[Dick Pierce[_2_]]

Gary Eickmeier wrote:

"Well, in theory a box-shaped sub enclosure will try to flex its wall inward
or outward as the cone moves in and out, with the idea being that the
pressure is trying to turn the box into a sphere.

Several things poop into my head when I read this:

"What's the difference between theory and practice?
In theory, theory and practice are the same. In
practice, theory and practice are different."

It's on the Internet, so it must be true.

A cylinder-enclosure shape, even with the cylinder walls only 3/8 inch
thick, cannot flex inward or outward (assuming the Sonotube fiber material
is decently strong in terms of stretch resistance, and believe me, Sonotube
is strong), because the surface is already curved to a circle.

Oh, be assured, kind people, it most assuredly can and quite
happily does. The assertion that it cannot is based on the
following assumptions:

1. That the wall material is infinitely rigid, has infinite
tensile strength, i.e., it can neither be stretched, compressed
or flexed. The 3/8" cardboard walls of a Sonotube are VERY
VERY far from that ideal,

2. That the internal forces on the walls are the same at every
point along the walls surface. With ANY musical material,
this is simply not the case.

3. That the cylinder is infinitely long, i.e., it has no
boundary conditions to deal with. Again, not the case
"in practice."

Consider two cyclinders of MUCH greater strength than that of a
20" Sonotube: the walls of the SRMs on a shuttle were 1/2" steel.
The ignition transient was predicted and was measured to cause the
walls halfway between the field joints to expand nearly a foot.
But because the field joints used a much heavier, thicker steel
ring, they did not bulge anywhere near that much. The result was
phenomenon called "joint rotation" and was crucial to causing
the failure of the field joints in the Challenger explosion.

Be that as it may, while "in theory" it seems like Sonotube is
an "ideal" way of building a subwoofer, "in practice"m most
of those teoretical advantages simply aren't there.

I had, for a former client, done quite a bit of measurements on
cardboard tube enclosure, most of which were MUCH more substantial
than Sontotube (minimum 3/4" wall thickness, higher density wrap).
Using acceelerometer measurements, wall flexing was NO better and
most of the time WORSE than a rectangular enclosure of the same
corss-sectional are. The unsuported walls of the rectangular
enclosure did no worse under the same conditions, and at the
corners, they behaved MUCH better.

At freuqencies where the wavelengths were smaller than either the
length or cricumerance of the tube, cyclinders tid MUCH worse
in almost every respect. We would see nodal vibration patterns
on the wall that had very high Q'. This is directly attributable
to the very dimensional symmetry of the shape. Siply because the
unsupported area of the "soundboard" was now the length of the
enclosure times its circumference: nodal patterns could be easily
set up and well established at MUCH lower frequencies than in a
similarly dimensioned rectangula enclosu the largest unsupported
surface in a box is substantially smaller than that of the cylinder.

And the argument that

"A cylinder-enclosure shape ... cannot flex inward or outward"

Makes the simplifying and quite incorrect assumption that there's
a single, in phase constant force on every square cm of its curface,
annd thus the wall MUST be compelled to move uniformly radially.
Sorry, but that just doesn't happen in loudspeaker enclosures. TO
get the walls to fles very nicely, all you have to do is get it
going outward at one point atv the same time it's going inward
womehere else along its circumference. And, as far as local
stiffness is concerned, Sonotube ain't so hot, especially 3/8"
Sonotube.

If you want to deal with this issue, consider getting TWO sonotubes,
on loerger in diameter (by maybe 1-2"): place them concentrically
and then fill in the space between the two with something heavy and
mechancially lossy, like dry sand (or, uhm, concrete!).

Yes, it is
not a sphere, but the top plate cannot flex outward or inward, because it is
mostly filled with the driver, and the bottom plate cannot flex much,
because it has the port tube in there reinforcing it and the thing is
usually too thick to flex much, anyway.

Wanna bet? The port tube DOES NOT reinforce it: the hole, in fact,
weakens it.

My bottom plates are each 2.25
inches thick,

Okay, so you "cheated." They don't flex because they are round, or
because they have a driver, or becasue they have a port, they
don't flex merely because they are 2 1/4" thick.

The inside of the Sonotube should be
layered with a couple of inches of fiberglass (mine is held in place with 3M
spray adhesive, and the fiberglass is actually furnace pipe wrap with the
foil removed), and the idea there is for the heating and cooling of the
fiberglass under pressure (which can occur at any frequency above the port
resonance) to help the driver to "see" a larger interior space than is
actually there. (Acoustic-suspension woofers and subwoofers also have
something like that in their cabinet interiors for the same reason.) The
effect is actually minimal, but in any case the fiberglass is not there to
control resonances or absorb noise. Resonances are minimal with a cylinder
enclosure, actually, due to the lack of glued-together speaker-wall joints.

Sorry, the very axial and radial symmetry means they have
the ability to suport quite nasty, highQ resonance. Take a
glass pie plate, fill it with water, and put a speaker near
it, driven by an oscillator: you'll find any enclosure with
a circular cross section support standing wave VERY nicely.

Or, for a dramatic acoustical demonstration of the effect,
toss a fire cracker into a cylinfrical storm drain: it
doesn't go bang! it goes Bonnnnnnnggggggg! for a VERY
long time. And, surprise, you find that the pitch of
the bong! it inversely proportional to its diameter.

And a cylinder also supports end-to-end standing waves quite
nicely as well. That's why flute, organ pipes and such are
made, largely, of cylinders.

In any case, in practical terms, a Sonotube enclosure should be the most
resonant-free enclosure possible for a subwoofer, although the light weight
is a theoretical problem - that I have not encountered.

Sorry, both theory and practice demonstrate otherwise.

If the port and driver are on the bottom, the designer
can have an attractive top on the cabinet that will allow it to function as
an end table, or the like.

And has a problem in calculating what the effective radiation
impedance seen by both the port and the driver a rest assured,
if you assume that it has no effect, you'll get it wrong.

However, the light weight of a sonotube enclosure will allow
the entire unit to vibrate up and down, which might make that
table lamp dance around a bit.

Please, do we have to bust this myth AGAIN (or did we miss the
smiley)?

Assume the mass of the driver cone is, oh 200 grams. Assume
also that the driver, enclosure and lamp together weigh 30
kilograms. By conservation of momentum calculate what the
worst case "bouncing" of the lamp will be. Show your work.
Get back to us when you have an answer.

--
+--------------------------------+
+ Dick Pierce |
+ Professional Audio Development |
+--------------------------------+