Making an Interior Dome

Most of the work was in getting the forward dome down. The next big effort in this area will be putting in the segments to replace it. I’m going aluminum all the way. Many “hidden &#*$(%*$&^#^#” rivets were found during this exercise. Urge to maim, but reached for the 2-Buck instead.

Click image for larger version</p>
<p>Name:	IMG_7752 dome down-s.jpg<br />
Views:	82<br />
Size:	148.5 KB<br />
ID:	51880Click image for larger version</p>
<p>Name:	IMG_7751 dome-s.jpg<br />
Views:	66<br />
Size:	123.0 KB<br />
ID:	51881

—————-

Today was the beginning of puzzling out the templates for the new dome panels. I started by assuming the middle panel would be 18″ wide at the top and 12″ wide where it joined the window frame. The other four would be 12″ wide along the shell end (the end of the panel away from the window) and 8″ wide at the window. The result is in the left photo. I didn’t like that very much, since the bottom panel didn’t even make it to the window, so I moved the lines over a little (not very obvious in the photo, I know, but it made a big difference), which you can see in the photo on the right. (the colors are photoshoped in to make the panel edges more visible.)

Click image for larger version</p>
<p>Name:	IMG_7773 colored wide panels-s.jpg<br />
Views:	61<br />
Size:	105.5 KB<br />
ID:	52187Click image for larger version</p>
<p>Name:	IMG_7773 colored panels-s.jpg<br />
Views:	63<br />
Size:	146.0 KB<br />
ID:	52186

Then I went back and studied JCFerguson’s work. I like his panel shapes much better. His what I am calling #6 panel really makes a difference in the “fatness” of the shell end of his panels. What I’m wrestling with right now is what the slope of the bottom edge of panel #6 ought to be–I’d like it to leave the shell at a more horizontal angle. Maybe that’s not possible.

Click image for larger version</p>
<p>Name:	carlos full dome.jpg<br />
Views:	58<br />
Size:	77.5 KB<br />
ID:	52188

One thing I didn’t quite understand in Carlos’ post was his explanation of how he matched the tar paper to the lines–I have to assume he drew the lines on the fiberglass end cap/dome, which is the light colored substrate under the tar paper in this image (this and the previous image are from his posts)?

Click image for larger version</p>
<p>Name:	carlos partial dome.jpg<br />
Views:	50<br />
Size:	36.2 KB<br />
ID:	52190

I’ll put outlines of all my panel shapes in a later post. The #4 and higher panels are much more scimitar shaped than you would expect.

—————-

Thanks, Malcolm. What confused me is that in his post in his thread he said something that made me think he didn’t have an end cap (or maybe he just didn’t have ready access to his trailer at the time). It looks to me like he used his flexible straight edge to draw the lines on the end cap itself. I have to assume he riveted the plywood strips to each other and to the end cap.

I’m not going to have the luxury of an in-place end cap. I know I can drill the panels in place and use clecos, then take it outside to buck all the panels together. I’m thinking of drilling 1/8″ pilot holes, then making a fixture to simulate the curve of the shell end and the window frame, then final drilling to 5/32″ and bucking the whole thing on the fixture to ensure I get the curve right. I’m also thinking I’ll have to put a 20 degree crimp along one long edge of each panel and maybe have to shrink that edge a little in order to get a smooth final product.

BTW, I did have the luxury of an end cap in the post above–it’s the empty back bedroom in a mid-bath Ambassador. I can’t wait to see how that relates to the front end cap in the Safari. Can the end caps in the 70′s models really be 3′ and 4′, not both 3′?

—————-

Today is “Dome Day.” Maybe I should say “Dome Eternity.” The objective is to replace the plastic forward dome with 11 aluminum panels. The dome should use solid rivets, except for the edges when it is installed to the interior skin and ribs. This requires the dome to be fitted, drilled, clecoed, unclecoed, removed, clecoed, riveted, the installed. (others have done this in a simpler fashion but, true to Andy’s comments elsewhere, I am willing to take extreme measures to achieve modest results see axle thread for proof of that!)

