Copper at about LN temperatures has about 1/5th the resistance of room temperature copper. Thus the same size coil can be 5x as strong. Seems reasonable to me that 2x WB6 scale windings with ~0.4x the packing factor and 5x the conductivity would result in ~8x the field which is what was required by contract.
Just another thought.
Recovery.Gov Project Tracker
Excellent!KitemanSA wrote:Copper at about LN temperatures has about 1/5th the resistance of room temperature copper. Thus the same size coil can be 5x as strong. Seems reasonable to me that 2x WB6 scale windings with ~0.4x the packing factor and 5x the conductivity would result in ~8x the field which is what was required by contract.
Just another thought.
Engineering is the art of making what you want from what you can get at a profit.
-
happyjack27
- Posts: 1439
- Joined: Wed Jul 14, 2010 5:27 pm
well i would imagine Lox and LN would cool a lot better than water, and to much lower temperatures.
also, maybe they're using it for something else, say, a vacuum pump. or maybe that's the fuel - maybe it's liquid hydrogen/duetrium/boron. there are other interesting possilbilities. Though I must admit getting 8x the field strength, esp. at continuous operation, doesn't sound easy. sounds like something you'd have to take some extra measures to achieve, such as the aforementioned liquid cooling.
also, maybe they're using it for something else, say, a vacuum pump. or maybe that's the fuel - maybe it's liquid hydrogen/duetrium/boron. there are other interesting possilbilities. Though I must admit getting 8x the field strength, esp. at continuous operation, doesn't sound easy. sounds like something you'd have to take some extra measures to achieve, such as the aforementioned liquid cooling.
A big tank of liquid deuterium? I doubt it. A welders size tank of compressed deuterium gas would probably provide enough deuterium to last them years.
I think the only real advantage of liquid nitrogen is the cooling of copper wires so that the conductivity increases 5-8X that at room temperature.. Water would also carry away the ohmic heat generated, and probably at much less cost, even when the additional resistance is considered. The advantage is that at liquid nitrogen temperatures, the increased conductivity allows for smaller wires, which allows for more windings, thus higher B-fields at the same current (heating). The reason high flow rates of water couldn't compensate for this is because of thermal conductivity issues. You have to get the heat out of the bundle of copper wires before you can carry it away in the liquid. IE: you would need to reduce the packing fraction- which means there is less room for wire windings.
[EDIT] Actually when I originally wrote this I was thinking of a linear relationship between the B- field (current * windings) and the heating. Actually it is a log-rhythmic (squared) relationship, so the above comparison is even more biased towards minimizing the current needed for any given B field generation.
http://www.ru.nl/hfml/research/levitati ... _solenoid/
Some liquid nitrogen could be used as a boost for vacuum pumping (cold trap), but I don't know if that is piratical at the vacuum levels they need, or if it would add anything to the presumed turbomolecular , or high capacity diffusion pumps they are using.
As far as steady state, with these research machines, a few tens of seconds of running the magnets may seem like an eternity for them. For real steady state machines (weeks on end) the cooling would need to be more robust. With short duration research machines the thermal mass gives them a buffer for a few seconds (? eg: WB6). Also, the outside thermal loads at these levels are trivial (neutrons, x-rays, electron impacts, etc., compared to the Ohmic heating of the magnet wires). Also, with WB6 I believe they had to wait for up to a few hours (?) between shots to allow the magnets to cool down in the vacuum chamber. With active cooling (either with water or liquid nitrogen), this recovery time time could be shortened to a few minutes. The vacuum pumping time limits (to get rid of the excess neutrals from the gas puffers or recombined ions) might be the limiting factor for time between tests). They might manage 20-30 tests per day rather than 4-5. they could accumulate data much faster, allowing for more statistical precision under same and varied conditions of operation. They might accomplish in a few days what it took EMC2 several months to do with WB6.
