DIY GFRP Sandwich Panel 1x15PR400

[will33]

Also, what the hell is a fillet? I'm thinking of cleaning fish here.....

 

[zac2944]

Nice to see diy composite builds alive and well. Nice work so far.

About the cutting without delamination. You can clamp thin plywood outside of your panels before cutting with a jig or circular saw and rotary cutting tools like dremmel work. Glass kills blades though. Diamond or abrasive cut off type saws work well too.

By far the easiest way to cut panels is with a utility knife when they are still green. Use a straight edge and a sharp blade and you get a really nice cut with no mess. Just gotta time it right.

Corners on cut panels are always a challenge. I prefer to glass the cab whole rather than panels, but there are certainly pros and cons either way. One thing you can do is carve out the core from your cut edge then fill it with a more durable material like filled epoxy or heavy two part foam that will hole a router shape.

 

[astack]

One thing I've never really "got".....how you guys do seams. It seems <---- the strongest would be to do the inside and outside fiberglass layers as one contiguous layer, (1 in and 1 out) but that has to be complicated, and, on the inside, a total PITA if not just impractical/impossible.

(I see what you did there.)

Yup. That's a tricky thing. I think in a closed box like this, it's not super critical.

But I'm thinking through this now... A corner can fail in maybe four ways. 1) It could bend and open up, so the inside stretches and/or the outside crushes. In that case, the FG isn't going to do much on either side - it's doesn't have leverage on the inside and it isn't that strong in compression on the outside. 2) The reverse of that -- bending inward -- the outside glass will help out. 3) The panels could shear/slide past one another. I think the epoxy matrix takes most of that stress. 4) It could crush.

In all of those failure modes, it's the epoxy fillet that's the most critical, I think, especially 1 and 3. For 2 and 4 it's the outside glass.

Also as for the contiguous layer everywhere, I could be convinced that it doesn't have to be. In designing reinforced concrete, for example, where instead of FG strands and epoxy matrix it's steel rebar and concrete: In that case, as long as the bar can develop it's strength in the matrix (meaning it's cut a few feet or so longer than it needs to be), you could have all discontinuous bars and be fine. So as long as there's a few inches of overlap between each layer of glass, it shouldn't be an issue.

PITA it is, no matter how you do it

Also, what the hell is a fillet? I'm thinking of cleaning fish here.....

Haha delicious... It's the built up glob of epoxy/silica on the inside face of the butt joint. (It's similar to a fillet weld.) Think of it like caulking the side joint. It gives a nice thick cross section of epoxy for load bearing and the bigger it is, the less stress that gets concentrated right at the corner. Sharp corners are not your friend.

Nice to see diy composite builds alive and well. Nice work so far.

Dude! Thanks for commenting Trying to keep the flame burning

About the cutting without delamination. You can clamp thin plywood outside of your panels before cutting with a jig or circular saw and rotary cutting tools like dremmel work. Glass kills blades though. Diamond or abrasive cut off type saws work well too.

By far the easiest way to cut panels is with a utility knife when they are still green. Use a straight edge and a sharp blade and you get a really nice cut with no mess. Just gotta time it right.

Yes, yes and yes. The plywood trick is about the only thing I could think of for a while, but it seems like a waste. Maybe I just need an attitude readjustment. If it works, it works and I know I've thrown away money on less during this build.

Cutting EVERYTHING (FG+foam+FG) while green, I had not thought of that. (If I understand you correctly) That would be slick. The timing bit has kept me away from trying to cut the green stuff -- feel like I'd just have to sit there all afternoon until it kicks. Hmm... but with fast hardener it shouldn't be terrible.

Corners on cut panels are always a challenge. I prefer to glass the cab whole rather than panels, but there are certainly pros and cons either way. One thing you can do is carve out the core from your cut edge then fill it with a more durable material like filled epoxy or heavy two part foam that will hole a router shape.

I've been reading a few different boards where fiberglass is a popular material (sorry TB!). The threads always start with a simple question about means and methods and then the first 10 responses are all different. It just comes down to what you're comfortable with as long as it generally will hold up.