First step is to get some trial templates. Luckily, I could use the dome over the rear bedroom of a Sovereign, so I had some thing stiff and of the right shape to lay the paper against.

Stiff paper seems like a good idea but the first conceptual error was that I could draw one edge per piece of paper, take them down, then transfer the other edge from the adjoining piece back to the first piece–I mean, once they’re out flat you can measure, right? Wrong. Well, not in theory, but if you want to drive yourself nuts, try it. You got to draw both edges on the same piece. Then make sure you lay up the next piece and copy the now underlying join line onto the next piece, plus it’s other side, etc.

But the paper just didn’t feel adequate (a better craftsman can get away with it, but I needed something stiffer). A sheet of thin plastic seemed the right answer ($20). Wow, much better, but the plastic still had the tendency to form a slight trough shape along the long axis. This doesn’t seem like a big deal until you see how much the join edges move with slight changes in how tight the template is to the dome.

Click image for larger version</p>
<p>Name:	IMG_7973 rosin paper-s.jpg<br />
Views:	41<br />
Size:	85.5 KB<br />
ID:	56432Click image for larger version</p>
<p>Name:	IMG_7984 standing-s.jpg<br />
Views:	51<br />
Size:	111.5 KB<br />
ID:	56433

After a lot of thought, I decided that panel #5 should have a horizontal lower edge, so all templates were shifted and re-cut to obtain a pattern where #5 meets the corner of the window and has zero width at that end.

I finally got the courage to cut some metal panels. They went up fine, but whoops, the overlap wasn’t sufficient. This could have been easily corrected by bringing the curve out from the shell just 3/8″, but I wanted to maintain the 1-5/8″ shell depth and dome shape as closely as possible. You can see the pink foam shims that I used to maintain spacing. These shims also provide sufficient support so that
the panels don’t flex when drilled.

Click image for larger version</p>
<p>Name:	IMG_8002 foam blocks-s.jpg<br />
Views:	49<br />
Size:	90.4 KB<br />
ID:	56434Click image for larger version</p>
<p>Name:	IMG_8000 4 panels standing-s.jpg<br />
Views:	55<br />
Size:	85.5 KB<br />
ID:	56435

The next step is riveting the the dome together. I don’t like depending on the clecos to pull the dome into the proper shape. I’d hate to get this thing together and then not be able to fit it to the shell. You can see some ripples in the edge of the panels, which tells me I also want to flange them slightly and then shrink the flanges to make the proper curve in each panel. Thanks to AEROWOOD, this may be practical.

Click image for larger version</p>
<p>Name:	IMG_8010 back shape-s.jpg<br />
Views:	47<br />
Size:	90.5 KB<br />
ID:	56436Click image for larger version</p>
<p>Name:	IMG_8009 edge shape-s.jpg<br />
Views:	50<br />
Size:	130.5 KB<br />
ID:	56437

Click image for larger version</p>
<p>Name:	IMG_8011 inside clecos-s.jpg<br />
Views:	57<br />
Size:	97.1 KB<br />
ID:	56438

In order to ensure a fit, I’m going to make a dome form from 3/8″ plywood to duplicate the perimeter and center of the dome. The following two details are posted to aid phone discussion with the expert…

Click image for larger version</p>
<p>Name:	IMG_8008 edge detail-s.jpg<br />
Views:	52<br />
Size:	84.6 KB<br />
ID:	56439Click image for larger version</p>
<p>Name:	IMG_8012 missed holes-s.jpg<br />
Views:	49<br />
Size:	56.3 KB<br />
ID:	56440

Every day with a cleco is a good day.

——————–

Spent the yesterday getting templates for the form and today making the durn thing. Turned out OK. All the layouts fit in both dimensions within 1/8″, so I’m pretty confident this form is very close and will really help in when I [solid] rivet the panels together outside the trailer.

