Upcoming development season..

[Ya-Nvr-No]

looks like a roots blower = paradoxic al gear

http://www.youtube.com/watch?v=-1Gfc1Iq0GY&NR=1

Back in the late 80's I helped startup the Eaton Superchar ger plant (Athens Georgia) where we made the 90 cubic inch Ford Thunderbi rd & Cougar unit, had a pair of 60 degree three lobe impeller using a pair of ground 53 tooth straight tooth gears to drive them. The trick was the cutter profile shape that was a developed mathemati cal formula generatin g the cutter path the grinder followed. The lobes were hobbed in with a flat multi inserted bladed cutter all set on an angle. The lobe clearance s were critical due to the heat build up that would grow the Aluminum Alloy till they came in contact and would seize. (at the same speed, the closer they came without touching the higher the boost/compression) An assembly machine was built to find the center of the lobe backlash, hold them and then press the two gears and the bearing plate on at the same time at their backlash center clearance distance. After assembly the units would be tested and analyzed for 1st, 3rd and 53rd vibration/noise at various speeds.  You ought to hear one when it seizes at top speed. The computer display would show the critical informati on in real time and show a pass/fail. The oil on the gear side was a very high priced synthetic .

Sorry, took me back to memory lane, thought some would be intereste d.

Not that I am going to build a superchar ger again but in some ways Id be intereste d also in this added feature.

[bosr]

Program crash!

I was just designing a spur gear to be used as a frying pan knob.  DP= 10, Teeth= 8, Stub= .2, Shaft size= 0.

The program crashed ungracefu lly when adding the gear to the project manager.

Randy

[Mooselake]

Ya-Nvr-No

Do you happen to remember the equations (or a source) to generate multi-lobe impellers and a housing for them?  It might be fun to prototype them with something like openscad.

Thanks!


Kirk

[Ya-Nvr-No]

Sorry I had to sign a non disclosur e statement, it was all pretty well protected project. Was a amazing program the guy wrote thou. Could calculate for just about any size unit and any number of lobes.

[ArtF]

Hi Bosr:

   Its the zero shaft diamter.. Fixed for next version, but try using a .1 and then deleteing it from your dxfs or Gcode till then.

thx for the bug..
Art

[Chuck]

Art,

 The Roots Compresso r is easily generated it is simply 2 meshing circular arcs forming a cycloid centered on a circle around the shaft of each rotor.

_rotor_co nstructio n_-_2_lobes.svg" target="_blank">http://en.wikipedia.org/wiki/File:Cycloidal _rotor_co nstructio n_-_2_lobes.svg

If you do add this to GM you should make the pitch circle and the tooth size adjustabl e.   This will change the compressi on ratio.

You will also need to have a clearance allowance and you may want to add turbulenc e seal groves to the rotor tips but that doesn't need to be in the generator program.

Of course the Roots MUST be driven by a 2 matched gears outside the compressi on volume.

If you make them spiral you must consider the angle, too much angle and they won't pump any air. 

Chuck

[ArtF]

Chuck:

 Thx. I may indeed look at those once an escapemen t is ready. Its about half done now and looking proper.
Its the simulatio n that will be a bitch.

Art

[ArtF]

That link seems bad though..


Art

[Chuck]

Art,
cycloid rotor page link..... lets try again.  I am not used to this posting system.

http://en.wikipedia.org/wiki/File:Cycloidal _rotor_co nstructio n_-_2_lobes.svg

I tested it this time..... .

Chuck

[Mooselake]

John Tribe has a page that mentions an escapemen t analyzer "Graham Deadbeat Escapemen t Analyser Version 1.3 by GM Smith does a fabulous job of simulatin g the critical points of interfere nce and allows you to change parameter s and try again" at

http://web.me.com/johntribe/Learning_Again/The_Process_of_Building_a_Clock.html

I was unable to track it down, but he may be lurking around here somewhere .

I found an article on deepdyve that discusses two lobe roots blowers, and has lobe design equations on pages 2 and 3, although I haven't tried to actually test them (

http://www.deepdyve.com/lp/hindawi-publishing-corporation/clearance-analysis-and-leakage-flow-cfd-model-of-a-two-lobe-multi-0PvF32WwgM
)

Thanks for the link Chuck.

Kirk

[ArtF]

Thanks Chuck:

   That explains it much better.. it isnt built of arcs but of epi and hypo cycloids. .

 Doesnt look hard to make at all..

Kirk:

   I DID find that anyalyser once.. I think it was 300 Euro's to get it..
Seemed a bit high priced to me, but then..it analysed the escapemen t with a microphon e..
It listened to the Tick and Tock to figure out how the escapemen t was working.. Sounds like fun..

Art
 

[gaston gagnon]

Hi Art,
Talking Escapemen t, I use Mark Headrick's document to rebuilt the Graham escapemen t of my clock :http://www.abbeyclock.com/escpdf2.html It is working very well.
Gaston

Chuck:

 Thx. I may indeed look at those once an escapemen t is ready. Its about half done now and looking proper.
Its the simulatio n that will be a bitch.

Art

[ArtF]

Hi Gaston:

 I used that document and a couple others to check the math and ensure all was good.
I now have the escapemen t completed . You can use any number of teeth and any span
of teeth in your specifica tion.  Its kind of fun to be able to simulate various deadbeat designs
and tooth spans to see the resultant escapemen t for it.

   Im now working on the simulatio n so one can check the binding and use that to set any desired
lock and drop angle for the pallets. Il post photos as I get closer to a version I can release..

Art

[Mooselake]

Kirk:

   I DID find that anyalyser once.. I think it was 300 Euro's to get it..
Seemed a bit high priced to me, but then..it analysed the escapemen t with a microphon e..
It listened to the Tick and Tock to figure out how the escapemen t was working.. Sounds like fun..

Art
 

If that's the Delphi program I've seen it too, but IIRC it was to allow you to adjust an escapemen t (hm, another arduino project, maybe with a GPS clock timebase).  I did some more googling and found a magazine article by  G M Smith discusing his program at

http://www.docstoc.com/docs/28628085/Program-for-a-Dead-Beat-Escapement

If I find a link to the actual program, or the author responds to my email, I'll let you know.

Kirk

[ArtF]

Interesti ng article.. thx for the link. It makes a few points that were unclear to me in the mechincs of it all.
Ill see how my simulatio n works out to see how close I come to his statistic s.

Art