Toroids with Wing-Like Internal Structure.

Billy Catringer · Post by **Billy Catringer** » Wed Mar 18, 2009 12:34 am

This is my first rough layout of a toroid with internal similar to the structure of an aircraft wing. It is also the result of me learning to use Blender's built-in render engine:

This shows a 5 degree segment of the LHe Jacket and the SC Conductor standing vertically. The center of the REACTOR is to your left. The center of the torus is straight down. Light blue designates 316SS. Brown designates the super conductor. Dark green is for electrical insulation.

This is to scale, but I am going to change a number of things, starting with the wall thickness of the jacket. Right now it is 9.5mm. I am going to change that to 10mm to take advantage of commercially available plate. This would be roughly the same as STD Wall pipe, or 3/8" wall.

There are a number of concerns when attempting this approach. Each torus will actually be two, not one wing. Each half-shell will need its own structure that renders that half-shell rigid. The two half-shells will be joined together by two seam welds to make the two pieces into a single rigid whole.

Fabricating the pieces will likely be a two step process. The plate will be cut out on a laser or plasma arc table and then sent to a CNC machine for final sizing. The inside of a torus presents us with a much more complex geometry than that found inside of an aircraft wing.

The internal structure will require a carefully written out set of assembly instructions and check list. If parts of the structure are installed in the wrong order, it will be impossible to weld some, possibly many, of the subsequent pieces.

Inspection will be time consuming and may hold up work on a single torus for hours at a stretch. It will be necessary to assemble the torii concurrently in order to balance manpower around inspection.

This kind of structure will in no way be in compliance with ASME B31.1, save possibly on the inspection procedures. This will complicate things for investors and underwriters.

As you can see, I still have a long way to go.

KeithChard · Post by **KeithChard** » Wed Mar 18, 2009 2:55 pm

Billy,

I would like to congratulate you on having introduced some engineering realism to the Magrid design. As you say, you have a long way to go: keep up the good work.

I have one suggestion, that you may not wish to incorporate at this stage, but which will need to be considered when you get a little nearer to a final design. The outer facing part of the circular cross section could do with a bit of flattening to ensure conformity with the magnetic field lines, particularly where the coils are most closely adjacent. Have you seen some of Indrek's simulatioms? This one is very relevant.
http://www.mare.ee/indrek/ephi/rect_vs_ ... gcolor.jpg

Non circular profiles can contribute to easing the total heat load on the magnets by reducing the area facing the "sun". Bulges on the back face may have uses in connection with the mounting structure and feeding the coolants without interfering with the field lines.

KitemanSA · Post by **KitemanSA** » Wed Mar 18, 2009 3:54 pm

KeithChard wrote:Billy,

I would like to congratulate you on having introduced some engineering realism to the Magrid design. As you say, you have a long way to go: keep up the good work.

I have one suggestion, that you may not wish to incorporate at this stage, but which will need to be considered when you get a little nearer to a final design. The outer facing part of the circular cross section could do with a bit of flattening to ensure conformity with the magnetic field lines, particularly where the coils are most closely adjacent.

It is even more significant than indicated in the URL listed because the internal electron pressure will flatten the inside face of the field too.

Edit: Sorry, got distracted. I also wanted to mention that this advise only applies to the outer TSP courses of course, not to the inner LHe ring.

KeithChard · Post by **KeithChard** » Wed Mar 18, 2009 6:52 pm

Modifying the core cross section might be the starting point for changing the outer profile. More field line calculations are needed.

Billy Catringer · Post by **Billy Catringer** » Thu Mar 19, 2009 5:52 pm

KeithChard wrote:Billy,

I would like to congratulate you on having introduced some engineering realism to the Magrid design. As you say, you have a long way to go: keep up the good work.

I have one suggestion, that you may not wish to incorporate at this stage, but which will need to be considered when you get a little nearer to a final design. The outer facing part of the circular cross section could do with a bit of flattening to ensure conformity with the magnetic field lines, particularly where the coils are most closely adjacent. Have you seen some of Indrek's simulatioms? This one is very relevant.
http://www.mare.ee/indrek/ephi/rect_vs_ ... gcolor.jpg

Non circular profiles can contribute to easing the total heat load on the magnets by reducing the area facing the "sun". Bulges on the back face may have uses in connection with the mounting structure and feeding the coolants without interfering with the field lines.

Many thanks, Keith. I apologize for taking so long to reply, but I have been rather busy wrestling with Blender and the inside of toroidal tubing.

For the time being I am going to proceed with circular cross sections. What is ovoid on the outside looks parabolic on the inside and I am not yet adroit enough with Blender to tackle that environment. Please be doubly sure that such shapes are necessary as they will give rise to a number of concerns.

First and foremost, it is a further departure from standard engineering practice and there are no tables that I know of for dealing with pressure and other stresses in oval tubes. Now that I have said that, someone will find one just to show me up. I hope that someone will because we may well need it. However, such departures from standard engineering further complicate the process of acquiring investment.

Secondly, such shapes greatly complicate fabrication. Believe me when I tell you that this is not a minor concern. I also suspect that ovoid sections will give the FEA boys, like Ed Kribbs, a very bad case of the heebee-jeebees.

MSimon · Post by **MSimon** » Thu Mar 19, 2009 6:16 pm

The field resolves to circular the farther you get from the coils. Look at some of Indrek's larger field simulations.

