Vacuum Box Design

Discuss the technical details of an "open source" community-driven design of a polywell reactor.

Moderators: tonybarry, MSimon

mvanwink5
Posts: 2143
Joined: Wed Jul 01, 2009 5:07 am
Location: N.C. Mountains

Vacuum Box Design

Post by mvanwink5 »

First, let me say that in my work group, we would say that seven complete redesigns is normal, then the fine tuning comes. In other words, now would not be the point to argue paint types. How to achieve significant drop box cost is the issue. Second, the goal, as I see it, is how to make the box plus vacuum pumps, total package, cheap (capital and operation costs) and serviceable. So, to start, I am going to try to drop the box cost and magnet support structure substantially. Here is the basic concept that I am proposing as a start to this process:

Preliminary: How big does the box around the mag grid need to be, assume a 100 megawatt thermal D-D polywell?

Make the vacuum box into two compartments.
1. The outer compartment is to be made of cheap materials with large doors that use rubber gaskets. The vacuum pump in this first stage is a roughing pump, with target vacuum in inches of mercury. In this compartment, out gassing of material should not be an issue, so I would think that epoxy painted steel would be used. Structural steel would be inside with the steel sheets on the outside.

2. Inner compartment vacuum would have the highest vacuum and would house the polywell mag grid. For a pB11 machine, the direct energy conversion blinds to collect the alphas would have to be in this compartment (I think unless alphas can be steered out of the chamber). It would seem that because the alphas would be the biggest vacuum issue on the inner chamber, that if their outward movement could be directed using magnets to collection chambers that their removal from the inner mag grid vacuum box would be facilitated and cost of vacuum pumping for the inner box reduced, (maybe substantially).

What materials are needed for the inner box (out gassing issues), construction methods?

Because of uncertainties with pB11 vacuum and direct energy collection designs, it might be best to go with a thermal D-D device. Thoughts? would it make a difference on the inner vacuum box design?

3. The mag grid magnet support structure has to be incredibly substantial due to the high B field forces, so its cost is also major. Where should it be located, in the inner compartment or outer vacuum compartment (part inner, part outer?).

4. With the two vacuum box design, what kind of vacuum seals would be needed for the inner box? Could reusable seals be employed?

I am not a mechanical designer and since cost is the central issue here, we need to get a rough idea on what we are talking about, so experienced mechanical design help is necessary. A basic sanity check would also be useful.

This kind of design process would be a team effort, so no idea is owned as in "my idea." The reason is that once that happens ego gets involved and people get stuck on "their idea." The goal is to drop the box cost.
Counting the days to commercial fusion. It is not that long now.

mvanwink5
Posts: 2143
Joined: Wed Jul 01, 2009 5:07 am
Location: N.C. Mountains

Re: Vacuum Box Design

Post by mvanwink5 »

Four years ago msimon posted this, and since then there has been no update. For sure magnets should be significantly cheaper and I can't see why the box is so much. Basically if one wants to be taken seriously, there needs to be some details to a design. Like I said in the previous post, because thermal energy conversion is well known, a first project would be best to be thermal. Here is a repost of msimon's "Just build it."

viewtopic.php?f=4&t=1258
MSimon wrote:I have been looking at stuff. And it seems to me that a 3T MRI based continuous operation machine could be built for around $15 to $20 million.

1. 3T MRI Magnets plus all supporting eqpt. $3 million.

http://www.ic.gc.ca/eic/site/mitr-crtim ... 00286.html
The cost will become very important since it will be compared with the top-of-the-line 1.5-tesla systems currently in all major hospitals in North America and Europe. The cost of a 3-tesla whole body system is more than US$3 million. The cost for the other systems is even higher. Most of the ultra-high systems, more than 5 tesla, have been integrated by the researcher group, who bought a magnet from Magnex and electronics from companies such as Varian and Bruker. An interesting development is the apparent involvement of Siemens in the 7-tesla system at the Massachusetts General Hospital in Boston.
2. 100 KV 20 Amp variable power supply - $5 million

that is 20 MW @ $.25 a watt = $5 million

http://www.divtecs.com/

3. Vacuum Chamber and Aux Eqpt. - $4 million

===

We could scale the magnets down to 1 or 1.5 T for some cost reduction. Which will cost us about a factor of 100 in power out. It may not matter.