The epoxy fill is what Jmac did on his honeycomb panels and worked great there. But epoxy's heavy

 

[zac2944]

Timing green cuts is not that tough. You usually have a good hour or two window of time, but it depends on the epoxy system you're using and temperature. Just a straight edge and a knife. Super easy.

Fwiw, I've been experimenting with molds and two part foam. You layup glass on the mold, assemble mold pieces, then fill the void with foam and watch it expand. Once it is all set up you pop off the mold pieces.
I've yet to build a full cab this way, but baffles, ports, handles, etc. You can do some really cool stuff. Also, with the right type of glass you get super peel strength between core and skins. Someday I'll put together a thread.

 

[astack]

Sounds awesome, Zac!



Soooo, disaster just struck, again (worse this time). Our place is a mess with packing and all right now and the speaker was still out of the box after yesterday's layout exercise... And well a chair leg found the cone.

@#$%^&*()(*&^%$#%^&*()(*&^%$#@!#$%^&*()(*&^%$#@$%^&*()_(*&^%$#@$%^&*()_(*&^%$#@$%^&*()(*&^%$#@$%^&*()_(*&^%$#@$%^&*()(*&^%$#%^&*()(*&^%$#@$%^&*(

I left it on the floor and I was the perp, so gonna have to take all the blame...

Now that I'm calming (slightly), wt* do I do now? It didn't puncture it, just left an indentation/crease. I'm guessing a recone is required? Maybe not?? So how much is this gonna cost ballpark? $100, $200, is it scrap metal now? I'll post a pic later...

 

[astack]

Anyone?

On a scale of good-to-go to fubar'ed, where would you score this? I figure I'll at least have to take it to someone to have it checked in person, right?


In other news, I got the box all cleaned up and round-overed today. I also made a bunch of returns on misc hardware and got $40 back. So that's cool. Glassing, etc. is gonna have to wait until after the move :/ Energy got zapped for the day after that mishap...

 

[JMac- Wiskerface Designs]

Whoa, sorry Man. That really sucks!
From the picture, it appears to have only creased the cone a bit. Did it damage the surround at all?

 

[astack]

Whoa, sorry Man. That really sucks!
From the picture, it appears to have only creased the cone a bit. Did it damage the surround at all?

Correct, it doesn't look damaged. So that's a good sign?

I'll say this, I'm shocked it doesn't look worse. I don't know what kind of abuse speakers can usually take, but I would have expected the cone to tear. I hope I'm done with damage tolerance testing for a while

 

[JMac- Wiskerface Designs]

AFAIK/AFAICT- As the damage seems very minimal, I would think it should be fine. Maybe some other Guys far more knowledgeable than I, will advise on what the best action would be. What I do know for sure, that if the creasing in the cone was bad enough it could potentially cause some sonic issues, as well as mess with the alignment of the voice coil, creating a "rub". (Again if I am wrong, someone please set Me straight)
IME/FWIW- I have seen drivers take a harder lick and still work okay, with some DIY repair attention(crushed dome, with a shoeprint in the cone). After all, if it's any consolation, it's not nearly as bad as "the old screwdriver through the surround".

Best of luck Man, I hope I was of some help. I know all to well, how darned discouraging something like this can be.

 

[1n3]

I've been lurking on this thread. This is not as bad as I expected. I would run it and see what happens; you have nothing to lose by trying. If I was going to run at high power, I'd probably reinforce (stiffen radially) the crease with a combination of fiber and polymer. You being an engineer, I'm sure you can come up with something light and stiff but not brittle.

If the crease becomes a tear, you can try heavier repairs, and if that's unsuccessful, you still have the recone and replacement options.

 

[astack]

Thanks a ton for the input guys! I feel a bit better about it now If there's no harm in using it as-is, I'll forge ahead. Phew.

As long as it doesn't burst into flames and melt the box... Does this mean I don't have to break it in haha

 

[will33]

It's not nearly as bad as it could've been.

A crease like that will compromise the structural integrity of the cone and you can't really get all of it back. It's the reason woofers that are creased from overexcursion can't really be "fixed", aside from a re-cone.

