Re: Electrode holder material for a high-power RSG

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Mark Broker by way of Terry Fritz <twftesla-at-uswest-dot-net>" <broker-at-uwplatt.edu>

Replies interspersed.


>
> looks like you do some very nice machining... 

thanks!  It's only good considering I was using a HEAVILY used WWII-era lathe. 
Tightening it up a bit is on my to-do list this week.


>
>    but...   consider using electrodes that are long enuf to be the total
path of current ( dont use 2 smaller electrodes per unit) it isnt that
difficult to grind tungsten electrodes to 2 thousandths accuracy.. make the
electrode holder as low profile as possible with as much electrode
extending above it. 1/2" is a minimal extended length at hi power.  ( at
15KV the arc isnt going to hit just the tips of the electrodes, at hi power
it will jump up to 1/2" or more to make contact)
> 
> also consider using socket set screws ( cup point style) to secure the
electrode to the holder and the holder in the disc. ( this will reduce air
friction) If you make the holes for the electrodes a light slip fit, you
will only need one set screw per electrode ( the 90 degree 2 set screw
pattern is needed when you have several thousandths clearance in the hole
/shaft assy. )
> 
> Since you hav!
> e the intention to round off the heads of the electrodes, try to make the
heads hemispherical instead of oblated ( bullet nosed).

I guess I forgot to mention that the tungsten carbide electrodes came in that
shape.  They are 8mm D, with a ~10mm long "shaft", and about 15mm overall
length.  They were free (thanks, Marc!), so I would like touse them as opposed
to purchasing some tungsten-carbide stock and grindingthem myself (although it
is still an option).

I also thought they were .312", not .315"(didn't have a caliper), so I reamed a
.312" hole.  I'll try heating the holder at school, and dipping the electrode
in LN2 (if we have any LN2).

I was planning on low profile set screws, most likely socket head set screws of
"appropriate" length.  I wanted to show an actual fastener in the CAD drawing,
and not just a hole.  I will be using only one set screw to hold the electrodes
in, since I intend to use the interference fit to increase safety.


>
> Another item to consider is the amount of force that is going to be
applied to the disc -at- 5000 rpm by the electrode assy. As the rpms increase,
the mass of the electrodes will want to shear out the thinner area on the
outside diameter of the disc, plus having a hole tangent to the disc for
securing the electrode assy in the disc weakens the disc even more. ( you
can either make the disc a bit larger in dia.((13")) and keep the 10" dia
for the electrodes or you can move the electrodes in another 1/2" (( you
want approximately the same thickness of edge
> measure as the dia. of the electrode holder... example if the electrode
holder dia. is 3/4" you want at least 3/4" of material from the holder to
the outside edge of the disc)).

This has been a concern of mine, as well.  I'm TRYING to get the ACAD model
into Algor FEA to determine if the 1/2" is enough at 5kRPM.  Unfortunately, my
former Algor Prof (who WROTE the textbook on Algor) doesn't know how to import
it (turns out the school doesn't have the proper interface package
installed.....).  I inted to get a feel for stress levels in this area before I
start drilling twelve 1" holes in G-10....

I'll eventually construct some sort of "blast sheield" around the rotor when I
start running RPMs higher than 1200RPM.


>
> What BPS are you looking for ?   12 electrodes at 5000 rpm will give you
1000 BPS ( good for magnifiers but you will need a BIG pole pig and a small
cap to make it effective). Consider reducing the number of electrode sets.
120BPS is a good starting point for SRSG's.  ( 4 electode sets at 1800 RPM
, you can also drill holes for 8 electrode sets and add the electrodes
later for 240BPS -at- 1800 RPM).

I was planning on using a 3/4HP 1200RPM motor that came from my parent's old
furnace to run 120BPS (remove half the electrodes) for starters.  120BPS -at-
1800RPM is also easy (four electrodes), if I had a 1800 RPM motor. 
<mailto:240-at-1200>240-at-1200 is all 12 electrodes....  12 is a very versatile
number for sync operation.  :-)

running higher breakrates than <mailto:240BPS-at-1200RPM>240BPS-at-1200RPM is just
sort of an unintentional bonus.  I don't really think I'll ever spin it up to
5000RPMs (yes, great for a maggie, even with only 2 electrodes (167BPS)), but
it's an option I'd like to keep open, so I'm REALLY building this one to last. 
My ideal is to build one REALLY nice RSG that I can use for YEARS.


>
> In your cad drawing I noticed that your mounting holes ( shaft flange
plate to disc) are a bit small in circumference. consider making the flang
plate about 4" in diameter. ( this will reduce the opportunity of the disc
to wobble ( develope run out).

Unfortunately, I'm limited by the free Al stock available to me.  I had 2"D
sitting on a metal stock shelf, so I used it.  I used it on a 10"x1/2" disk I
used on my 1800 RPM, 120BPS RSG without a problem (other than imprecise
machining....).  I, too, am a little concerned about wobble, and will continue
to sift through the Al stock to find some 3 or 4"D stock (have a 1"long piece
of 4"D, though....)


>
> hope this helped 

Helped greatly, thanks.

Mark