Seeking the minimum distance between the rings.
The "magrid" is short for "magnatic grid" - the magnets protecting the grid from collision and the grid itself.mattman wrote:Is this what the MagGrid is? When Bussard states he is: turning on the MagGrid, does this mean he is switching on the voltage drop between the rings and the cage?
The "beta=1" condition Bussard refers to is when the expected velocity of electrons gives enough deflection in the magnetic field to them that they cannot actually reach the charged surface of the grid.
This cloud of electrons kept from neutralizing is the polywell.
Wandering Kernel of Happiness
He should have just called it "the rings". MaGrid is a holdover from the fusor.
We will need new terms.
===================
So based on his IAF paper here are the steps for starting up WB-6:
1. Pump chamber down to <1E-7 torr. This means there is trace air in there.
2. Bleed in D2 Gas. This raises pressure to 3E-4 torr.
3. Switch on voltage (12.5 Kv) between rings and outside cage.
4. Turn on electron emitters. Electrons fall down voltage and fly into rings/center. These electrons fly past D2, striping off their electrons making them ions. This ionizes the gas.
5. Turn on electromagnetic rings. The electrons ride the magnetic field, but pile up in center. About 5.5E11 pile up. This makes a (10 Kv) voltage drop.
6. The electron cloud is a swarm of moving electrons. The moving charge makes a magnetic field. This internal field aligns against the ring field. The inside field pushes back. This pinches off the holes in the ring field. This increases confinement, efficency. This is the “Whiffleball”.
7. Ions pile up in dead center. They are hemmed in by the electrons. This is the “virtual anode”.
8. Some ions fall down the 10 Kv drop, they reach 10 KeV of energy. They hit. They fuse.
9. The high energy products fly out of reactor. They do not initiate new reactions.
10. They hit walls, heating the walls. There are fluid tubes in the walls. Fluid heats up, makes steam, turns turbines, we solve energy crisis.
We will need new terms.
===================
So based on his IAF paper here are the steps for starting up WB-6:
1. Pump chamber down to <1E-7 torr. This means there is trace air in there.
2. Bleed in D2 Gas. This raises pressure to 3E-4 torr.
3. Switch on voltage (12.5 Kv) between rings and outside cage.
4. Turn on electron emitters. Electrons fall down voltage and fly into rings/center. These electrons fly past D2, striping off their electrons making them ions. This ionizes the gas.
5. Turn on electromagnetic rings. The electrons ride the magnetic field, but pile up in center. About 5.5E11 pile up. This makes a (10 Kv) voltage drop.
6. The electron cloud is a swarm of moving electrons. The moving charge makes a magnetic field. This internal field aligns against the ring field. The inside field pushes back. This pinches off the holes in the ring field. This increases confinement, efficency. This is the “Whiffleball”.
7. Ions pile up in dead center. They are hemmed in by the electrons. This is the “virtual anode”.
8. Some ions fall down the 10 Kv drop, they reach 10 KeV of energy. They hit. They fuse.
9. The high energy products fly out of reactor. They do not initiate new reactions.
10. They hit walls, heating the walls. There are fluid tubes in the walls. Fluid heats up, makes steam, turns turbines, we solve energy crisis.
My recollection is more like this.
1 ) Open the copper switch.
2 ) Ensure gas solenoid switch is closed. (tank pressure < 1E-6 torr)
3 ) Enable all recording instrumentation.
4 ) Charge section of gas tubing to desired value (usually 300mtorr).
5 ) Activate heating current for emitters, via Siemens switch.
6 ) Enable the Hipo power supply, and charge capacitors to desired voltage.
7 ) Disable Hipo, via “Stop Charge”, while leaving caps charged.
8 ) Turn on magnetic field to desired value, via battery bank IGBTs.
9 ) Close copper switch, thus allowing emitters to start emitting electrons
10) Close relay for gas solenoid; gas into system, discharge cap bank.
11) Turn off uncooled magnets, once sure caps have discharged.
12) Secure emitters, gas line, high voltage, and stop recording data.
1 ) Open the copper switch.
2 ) Ensure gas solenoid switch is closed. (tank pressure < 1E-6 torr)
3 ) Enable all recording instrumentation.
