Bussards description of WB-6 Startup, explained.

[mattman]

Bussards’ IAF paper provides us some idea of how to startup WB-6. The machine was placed in a vacuum chamber. The chamber was pumped down to <1E-7 torr. The rings are now in a vacuum environment. First deuterium is puffed into the chamber. There has been talk about using ion guns to inject deuterium. However in WB-6 this injection was “gas puffed” into the chamber. Puffing gas into the vacuum raises the vacuum pressure. It is not clear if it raises the pressure to 3E-4 torr, but this startup calculation is done for that pressure. This may be some kind of error. If an expert would like to weigh into clear up this confusion we would appreciate it. This description is lifted from reference [8] on bottom of page 11. The startup calculation is done for a pressure of 3E-4 torr. However, Bussard states and shows that the WB-6 experiment started at 1E-7 torr.

Now high energy electrons are injected. They are injected using an electron gun. This “gun” makes a voltage drop from the electron source to some other point in the chamber. The electrons see the voltage, experience a Lorenz force and fall down the voltage. They pass through the gas cloud in the center. Bussard now describes a mechanism to ionize the cloud. The fast moving electrons ionize the gas inside WB-6. The fast moving electrons strip off electrons from the deuterium. This increases the number of free electrons and free ions flying around. It separates material inside the machine. Ideally, both clouds of electrons and ions have too much energy to recombine. It would not take much energy to strip off these electrons. The ionization energy for deuterium is 14.9 eV. Bussard stated in his Google presentation that the drive voltage for WB-6 was 12,500 volts. Hence, a high energy electron might come into the cloud with 12,500 eV of energy. With so much energy if it hits a deuterium atom, it will heat it up. If the deuterium is now “hotter” than 14.9 eV, the electron flies off. The gas is ionized. Now you can see one reason why the Navy wants to get several 10,000 volt electron guns for their reactor [10]. These drive beams seemed to all be within this range. Bussard stated in his Google presentation that the e-beam voltage for the PZL-1 machine (built in 2003) was 15,000 volts [19].

The number of electrons is growing inside the machine. As the electron cloud grows the cloud of ionized deuterium grows. Bussard estimated that this process happens on the order of several microseconds. He estimated that all the electrons were stripped off the deuterium in about this amount of time. At this point the beta ratio is 0.01. The beta ratio is the ratio of the plasma pressure to the magnetic field pressure. It sets the limits to this process. When the beta ratio is 1 the plasma pressure has equaled the magnetic field pressure outside. The device is full. The cloud in the center cannot handle any more electrons and ions. Past this amount and the cloud becomes unstable. This instability comes from magnetohydrodynamics. As I understand it, there is a beta ratio for the electrons and the ions. When the electron beta ratio is 0.01, Bussard estimated that the electrons are at 100 eV [8].

Now Bussard proposes that there is electron heating. Let us assume that the hot beam of electrons is still being injected. This is possible because the center is not filled up yet. In fact the magnets are not turned on. Bussard turns the magnetics on. More specifically, he starts pumping roughly 4,000 amps of current through the rings. This generates six fields. These six fields are all pointed into the center. If they were bar magnetics it would be like 6 north poles pointed into the middle. In the center there is a zone of no magnetic field, a pocket. It is a star structure with 14 points; one point pointing to each corner and each side. It may be possible that the magnetic fields reconnect. This is a totally open question. Someone needs to examine if plasma fields, under such pressures and energies would even be in the right operating conditions for reconnection. Reconnection is when magnetic fields in the presence of plasma recombine. It is a strange concept for many to understand. We are taught that magnetic field lines can never cross or reconnect. Someone please look at this: would we even expect this to happen? If it did happen, what would this mean for containment?

The electrons are coming in at 12,500 eV, and the average electron is still 100 eV and the magnetic fields are switched on. Now Bussard argues that electron heating occurs. The argument is the hot electrons hit the cold electrons in the center and impart energy. This continues until the average electron energy is 2,500 eV. Bussard calculates that the time scale for heating as 1 microsecond. Within a few microseconds, the gas is entirely ionized and in about 20 microseconds the machine is full of electrons at an average temperature of 2,500 eV. For the WB-6 test this was about a fourth the temperature of the ions. Rider would disagree here. He predicts that the ions cannot have more than 5% variation in temperature [17]. Also, the ion and electron energies supposedly equilibrate. The rate of transfer of energy essentially depends on the ratio of the number of electrons moving more slowly than the ions [18]. This is shown below.

