The Core

11 04 2009

I’m preparing to have the core fabricated. I have a number of considerations to consider.


Welding. We have to weld the lids to the chassis.

Maybe TIG welding will work. My concern is that heat affected zone will damage the SC coils inside. We have ~2.5 mm from the surface to the coils. Laser welding has a much smaller heat affected zone. TODO: get a quote for laser welding from EB Industries. Can anyone comment of the viability of TIG welding for this sitation?

Surfacing. The product that comes back from prometal has a rough surface which we need to machine so that the lid mate well.

Previously I tried wet sanding. This worked decently. However, I wasn’t able to get the deeper surface imperfections, it took a lot of sanding. The outer ring of the torus half saw more material removed than the inner ring, which means that the inner rings mate very tightly, but the outer ring has about a 0.25 mm gap.

There is a surfacing machine here in the shop. It’s large enough to accommodate the lids, but not large enough to accommodate the chassis. The surfacing machine uses a magnetic vise, so the work piece must be magnetic. The sample parts we ordered from prometal are magnetic, however the next parts will be made with a less magnetic stainless steel alloy (the chassis should not be magnetic).

We may need to take surfacing into consideration for the design of the part. ie, we may need to include some extra material on the prometal part, so that after we surface it, we have a perfect half torus.

UPDATE: Stuart told me about Lapping which seems to be an advanced for of wet-sanding.

CAD problems

I’m using BRL-CAD to generate my parts. Lately I’ve been getting this error when I try to export to STL: class_lu_vs_s: loop transits plane of shell/face?  I can’t proceed until I overcome this bug.

Even when the STL export works, it takes forever to render an STL with the resolution I need for production (I’m talking days here). This is really cramping my flow.

Permeability of the Core

We are building a superconducting core. There will be liquid nitrogen at atmospheric pressures inside the core (and connected to outside of chamber via a fluid feedthrough). The core can’t be so permeable as to leek nitrogen into the vacuum which would poison the reaction. Speaking of pressure differentials, the core must withstand the pressure from the inside. To calculate this pressure, I think we need to know the internal surface area of the core.



3 responses

11 04 2009

Just made some progress on the STL issue. It had to do with tolerance distances. Not sure on the specifics.

I also added some code to help me “see” the STL files. Basically I’m converting the output STL files back to .g format, then I do a render. This lets me see how faceted the STL object is.

12 04 2009

The stress from your nitrogen pressure is basically determined by the inner diameter of your tube; you’ll probably have to use the thick walled equations, but the general idea is here:

Give yourself a generous safety factor to compensate for the welds and intersections. I wonder if running a copper tube of nitrogen next to the superconductor would keep it cool enough; the heat transfer would suck, but it would be easier to keep sealed. Seems like there’s going to be a lot of mechanical stress on that tape as well; I imagine it will need to be secured pretty well.

Are you running BRL-Cad on your new Mac? Bought myself my first Mac recently, but I wasn’t able to get BRL-Cad running on it. I’ve been running CoCreate and Alibre on a virtual XP machine.

14 04 2009

ug. This STL has been processing for days, then I get this:

class_eu_vs_s: edge didn’t get cut

FAILED in Boolean evaluation: /cutout
0 triangles written

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: