I had a set of Aluminium etrusions that were perfect for a 2.1 litre µFonken, unfortunately they dissapeared diring the purge (while i wa sin hospital). It is enlightening how much skinny material (4mm) makes to the gross size of a small box.
dave
I’ve used constrained layer to good effect. The key is to get a good glue joint between the flat panels. I use tightbond between 12mm MDF and 12mm birch plywood, I set gallon milk bottles full of water from the center of the panels out to the edge giving the glue time to squeeze out. Gravity works better than clamps! You don’t want any bubbles of glue between the panels and you don’t want a thick layer of glue.
This technique seems to filter out the dullness of solid MDF and the brightness of solid birch ply construction. Then I cover the enclosure with thick 1/8 to 1/4” veneer of something my wife likes. You still have to brace and damp the interior of the enclosure but the box will be remarkably dead.
And what about the midrange?
Really, I find It disconforting when people subdivide the audio in bands and make diproportions in attributing a special value in each. Other than the 'evolutionary' peak of sensitivity of our hearing in the 2-3 kHz range that I read about. Probably it's the logarithmic scale that represents the frequencies ....
As a mechanical engineer, have you not considered starting with an inherently more suitable design/shape which is fit for purpose, rather than one which is fundamentally flawed and then requires 'fixing' by using excessive materials/bracing to make it function properly? For example, I'm sure you would not choose to design a lightweight 250PSI pressure vessel using flat plate only to have to reinforce the life out of it to stop it bulging at even a few PSI...
12mm ply and minimal 6mm bracing is more than adequate for my 50l 700W 18" subs, with the first resonance pushed way above the crossover point due to inherent stiffness where the largest panel is only 130mm wide.
Hydroforming
Get my drift?
My cuboctohedron speakers were a fun project to build. A bit of 3d Pythagoras to figure out the cutting angles for a set of 6 squares and 8 triangles per box.
Lesson learnt: you can never spend "too much" time cutting accurately because the alternative is more sanding later (j/k! I still haven't learnt that one!). Version 2, if I ever build up the courage, will surely have box joints, even if it's just the bare minimum for the panels to lock together. Apart from making the shell a lot stronger, it will help with gluing as clamps are tricky.
A minor regret is that I didn't go bigger with a 4-5" full-range instead of the 3" CHN-50 (Mark Audio), just for a bit more bass. I like 'casual' listening from anywhere in the house, so a smooth off-axis response without so much diffraction from 90 degree turns seems to help.
Lesson learnt: you can never spend "too much" time cutting accurately because the alternative is more sanding later (j/k! I still haven't learnt that one!). Version 2, if I ever build up the courage, will surely have box joints, even if it's just the bare minimum for the panels to lock together. Apart from making the shell a lot stronger, it will help with gluing as clamps are tricky.
A minor regret is that I didn't go bigger with a 4-5" full-range instead of the 3" CHN-50 (Mark Audio), just for a bit more bass. I like 'casual' listening from anywhere in the house, so a smooth off-axis response without so much diffraction from 90 degree turns seems to help.
I had a similar thought last night, why start with the traditional rectangular box? Well, the answer should be somewhat obvious. After I had a teardrop shaped tower design done I had to figure out how to build it with typical garage shop tools. My calculations took me to six months of routing plates to stack. So, unless you have a CNC rig it’s really impractical to cut so much material, and make so much MDF dust!
Reference Randy’s concrete suggestion, I thought about making a concrete teardrop from two bent concrete (Sono tube) forms. The plan was to use egg shaped forms inside an 8” tube inside a 10” tube with spacers to maintain the egg shape.
I gave up and made a box. Maybe someday I’ll make some concrete towers.
Be interesting to see how the curved wall design holds up
in the Steady State Dynamic modeling.
Likewise see a lot of symmetrical bracing in many designs.
Wouldn't it be better to divide panels with braces
into non symmetrical sections.
Recent project I just built a shell, no bracing.
tapping the box I found the lowest tone of the ping.
Added a brace there. Location wasn't symmetrical.
after adding the brace found 2 more areas with lowest pitch.
(now higher) added braces there.
No guessing or assumptions, actually put the braces according
to the actual sound of the box as it changed brace by brace.
in the Steady State Dynamic modeling.
Likewise see a lot of symmetrical bracing in many designs.
Wouldn't it be better to divide panels with braces
into non symmetrical sections.
Recent project I just built a shell, no bracing.
tapping the box I found the lowest tone of the ping.
Added a brace there. Location wasn't symmetrical.
after adding the brace found 2 more areas with lowest pitch.
(now higher) added braces there.
No guessing or assumptions, actually put the braces according
to the actual sound of the box as it changed brace by brace.