Electron-ion Heating Rate ~ Electrons with lower energy than ion mean/ Electrons with higher energy than ion mean

Hence, Rider argues, the cloud must be at roughly one temperature.


However, if you continue with Bussards description the electron cloud is at 2,500 eV and it has created a 10,000 voltage drop in the center of the device. Using Gausses law you can estimate that the cloud in WB-6 had about 5.5E11 net electrons inside WB-6. The ions are flying in at 10,000 eV and fusing. We know that if they hit with that much energy they are in the right ballpark for fusion. NIF and ITER both are trying to get their average cloud temperatures to around 10,000 to 20,000 eV. It may be that Rider and Bussard are both correct. The average temperature of the ions could be 2,500 eV, the same at the electrons, except that the ions have a bell curve of energies. If part of that bell curve of ion energy contains ions at 10,000 eV, then fusion seems likely for a small amount of material.

[D Tibbets]

I think you are off on some concepts. My understanding is that the magnets are turned on, the electron guns are turned on and then the neutral deuterium gas is turned on through a pipe just inside the magrid. The initial pressure is ~10 ^-7 atmospheres. The gas at ~ room temperature has a thermalized average speed of around perhaps 100 M/s or ~ 0.1 mm/ microsecond. This is a guess. This means that if the neutral molecule is not disturbed it will transit ~ 300 mm in ~ 3 milliseconds,and then exit the volume within the magrid and accumulate in the external spaces within the vacuum chamber. When this results in an external pressure above ~ 5 * 10^-6 atmospheres Pashin discharge (arcing) begins and voltage on the magrid cannot be maintained against this resulting current to the walls. The 10^-4 pressures refer to the Wiffleball trapping of charged particles at ~ 1000X levels compared to neutrals.

I'm not sure where you get the ~ 1200 eV electrons. It may take only ~ 14 eV to ionize / strip off an electron. But at ~ 100eV the ionization crossection is maximum, so most of the striped electrons will be at this initial energy. These secondary electrons are doing two things, they are being further heated by the hot injected electrons, and they are causing secondary ionizations of further deuterium molecules/atoms. I think this occures on time scales of a few micro seconds. It is an exponential process, each few micro seconds the number of ionized deuteriums are doubled- 1 to 2 to 4 to... 64... 1000,,, 1,000,000, etc. The high speed injected electrons cause the first few ionisations, but the cascading secondary electrons quickly become the dominate mechanism. I once was confused how these secondary electrons were heated to near the energy of the injected electrons if they are the dominate source of electrons in the near neutral plasma. The life time of the electrons vs the lifetime of the ions is the key. The electrons,injected or secondary, may last ~ 10,000 passes (before recirculation) or a couple hundreds of microseconds. The ions may last 1,000,000 passes or more. With the speed ~ 60 times slower than an electron at the same energy, this would result in confinement times of ~ 100 milliseconds. Actual life times would be shorter due to upscattering losses, but the ion lifetimes are at least 1-2 orders of magnitude more than the electrons. Thus the secondary electrons that come from ionization are only a small proportion of the injected electrons, so they can be heated to near the injected electron energies. There are perhaps ~ 10-100 injected electrons for each secondary electron that came from ionization.

The problem with gas puffing into the magrid space is that while the the primary and cascading secondary ionizations are fast, there is a finite amount of time involved and not all the neutral atoms will ionize. This is dependent on the speed of the neutrals (which may be increasing due to non ionizing collisions and recombinations) and the distance they have to travel to escape out the other side , This is why Bussard said larger machines are much better for this mode of operation.

A ~ 10,000 eV electron may be traveling at ~ 10,000,000 M/s. Without confinement an electron would traverse and escape the 30 cm magrid in only about 0.03 microseconds. With cusp confinement this is increased to ~ 2 microseconds. As excess electrons (electrons and ions?) are injected the Wiffleball forms,as Beta increases. Final electron confinement may reach ~ 10,000 passes or ~ 200-300 microseconds. Recirculation accounts for the remaining gain claimed for WB6.
So the test procedure in WB6 was.:

Charge the high voltage capacitors.

Vacuum pumping.

Turn on the magnets powered from batteries.