Dan Tibbets
I think the only real advantage of liquid nitrogen is the cooling of copper wires so that the conductivity increases 5-8X that at room temperature.. Water would also carry away the ohmic heat generated, and probably at much less cost, even when the additional resistance is considered. The advantage is that at liquid nitrogen temperatures, the increased conductivity allows for smaller wires, which allows for more windings, thus higher B-fields at the same current (heating). The reason high flow rates of water couldn't compensate for this is because of thermal conductivity issues. You have to get the heat out of the bundle of copper wires before you can carry it away in the liquid. IE: you would need to reduce the packing fraction- which means there is less room for wire windings.
[EDIT] Actually when I originally wrote this I was thinking of a linear relationship between the B- field (current * windings) and the heating. Actually it is a log-rhythmic (squared) relationship, so the above comparison is even more biased towards minimizing the current needed for any given B field generation.
http://www.ru.nl/hfml/research/levitati ... _solenoid/
Some liquid nitrogen could be used as a boost for vacuum pumping (cold trap), but I don't know if that is piratical at the vacuum levels they need, or if it would add anything to the presumed turbomolecular , or high capacity diffusion pumps they are using.
As far as steady state, with these research machines, a few tens of seconds of running the magnets may seem like an eternity for them. For real steady state machines (weeks on end) the cooling would need to be more robust. With short duration research machines the thermal mass gives them a buffer for a few seconds (? eg: WB6). Also, the outside thermal loads at these levels are trivial (neutrons, x-rays, electron impacts, etc., compared to the Ohmic heating of the magnet wires). Also, with WB6 I believe they had to wait for up to a few hours (?) between shots to allow the magnets to cool down in the vacuum chamber. With active cooling (either with water or liquid nitrogen), this recovery time time could be shortened to a few minutes. The vacuum pumping time limits (to get rid of the excess neutrals from the gas puffers or recombined ions) might be the limiting factor for time between tests). They might manage 20-30 tests per day rather than 4-5. they could accumulate data much faster, allowing for more statistical precision under same and varied conditions of operation. They might accomplish in a few days what it took EMC2 several months to do with WB6.
Dan Tibbets
To error is human... and I'm very human.
Two possibilities.ladajo wrote:Another issue with packing loose to promote heat transfer is shock leaping the coil wires will do when the machine is fired. This was failure mechanism for WB6, specifically the wiring at the nubs rubbed on the casing during shots, and shorted out.
* Put a single pipe thru the middle of the coil running LN thru the pipe. Pack wire tightly around the pipe so that you can cool things down by conduction for a pulsed run.
* Use tubing (my preference is square tubing) in lieu of wire and pack it as tight as you can. Run the LN thru the tube. Might allow for longer runs but may have nasty problems if it heats too quickly. Kind of a magnified WB6.
I liked the tubing idea as well, specifically small diameter copper tubing with LN cooling. That way you can get multiple turns, and the benefits of 5 to 8 conduction, as well as longer runs.
They used large bore copper tubing in the water cooled machines early on. Why not shrink it down for some more turns and internal cooling. It is not so hard to dip and bake the assembly either, then even ceramic it in total to help control outgassing. Place that inside a torus shell, and even leave an outer gap for more cooling flow between the winding assembly and the casing. Obviously, more complicated equals more cost, but I think this approach is not that costly.
As a side note, famulus is playing with tape loading configurations, but in his concepts is not leaving a lot of space for cooling, so I remain (as I have for a while now) curious as to how he will cool for runs.
As another aside note, I had a chat with a senior gov official that watches S&T stuff for a living, and he was unaware of navy efforts with polywell. We talked on implications of implimentation, and his observation was that oil would take a dive. Good back and forth on this.
They used large bore copper tubing in the water cooled machines early on. Why not shrink it down for some more turns and internal cooling. It is not so hard to dip and bake the assembly either, then even ceramic it in total to help control outgassing. Place that inside a torus shell, and even leave an outer gap for more cooling flow between the winding assembly and the casing. Obviously, more complicated equals more cost, but I think this approach is not that costly.