4 ) Charge section of gas tubing to desired value (usually 300mtorr).
5 ) Activate heating current for emitters, via Siemens switch.
6 ) Enable the Hipo power supply, and charge capacitors to desired voltage.
7 ) Disable Hipo, via “Stop Charge”, while leaving caps charged.
8 ) Turn on magnetic field to desired value, via battery bank IGBTs.
9 ) Close copper switch, thus allowing emitters to start emitting electrons
10) Close relay for gas solenoid; gas into system, discharge cap bank.
11) Turn off uncooled magnets, once sure caps have discharged.
12) Secure emitters, gas line, high voltage, and stop recording data.
I have not seen the term "Magrid" used in fusor talk. Occasionally some mussings on magnetically shielded grids are discussed.
I'm unsure about the pressures mentioned. It is important to note that the deuterium is introduced via a gas puffer at the inner edge of the the magrid. Volumes have been mentioned, but without working the numbers. It is safe to say a quantity of gas is introduced into the magrid. And this stock of gas is enough to fill the resulting Wiffleball volume to a density of approximatly 10^19 particles / M^3 (in WB6). In a working Polywell, this internal density may reach as high as 10^22 particles / M^3. This would be
pressures of ~ 1 Micron and 1000 Microns respectively. The starting pressure of ~ 0.1 Micron (10^-7 atm) is sort of the porest baseline tolorable. Starting pressures 10-100 times less would be perferable.
There are two important points. From the starting internal pressure, the Wiffleball trapping factor can contain for adiquate times a charged particle density/ pressure ~ 1000 times greater in WB6, and perhaps 1,000,000 times greater in a working Polywell. That would be a pressure of ~ 700 Microns or ~ 10^-3 atm. This assumes a starting pressure of 0.1 Microns.
The important point is the pressure of charged particles that can be contained, not nessisarily the starting pressure, except for arcing concerns.
Arcing to non magnetically shielded surfaces will start at ~ 5 Microns. So the pressure outside the magrid cannot exceed this. This in relation to the neutral gas puffers and the starting pressure and the magrid radius determine the experimental limits. The neutral gas has a starting velocity, and will travel across the magrid and escape on one pass, unless it is ionized. This is a time dependant process and limits the test time to ~ 1 millisecond in a 30cm machine. Bussard expected this issue to be eased considerably in larger machines. The ionization process is logrhythmic, so a 60 cm machine would have ~ twice the transit time for the neutral gas and this would allow the ionization efficiency to increase proportionatly more. Eg: if 30 cm allowed for 90% ionization, a 60 cm machine might allow for a 99% efficiency, etc. The larger machine would also allow for a greater percent of the ionization to occur relatively closer to the edge of the wiffleball and this would increase "monoenergetic" properties.
An ion gunned Polywell has considerably different dynamics and presumably could allow for much greater test times, delaying for modest to extream times the onset of arcing. Ideally this could be extended to steady state times of hours or more. Other issues like sputtering and vacuum pumping becomes critical.
D Tibbets
I'm unsure about the pressures mentioned. It is important to note that the deuterium is introduced via a gas puffer at the inner edge of the the magrid. Volumes have been mentioned, but without working the numbers. It is safe to say a quantity of gas is introduced into the magrid. And this stock of gas is enough to fill the resulting Wiffleball volume to a density of approximatly 10^19 particles / M^3 (in WB6). In a working Polywell, this internal density may reach as high as 10^22 particles / M^3. This would be
pressures of ~ 1 Micron and 1000 Microns respectively. The starting pressure of ~ 0.1 Micron (10^-7 atm) is sort of the porest baseline tolorable. Starting pressures 10-100 times less would be perferable.
There are two important points. From the starting internal pressure, the Wiffleball trapping factor can contain for adiquate times a charged particle density/ pressure ~ 1000 times greater in WB6, and perhaps 1,000,000 times greater in a working Polywell. That would be a pressure of ~ 700 Microns or ~ 10^-3 atm. This assumes a starting pressure of 0.1 Microns.
The important point is the pressure of charged particles that can be contained, not nessisarily the starting pressure, except for arcing concerns.