If we run D-D with the neutron flux = to 100 MWf power the coils might last an hour or two until loss of superconductivity. At 1 MWf that could translate into 50 hours of operation.

That is 18,000 10 second pulses.

===

4. Radiation Shielding - $3 million

This could be high or low. Any thoughts?

==

5. Mad Money - $2 million

For anything unexpected.

==

1. 3T MRI Magnets plus all supporting eqpt. $3 million.

2. 100 KV 20 Amp variable power supply - $5 million

3. Vacuum Chamber and Aux Eqpt. - $4 million

4. Radiation Shielding - $3 million

5. Mad Money - $2 million

That is $17 million total This is a bottom of my pants (BOP) estimate and worth every penny you paid for it. Any one else have some better estimates?
Counting the days to commercial fusion. It is not that long now.

happyjack27
Posts: 1439
Joined: Wed Jul 14, 2010 5:27 pm

Re: Vacuum Box Design

Post by happyjack27 »

could the vacuum chamber (and surrounding equipment) be immersed in an inert liquid to help seal off any leaks?

mvanwink5
Posts: 2143
Joined: Wed Jul 01, 2009 5:07 am
Location: N.C. Mountains

Re: Vacuum Box Design

Post by mvanwink5 »

Use a liquid for seals? I have seen that used when a stirrer shaft penetrated the vacuum vessel wall. It worked, but was not cheap.

I suspect the inner vessel is expensive due to material costs and seal costs, so just trying to make the inner vessel significantly cheaper by using a cheap outer vessel. For a cheap outer vessel, steel could be used, maybe stitch weld the plates together or bolt them and seal the seems with RTV rather than seal weld it all? Steel costs much less per pound, but as I understand it, steel is too gassy for use on ultra high vacuum of the inner vessel, where more expensive material is needed. Just trying to drive the cost of material and construction down considerably.

Also, how good do the seals have to be for the inner vessel? Vacuum pumps for the inner vessel have to deal with gasses from the polywell reaction, so how much leakage can be dealt with from the inner vessel seals without seal leakage being an issue?
Counting the days to commercial fusion. It is not that long now.

hanelyp
Posts: 2261
Joined: Fri Oct 26, 2007 8:50 pm

Re: Vacuum Box Design

Post by hanelyp »

I've not heard of steel being "gassy" in vacuum vessels. If that was a problem, vacuum cast plates of low alloy metal would seem an answer to dissolved gases in the metal. If that's not enough, vacuum vapor deposit of a coating might help.

As for seals between plates, seems to me that nothing would beat a professional welding job. My impression is that metal on metal seats are a well established method for access hatches, the selection of seat metal and good surface finish being important.

In general, I'm aware of only one engineering difficulty with the outer vacuum vessel without an off the shelf solution, which is specific to the radiation environment. Direct conversion with an aneutronic reaction would seem to greatly reduce the radiation difficulty. The magrid shell, with constrained volume and access, and possibly much higher heat flux, seems a tougher engineering task.
The daylight is uncomfortably bright for eyes so long in the dark.

mvanwink5
Posts: 2143
Joined: Wed Jul 01, 2009 5:07 am
Location: N.C. Mountains

Re: Vacuum Box Design

Post by mvanwink5 »

The inner shell will also have to be non magnetic due to high B fields. SS water wall tubes could be used, maybe.
Counting the days to commercial fusion. It is not that long now.

happyjack27
Posts: 1439
Joined: Wed Jul 14, 2010 5:27 pm

Re: Vacuum Box Design

Post by happyjack27 »

I suppose the question is: what makes a vacuum chamber expensive? I see three basic parts to the construction: materials, molding, and welding. Now each of these of course has complexities.
But I think nonetheless its helpful to break it down. How much does e.g. Steal really cost? How much to melt it down and mold it, etc. then we get to what makes up the cost of a large vacuum chamber, which seems to me to be one of the bigger expenses.

mvanwink5
Posts: 2143
Joined: Wed Jul 01, 2009 5:07 am
Location: N.C. Mountains

Re: Vacuum Box Design

Post by mvanwink5 »

Somebody who does metal for a living, who can cost out the labor and materials, do some structural calculations is needed to make headway here. When I look for metal prices on the internet, these guys don't show their prices. I suspect that the steel vs alloy costs are about a factor of 5 to 10 depending on what alloy is needed. Holding the magnets would require some serious steel or alloy structure (weight). So it seemed to me that doing the heavy stuff with steel would be worth it. Hence, an outer chamber for the 15 psi pressure drop. The inner chamber could be real thin gauge alloy sheet work, all seal welded (light weight). Without a designer with real prices, though, we are nowhere.

