I'm not sure the diameter is exactly right, but opening up the inside of a roller blade wheel on a lathe should give you what you want.  Slide it on the end of the tube, match drill 4 holes, and pop some rivets.

If the throat of your lathe is large enough, you could also chuck the tube, and spin the flange onto it.  3:30 into this video for the idea.

http://www.youtube.com/watch?v=S0HGAfVJMcg


On Feb 17, 2013, at 3:41 PM, Charlie England wrote:

>
> Now that everyone's awake again, I thought I'd send a pic of my intake progress (RV-7 Renesis with James Lyc cowl). I played with a couple of different materials for the bell mouths. I tried gluing up some 1/4" plexi from an old windshield, but used super glue instead of proper plexi cement (which I wasn't able to find locally in a hurry). The 1st try popped apart on the lathe; the 2nd turned out ok. Next effort was with 3/4" MDF (medium density fiberboard). That went well, until I put a little too much side pressure on the ring (homemade cutting tool) after undercutting the center section. Overall tube lengths will be ~11 3/4" block surface to bell ends. The plenum is *much* bigger than most tuning sites recommend. I figure that I can experiment with plenum size by just stuffing it with rigid foam to take up some volume, if needed. Going the other way wouldn't be so easy. :-)
>
> Since I don't have Mark's TIG skills, I thought I'd ask what others have used in joining thin wall tubing to 1/4" aluminum plate. Aluminum brazing rod? High temp epoxy? JB weld? I do intend to add bracing from the plate to the plenum assembly to take some of the cantilever & vibration stress off the tubes.
>
> I'm using this length and concept because Tracy has had great luck with both HP & BSFC on his Renesis with a similar configuration. However, I'm curious about how others have adapted the common Helmholtz intake tuning formulas to the rotary. Would anyone care to 'show their (math) work'? Renesis users would be better for me, but any calcs would do. When I tried to adapt the common formulas to a rotary, I was getting 'interesting' results, so I'd like to know if I got lost somewhere while trying to plug rotary 'valve' timing into the formulas.
>
> Charlie
> (Sorry for the sideways iphone pic; I guess you can pretend that you're looking down on it...)
> <photo.JPG><Attached Message Part.txt>--
> Homepage:  http://www.flyrotary.com/
> Archive and UnSub:   http://mail.lancaironline.net:81/lists/flyrotary/List.html