PDA

View Full Version : Machining impellers



loose nut
12-13-2010, 03:46 PM
Some time ago I found (and lost) a site on the web that showed how to manually machine a centrifugal compressor impeller, out of bar stock, of a type that would be used on a small home built gas turbine engine. Does any one no where that kind on info can be had. Info on casting one would be equally useful. Thanks.

alanganes
12-13-2010, 03:49 PM
No idea if it is the one you are thinking of, but check the "jet engine" post in the "networking" section of this BBS. Some great photos there and I'm sure the OP can direct you to the source of his design.

-Al

whitis
12-13-2010, 06:30 PM
If you need to machine curved vanes using manual machines, Tabletop Machining (sherline) shows a setup with a sing arm "table" mounted on top of a rotary table and an outrigger support and another rotary table mounted on the swing arm table. There was also a thread here fairly recently using a single rotary table with various alignment holes used to move the work.

You are limited to circular arcs, this way, but you can do pretty well with a well placed circular arc, particularly if you are more interested in pressure than flow rate:
http://www.ckdnoveenergo.cz/new/data/files/102/000000120_f.pdf

macona
12-13-2010, 07:29 PM
The ones in rockets and the like were done on tracer mills back in the old days. Modern turbocharger impellers are just cast and balanced.

I would just find one from a turbocharger like everyone else does.

.RC.
12-13-2010, 09:15 PM
http://img.photobucket.com/albums/v606/OzRinger/1945_10126Custom.jpg

Evan
12-14-2010, 12:52 PM
Look in the networking forum. Or the helping hands forum.

.RC.
12-14-2010, 03:45 PM
Do you want a bigger version.... The original I have is 3MB in size. Much clearer

gwilson
12-14-2010, 04:53 PM
That picture is quite the rig!!! But I'll bet it took a LONG time to mill that impeller. Especially with that LONG endmill hanging way down from the M head.

I have a Mimik tracer head for my Bridgeport mill. It also takes EL-MUCHO back and forth table work to make a decent surface that is ready for the riffler files.

macona
12-14-2010, 04:58 PM
That pic, plus many others if from archive.org

lazlo
12-14-2010, 07:56 PM
http://img.photobucket.com/albums/v606/OzRinger/1945_10126Custom.jpg

Oh, man -- now that's porn! :D

.RC. -- I'd love the full size version!

It looks like there's two tracer styli -- how does that work?

Liger Zero
12-14-2010, 08:15 PM
Oh, man -- now that's porn! :D

.RC. -- I'd love the full size version!

It looks like there's two tracer styli -- how does that work?


I assume one controls the rotary axis somehow.

A video of one of these in action would be cool.

Tinkerer
12-14-2010, 09:07 PM
http://img.photobucket.com/albums/v606/OzRinger/1945_10126Custom.jpg

It looks like there's two tracer styli -- how does that work?
To me it looks like the silver one is for the profile used first... the black one is for the impellers gullet... space between blades.

Toolguy
12-14-2010, 11:37 PM
It looks to me like they are tracing X,Y,and Z simultaneously, with a separate stylus for X and Y and both of those doing Z on the patterns. A 3 axis CNC mill. Crank N Cuss. I did a lot of Bridgy tracing with 1,2 and 3 heads when automatics were pin drive, then paper tape drive.

.RC.
12-15-2010, 01:34 AM
macona has it, I found a link here

http://www.archive.org/details/C-1945-10126

There are others of a room full of 10EE's and other lathes all in use..