Arcing to non magnetically shielded surfaces will start at ~ 5 Microns. So the pressure outside the magrid cannot exceed this. This in relation to the neutral gas puffers and the starting pressure and the magrid radius determine the experimental limits. The neutral gas has a starting velocity, and will travel across the magrid and escape on one pass, unless it is ionized. This is a time dependant process and limits the test time to ~ 1 millisecond in a 30cm machine. Bussard expected this issue to be eased considerably in larger machines. The ionization process is logrhythmic, so a 60 cm machine would have ~ twice the transit time for the neutral gas and this would allow the ionization efficiency to increase proportionatly more. Eg: if 30 cm allowed for 90% ionization, a 60 cm machine might allow for a 99% efficiency, etc. The larger machine would also allow for a greater percent of the ionization to occur relatively closer to the edge of the wiffleball and this would increase "monoenergetic" properties.
An ion gunned Polywell has considerably different dynamics and presumably could allow for much greater test times, delaying for modest to extream times the onset of arcing. Ideally this could be extended to steady state times of hours or more. Other issues like sputtering and vacuum pumping becomes critical.
D Tibbets
To error is human... and I'm very human.
The Beta= 1 condition does not mean the above. The electrons reaching the surface of the magrid is dependent on the KE of the electron and the strength of the magnetic field. But this has little importance in the Polywell. It may be more significant in a Tokamak or other cuspless machine where these losses - on a single pass, or through a cumulative random walk process (ExB drift) is the major loss mechanism. So long as a minimum relationship is maintained, hitting the magnets with electrons is trivial in the Polywell.WizWom wrote:The "magrid" is short for "magnatic grid" - the magnets protecting the grid from collision and the grid itself.mattman wrote:Is this what the MagGrid is? When Bussard states he is: turning on the MagGrid, does this mean he is switching on the voltage drop between the rings and the cage?
The "beta=1" condition Bussard refers to is when the expected velocity of electrons gives enough deflection in the magnetic field to them that they cannot actually reach the charged surface of the grid.
This cloud of electrons kept from neutralizing is the polywell.
What is important is how increasing Beta towards one effects the cusp geometry. The cusp loss surface area is effectively decreased compared to the total surface area of the Wiffleball. At low Beta, the Polywell has a cusp leakage rate such that an electron may make ~ 60 passes before hitting a cusp and escaping ( from the patent application). This equates into the cusp loss surface areas of ~ 1/60th of the total surface area of the low Beta surface area bounded by the magnetic fields. With Beta at 0.99999 The cusp loss surface area is reduced to ~ 1/3,000 of the total magnetically contained plasma surface area, now called the Wiffleball surface area. The surface area of the plasma ball increases some, but most of the change comes from large decreases in the cusp loss surface areas.
Thus in a Polywell the Beta is important because it limits losses through cusps. The cross field losses are almost irrelevant.
In a Tokamak, if macro instabilities can be eliminated, the containment time is governed by the drift losses, mostly ExB drift. In a collisional plasma (essential if you want fusion collisions) this drift is governed by the square of the density, the temperature, B (square of the B field?) field strength and the distance to the wall. This is why some say that magnetic confinement of plasmas (specifically neutral plasmas) is poor. Still with sufficient size and limited densities it can be made to work. Confinement times of hundreds of seconds is possible for magnetic confinement of ions provided the density is low enough. It is my understanding that this is why Tokamaks only work at low Beta (relatively low densities) With Polywells and their non neutral plasmas, the electrons can be maintained at substantially higher densities because of their significantly smaller gyro radii- this plays a large role in the random walk process as when electrons collide with each other in a magnetic field the particle always jumps one gyroradii. It takes many more collisions for the scattered electrons to reach the magnetic container wall. Still, in the Polywell this electron ExB drift may lead to non cusp magnetic confinement times of perhaps a second. This is much worse than a tokamak, but remember this is at ~ 1000 times greater densities and ~ 1,000,000 times greater collision rates. When cusps are considered, electron magnetic confinement times drop to~ < 1 millisecond. This implies that cusp losses dominate over drift losses by a factor of ~ 1-10 thousand, even with the Wiffleball effects. In the patent application, it is mentioned that recirculation may reduce this factor to ~ 10-100X, but still, the cusp losses dominates the electron magnetic field drift losses.
