Since our new experiments will start soon, we thought this is a good time to summarize for everyone, including those who have only started following us recently, what our plans for the experiments are, what we hope to accomplish and why we think we can get good results. The basic plan is to increase the density of the plasmoid and thus the fusion yield by greatly reducing the heavy-metal impurities. So let’s answer some questions about this plan.
1) First, why do we think that reducing heavy-metal impurities will increase fusion yield? We have solid experimental evidence for this hypothesis. Over the history of our FF-1 experimental device, whenever we reduced impurities, yield went up. Whenever we failed to reduce impurities, yield did not rise and whenever we inadvertently increased impurities, yield went down. For example, back in 2011, a mis-fitted part created heavy arcing inside the vacuum chamber, releasing large amounts of vaporized copper and steel impurities into the plasma. Fusion yield plunged all the way down to zero. This was the first big clue that led us to believe impurities were a key problem. When we fixed the arcing, and impurities went down, yield went back up.
Again, in 2015 we installed the new monolithic tungsten electrodes that eliminated the arcing between metal pieces inside the chamber. However, we observed a bright golden substance when we started firing and we got very low yield. The golden substance was tungsten bronze, a compound of tungsten hydrogen and oxygen that was very fragile. It was releasing large amounts of tungsten into the plasma—again resulting in high impurities. When we then reduced the oxygen level by baking out the vacuum chamber, the fusion yield in 2016 tripled to a record high. In addition, when we could not pull the oxygen levels further down in 2017, we were unable to increase yield. The pattern of experiments gives strong evidence that high impurities cause low fusion yield and that thus reducing impurities will increase yield.
2) How much yield increase do we expect initially, if we do decrease impurities? The dense plasma focus device (DPF), for solid physical reasons, has a fusion output that increases sharply with electrical current—approximately as current to the fifth power. In other words, if current goes up by 2, yield goes up by 25 or 32. (Fig. 4). This scaling law, which works for smaller DPF devices, has been interrupted for larger ones—they don’t get the yield expected from the scaling law. We think that is due to the larger impurities that powerful DPFs have produced. So from the results we obtained at lower currents, we anticipate that with low impurities our initial experiments with pure deuterium should get our fusion yield up from about ¼ J—our best result with tungsten electrodes—to over 2 J.3) Why are we confident that in this experiment we will radically reduce heavy impurities?We are eliminating, almost entirely, all heavy metals from the experimental chamber. There are both strong theoretical reasons and abundant experimental evidence that impurities affect plasma characteristics, such as electrical resistivity, in proportion to the product fz2, where f is the fraction (by number) of ions with an atomic charge z. We are switching our electrodes from tungsten, with a z of 74, to beryllium, with a z of 4. This means that, when fully ionized, each beryllium ion in the plasma has 340 times less effect than each tungsten ion. We don’t expect a lot more beryllium ions to be vaporized, because the energy to vaporize and ionize one beryllium ion is already ¾ the energy needed for one tungsten ion. So the contribution of the electrodes to impurities will be hundreds of times less in the new experiment. In addition, we are being extremely careful to ensure that the other materials in the chamber, located where the plasma will be at lower temperatures, but will still be hot, can resist vaporization at the temperatures they will be exposed to. That includes the AFC mentioned above.
Figure 4. From the scaling law of plasma focus yield vs peak current, we expect to be able to increase yield from 0.25 J (top large orange square) to the scaling predictions (black line and large triangles), lifting yield first to 2 J and then 10 J with deuterium fuel. More current and proton-boron fuel will be needed to go still higher.
In addition, reducing impurities to the level we expect, with the product fz2 due to impurities reduced to less than 0.1, will mean we are using plasma as pure as that observed in the sun and other astrophysical objects. These plasmas are mostly hydrogen and helium with only small amounts of carbon and heavier elements. We know from abundant observations of phenomena like solar flares that plasma filaments are the first stage of compression that lead to dense plasmoids. With pure plasma, we can confidently use these abundant astrophysical observations to predict the behavior of our plasmas. Just like the Wright Brothers used observations of birds to develop controlled flight, we use observations of the natural behavior of plasma in the Universe to control fusion.
There are good theoretical reasons to believe that if the compression of the plasma starts at higher densities, because the filaments have produced an initial compression, it will also end at higher densities, leading to greater fusion yield. As early as the 1970’s plasma focus pioneers Dr. Winston Bostick and Dr. Vittorio Nardi published extensive experimental studies showing that the filaments led to the production of the densest plasmoids.
In our previous experiments, heating due to impurities destroys the plasma filaments that are the first stage of compression of the plasma. We observed, through our fast camera ICCD images, back in 2013 that the filaments that existed in smaller DPF devices did not exist in our device at the end of the run-down phase, when compression started. This year, we obtained images showing that the filaments formed early in the shot, but blew up at around 500 ns into the shot. Theoretically, we know that impurities, by increasing the resistivity of the plasma, will make the filaments heat up faster and expand before the magnetic pinch forces can build up to compress them. With low impurities, we confidently expect the filaments to survive until the compression phase, leading to higher density plasmoids. Our new experiments will enable us to definitely confirm or refute the hypothesis that the filaments are needed for the best compression. The results will then guide us to the optimal conditions for fusion yield.
Following our initial experiments with pure deuterium, we will introduce a mixing gas, either nitrogen or neon, to start simulating the mixture of gases that we will have with our ultimate hydrogen-boron fuel. We expect that this mixture will lead to higher fusion temperatures than with pure D, as the heating mechanism involves the viscosity of the plasma, which also increase with atomic charge. These experiments will be a bit trickier to optimize, as too much higher-z mixing gas will cause the filaments to blow up again. So we will need to get to the “Goldilocks” point here. If we can, we expect fusion yields to rise above 10 J.
Our further plans, for the second half of 2019, include upgrading our switches to improve our peak current, again increasing fusion yield. We can now do this without opening up our redesigned vacuum chamber. Then, in the fall or beyond, we will start introducing our experiments with hydrogen-boron, pB11 fuel. Since this fuel burns faster and more energetically than deuterium, that will again boost our fusion yields and put us on the track to our goal of getting more energy out of the device than we put into it—net energy.
