I cut a panel to approximate the wing window size from 2024-T3. As noted above, the full sheet wouldn’t even come close to curving into the shape of the shell. I was thinking maybe the smaller piece would do better. Before I could try it out, Aerowood loaned me his roller tool and suggested I give it a try on a shot bag. Not having a shot bag, I settled on a piece of my firm rubber floor mat.
Viola! In about 10 minutes of rolling and a few more minutes running in and out to the Overlander to see if I was getting and effective compound curve, it looks like it’s working. I won’t be able to fit it to the Safari until Wednesday.
The technique is to put your full weight on the roller and move it rapidly (maybe speed has nothing to do with it) back and forth along the lines shown in the photo. Actually, there was a fair amount of orthogonal rolling cross ways and length way, too. The objective was to get the sheet to deform outward [very slightly] approximately where the gray lines converge. This would force the upper outside corner to bend inward the 1/2″ that was required.
From the photo it may appear that the panel is only single curved, but even though it’s curved more in the lengthwise direction, it’s also curved cross-wise, just not as much. Once I can put the panel against the shell I’ll see how much more rolling is needed, if any.
The green lines show where the panel edges conformed to the shell without any rolling at all. As you can see, the area of curvature is quite limited.
Unbelievable, but the panel fits almost exactly. Normal spring type clecos were sufficient to pull the panel tight to the shell. Kip, my hat’s off to you for your tool and your suggestion–I didn’ tthink it was possible. Here is the untrimmed panel (1/4 to 1/2 inch extra on three sides) installed with clecos:
Not wanting to get back to [ugh] sanding the frame, I took some time to admire the panel–sorta ick. I am now thinking that maybe it would be a great esthetic to cut the panel in half along the line shown below (with suitable overlap, etc.) and to extend it at the corners to look more like an original panel than a window cover. The rivet line would be a natural extension of the existing dome joint and might attrack the eye in a positive way. It’s no Buck Rogers helmet, a la a Wee Wind, but it borrows from that tradition. What thinks the peanut gallery?
Remember post #58 above? Well, here’s the final wing window conversion. Looks tacky as hell, maybe polishing the shell will help. But the rivet line is perfect.
In order to put the patch under the dome skin, the edge had to be trimmed to get rid of the ragged edge and the window frame rivet holes. So there’ s no going back to the window shaped single patch panel.
This patch, even though it is in two pieces, still had to be rolled to get a slight compound curve.
I got the wing window skins riveted in. UWE had suggested that taking the cut line down straight to the corner of the front window might be improved by curving the line. For those who might be contemplating this “fix” to their wing windows, here’s a comparison of the look, Overlander on left, Safari on right:
I think maybe UWE was right, although the projected joint on the dome skin could really go either way. I think esthetically you eye wants to keep the curve going.
Here’s a frontal view so you can see the rivet line along the edge of the front window.