For thermal conversion, we would need temperature and pressure for water, which is what made me think of the waterwall tubes (commonly used in boilers). For 100 MWe we would likely need 300MW thermal. Find someone's coal boiler they need to shut down? can you line it with lead (big dollars?) to absorb the neutrons? The more I think about it the more I like direct energy conversion, do some R&D! :lol:
Counting the days to commercial fusion. It is not that long now.

D Tibbets
Posts: 2775
Joined: Thu Jun 26, 2008 6:52 am

Re: Vacuum Box Design

Post by D Tibbets »

Two vessels? So long as there is any communication between the vessels (any gas flow possible) they will be at ~ the same pressure, within small limits related to the vacuum pumping speed in each chamber.
Stainless steel welded construction is the material of choice,. The stainless steel needs to be a low zinc content alloy as zinc has too high of a vapor pressure. The chamber pressure will need probably less than ~ 0.1 Microns. That will require roughing pumps down to ~ 50-100 Microns and then oil diffusion pumps or turbo molecular pumps. Seals are often copper, though selected rubber gaskets might work if they are protected from exposure to the plasma.

A reasonable chamber diameter may be ~ 2-3 times that of the magrid. The Faraday cage was ~ 3 times the diameter of the 30 cm magrid. in WB6 testing. The WB7 chamber was ~ 1 meter.

Supporting the magnets is a miner issue. The tensile strength of the nubs in WB6 and 7 was adiquate. Mounting the magnets on standoffs would load the supports in compression. Again a minor engineering challenge. The advantage of supporting each magnet separately is that youn can thus feed each magnet seperatly with power and coolant. This eases concerns about coolant flow.

If each magnet is attached to a removable wall plate, then there are advantages, though there would need to be a seal for each. The magnets could be mounted separately, but to an internal cage. This would allow for one door/ seal, etc.

Superconducting magnets are challenging. Unless you can salvage superconducting magnets from MRI machines (as M. Simon suggested) the cost would be stupendous, and a separate engineering path would be needed to develop the super conducting magnets.

There was a reason Bussard planed to stick with copper conducting magnets, even for a Demo machine. WB4 could handle ~ 3,000 Gauss generating amp turns with water cooling, at least for a few seconds(?).
Cooling copper to liquid nitrogen temperatures could increase the amp turns capacity ~ 6-8 fold.That would translate to ~ 2.4 Tesla fields. Cooling to liquid helium temperatures would be more expensive but it would allow for a further ~ 2.5 times greater improvement in the conductivity of the copper wire. So you could improve the magnetic generating amp turns by a factor of up to ~ 20 fold. In a 30 cm machine that would be ~ 6 Tesla. A one meter machine might reach ~ magnetic field strengths of ~ 18 Tesla, at least in theory.
Super conductors have two problems, quenching and current capacity. Some of the liquid helium superconductors might be able to carry upwards of 8,000 amps. That is not much more (or perhaps less) than the capacity of super cooled copper wires. WB6 had 200 windings of ~ 8 gauge (?) copper wire and up to ~ 2000 amps flowing through them. Super cooled copper wire of the same size may be able to carry up to ~ 40,000 amps (20X conductivity or 1/20th the resistance). You might need ~ 5 times the length of an optimistic superconducting wire to match the performance, and the cost per foot is considerable compared to copper wire. In the long run you might save electricity costs, but only if you maintained the cryostats 24 hr per day. For production power plants this is desirable, but for intermittent research reactors it may be a big disadvantage. Liquid nitrogen or even water cooling of large terrestrial plants may be more economic, especially if superconductors have a significantly shorter life time under neutron/ gamma ray/ x-ray bombardment. The superconductors might be reserved for mobile small reactors. Of course,if cheap robust room temperature superconductors with huge current carrying capacity are developed, the picture changes considerably.