As a side note, famulus is playing with tape loading configurations, but in his concepts is not leaving a lot of space for cooling, so I remain (as I have for a while now) curious as to how he will cool for runs.
As another aside note, I had a chat with a senior gov official that watches S&T stuff for a living, and he was unaware of navy efforts with polywell. We talked on implications of implimentation, and his observation was that oil would take a dive. Good back and forth on this.
The previous post seems unclear. The magnetic field is dependent only on the amp turns. The magnetic field obtainable is dependent on the amp turns you can pack into a given volume and adequately cool.
If you pass 1 amp of current through a single loop of wire that is one meter long you have 1 unit of heating. If the resistance is decreased 5 fold, only 1/5 units of heating would be produced. If the wire is reduced in cross sectional area to 1/5th the original wire, you could pack 5 loops into the same volume,.
At 1 amp the magnetic field would be 5 times as strong and the heating would be the same as the original wire (resistance/ meter of Room temperature copper * 1 meter / resistance /meter of LN cooled copper * 5 meters).
The relationship is linear.
OK, now I've succeeded in thoroughly confusing myself. Where does the square relationship come into play? The only thing I can think of is the surface area vs the internal volume of the wires. The smaller wire has a larger ratio. This means it would be easier to extract the heat from the wire. To get the same amount of heat carried away from the larger wire, you would need higher temperature differentials ( coolant flow rate). So, this heating mismatch may be more practical than theoretical. Does that make sense, am I missing something?
Dan Tibbets
If you pass 1 amp of current through a single loop of wire that is one meter long you have 1 unit of heating. If the resistance is decreased 5 fold, only 1/5 units of heating would be produced. If the wire is reduced in cross sectional area to 1/5th the original wire, you could pack 5 loops into the same volume,.
At 1 amp the magnetic field would be 5 times as strong and the heating would be the same as the original wire (resistance/ meter of Room temperature copper * 1 meter / resistance /meter of LN cooled copper * 5 meters).
The relationship is linear.
OK, now I've succeeded in thoroughly confusing myself. Where does the square relationship come into play? The only thing I can think of is the surface area vs the internal volume of the wires. The smaller wire has a larger ratio. This means it would be easier to extract the heat from the wire. To get the same amount of heat carried away from the larger wire, you would need higher temperature differentials ( coolant flow rate). So, this heating mismatch may be more practical than theoretical. Does that make sense, am I missing something?
Dan Tibbets
To error is human... and I'm very human.
A third possibility would be a Bitter magnet arrangement, which has been mentioned before.KitemanSA wrote:Two possibilities.ladajo wrote:Another issue with packing loose to promote heat transfer is shock leaping the coil wires will do when the machine is fired. This was failure mechanism for WB6, specifically the wiring at the nubs rubbed on the casing during shots, and shorted out.
* Put a single pipe thru the middle of the coil running LN thru the pipe. Pack wire tightly around the pipe so that you can cool things down by conduction for a pulsed run.
* Use tubing (my preference is square tubing) in lieu of wire and pack it as tight as you can. Run the LN thru the tube. Might allow for longer runs but may have nasty problems if it heats too quickly. Kind of a magnified WB6.
Dan Tibbets
To error is human... and I'm very human.
.gov site updated for last quarter in case it has not been noticed. It is on the earlier site, link:
http://www.recovery.gov/Transparency/Re ... =Contracts
http://www.recovery.gov/Transparency/Re ... =Contracts
Counting the days to commercial fusion. It is not that long now.
Interesting.mvanwink5 wrote:.gov site updated for last quarter in case it has not been noticed. It is on the earlier site, link:
http://www.recovery.gov/Transparency/Re ... =Contracts
Quarterly Activities/Project Description:
WB8 is fully under construction, progress made on Theoretical modeling of the Polywell.
Engineering is the art of making what you want from what you can get at a profit.