Any details like cost, yield (Ratio of good to bad) etc would also be appreciated...

The frame looks quite clever...

Is that angle iron stiff enough to completely eliminate flex? If you pour plastic around the cells and something flexes down the road, those cells will certainly crack.

You solder them in series, the front of one to the back of the next one, all the way down the 36 (or 35 in your case) cell line. When you get to the end of a row, you build a little bridge out of leftover ribbon or something to go to the next row of cells.

Here's the back of one of my panels, maybe this will help you undestand the connections. My arrays are only two cells wide and I have two of them per panel.

I found it easiest to tin the pads of the cells first and then connect the ribbon. The flux of the rosin core solder helps a lot to make the solder stick to those pads.

If I could do mine all over again, I'd try really hard to find a way to laminate the cells between plastic and tempered glass, then build a sturdy frame around it. I got an old Siemens 36 watt panel out of the trash the other day that had a bullet hole or three in it. It looked to me like the cells might have been sitting on a thin layer of foam when they laminated the panel.

What sort of "pourable plastic" are you planning on using?

Volvo Farmer

Stop.
Measure your polycarbonate sheet indoors to the nearest 1/64th inch.
Take your polycarbonate sheet outside into midday full sunshine.
Place it in square track with suns light and allow it to heat up.
Support the polycarbonate w/ dry plywood or stiff cardboard backing, it softens alot.
After 20 or 30 minutes measure the dimensions again.

I hope you take the time and post back your findings, the steel frame lengthens also but not nearly as much as the plexiglass. I think you will do best if you do not bolt the plexiglass at all, assemble array strings and seal into a coverglass-backing assembly and then set into loose fitting frame w/ small angle brackets scattered around periphery that give mechanical grip w/ small rattle room against wind but will allow 'assembly' to heave and shrink w/ temperatures.

http://www.fieldlines.com/story/2005/5/14/203536/676

"Whenever screws or bolts are used through holes drilled in the Plexiglass as for the securing of an aircraft windshield or window, consideration should be given to the expansion or contraction of the Plexiglass as effected by changes in temperature. Plexiglass expands or contracts with temperature at the rate of 1/16 inch per foot per 100 degrees F. temperature change. For example a line of holes at the top of the windshield 36 inches across using a 1/8 diameter. screw would require that the holes be 5/16 inch in diameter to take care of a temperature spread of 100 degree F.. When holes are drilled with insufficient clearance excessive strain is placed on the windshield and may cause the holes to crack out at high or low temperatures."

Also if your not using a drill bit designed for plexiglass your can use a standard angle drill bit if you flatten & dull the cutting edge so it scrapes its way through instead of biting too deep and cracking plastic.  Change cutting edge _  to |_ for the 1st 1/8th inch.

Soldering - http://www.fieldlines.com/story/2005/5/12/03013/2634

Remember to keep solder points as smooth and evenly low as possible - a ball of solder will act as pressure point for every vibration from here on out - the ebay 'evergreen' cells are laced w visible & invisible flaws and crystaline structures that want to snap apart more then they want to stay together. If the solder stands tall the expansion/contraction cycles of the epoxy resin over time may 'toggle' the bond into open circut. Just the weight of other cells on top during handling and assembly can cause whole 1/4 segments to snap off.

http://www.epoxyproducts.com/25points4u.html

One gallon of epoxy will fill 6.5 square feet 1/4" deep. 36 cells at 18 sq inches each equals 4.5 sq foot, with borders and spacing your're looking at 5 sq ft so you'll get about 4/10th inch coverage.

If the resin is two-part look at the specs for shrinkage - using extremly small amount of hardener and allowing 24 hours + cure time, some epoxies even by placing assembly in protected tent in sunshime get suns heat to help set it up, keep epoxy cracking down and tension down in silicon wafvers by slowest cure you can get. PV cells do not do well under tension.

If your resin has VOC's (voilatile ingrediants) it is sure to shrink.

I hope you're not planning on bonding the cells directly onto the lexan - there is too much movement and the cells will eventually be torn apart. They can stand zero flex. Using glass is advised for encapsulated planels unless you make the epoxy the top surface and have a stiff weatherproof board backing - I've seen marine buoy arrays layed up on top of 3/8" dense fiberglass board, like a circut board on steroids that has nill flex to it, so the cell surfaces are protected by a layers of epoxy built up to 1/8" or so. Easier to renew the epoxy surface every 6 months then trap marine saltwater under glass, capillary wicking will kill cells so multiple layers front and back may be the best bet, copy a design meant for worst conditions.

More reading:

http://www.fieldlines.com/story/2005/4/18/231236/457

My 50lb lot of cells - I'd decided on $1000 of storebought panels already so I spent the $1000 on raw cells. I am still working on best design for me, I'd hate to screw up dozens of panels cause I got impatient!

I have found that expansion over time causes the top ribbons (tabbing) to come loose on cells that do not have the tabbing perfectly aligned and well attached to the top solder strips. I find I've had a few panels go dead on me for two reasons. One is because of loosened tabbing. The other is because of corrosion in the solder joint between the main copper buss and the bolt that brings the power out through the back of the panel (a poor design I used on a few panels). As a result, I've had to do a little repair work resoldering the tabbing on the tops of the cells (a real pain) and redoing some of my buss bars and bolts (on the next one, I'm trying stainless steel bolts.

Bear in mind that in your setup, you're going to have expansion problems, and your panels will likely be impossible to take apart. Make sure the top tabbing is soldered down good before you seal it up.

I'm real curious to see how that perforated angle stock works out for you. I've been using 1 x 1 x 1/16 aluminum angle stock to brace the backs of my plywood base. I don't use it as a frame in an attempt to shave costs. Your material might be a little cheaper, yet be just as strong, if not stronger than what I'm using.