Talk-Polywell.org

kcdodd did an analysis which I think he called a "bag" analysis (see the link below) which appeared to calculate the effective wiffleball surface. If so, one possible figure of merit would be to determine the deviation from a sphere that such a surface has and compare them. Just a thought.


viewtopic.php?p=8499&highlight=#8499

rjaypeters wrote:Randy,
I don't have Mathematica, not sure I can convince my spouse to spend $2500 (or even $250 for M. Home which might not work?). Drat.

...been there dude. But family always comes first... The rest is just your hobby.

~Randy

I use Maxima, available under the GNU General Public License (GPL) here:
http://maxima.sourceforge.net/
I don't know for a fact that Mathematica code will run on Maxima, but I understand it to be so. I do know that the code I run was originally written for Mathematica and transferred to Maxima. The part I don't know is how much re-write was involved in the transfer.

happyjack27,
You don't want much, do you? I'm going to have to root around for analysis software that's cheap and I can wrap my head around (mutually exclusive requirements?). Aero gave me a hint on Maxima.

It's my understanding the extended single coils are possibly non-starters for engineering reasons (can't keep them cool), so I'm leery of looking at those. Willing to be convinced otherwise.

Perhaps it is possible to prioritize your list? Also, some of the concept pictures are stick figures I don't understand, sorry.

R. Peters,

I tried to run all three versions of Radia on my copy of Mathematica but my version will not work with any of the available Radia distributions.

Oh well...

~Randy

rjaypeters wrote:Perhaps it is possible to prioritize your list? Also, some of the concept pictures are stick figures I don't understand, sorry.

it's already prioritized, in the order i listed them. saw that one coming.

(though actually i'd move cusp disruption / conversion testing up to the second spot. i really only put it third 'cause i was anti-favoring it 'cause it's (half) my own.)

i made the list as long as i did cause i was thinking other people would make lists and you'd then take the ones common to all lists, which would result in a much smaller set.

re single coil, i suppose if you're running fluid along the coil i see what you mean. the interesting thing about a single coil design though is that a) they have smaller line cusps, and b) the north out fields and north in fields are perfectly balanced/symmetrical. it would be interesting to find out what effect balanced/symmetrical fields has on the device. could possibly result in better containment.

and i suppose the single coil designs could be psuedo-single coils. i.e. the electric current / magnetic field would be the same, but you can pump coolant or whatever at different point on the coil as if you were splitting it up into multiple coils.

but if you really want multi-coil, tombo's got multi-coil designs w/symetric fields on pg 5. (i'd go for the black and gray one.) though they might leak charged particles through the top and bottom. would be interesting to see. if my list was one longer i would have put that one on it. though i think it would be nice to have one good perfectly field-symmetric single-coil in the list so we can see if anything particularly interesting happens in such a case.

icarus wrote:

From a physics stand point your design is intriguing, but it would probably be difficult to engineer, it's a shame that we have to account for gravity and heat .

I similar, but simpler approach, would be to have for segments that arc 90 degrees down, bending 90degrees right and bending 90degrees down again. The electrical polarity, would be alternated, with each segment (segments 1 and 3 positive up negative down, and segments 2 and 4 positve down and negative up).

I'm sorry, that I don't have the means to draw it, but it would be like a spherical verison, of the this early Idea of Tombo's:

rjaypeters wrote:Randy,
I don't have Mathematica, not sure I can convince my spouse to spend $2500 (or even $250 for M. Home which might not work?). Drat.

Possibilities... http://en.wikipedia.org/wiki/Category:F ... s_software

Everyone,
I've been looking at Radia and it's as complicated (to my simple understanding) as Randy claims. I've made a cursory look at the free software available, so fun none looks as good a fit as Mathematica+Radia.

Weird thought: since EMC2 is developing simulation software, why don't we hire them to run the simulations we like? We'd have to do messy things like figure out priorities (probably he/she who pays the piper calls the tune) and raise money.

Imaginatium,
I'll model the spherical Tombo you showed later.

P.S. I've fixed the pictures of the Tibbets triples with Icarus splits, with and without twists. See previous page.

rjaypeters (from previous page):

Since this concept has six magnetic elements, is it the closest to the current wiffleball prototypes?

I would say no. In this current 'zoo' of configurations the closest would be a set of six 'square' coils, with rounded corners, and bowed sides that lie on the surface of a sphere.

Basically, map a cube onto a sphere and the edges of that figure form the line cusps. The coils' sides run parallel to these edges and tangent to the sphere, with some small space between them to allow for recirculation, and rounded corners for a smooth mag-field.

Here's a basic pic of a cube mapped onto a sphere,

viewtopic.php?t=650&postdays=0&postorder=asc&start=30

(scroll down to the big beach-ball image.)

imaginatium,
Here is a Tombo spherical:

EDIT: This is a "false" Tombo spherical. Later, I will complete a correct Tombo spherical.



Including simplicity, the thing I like about this concept is one could use the repulsive forces to distort the magrid into the ideal shape for confinement (and other purposes) and possibly minimize or eliminate other support structures.

Icarus,
A "zoo", is it?

I'll model the spherized square ring magrid later. Thanks.

Six square spherized rings:

Pretty!

Will need nubs or lots of legs. Looks like tombo's but in color and without the "real triangle magnet" cross pieces.

This, without the bowed sides, is what Dr. B. called WB7.

rjaypeters wrote:imaginatium,
Here is a Tombo spherical:

Including simplicity, the thing I like about this concept is one could use the repulsive forces to distort the magrid into the ideal shape for confinement (and other purposes) and possibly minimize or eliminate other support structures.

Your moment arms for the magnetic reactant force are huge. You'd have shear forces on the supports that would be VERY hard to manage.