Back in November 2013, I got the first order of new tinted glass for our Roamer. Those panes were for the aft deck enclosure, the original helm windshield, and a couple of salon windows that shattered by strong winds and careless stacking back when we first started the project. If I’m going to splash the boat this year it needs to be waterproof, and that means I need new glass for the portholes, the helm side windows, and a few other places. The old salon glass has been etched over time, and the new tinted glass looks really good with the Matterhorn White Awlgrip paint. So I’m also ordering replacements for all of the salon glass.
On the first order of glass, I made patterns from 1/4″ luan plywood. What I didn’t think about at the time was that the glass shop would measure the patterns I supplied and plug the dimensions into a huge CNC glass cutting machine. When I got the invoice, I realized that the measuring charge per pattern is almost as much as each pane of cut glass itself. So I decided to measure the windows myself, lay them out in the freeware Sketchup CAD program, and send the glass shop 2D renderings with all of the dimensions listed. For regularly shaped windows–rectangles and circles–that’s a breeze. But there are few regularly shaped windows on my boat. Even the ones that look square don’t have 90° corners, and the longer the piece of glass, the more precision you need when measuring.
I spent a lot of time looking for an accurate digital protractor. The good ones are expensive, but this looked like a situation where it would pay for itself. I almost settled on the Bosch DWM40L, which indicates an accuracy of ±0.1°, but some online reviews suggested it’s inconsistent even at that level. None of the other angle finders came even close. At least one had a display that went out to 0.01°, which sounded great, but reviews indicated that they weren’t even capable of 0.1° accuracy. Ultimately, I couldn’t justify buying a tool that might cause expensive problems for me. So…I improvised.
The Starrett miter protractor is intended for miter cuts. So 0° indicated by the inner arrow = a 90° corner. With 2° per tick mark, is that inner arrow pointing at 1.00° or 1.10°…maybe 1.15°? In any case, it’s indicating somewhere around a 91° angle.
So, rather than relying on the protractor to try and get the angles right, I’ll use framing squares, rulers, and measuring tapes to get the dimensions as if the windows were square, which gives me a 90° right angle to work with. The edge of the glass will become the hypotenuse, and I can use trigonometry to get accurate angles from the length of the opposite leg of the triangle–that 9/16″ gap in the pic above. Better still–with known lengths of the hypotenuse and opposite leg, Sketchup will do the trig for me and give me the angles.
The Bosch laser measurer was especially useful for measuring from the bottom of the helm side window tracks, which are too narrow for the tape measure.
I don’t want or need three inches of overlap between the fixed panes and sliders, so I’ll adjust my measurements to reduce it.
The Sharpie markings reflect the dimensions I want for the new glass.
The bottom edges fit fine into the window tracks, but the back edge and the corner of the top edge just barely enters the track. That lets cooled or heated air inside the boat escape outside. In winter, the breezes that can come through those little gaps is enough to blow out a candle. I need to add 3/8″ to the top edge and 1/4″ to the back edge at this corner to properly seal the new glass.
To make the right triangle trig work out, I need to start with a framing square along the top edge of the glass. I used a small clamp to hold that in place then used another framing square lined up with the first one to pull the dimension for the opposite leg of the triangle. But because my framing squares are standard size, they’re not long enough to extend all the way along the adjacent leg of the triangle. So I used a metal ruler to extend the line representing the adjacent leg and clamped another metal ruler 1/4″ off the end of the glass (adding that extra 1/4″ I need to the pane) to identify the point the new glass should exend to. I did the same thing for the top right triangle, positioning the second framing square and ruler to give an extra 3/8″ of height and 1/4″ of length to the top corner. It took hours to set this jig up.
After pulling all of the dimensions, I took the Rube Goldberg measuring device apart and then put it all back together again. The measurements were slightly off, so I repeated twice more until I got consistent dimensions. On the last go-round, the setup went pretty quickly. Practice makes perfect, I guess. Finally, I used the Starrett protractor to record rough angles, just as a check, and then headed home to fire up Sketchup and started plugging in numbers.
Hopefully, the glass order will be delivered in the next couple of weeks.