But, yours doesn't look to be that bad. I might be inclined to put a smear of rubber cement along both sides of the crease (front and back), and just keep an eye on it over time. The constant movement from playing/flexing may want to weaken that area more over time, or, with the cone being weaker in that particular spot, it may want to rock at higher power levels.

For now, I'd just keep an eye on it. If it seems to weaken in the future, then maybe dream up some sort of flexible fiber type stuff to maybe mix with a rubber cement layer to help stiffen it some.

All is not lost. For now, it's just one more variable.

And thanks for the "fillet". The weld analogy I can totally understand.

 

[chadds]

Also, what the hell is a fillet? I'm thinking of cleaning fish here.....

From the title, I'm thinking of a Sanny!

 

[astack]

From the title, I'm thinking of a Sanny!

Ha, om nom. Yeah, I'll take that "Most Cryptic Thread Title" award now

Man, can't thank you all enough again. Still embarrassed by that snafu, but getting the answers makes it worth it The reason I'm so anal about projects is that stupid stuff like this always happens, so just trying to make it through with as few of those as possible. I'd guess I've got one more coming...

Alright, enough of this love fest. Back to bass.

We're ahead of schedule packing, so I'll take the extra time for another info post (don't worry - there're pics!)


Tonight's topic: Box dimensions and proportions

I already mentioned I got the Vb from WinISD (I guess I should go back to that eventually...) and my desire to keep it on the smaller side. So how did I turn that into dimensions? If you guessed a spreadsheet, you get a prize!

Gray boxes are inputs, yellow are outputs/important. The cells in K9 to M14 are proportions I got from that "Great Sound Stereo Speaker Manual" book.

The main concept is that for an even response, you don't want the box to resonate. That includes panel resonance and air resonance inside the box. A perfect cube will resonate more than a funkily proportioned box because the distance between each surface is the same and therefore the "fundamental frequency" or mode of each dimension will be excited at the same time Yeah like that. (It's the same reason you don't want your mastering studio to be a cube either.) Hmm, or like with a finger creating a node at a harmonic, the string will still ring. At a funky interval, nothin. That concept applies to 3D, too.

OK, so this is a 2nd order effect, if not even less critical. But if I'm making it from scratch, why not try? And I have no other rationale for other dimensions, so that's what I did.

The proportions in the book are just common "rules of thumb" proportions that some people dig for this stuff.

So the numbers are great, but I'm a visual person, so I went to Google's trusty SketchUp to see what these looked like:

With "Sang" and my "is that an Aggie in your pocket" TH500 for scale.

I love SketchUp

I don't think those were the final sizes, but you get the idea of the proportions. I dig the tall and shallow for aesthetics (think the old 2x12 Bassman cab) and portability. I think I've read that a deeper box is better acoustically, but again trades offs...

I went with the one with the ports (3rd from the right).

With the general proportions down, I went back to the spreadsheet and "fine tuned" the numbers.

A note about precision: It cracks me up when I read on here someone saying "3.025309344 cu. ft." Honestly, any manufacturer is lucky to get "3.02 cu. ft." There's a reason I'm using 3.1 cu. ft. and not 3.1000. Anyway, engineer out.

So, I got it close enough. I could never find the actual volume of the Faital. From my rough calcs and some stuff I read online, I got 0.08 cu. ft. Looks great. I don't know what the bracing will end up as, so I just earmarked some 2 in. sq. cross pieces in all directions.

Now for the cool part: Next I went into SketchUp to study how this thing goes together. First, I modeled all the pieces and grouped them into components:

Then I could "build" the box out of components. If something changed, it's really easy imo to edit the individual components rather than the big mess of a box. (Edit: Finally, I checked that the shallow box wasn't too shallow for the ports to be in the front (or back). Same minimum one port diameter clearance rule of thumb.)

And we have a virtual box! As modeled by Bryce:

Man, he makes 'em look so light!

(Note: Tolex and grille cloth TBD)

 

[1n3]

I dig the tall and shallow for aesthetics (think the old 2x12 Bassman cab) and portability. I think I've read that a deeper box is better acoustically, but again trades offs...