Of course, the ions magneti drift losses are ~1/ 60X (?) [EDIT- The ions gyro radii is ~ 60 times greater than the electrons, so the magnetic confinement time for the ion is ~= to or < than one sixtieth of the electrons ] that of the electrons due to their greater gyro radii at the same energies, But this is where the non neutral plasma comes in. The resultant ion electrostatic confinement make the ion magnetic loss issues mostly trivial. The exception to this is with fusion produced ions. These high energy ions are not contained by the potential well, so the magnetic confinement of these ions dominates. Because of their low collision crossection , they depart the reaction volume relatively quickly without heating the plasma.
The above drift consideration about the gyro radii applies to these fusion ions. If the B field is not strong enough , and/or the distance to the wall is not large enough,then the ion will hit the magnetic can wall on a single pass.
This led to R. Nebel's point that to prevent this a Polywell would need a minimum of ~ 3.5 Tesla magnetic fields (presumably in a ~ 3 meter diameter machine). Once this condition is met, most of the fusion ions (from P-B11 fusion) would escape via the cusps and direct energy conversion becomes a possibility. Other engineering concerns about collisional heating of the Magrids is also relaxed.
Dan Tibbets
Last edited by D Tibbets on Tue Jul 10, 2012 9:21 pm, edited 1 time in total.
To error is human... and I'm very human.
Just to clarify:
Cap discharge ==> Electron Emission.
Puff gas-discharge ==> Deuterium (the ions) emission.
MaGrid ==> the rings.
The chamber is 1E-7 torr. The gas emerges at 3E-4 torr, is also about the gas pressure inside the rings.
Based on what kiteman said, the order is:
1. Pump down tank.
2. Turn on voltage between rings and cage.
3. Turn on rings.
4. Emit electrons.
5. Emit ions.
Cap discharge ==> Electron Emission.
Puff gas-discharge ==> Deuterium (the ions) emission.
MaGrid ==> the rings.
The chamber is 1E-7 torr. The gas emerges at 3E-4 torr, is also about the gas pressure inside the rings.
Based on what kiteman said, the order is:
1. Pump down tank.
2. Turn on voltage between rings and cage.
3. Turn on rings.
4. Emit electrons.
5. Emit ions.
No, not quite. The "cap discharge" is the "apply voltage between the MaGrid and chamber which then accelerates the electron into the MaGrid rather than having them just fly off in ALL directions. So, by your numbers above, it ismattman wrote:Just to clarify:
Cap discharge ==> Electron Emission.
Puff gas-discharge ==> Deuterium (the ions) emission.
MaGrid ==> the rings.
The chamber is 1E-7 torr. The gas emerges at 3E-4 torr, is also about the gas pressure inside the rings.
Based on what kiteman said, the order is:
1. Pump down tank.
2. Turn on voltage between rings and cage.
3. Turn on rings.
4. Emit electrons.
5. Emit ions.
1
3 (I.e. turn on magnets)
2 & 4 & 5 simultaneously.
Note that they don't actually emit ions, they emit neutral gas which ionizes in the core.
The sequence of events is some what arbitrary. Certainly the chamber has to be pumped down to ~ 10^-7 to -8 atms befor anything else can happen. The magnet current , e- gun current can then be turned on and run for a few 10s of milliseconds. Depending on how you are building the Wiffleball, the sequence may vary some (Bussard described two methods of Wiffleball formation)
The critical elements are dictated by the capacitor discharge that powers the coil surface electric field. And, the absolute limit is the arc discharge that follows the gas puff. So this is possibly the last event in the chain. There is some tolerable variation depending on the setup. A current limited power supply replacing the capacitor, and tighter control of the gas puffer may change the optimal sequence. Ion guns of course also changes the picture. As for WB6, Kitemans' list is consistent with the WB6 final report. This does not mean that WB7, or WB8 followed this sequence to the letter, nor that it was the optimal sequence from a physics viewpoint. It was chosen largely because of the limits on the equipment that was available for WB6 testing.