Getting the form shape turned out to be easy, but time consuming. I’ve done a lot of interior oodwork fitting and always took 50 trips back and forth from the bandaw and belt sander. This time I used narrow (like 10″) strips of luan that spanned about half the desired curve, then clamped them together when the shape got close. Took about three tries. Marvelous! Why didn’t this occur to me 9 years ago when I was younger and had consumed less Two-Buck Chuck?

Click image for larger version</p>
<p>Name:	IMG_8021 curve pieces-s.jpg<br />
Views:	38<br />
Size:	80.9 KB<br />
ID:	56546

Here’s the finished form. It can be disassembled and reassembled inverted to provide a form for the the other side of the dome.

Click image for larger version</p>
<p>Name:	IMG_8022 dome form-s.jpg<br />
Views:	41<br />
Size:	106.4 KB<br />
ID:	56547

———————

here’s the form in action. These are “trial” panels that are pictured in a previous post with the wrinkles” on the back side flange. Now that they are in the form and the right shape, the wrinkles are essentially gone.

Click image for larger version</p>
<p>Name:	IMG_8029 inside form dome-s.jpg<br />
Views:	39<br />
Size:	91.1 KB<br />
ID:	56566Click image for larger version</p>
<p>Name:	IMG_8027 outside dome-s.jpg<br />
Views:	48<br />
Size:	83.3 KB<br />
ID:	56567

—————–

It’s been several days of trial fittings, flanging edges, shrinking the flanges to get the right curve, then more trial fitting and finally hole drilling. Yea! After all the problems fitting the curb side, the street side went quickly. The dome shape allowed all panels to be stable with only clecos in the window frame and the shell ends of the panels.

Click image for larger version</p>
<p>Name:	IMG_8042 street side fitting-s.jpg<br />
Views:	53<br />
Size:	87.7 KB<br />
ID:	57043

Here’s the final fit and drilling of the curb side. Rivet spacing is 2″ in the center panel and 1-1/4″ in all the others. I think it’s been over 2,000 cleco in-outs so far in getting the panels fitted, so my grip is getting better…

Click image for larger version</p>
<p>Name:	IMG_8049 curb dome fitted-s.jpg<br />
Views:	60<br />
Size:	138.8 KB<br />
ID:	57044

(no, that’s not a large HD TV under the dome!)

It turns out the metal dome is a slightly different shape from the plastic dome–the plastic dome only came down to the top of the wing windows and to the main window box top rib above the center window, but not to the center window. That’s about two inches difference. So the last (lower) panel’s lower edge starts off from the wall on the horizontal (at the same place the plastic dome met the wall), but then slopes down as it turns, to meet the top corner of the window. This panel actually comes to a point right at the corner of the window frame.

Also, the inner metal dome is nearly a constant 1-5/8″ from the outside dome (as constant as I could make it). I think the plastic dome had more space in many places between it and the outside shell. I surmise this from the poor fit of the templates that were made against the dome in the rear bedroom of the Sovereign.

With only half the dome installed it is unbelievably stiff. What a great improvement over the plastic dome. Now I’ll take it down, put it back on the form, do the final hole drilling, and rivet the five curb side panels together.

I have to admit to a late epiphany today. I know why the older Airstream domes are interleaved the opposite direction, eg, the upper edge of the panels are on the inside, versus mine where the lower edges are inside–with the Airstream design you don’t have to flange and shrink the hidden mating edge. This would have saved me at least half the hours I’ve spent so far. But esthetically, I like my interleaving better.

Speaking of flanges, here’s the tool (another AEROWOOD special) that does it. I had no clue about this tool until Kip showed it to me–now I see a number of different varieties of the same tool in most sheet metal tools web pages!

Click image for larger version</p>
<p>Name:	IMG_8033 flanging tool-s.jpg<br />
Views:	53<br />
Size:	66.5 KB<br />
ID:	57045