The field is going to vary around the conductors - it can't be helped. I don't think that is a good enough argument for deforming the coil containers. If there is some other advantage to be gained - excellent.

KitemanSA · Post by **KitemanSA** » Thu Mar 19, 2009 9:53 pm

I think there is no doubt that the field in places where the coils are in close proximity will not be circular. This gives us a bit of lee way. In said locations, it may be feasible to flatten the outer courses of the TPS (maybe eliminating the outermost) to provide cusp room and not have a problem because of 1) the non-circular field, and 2) the shadowing effect of one coil on the other.

Out in the middle between junctions, the field will be quite round (except perhaps for a flattened area inside the sphere).

MSimon · Post by **MSimon** » Thu Mar 19, 2009 10:44 pm

KitemanSA wrote:I think there is no doubt that the field in places where the coils are in close proximity will not be circular. This gives us a bit of lee way. In said locations, it may be feasible to flatten the outer courses of the TPS (maybe eliminating the outermost) to provide cusp room and not have a problem because of 1) the non-circular field, and 2) the shadowing effect of one coil on the other.

Out in the middle between junctions, the field will be quite round (except perhaps for a flattened area inside the sphere).

The problem is that flattening in one area must lead to a sharpening in another. There is a minimum allowable radius due to electrostatic considerations. And there are thermal considerations as well.

KitemanSA · Post by **KitemanSA** » Fri Mar 20, 2009 12:12 am

MSimon wrote:The problem is that flattening in one area must lead to a sharpening in another. There is a minimum allowable radius due to electrostatic considerations. And there are thermal considerations as well.

You bet. It will be an interesting balancing act.

KeithChard · Post by **KeithChard** » Fri Mar 20, 2009 12:10 pm

Billy,

Keep to the circular model until you have finished it. Complications are only for action later, and then only if they are really needed. We may discuss them, but what you are doing is the basic design which should be followed right through.

Billy Catringer · Post by **Billy Catringer** » Fri Mar 20, 2009 2:02 pm

KeithChard wrote:Billy,

Keep to the circular model until you have finished it. Complications are only for action later, and then only if they are really needed. We may discuss them, but what you are doing is the basic design which should be followed right through.

Will do, Keith. At the moment I am trying to poke holes in Blender meshes. You'd be suprised how hard that is to do without making an unholy mess of it.

Billy Catringer · Post by **Billy Catringer** » Sat Mar 21, 2009 2:16 am

Okay, here is my second pass. This one is a lot smoother with the electrical insulation safely trapped between the metal and the SC Core. The empty slots will be for additional support rings. The two ribs you can now see still need more work. I am going to make some minor changes to their shape and will add holes to them during the next pass. We can't use solid pieces anywhere in this thing because that would block the flow of the coolant.

The beveled edges are where welding will be done.

KitemanSA · Post by **KitemanSA** » Sat Mar 21, 2009 4:05 am

I presume the solid block of coppery color is a kind of "here there be super conductor" place holder, not anything more. I suspect the actual design of the surrounding structure will have to be tuned to provide flow thru the core. Maybe not, but if we use High Temp SC, the field connections are not superconductive. They have a low but real resistance and at least those will have to be cooled, not just protected from outside warmth.

Billy Catringer · Post by **Billy Catringer** » Sat Mar 21, 2009 4:31 am

KitemanSA wrote:I presume the solid block of coppery color is a kind of "here there be super conductor" place holder, not anything more.

That is correct. As soon as I find the way to turn different objects on and off for a particular render job, I will produce a picture without the core and without the electrical insulation.

KitemanSA wrote:I suspect the actual design of the surrounding structure will have to be tuned to provide flow thru the core.

Also true, but bear in mind that I am already working on some of that. Nearly all of the metals depicted in this schematic will have a series of holes in them. I would have put them in this time except that I am up against the stops on my knowledge of Blender.

As far as I know at this time we are going to have either a coil of superconducting tape, or we are going to have Bitter Plates with sizes close to the 29cm x 29cm core I have been designing everything around. If it turns out that we can get by with using a smaller SC core, I will be tickled pink and will not mind changing the model in the least. BTW, I will soon be producing drawings of the individual parts.

KitemanSA wrote:Maybe not, but if we use High Temp SC, the field connections are not superconductive. They have a low but real resistance and at least those will have to be cooled, not just protected from outside warmth.

I have been meaning to ask about this. I have to bring the LHe piping out of the magnet structure anyway, so I had planned on a series of step-downs for the jackets once the piping is outside the vacuum vessel. There are actually two questions now that I think about it. One, what about the terminations to conventional wiring and, second, what do we do about closing the loop inside the magnet? In other words, you get the current started and then close the circuit so that it continues to flow without further input of energy, right? Wrong? If wrong, no big deal If right, the plumbing will have to be designed to accommodate a switch of some kind.

I would very much like to drop the 316SS in favor of titanium, but there is a snag. 4He is liquid only at something around 4K. Titanium is only good down to 60K without special heat treatment. I do not yet know just how far down in temperature you can go with titanium. I also understand that the SC needs to be down around 30K. Maybe we can use Ti for the core tube and maybe we can't. I'll have to do so more digging.

The suppliers of Ti say that it does not hold up very well at all to neutron bombardment. This means that I will have to stick with Inconel 690 for the outer jackets.

Billy Catringer · Post by **Billy Catringer** » Sat Mar 21, 2009 4:56 am

Same thing with the parts spread out a bit and with the SC Core and the electrical insulation removed.