Dan Tibbets
The critical elements are dictated by the capacitor discharge that powers the coil surface electric field. And, the absolute limit is the arc discharge that follows the gas puff. So this is possibly the last event in the chain. There is some tolerable variation depending on the setup. A current limited power supply replacing the capacitor, and tighter control of the gas puffer may change the optimal sequence. Ion guns of course also changes the picture. As for WB6, Kitemans' list is consistent with the WB6 final report. This does not mean that WB7, or WB8 followed this sequence to the letter, nor that it was the optimal sequence from a physics viewpoint. It was chosen largely because of the limits on the equipment that was available for WB6 testing.
Dan Tibbets
To error is human... and I'm very human.
publically available.
I would also add that several references have been made a number of times to statements by project team members about confinement/wiffleball.
I would also add that several references have been made a number of times to statements by project team members about confinement/wiffleball.
The development of atomic power, though it could confer unimaginable blessings on mankind, is something that is dreaded by the owners of coal mines and oil wells. (Hazlitt)
What I want to do is to look up C. . . . I call him the Forgotten Man. (Sumner)
What I want to do is to look up C. . . . I call him the Forgotten Man. (Sumner)
There is a paper that describes the Wiffleball formation and the methods of achieving this.
And, there seems to be some confusion about the Magrid. There are two systems associated with the magrid.
The magnetic field creating windings are contained within the magrid cans, and insulated from the metalic surface. This insulation breakdown is what lead to failure of WB6. The current is turned on to the wires at relatively low voltage and high current . Perhaps 1-2 thousand Amps and Voltage of perhaps 12-24 volts. This is powered from Marine batteries. I havn 't calcultated the voltage nessisary, but it would be easy to do so using Ohm's law once the wire gauge/ thickness was known. along with the length ( ~ 2,000 meters). There have been mention that the wires in this electromagnet will heat up , but it takes one or more seconds, so the relative time duration of this system is huge compared to other systems, so they are turned on early, once the chamber has been pumped down.
The metal surface of the magrid (in WB6) was the high voltage electrode. The voltage stored in the capacitor is connected to this surface. So long as the pressure is below the Pashin breakdown voltage there will be little current flow, so the switch can be closed. The magrid surface thus provides the accelerating potential. The electron current comes from the electron guns. There may be perhaps ~50 amps of current at low voltage (12V?) so only ~ 60 Watts of power is coming from the E- guns at this Wiffleball maintainance time. This electron current multiplied by the accelerating magrid potential of ~ 12,000 volts gives the final power of ~ 500,000 Watts in WBg during the ~ 0.2 milliseconds that the Wiffleball was maintained.
The gas puff of ~ 1-2 cc of deuterium gas at ~ atmospheric pressure (?) will quickly expand to fill the aviable chamber space, and this
This is will occur within a few milliseconds. It takes longer in a larger machine. I've not read what percentage of gas was not ionized and traped as it passed within the magrid. I generally use an arbitrary 90% ionization for WB6. This leaves 10% to expand throughout the chamber. Once this external average gas pressure reaches about 1-10 Microns, arcing begins As additional gas exits the magrid volume and increases the density outside the magrid, this arcing quickly builds proportionatly to the gass pressure* The capacitor supplied voltage droops inversely proportionatly to the current. The potential well is destroyed and, recirculation of electrons is destroyed, and ion electrostatic confinement time is destroyed (plasma quickly becomes neutral), so the containment is lost through the magnetic cusps without the electrostatic modifiers.
This pressure outside the magrid is what limits the test durationand supsuquent rapid draining of the capacitors
* The volume of gas puffed has been stated, but I am too lazy to look it up. Also the dynamics of the valve opening and closing play a role, but I will ignore them here.
