[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: FEA analysis of 12" rotary spark gap disk

Original poster: "Dr. Resonance" <resonance@xxxxxxxxxx>
That's a nice analysis. It's always amazing to me to note all the different and excellent specialty training different members of this great list has to offer.
I have never used setscrews in the past 15 yrs in any of my RSGs and they always hold --- never loosen up. We always use 1/2 inch dia. brass feedthrus x 1.5 inch long thru the 1/2 inch dia. G-10.
I've noted the 3450 rpm motors with 10 electrodes always seem more exciting and violent in the discharge sparks than the 1725 rpm motors --- the lower break rate seems more sedate and just not as exciting as the 3450 hummers. These really generate a lot of strong airflow to aid in quenching via better cooling. Good idea to balance them at this speed though. 

Dr. Resonance


Original poster: "MakingLightning" <MakingLightning@xxxxxxxxxxx>

I am designing a 12" rotary spark gap.
I really wanted some hard numbers so I knew where I stood on the safety margins. I also wondered how good a press fit electrode was held in. I had it 3D modeled in SDRC IDEAS and then analyzed it with their FEA package called Visualizer. It was calculated and the numbers matched what the FEA came up with. It was really nice to see how the press fit stress reacted with the spinning disk.
Chip posted graphic representations of the results for me at: <http://www.pupman.com/kevin>http://www.pupman.com/kevin 

The disk I will be using is:

12" diameter 1/2" thick G-10 fiberglass laminate.
8 - 1/2" diameter x 1.5" long pure tungsten electrodes Top speed 4000 rpm which equals 533 BPS, way more than I anticipate using. 

G10 Fiberglass:
<http://matweb.com/search/SpecificMaterial.asp?bassnum=PGLAM04>http://matweb.com/search/SpecificMaterial.asp?bassnum=PGLAM04

Tungsten:
<http://matweb.com/search/SpecificMaterial.asp?bassnum=AMEW000>http://matweb.com/search/SpecificMaterial.asp?bassnum=AMEW000
In my presentation that chip posted for me you will see that the top shot is the disk with the hole in it, at rest. 
The one next to it is spinning 4000rpm.
The next set starts with the 0.5" electrode press fit into a hole 0.0005" undersize hole, as Dr Resonance recommends. On the left is the disc at rest and the right is spinning at speed. 

The last set is stretched out and the scale increased.
You can see that with the press fit, there is over 9600lbs of force holding the electrode in. The set screw added could hurt, help, or just not be needed. If someone can get me good data on the material properties for putting something into the side of the material, that can then be simulated. The mfg does not give that info (see data above), as that is not what it is designed for.
As far as the material giving out when my disk is at speed, I think there is plenty of safety margin, there is only 1600lbs of force pushing the electrode out and the material is good for 38,000 lbs/sq in. I feel comfortable proceeding now. I will still build a Lexan enclosure for it. 

Interesting links:

<http://www.theodoregray.com/PeriodicTable/>http://www.theodoregray.com/PeriodicTable/

http://matweb.com/
I have to give credit for the FEA that was done on this by my fellow coworker, Todd List. 


Kevin