Talk-Polywell.org

rjaypeters wrote:Now if we could just find a way to hold the close approaches together without messing up recirculation or contaminating the vacuum.

Well, happyjack27 had an idea* about what I call chain-link. I've just added support extensions:



Leaving the problems of how to power and cool the six primary coils. I might think about SC and coolant paths between the primary coils and support coils.

Only eight extensions are shown, but it could easily be twelve.

*Taking a bit of my own advice, I've been reviewing the earlier posts on this thread. I don't think this a completely new idea.

rjaypeters wrote:Only eight extensions are shown, but it could easily be twelve.

Here are the twelve, alone:



Reminds me of a Tombo'd octahedron. So, why wouldn't we do it this way, but shoved closer together?

Does this not create more point cusps?

I think so, but no need for nubs (nurbs, etc.). If the support legs are splayed (which I will show later), there is no need to keep them separated by interfering matter, either.

As promised:


The "splay" angle is about 13.5 degrees. Twenty-four close approaches v. twelve for the WB-6, 7 & 8. It is worse from a point cusp perspective, but I think the structural simplicity and the confinement from the octahedron shape might make it worth the trouble.

I know this looks pretty, but I still think it will lead to more transport leakage. Too many field seams.

A stellation of the truncated icosahedron:

Top View:


Half of Top View:


Left View:


Pretty, but it has issues:

Big holes:


A really big hole:


And pipes that run into each other (detail of left view):

Magrid Development Platform:

It’s been ‘bout a year or two since my last post here guys. Anyway, here’s some screenshots from the latest version of my 3D magnetic field line rendering program. I couldn’t find any example programs on the internet which showed me how to plot the 3D magnetic field lines (magnetic flux density lines – B lines) generated by 3D electrical current conductor sources. So I started from the only 2D example code(s) I could find and figured out the rest from there. With this program, you can drag the 3D conductors and their associated 3D magnetic field line images with your mouse to see them rotate together about the origin (core) in real-time. The program is still in development mode. I’ve discovered that 3D magnetic field line rendering is in “NO WAY” a science yet – it is still an art form… The exceptions you have to make to reasonably render a recognizable magnetic field image are enormous and almost always different for each rendering. You simply cannot plot all the information at once – it’s too mind-boggling. You’re better off learning how to plot only certain subsets of the 3D magnetic field lines in question. In 2D, magnetic field lines (B-lines) form closed loops. In 3D, magnetic field lines (B-lines) generally – never end! What I’ve learned so far is that the field shape you really get doesn’t look like what you thought it would - there are subtle and important differences. But after you study what you really get (for a while) it makes total sense. ~Randy

Images:
Proper Current Flow:
http://i115.photobucket.com/albums/n296 ... ntflow.jpg

Wrong Current Flow:
http://i115.photobucket.com/albums/n296 ... ntflow.jpg

Spherized Octahedron:
http://i115.photobucket.com/albums/n296 ... hedron.jpg

Don't know what I did wrong. But only the links to the images posted. Maybe someone here can re-post the images or instruct me on how to properly post such things from 'Photobucket'. It's been a long time since I posted stuff from photobucket. ~Randy

Well, for one thing, you didn't wrap them in [img] tags.

Here ya go

Randy wrote:Magrid Development Platform:

It’s been ‘bout a year or two since my last post here guys. Anyway, here’s some screenshots from the latest version of my 3D magnetic field line rendering program. I couldn’t find any example programs on the internet which showed me how to plot the 3D magnetic field lines (magnetic flux density lines – B lines) generated by 3D electrical current conductor sources. So I started from the only 2D example code(s) I could find and figured out the rest from there. With this program, you can drag the 3D conductors and their associated 3D magnetic field line images with your mouse to see them rotate together about the origin (core) in real-time. The program is still in development mode. I’ve discovered that 3D magnetic field line rendering is in “NO WAY” a science yet – it is still an art form… The exceptions you have to make to reasonably render a recognizable magnetic field image are enormous and almost always different for each rendering. You simply cannot plot all the information at once – it’s too mind-boggling. You’re better off learning how to plot only certain subsets of the 3D magnetic field lines in question. In 2D, magnetic field lines (B-lines) form closed loops. In 3D, magnetic field lines (B-lines) generally – never end! What I’ve learned so far is that the field shape you really get doesn’t look like what you thought it would - there are subtle and important differences. But after you study what you really get (for a while) it makes total sense. ~Randy

Images:
Proper Current Flow:

Wrong Current Flow:


Spherized Octahedron:

Thanks Stubby.

Is there any place where this can be downloaded?


Currently having a hell of a time modeling the B Field.


The values are correct but the direction vectors are all messed up.

mattman

I just uploaded the program source code to ‘Planet Source Code’. However, you will have to wait a few days for the PSC moderators to approve and post the code. Just keep checking the PSC site and the code will become available in a few days.

* Go to http://www.planet-source-code.com/
* You will have to register at Planet Source Code to download files.
* Enter: RandyT_CS in the Search Term… box and click the search button.
* Select the Magrid 1.0 file from the list that comes up and download the source code from there.

Hope this helps you. ~Randy

mattman

The source code post is up.

May I ask what programming language you develop with?

I may be able to help you quickly if you develop in Visual Basic 6.0 or MinGW 'C'.

I'm stuck with a dial-up connection and only have access to Windows platform development.

~Randy