The WB6 Faraday cage defined the volume within the chamber. Assume it was ~ 1 meter in diameter so the working volume was ~ 1 cubic meter. The amount of gas puffed into the chamber was (assume) ~ 1 cc at atmospheric pressure. Expanded into the 1 cubic meter volume resulted in a pressure of ~ 1cc / 1,000,000 cc. yields a pressure of ~ 0.0000001 atmospheres, of ~ 1 Micron. This would seem to be within acceptable margins, especially if a substantial portion of this gas was quickly ionized and contained within the Wiffleball. But, the actual puffer volume was, I think several times greater, there were structures closer to the magrid than one meter (like E-guns, gass puffer tube, etc.) and once any arcing (glow discharge started, there would be sputtering and significant outgassing from the walls. This may be a consideration of why Nebel declined to consider smaller test machines. Outgassing is proportional to the surface area of the walls, etc. And this becomes less dominate as the machine grows in size. The surface area scales slower than the volume.
Also, the electrons from the e-guns that do not enter the magrid need to be accounted for.
Dan Tibbets
And, there seems to be some confusion about the Magrid. There are two systems associated with the magrid.
The magnetic field creating windings are contained within the magrid cans, and insulated from the metalic surface. This insulation breakdown is what lead to failure of WB6. The current is turned on to the wires at relatively low voltage and high current . Perhaps 1-2 thousand Amps and Voltage of perhaps 12-24 volts. This is powered from Marine batteries. I havn 't calcultated the voltage nessisary, but it would be easy to do so using Ohm's law once the wire gauge/ thickness was known. along with the length ( ~ 2,000 meters). There have been mention that the wires in this electromagnet will heat up , but it takes one or more seconds, so the relative time duration of this system is huge compared to other systems, so they are turned on early, once the chamber has been pumped down.
The metal surface of the magrid (in WB6) was the high voltage electrode. The voltage stored in the capacitor is connected to this surface. So long as the pressure is below the Pashin breakdown voltage there will be little current flow, so the switch can be closed. The magrid surface thus provides the accelerating potential. The electron current comes from the electron guns. There may be perhaps ~50 amps of current at low voltage (12V?) so only ~ 60 Watts of power is coming from the E- guns at this Wiffleball maintainance time. This electron current multiplied by the accelerating magrid potential of ~ 12,000 volts gives the final power of ~ 500,000 Watts in WBg during the ~ 0.2 milliseconds that the Wiffleball was maintained.
The gas puff of ~ 1-2 cc of deuterium gas at ~ atmospheric pressure (?) will quickly expand to fill the aviable chamber space, and this
This is will occur within a few milliseconds. It takes longer in a larger machine. I've not read what percentage of gas was not ionized and traped as it passed within the magrid. I generally use an arbitrary 90% ionization for WB6. This leaves 10% to expand throughout the chamber. Once this external average gas pressure reaches about 1-10 Microns, arcing begins As additional gas exits the magrid volume and increases the density outside the magrid, this arcing quickly builds proportionatly to the gass pressure* The capacitor supplied voltage droops inversely proportionatly to the current. The potential well is destroyed and, recirculation of electrons is destroyed, and ion electrostatic confinement time is destroyed (plasma quickly becomes neutral), so the containment is lost through the magnetic cusps without the electrostatic modifiers.
This pressure outside the magrid is what limits the test durationand supsuquent rapid draining of the capacitors
* The volume of gas puffed has been stated, but I am too lazy to look it up. Also the dynamics of the valve opening and closing play a role, but I will ignore them here.
The WB6 Faraday cage defined the volume within the chamber. Assume it was ~ 1 meter in diameter so the working volume was ~ 1 cubic meter. The amount of gas puffed into the chamber was (assume) ~ 1 cc at atmospheric pressure. Expanded into the 1 cubic meter volume resulted in a pressure of ~ 1cc / 1,000,000 cc. yields a pressure of ~ 0.0000001 atmospheres, of ~ 1 Micron. This would seem to be within acceptable margins, especially if a substantial portion of this gas was quickly ionized and contained within the Wiffleball. But, the actual puffer volume was, I think several times greater, there were structures closer to the magrid than one meter (like E-guns, gass puffer tube, etc.) and once any arcing (glow discharge started, there would be sputtering and significant outgassing from the walls. This may be a consideration of why Nebel declined to consider smaller test machines. Outgassing is proportional to the surface area of the walls, etc. And this becomes less dominate as the machine grows in size. The surface area scales slower than the volume.
Also, the electrons from the e-guns that do not enter the magrid need to be accounted for.
Dan Tibbets
To error is human... and I'm very human.