I generally think of copper wire cooled to liquid nitrogen temperatures as this is the cheapest cooling, and a WB4 configured machine might be able to generate ~ 2.5 Tesla fields for multiple second tests. I use the WB4 model because it had liquid (water) cooling and translating this into liquid nitrogen plumbing cooling / packing fraction should be similar. As size is increased, the available internal volume increases as the square of the diameter, while the run length of the wires scales linearly. So doubling the diameter to 60 cm would result in a 4 fold increase in internal volume, with twice the length of copper windings. Thus current/ heat load could double, or conversely the windings could double. So the B field strength could double. Bussard extrapolated this to a 3 meter machine with water cooled copper electromagnets. This could have ~ 100 times the windings, thus ~ 10 Tesla, or even 30 Tesla if you use the WB4 baseline, without exceeding doable cooling requirements. Going to liquid nitrogen cooling eases the B field vs size constraints considerably*. I suspect things would overheat but not for a few seconds, which is forever from a physics verification standpoint. For a sustained operation the magnetic field strength (amps) would need to be scaled back and the coolent flow may need to be increased. Just how much is unknown (by me).

High fusion yields with D-D fusion and possibly P-B11 fusion will produce extremely dangerous radiation- neutrons or gamma rays. Shielding for these and monitering the system would be expensive and potentially politically touchy. A dedicated certified radiation officer would be needed. WB 6 produced ~ 1 milliwatt of D-D fusion energy, and this resulted in ~ 1 micro REM of neutron radiation per hour (?) which is way below background radiation. But a MW of output would be ~ 1 REM per hr, and 100 MW ~ 100 REM per hr. You want good shielding and short run times. Then there are the X-ray concerns...

* WB 8 may be ~ 60 cm (or less) in diameter. It may have stayed at 30 cm, and only the vacuum chamber diameter was increased. And it is apparently cooled to liquid nitrogen temperatures. This suggests that the B field strength capacity may have increased to 0.8 to 1.6 T. It depends on several factors including the minor radius of the magnet cans. WB6 had ~ 17% of the total radius, WB4 had ~ 25%. Using one set of assumptions for WB 8 is that the diameter is doubled, the minor radius is maintained at 17% and ~ 1/2 of the internal volume is taken up by the coolant plumbing. This means that the amp turns could have been doubled,and the nitrogen cooling improved conductivity 6-8 fold. Thus 0.8 Tesla operating B fields may have been a conservative goal, perhaps related to selected liquid nitrogen coolant flow during the tests and the duration of the tests.

Dan Tibbets
To error is human... and I'm very human.

D Tibbets
Posts: 2775
Joined: Thu Jun 26, 2008 6:52 am

Re: Vacuum Box Design

Post by D Tibbets »

A point about out gassing. A remarkable amount of gasses can absorb into metal surfaces. This is used for hydrogen as a possible storage method. Water absorbs to metal surfaces also. It is the major contaminate in vacuum systems because of this. It takes hours if not days for most of the water to out gas from the walls under high vacuum conditions. Heating helps. Even a few moments of the chamber being exposed to normal humid air can reload the walls to a large extent.

The supply of out gassable water or other contaminates is dependant on the surface area of the walls. As the volume increases as the cube of the size (diameter) and surface area increases as the square of the size, the issue becomes less of a pain as chamber size increases. The obtainable signal to noise ratio improves. This is why Nebel dismissed using smaller machines (less than 30 cm) for research. The noise caused by out gassing interferes with the operation/ measurements of the system to a greater extent. And WB6 neutron productionwas already near the signal to noise floor.*

Bussard, etel. pointed out that as the magnets heated up from the Ohmic heating of the magnet copper wires, the out gassing of absorbed gasses increased rapidly. This increased the burden of neutral gasses in the chamber, which led to terminal arcing. It was not just the escaping neutral gas from the puffers. As mentioned above, increasing the size not only lessens the gas burden from the neutral gas puffer (a greater percentage is ionized before it can escape). It also lessens the contribution of out gassing on the build up of pressure inside the chamber because the internal volume increases faster than the surface area.

* Still, I think, measurements of certain perimeters such as Beta, Wiffleball formation, and various plasma parameters is addressable in smaller machines.

