with the distance between chambers down to around 3 feet. Although you could clearly see each other it was NOT possible to talk through that tube!.Now THAT is sound absorption!. Yelling, sounded like something far off in the distance, barely heard.

sound in a

is

ign a

peration

erent

ent).

that

own.

undary

s, so

re

the

etting a

l

he

e phase

nce k_z

igh

e group

less

ition.

d

ls.

ation

to

n free

ne

's some

is

d that

aviour

may

ains

ame as

Did you

...

er

.

sed

no

he

e

is

me

uth

he

.uk-Hidequoted text -

xhibits/echo-tube/

OK I was lead to a nice article by Frank Crawford in AJP

The sound waves in a long pipe are simialr to TE mode in a waveguide. There is some maximum wavelength (lamda max ~ 2*D (D =3D diameter)*) And a group velocity vg =3D c*sqrt(1-(lamda/lamda max)^2).

Frank also points out that you can have plane waves propigate down the tube with no cut-off because of the different boundary conditions. (So you still can have long wavelength waves travelling in the tube. Waves beyond cut-off)

Fun... George H,

*the 'real' factor has one of those pseky Bessel functions in it and the cut-off really is 2D/1.172

reply)

text -

this,

Help me with this, George...how do you have both a well-defined cutoff wavelength _and_ low frequency response???

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510
845-480-2058

hobbs at electrooptical dot net
http://electrooptical.net

I am fairly sure that sound speed in a tube is essentially frequency independent and the same as in free air for your purposes. If your tubes were so small that the film layer (where displacement is reduced due to friction with the tube wall) approached 1/2 the tube diameter, and your tubes were very long, the answer might be a bit different.

Are you talking about the directional "line microphone" design discussed on page 322 of Harry F. Olson 'Acoustical Engineering', 1957, referencing Mason and Marshall, Journal of the Acoustic Society of America, Vol 10 No. 3 p. 206, 1939? I doubt if anyone has done a more rigorous analysis on this design due to poor directionality compared to parabolic reflector microphones.

If you are serious about the physics of acoustics you might want to retire Olson to your antiques bookshelf and get a copy of Fundamentals of Physical Acoustics by David Blackstock:

Which has a good analysis of sound in waveguides (partial differential equations required) as well as a frequency response model of the sound tube microphone, where a tube connects a microphone to a remote sound field, referencing the manual for B&K 1/2 inch microphone as a source of a typical sound tube microphone frequency response plot.

The analysis of the line microphone is complex enough that measurement is your friend :-).

Glen

in a

operation

boundary

phase

free

some

that

behaviour

as

reply)

text -

this,

Only sort of similar. EM waves don't move mass like sound waves do, and don't have viscosity or shear forces. You can have continuous smooth (or even turbulent) mass flow in a pipe, but there's no equivalent in a waveguide.

At low frequencies, in a skinny pipe, flow behaves sort of like a lossy transmission line. Step response will be vaguely similar.

--

John Larkin, President
Highland Technology, Inc

jlarkin at highlandtechnology dot com
http://www.highlandtechnology.com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom laser controllers
Photonics and fiberoptic TTL data links
VME thermocouple, LVDT, synchro   acquisition and simulation

ound in a

n a

ration

ent

t).

that

wn.

ndary

, so

e

the

tting a

e

phase

ce k_z

gh

group

less

tion.

s.

tion

to

free

e

s some

s

that

viour

may

ins

me as

Did you

...

er

.

used

s no

the

ne

his

ome

outh

the

o.uk-Hidequotedtext -

u/exhibits/echo-tube/

d

Gee, I count on you for the help. :^)

Have you ever heard the 'twang' from a long pipe? (rule one, if something happens it must be possible.)

So I guess both modes are possible. Plane waves moving down the pipe at c, and also waves bouncing off the side's with a cutoff.

In the AJP article Frank C. made a 'bongo drum' on the end of his long pipe and was amazed to be able to hear the low wavelength thud first and the the long twang with a cut off later.

How about this for an analogy (also from the article) cylindrical waveguide has a cut off, but a piece of coax doesn't. So would waveguide with piece of wire running down the center have more than one propigation velocity?

George H.

t -

I didn't measure it, but it ought to be 90 degrees if the cutoff slope is 6dB/octave (which it might not be).

And I thought I was doing a strange thing when I clapped my hands in a canal tunnel.

So how do speaking tubes on ships work?

--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

That was the microphone I had in mind but I don't have Olson's book and was using similar information which I found in "Microphones" A.E. Robertson (Iliffe 1951)

I have a number of good acoustics books, but not that one. It surprised me that even Kaye & Laby didn't have a clear staement of the velocity of sound in tubes (other than saying it was different from free air).

I think I need to build one and test it. When I have discovered its shortcomings, that will be the time to analyse what is actually happening.

--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

m

tube

same

.co.uk

ld

ut

t 3%

at

uld

ing

and

be a

at

ube

til

ual

o.uk

ge

ign;

y d

es

h

uk

sorry, not to be clear.

Nothing but snow EVERYWHERE. nothing physical added. We were actually trying to make a tunnel through this huge snow pile [can't believe it didn't collapse on us] and part of the construction was to punch a communication hollow between our two sections, to make certain the two paths lined up, not gone wonky directions. It was after punching the hole through the snow and trying to talk, we discovered how NO sound would go through that tube! Absolutely amazing to us.