I also like the tall and shallow aesthetics, and have wondered about the acoustic disadvantages. Never have seen a good discussion of that.

I could never find the actual volume of the Faital. From my rough calcs and some stuff I read online, I got 0.08 cu. ft.

I once asked Eminence for the displacement of the 3015; it's 155.5ci, which is .09cf. I assume from the similarity of construction that they have similar volume. Published weights are nearly identical.

Speaking of volume, Sketchup will give precise volume for perfectly sealed objects. If you take a complete cabinet model, delete all but the interior surfaces, and seal the inner port ends and speaker openings, it will report your interior volume. I've found that useful.

 

[astack]

I once asked Eminence for the displacement of the 3015; it's 155.5ci, which is .09cf. I assume from the similarity of construction that they have similar volume. Published weights are nearly identical.

Awesome, I'll take it. I thought about asking around about it, because there have been a good number of discussions on here about this speaker but nobody's mentioned the actual volume. And I don't speaker italian...

Speaking of volume, Sketchup will give precise volume for perfectly sealed objects. If you take a complete cabinet model, delete all but the interior surfaces, and seal the inner port ends and speaker openings, it will report your interior volume. I've found that useful.

Oh yeah, totally forgot! I could go back into my model and explode one of the boxes, delete all the outside stuff as you said and get it. Best of both worlds. I'll repeat: I love SketchUp.

 

[1n3]

Awesome, I'll take it. I thought about asking around about it, because there have been a good number of discussions on here about this speaker but nobody's mentioned the actual volume. And I don't speaker italian...

Faital has a US rep, though I've not had any contact. Don't know why manufacturers don't publish that info on spec sheets.

I could go back into my model and explode one of the boxes, delete all the outside stuff as you said and get it.

Yeah, that's what I did. With an interior model in hand, you can work backwards, stretching to reach a desired volume, then extruding to recreate the outer shell. For those averse to spreadsheets...

 

[will33]

One of the tabs in WinISD Pro Alpha has a thing where you can enter all your basket dimensions and calculate a volume for the driver but I've never used it. I just take the diameter of the basket where it comes through the inside of the baffle, the diameter across the back of the magnet, average the 2, and use that as the diameter for figuring the volume of a cylinder. It's close enough.

There's also this...

http://www.calculatorsoup.com/calculators/geometry-solids/conicalfrustum.php

+1 to the 3.032548762 as being 3 cubic feet.

I call my 12/6-112 pair 2.27 cu.ft. because that's what I measured after they were done, even going so far as to add about a 9 cubic inch piece of solid foam inside one of them to make them "perfectly identical", but there's still the woofer fudge factor there. In reality, they're "about 2-1/4".

 

[mulchor]

Seems like you'd do reasonably well estimating the volume of a speaker as a truncated cone. If you don't have a high schooler in geometry around, a cone's volume is 1/3 pi r^2 h, where r is the radius and h is the height. So subtraction and some simplifying yields 1/3 pi h (r^2 + R^2 + rR) where R would be the radius of the driver which is about 6.7" for a 15" and r is the radius of the base of the speaker.

It's about the same as pretending it's a cylinder, but there's no reason not to use the correct model.

edit: or follow the link above, duh, it's got the proper formula np

 

[will33]

Seems like you'd do reasonably well estimating the volume of a speaker as a truncated cone. If you don't have a high schooler in geometry around, a cone's volume is 1/3 pi r^2 h, where r is the radius and h is the height. So subtraction and some simplifying yields 1/3 pi h (r^2 + R^2 + rR) where R would be the radius of the driver which is about 6.7" for a 15" and r is the radius of the base of the speaker.

It's about the same as pretending it's a cylinder, but there's no reason not to use the correct model.

edit: or follow the link above, duh, it's got the proper formula np

You're very right about that. There's still some variances between different drivers as to how far the magnet sticks out behind the basket. That little part would be an actual cylinder. Some have basket arms that have an angle in them, others are straight, etc.

Always best to be precise, but, difference of a couple/few cubic inches is well into the area that is "no audible difference". Port tuning difference of some fraction of a hz are the same way. It is cool to know exactly what you have though.