Dan Tibbets
To error is human... and I'm very human.

happyjack27
Posts: 1439
Joined: Wed Jul 14, 2010 5:27 pm

Re: Vacuum Box Design

Post by happyjack27 »

If the seams don't pose a leak issue, and welding / sealing costs are not prohibitive or can be reduced, it may be more economical, as far as molding goes, to construct the vacuum chamber from really thick "plates" sealed together at the edges to form a polyhedron. Especially for very large chambers, were the volume of a molding cast - representing e.g a top / bottom half grows with the 3rd power of the radius, whereas molding a side at a time would only grow with the second power. And the sealing surface area would grow proportionally to the 1st power, times the side thickness, times a little less than the number of faces.


Sounds to me like a potentially better scaling law for construction costs, if the scaling law for leaks (similiar to sealing surface) is not prohibitive.

mvanwink5
Posts: 2143
Joined: Wed Jul 01, 2009 5:07 am
Location: N.C. Mountains

Re: Vacuum Box Design

Post by mvanwink5 »

You guys have caught on to the idea process. Key here guys is to put real cost $'s to the ideas. What I am thinking is that if we drop the capital cost for a continuous operating polywell, that produces enough, serious net power, that it will be a game changer for, let's say making aluminum. I remember back, maybe 6 or so years, that an entire major aluminum smelting plant was shut down due to power costs. I think the EPA shut down some coal power plants, or blocked them. The point is that there are factories that become virtually worthless, that with a reasonably priced polywell would be worth a company's future. If a polywell is financed and is put together, say in two years, is up and demonstrates serious power. The financiers would make serious money, enough to take a reasonable risk. So, cost, time, and risk must together be managed. That means an upfront smart design that factors in costs. The design must be developed, and incorporate smart shopping.

So, are you guys up for the crazy? If so, we need to put in some reality calcs, to see ball park where we are...I think we can beat MSimon's SOP numbers, but maybe not, how creative are we?

Sure beats talking about some of the things that have dominated this blog for the last year...
Counting the days to commercial fusion. It is not that long now.

hanelyp
Posts: 2261
Joined: Fri Oct 26, 2007 8:50 pm

Re: Vacuum Box Design

Post by hanelyp »

Any one here with a background in steel ship construction? Might be some insight there for constructing the outer vacuum chamber.

Today I ran across http://en.wikipedia.org/wiki/Vacuum_flange, which might apply to some details.
The daylight is uncomfortably bright for eyes so long in the dark.

Stubby
Posts: 877
Joined: Sun Aug 05, 2012 4:05 pm

Re: Vacuum Box Design

Post by Stubby »

Tri clamps
Run across those all the time in high purity plumbing or lyophilizers.
Very effective.
Everything is bullshit unless proven otherwise. -A.C. Beddoe

mvanwink5
Posts: 2143
Joined: Wed Jul 01, 2009 5:07 am
Location: N.C. Mountains

Re: Vacuum Box Design

Post by mvanwink5 »

Ok, 100,000 gallon water tank, used, basically for the cost of dismantling and transport.
http://www.govdeals.com/index.cfm?fa=Ma ... cctid=3536

Used bolted water tanks are also available (easier to move and set up). It is a matter of shopping and time.

We could put a water tank inside of another water tank and fill the space in between with water to give neutron shielding? This would mean three tanks, but the outer two are cheap, the inner of these two would need a steel roof (under rough vacuum). Then the polywell tank would go inside of those two tanks. We could put the tank in the ground and use the ground as shielding, but above ground might be actually cheaper using a cheap exterior tank for shield water. How many feet of water shielding would be needed?

The 1 micrometer vacuum tank (the expensive tank) could be made of thin gauge (26 gauge?) alloy metal, all seal welded (certified welding), special high vacuum flanges. I suspect these flanges will cost more than the tank itself, so they need to be minimized. How often does the polywell tank need access via large flanges? Maybe use small instrument flanges only, construct the tank around the magnets? After all, it is thin gauge metal.

Polywell magnets would be loaded into the center tank from the top (put a hatchway for moving large items in and out, also, a man way hatch for access to the polywell and tank.

Roughing vacuum pump, good for 3 to 4 inches of water vacuum should be cheap, as size is dependent on pump down time and leaks. The diffusion pumps for operation are the expensive equipment. That is dependent on gasses generated by the polywell operation.

So, what do you think we are talking about ($) for the alloy polywell tank and diffusion pumps for operation, the rest might be $100k-$200k (if installation is included)?
Counting the days to commercial fusion. It is not that long now.

Post Reply