Welp boys and girls, made some real progress today, and then effed a bit of it up. So 2 steps forward and all that.
Let's back up before we get to that and talk sequencing. One reason this stuff is trickier than ply is there are a lot more ways one could build a box. Zac and Rob epoxied all but one plain foam pieces together and then glassed. Then they plopped the final piece in and sealed it. The other extreme is making all the panels separately or laying up one big sheet to cut up. This method goes a lot more like ply - basically how Jmac did it with his prefab composite panels. Petebass went with a hybrid approach of glassing the inside faces of the side panels and both faces of the front and back panels (as in the second method). Then he epoxied it together and finally wrapped the final glassing around the whole mess (as in the first method). I liked his thinking and it seems like a good middle ground in ease of construction and strength. So I'm doing it that way.
So that means there are 8 sides to glass before this thing starts to look like a box. Last week I got through 5 and today my plan was to get through the last 3 (inside left, outside baffle and outside back) so I could make a box tomorrow. Spoiler: didn't happen.
Today started out great. I waited for the cool morning fog to lift then got to work cutting glass. I'm using a 4'x2' sheet of melamine as a cutting board. I've found it's much easier to make a clean cut using a cutter and straight edge compared to scissors.
So clean.
Once I got 2 pieces for each layup, I moved on to preping my "molds." I've been testing out using melamine sheets as a mold surface and a poor man's vacuum bag. (Epoxy doesn't stick to the stuff.) Plus I've designated one side as the cutting board shown above, so double duty. I'm not sold on it just yet, but I like the idea of it maybe consolidating everything. The surface, however, unless I used way more epoxy than is needed doesn't come out all that great. Today, I stuck with the plan though (mostly...) and used them. Even though epoxy isn't supposed to stick to it, I don't want to risk it, so I waxed the surfaces before moving on:
Instructions: Wax on. Wax off.
I've also experimented with using mylar plastic for the same purpose, but instead of adding weight, you can either use a hard roller (like used with tolex) over the sheet to work the epoxy down, and then you're left with a nice shiny surface after it cures. Nice in theory, but I wasn't digging it as much as I thought I would. I think for the outside layups, I'm not going to mess with mylar or melamine. The print through of the fabric weave will get covered and there are less surprises the next day without a surface mold.
With all that prep work out of the way, I can get down to the real work, the fiberglass layup! Here's everything assembled at ground zero.
The process is part relaxing part cathartic and part hectic when you're in a race against the epoxy, but having everything ready ahead of time moves that balance to the relaxing side, hopefully.
The other trick I use to keep my mind clear, is this handy little spreadsheet to tell me how much epoxy to weigh out based on the panel size, glass weight, number of layers, epoxy ratio and the density info from the MSDS:
I can updated it on the fly if I think it's looking a little dry/wet. Man, I love spreadsheets.
The layup itself is just something you have to get a feel for. Here's what I do: after mixing the epoxy for a minute or two, I pour 95% of it on the panel. Then I use that yellow squeegee to evenly coat the surface. Carefully, so I don't distort the weave too much, lift and drape the glass down. I typically use some fancy bubble-busting roller to consolidate everything at this point, but they don't last forever and unfortunately, and mine are shot ATM (see handles sticking out of that bag above). No worries, I just carefully used the squeegee instead and a chip brush. It worked out just fine, maybe better because I had to be more careful. That last 5% of epoxy is used where it's needed after you've worked over the panel once. Now give it a final passing. Here's what it might look like:
Here's where it got interesting. With all the pieces laid up, I went to arrange my curing setup. The original plan was to use my two sheets of melamine on two of the day's panels and then the last of the mylar on the front of the baffle. I layer all these together to add weight to the previous. Here's the beginnings of it:
But somewhere in the middle I got clever and thought I could use one sheet of mylar for two panels (front and back outside faces). It worked. Kinda. Until I noticed the panels started to slide past each other. At this point I tried to gently persuade them to line back up. It worked once. Then they just kept slipping. So I lost a third of the day's work when I had to separate them. Mega sad face.
I decided against giving it a second go after that, as I was just out of patience for the day and it was getting late. Anyway, moral of the story is, stick to the plan and don't try to use one sheet of mylar for two surfaces at the same time.
Tomorrow I'll try that panel again and hopefully trim up the rest of the pieces in preparation for assembly
)
One of the biggest question marks I've had is about the round over. I've read that corners tend to have the wrong internal radius spec'd, so I couldn't really test anything until now. That was correct - the 1/2" R is closer to 3/4" which I'm actually happy about. 
I mainly just wanted to put it as high as possible. As for the ports, I made sure to keep one diameter clearance (3"). Oh and I chose to front port. From what I've read, it doesn't make much difference where you locate them, unless they're shooting at a boundary within 1/4 of a wavelength. I don't want to think about that when I have to set up on a tiny stage, so I was anti-rear porting from the get go. Finally, there's enough clearance for a grill frame around the edges.
Glad I'm through with that step. Gotta figure out a better method for cutting out the jack plate...
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.