Turn on the E-Guns powered from batteries.


Turn on the capacitor discharge onto the magrid (10-12,000 volts).

Turn on the gas puffer.

Make what measurements you can while waiting for neutrals (and charged particles) to build up outside the magrid to the point where arcing occurs and terminates the test. The hopefully much higher densities of charged particles inside the magrid do not arc because of rounded corners and the magnetic shielding.

Dan Tibbets

[mattman]

Well this is why I post this here. So we can get some consenus on it. I plan to go through your corrections and weave them into the explaination. Thanks for the input.

[D Tibbets]

Your welcome, but keep in mind that my understanding is not gospel. I may have the advantage that I have read the WB6 final report. This short paper was on the EMC2 web site for a while but was subsequently withdrawn.

Some other details are that the high Beta, inflated Wiffleball condition lasted for only ~0.25 milliseconds. Bussard claimed that this was steady state from a physics perspective as the electron dynamics occurred over microsecond time scales. I'm not sure this can also be said for the ions, and thermalization issues. There were also design and instrumentation limits that confounded the results. The gas puffers were one example. Another was the neutron counters. Apparently their sampling interval was somewhere around 1 millisecond (?). So they could say that all of the neutrons was produced in a one millisecond interval and that the high Beta condition occured during this time, but that says nothing about if the neutrons were produced during high beta, or as a spikes at the beginning or end of high beta or some some other time within the sampling time frame. Also, current measurements from the e guns showed the input current during high beta, but I at least worry about the precision of these measurements with the few brief tests. With hindsight, and presumably better instrumentation and perhaps better power supplies, and vacuum control, I expect that WB7 tests not only confirmed WB6 results but significantly increased the precision and statistical confidence of the results.

Dan Tibbets

[ladajo]

yup. That was the whole point of WB7.

[DeltaV]

I may have the advantage that I have read the WB6 final report. This short paper was on the EMC2 web site for a while but was subsequently withdrawn.

No, still there.
http://www.emc2fusion.org/RsltsNFnlConc ... 120602.pdf

[KitemanSA]

Not that one, the other one. You know the one that's not that one?

[DeltaV]

Know not that one.

[TallDave]

I think you are off on some concepts. My understanding is that the magnets are turned on, the electron guns are turned on and then the neutral deuterium gas is turned on through a pipe just inside the magrid.

Looking at the graphs and description on p12, it looks like the e-guns are turned on first, and stay at very low current (I assume because space charge limits prevent it from confining many electrons and there isn't much current through the machine without a plasma). Then the gas is puffed and gets ionized, and the ions allow the machine to draw 40A of current. Eventually the gas spreads out into the area around the machine and it arcs 4000A to the tank walls.

[ladajo]

As I get it:

Magnets on.
Gas puffs and then quickly thereafter; E-guns fired.

I base this on the fact that the timing is very tricky, and the gas puffs much slower than E-guns firing.

[KitemanSA]

Wouldn't you want to have the well BEFORE puffing the gas? Without the well, the gas and the ions just zip around getting everywhere, no?

[ladajo]

I was thinking this, but then wondered how long they could keep the magnets hot, fire the e-guns, and puff the gas.
It really does sound like a fun timing adventure.
Comparatively, the gas is the slowest bit.

in physics terms:
Magnets:on
e-: injected
Well: formed
ions: puffed
and then zippy zoom and whammo. fusions.

But in real terms, could it be that the slow puff demands trigger before the fast e- injection? 250msec...

Dunno for sure. Just thinking out loud.

[mvanwink5]

Maybe the reason for the high voltage e-guns is for start up only, to establish the well, given there are no ions?

[mvanwink5]

Here is the relevant language for you consideration:
"To solve these anomalies, the additional effort will require the incumbent contractor to further their studies by employing independently powered electron gun arrays operating at up to 10 kilovolt (kV) to inject high energy electrons onto the Plasma Wiffleball 8 core and control the WB formation process."

Best regards

[TallDave]

Wouldn't you want to have the well BEFORE puffing the gas? Without the well, the gas and the ions just zip around getting everywhere, no?

From the nonzero current, I gather the answer is yes.

But I wonder how that initial well compares to the one formed by the 1/1E7 electron-rich plasma. Obviously there are a LOT fewer electrons before the ions arrive, something on the order of a million times less. I expect the pressure is very low initially and hence no WB.