Window Installation

Here's a phrase that seems to be repeating more and more: "Installing/building (fill in the blank here) took way longer than I had anticipated."  Window installation was no exception.   I actually started the prep work for window installation in September 1999, but then I turned my attention to building the glareshield.  If you're reading in sequence, however, you already knew that.  I'm writing this in May, 2000, some 8 months later.  Yikes, where does the time go?

The first thing to point out is that the windows in the SQ2000 get installed from the inside.  While I don't have direct knowledge of other composite kitplanes out there, it is my understanding that their windows get installed from the outside.  In a perfect world, installing windows from the inside kind of makes sense: by leaving the outer skin intact, you would, in theory, have little or no detail work to do around the outside of the windows.   You would just glue it into place, apply your overlay ply of fiberglass on the inside, and call it a day.  Click the image at left for a nice view of the theoretical interface.

Although I made the basic window cutouts long ago, more work was necessary before the windows could be fit into place.  The first step was to hold the windows in place and mark around them on the inside of the fuselage with a Sharpie.   This is really a 2-person job, especially for the windshield.  Luckily my Dad was in town for the weekend to help me out with that. 

Step 2 was to take the grinder to the just-marked areas of the window cutouts and remove inner plies--and in some cases core too--until just the last 2 OML plies were left.  This was a delicate operation; a little slip of the grinder and oooops, there's a hole in the skin.  It was tough, in some instances, to judge the skin thickness too.  I did a pretty good job with the grinding, but had a couple little break-through areas that I later fixed.  I started off using my pneumatic grinder with a sanding drum, but later found that I had better control with a Dremel Tool.  I also found that this job was much easier with the fuselage upside-down.

With the window frames prepared, I could now do a trial fit of the windows.  Here's where theory met reality. Due to manufacturing tolerances in the fuselage, windows or both, I was a little disappointed to see quite a few gaps where the windows met the skin cutouts.  The graphic at left illustrates more or less what I saw, although I've exaggerated the condition just to make my point.  Having worked on multi-billion dollar stealth bombers where we measured things to .0001" accuracy, perhaps my expectations were a bit too high.  Be that as it may, I now had to come up with a plan.

I came extremely close to just installing the windshield, thinking "screw it, I'll just install it with a thick bond line and deal with the resulting OML step later."  But I couldn't bring myself to do it.  That would be the "half-assed" solution.  Instead, I decided that adding a couple of very narrow plies to the periphery of the window cutouts would work: it may not completely cure the gaps, but it would certainly help.  If you're having trouble picturing this, pick the image at right.

"Adding a couple of narrow plies" doesn't sound too hard, but working with ~5/16" strips of glass is pretty tricky: they want to distort, separate, fall apart, stretch, you name it.  But I managed to get them in place, cured, sanded smooth and tapered.  Subsequent window fittings found that my plan worked.  With the exception of one troublesome corner on the #2 window, all gaps had been eliminated.  My time was well spent.  The last little job to complete before actually bonding the windshield in place was to peel back the protective Spraylat, mask around the faying surface of the windows, and scuff them up with a sanding block to provide better adhesion.   "Masking" is another of those "simple" jobs that ends up taking a lot longer than you'd think.

OK, now I was finally ready to install the windshield.  This is definitely a two person job as the windshield is a fairly large piece of plastic, and mishandling it could easily result in irreparable scratches.  Luckily The Herminator was willing and able to help me with this task.  The bonding process was fairly straight forward;  I simply mixed up a "wet" batch of epoxy and cotton flox, spread it evenly around the inside of the windshield frame, and dropped the windshield in place.  The fuselage was then lowered onto its nose and sandbags were placed on the IML to apply positive pressure during cure.   Meanwhile I removed all the flox squeeze out on the outside of the joint.  Then I placed my electric heater inside and sealed up all the fuselage openings to keep the temperature elevated during cure.

After cure of the primary bond, it was time to work on the inside joint.  This amounted to filling in the gaps around the windshield with flox and laying up a ply of 2.5" wide fiberglass tape over the joint as depicted in the first graphic on this page.  Of course I had to mask off the inside and scuff up the faying surface of the plastic first.  The hardest part about this job was having to work around the instrument panel bulkhead.  Oh yeah, working hunched inside the fuselage for long periods of time is also a bit hard on the back... I think I've mentioned that before.

The last step was to de-mask.   Because the epoxy tended to run down the masking, I had to use a flat-tipped hobby knife to cut it free... another time-consuming task.  That was it.  Windshield complete (almost).

Now I just had to repeat the process 4 more times for the side windows.  Since they were smaller, installation went a bit faster.  The only hurdle here was figuring out how to apply positive pressure to a side window.  I had 3 different thoughts: (1) draw a vacuum on the outside, (2) do two windows at once and use some sort of pole device to force them apart, or (3) somehow balance the fuselage on its side and use dead weight on the inside of the window.  With the help of my hand 3-legged leveling jig (and Herms again), option 3 worked just fine.  The fuselage wasn't too stable in this configuration, but happily no earthquakes hit Seattle and everything worked out.

Once all windows were installed, internal plies cured, and everything de-masked I applied a coat of Spraylat to the inside of the windows.  Spraylat is a milky substance that is sprayed or brushed on and cures to a colorless, pliable, protective film.

The last operation was for appearance and aerodynamics.  The outside skin was tapered to meet the windows, but a step was noticeable.  Not a big deal, but I figured I could make the installation look a bit better--and cut down on wind noise--by fairing onto the windows about .25".  With some careful masking and small applications of micro, I faired out the steps to a smooth transition.  That's about it.  Pretty cool, eh?

I have to say though, that after going through all this, I think it would have been much easier to install the windows from the outside.  One addition concern I had was, "what do I do if I ever have to replace the windshield 10 years from now?"  I presented that question to the factory.  Their response: "Don't ask."