let’s imagine that if there’s even a single atomic layer of erosion per shot, that it would be in nanometer scale thicknesses; then a million shots later, we’re up to millimeter scale, and that could be quite significant.

Most of the anode material loss seems to come from the e-beam damage on the bottom of the anode. The outer anode walls tend not to erode and in some cases they will actually increase in diameter as cathode material will deposit on the anode faster than the anode material erodes. I’ve done tests with copper anodes and stainless steel cathode rods. The anode looked like stainless steel in less than 1000 shots on the outer diameter at 140 kA. The cathode erosion is probably more substantial on the inner diameter but there is plenty of cathode material to burn so it is a minor problem. I’ve burned up plenty of anodes and still use the same cathode rods.

Beryllium should do well compared to copper on a per unit mass basis by comparing the heat of vaporization. However, copper and beryllium are very similar on a per mol basis.

One could make the anode central hole very deep so it enters into a secondary vacuum chamber to manage the e-beam by spreading it out so that it cannot vaporize the beam dump area. The key problems to date have been the small expansion of the e-beam (~3 deg half angle) so if you let the beam grow naturally, it could take meters before you get to a point where the beam cannot vaporize the beam dump. Most experiments don’t have that kind of space. The alternative to is allow the beam into a small vacuum region well away from the anode and use magnetic fields and electric fields to spread out the beam. The key concern is to keep these fields far from the pinch region so it doesn’t screw up the PF dynamics. It is possible to do the second on most experiments but few groups have serious problems with anode erosion because they fire so few shots before changing the anode.

The plasmoids ability to contain the e-beam needs to be demonstrated before one can seriously speculate about the reduction in damage. It is probably prudent to plan for the non-plasmoid e-beam dose and know that a reduced beam is easier to handle.

willit wrote: maybe it should be tested to get to Q>1 before bickering about materials science. there are a lot of engineers out there that would fight to be on the project once this milestone is met.

run it till it breaks then make that part stronger. repeat.:cheese:

Just sharing information from experience. Q>1 is not required to burn up anodes, tax switches or improve pulse power. Some plasma focus operators don’t aspire to Q>1 but we are working on these problems for other applications. Why not design around a problem before you have it?

Burning something up has a cost to it as well and redesigning has a cost as well. The cost of choosing bad switches are already well discussed on this board. I’m not suggesting significant time and money be spent on a massive redesign. I agree that at a few shots per day it isn’t worth addressing some problems. It was a suggestion from my experience operating a rep-rate PF that that anode erosion be addressed and the sources that I observed in my experiments.

Vansig: I already presented my solution a few posts back in this thread. The solution is derived from more than 250,000 shots fired on a single anode and the impact it has on the anode. We revised our design and fired another 100,000 shots to find more problems. We are on our third revision to address thermal management and e-beam management. To be honest, the e-beam problem isn’t the hard problem. The hard problem is figuring out if the anode temperature is an important parameter in optimizing the fusion yield of a pinch and if this temperature is useful/allowed for a given application. There is little to no data in published literature about this subject. It might be relevant to achieving Q>1 as we’ve observed increases in neutron output as temperature increased and then it declined after a peak temperature. The increase is 4-5X in some cases.

The temperature that seemed optimum for us was around 300 C. I don’t know the exact cause but I speculate it has to do with the local gas pressure changing the mass carried by the plasma. Our anode heats due to repetition rate. When it is cold, the implosion time (time between current start and minimum in the dI/dt) is ~700 ns. As the run progresses to steady state, the implosion time drops to ~550 ns. At 550 ns, the neutron yield is optimized. Even at the optimum implosion time, the hot anode performed better than a cold anode.

It is important to note that we use SS304 as the anode. SS304 has poor thermal conductivity. We used emission spectroscopy to look for impurities and found none. I want to do a scan with a mass spec on the fuel gas after the run but it’s not in the cards right now. When we switched to a Moly anode, we noticed a drop in yield of a factor of two. Moly has a thermal conductivity about 1/4 copper. We know from measurements that the SS304 has a large temperature gradient relative to Moly. Is that gradient important? Don’t know. We know that all metals take up deuterium to some extent. We fired a number of shots on an anode in deuterium and then switched to argon. We recorded neutrons during the argon run. The temperature might be the optimum to release hydrogen from the metal and allow it to be replaced with D2. Over several runs you go into saturation. Burns reported a similar hypothesis on DT shots. It is hard to see the difference between H2, D2 and T2 using spectroscopy. We’ve found that operating at 300 C is difficult with Moly because it is nearly isothermal and our o-rings suffer from the temperature. Therefore, we actively cool the anode. In the process, our yield went down. The parameter space grows more complex as the temperature increases because the chemistry of the electrodes and possibly the chamber walls becomes important. These are problems addressed in plasma reactors for deposition and etching but they are seldom thought about in the context of a PF. You mentioned sulfur in SS304 in a past conversation…we see no evidence of sulfur. Based upon conversations related to other work, sulfur is just a bad in copper as SS304. In our experiments, the most likely contamination comes from the vaporization of anode metal. The cloud of vapor expands from the anode base and into the pinch region. Is some tiny mass present before the next shot that enhances the neutron yield as is described in lit with high Z noble gases. I know in our experiments with fuel mixing that we could not see any lines of argon additive at the optimum neutron enhancement. The same could be true for Fe, Moly or tungsten. The metal should plate out so I wouldn’t expect you to see this effect firing a few shots an hour or day. It might be that the vaporized metal releases the impurities trapped in it. Our solution to volatile products is a low base vacuum. We don’t start operating until our base vacuum is ~1E-6 Torr. The machine works better at base pressures of ~5E-7 Torr.

annodomini2 wrote: http://www.liquidmetal.com/

Don’t know if this is of any help, but something to consider.

Too much resistivity. By far. Not a good conductor.
190 microOhms per centimeter. Compared to Aluminum 380 which has 6.5 microohms per centimeter. Copper wasn’t even mentioned.