LPP looks back on 2010, ahead to fusion breakthrough in 2011February 5, 2011
An overview of accomplishments, challenges, and goals for LPPhysics:
Progress over the past year—Where are we and where are we going?
You can find additional background with this report at the Focus Fusion Society.
In this web write-up, we will be taking a longer view than our monthly reports, looking back over the entire past year’s work: how the project is progressing towards its goals, what our challenges and questions are, what we have learned, and what we still don’t know. The report is divided into five sections to cover scientific results and comparison with theory, the FF-1 device itself, instruments, simulation, and personnel—the people who make everything else happen.
1. Scientific Results and Theory
We started the experiment in 2009 with five scientific goals and three goals for the device itself. We will deal with the device in the next section, but in the past year, we have partially achieved two of the five scientific goals. With these goals, we aimed to confirm and duplicate the high ion energies and high densities first achieved in our Texas A&M experiments, while at the same time greatly increasing the efficiency of energy transfer into the plasmoid. We also aimed at testing our general theory of DPF functioning and, in particular, to test the theory that a small axial field will greatly increase energy transfer into the plasmoid.
Since the beginning of 2010, we have achieved much of these two goals. We showed repeatably that the DPF can confine ion energies well beyond 100 keV into tiny regions about 100 microns in radius and 1 mm in length. These are the energies we need to initiate p-B11 reactions, and we demonstrated that this tiny plasmoid lasts for tens of ns, more than enough time to burn p-B11. We have also proved that the fusion reactions we observe come mainly from the hot ions trapped in the plasmoid, not primarily from the powerful ion beams that the plasmoids emit, thus helping to resolve a long-standing debate in the field.
In addition, as detailed in the January 2011 monthly report, we have demonstrated by independent sets of data that we have vastly increased the efficiency of energy transfer into the plasmoid, and are now putting over 10% of the total bank energy into the plasmoid and getting about the same energy out in the beams. This compares with our estimates of only 0.01% or less energy transfer in the Texas machine. In this area, we have made great strides and are in the range of where we need to be for demonstrating the scientific feasibility of focus fusion.
As a result of our high ion energies and high efficiency, we have achieved fusion yields that are well above (by a factor of at least four) the trend line of other DPFs’ scaling with energy and current. The beam power we have observed (around 100GW) may well be enough to generate over 100 J of hard X-ray energy when the electron beam hits the anode. We still need to test this by further observations, as we don’t know right now how much electron beam energy escapes the plasmoid. If we can generate this much X-ray power, we will have the basic technical achievement needed for our X-Scan spin-off non-destructive inspection technology.
Theoretically, we have confirmed the basic outlines and a good many details of our theory of DPF functioning. We have shown that the plasmoids form by a process of kinking the current filament, and that they are indeed tens of microns in radius, not many millimeters as other researchers have thought on the basis of lower-resolution data.
The first concrete result of this progress was the publication of our first peer-reviewed paper in the Journal of Fusion Energy, which appeared online January 28th, 2011. This publication will add greatly to our credibility and will hopefully be followed with a series of such papers in ever more widely circulated journals.
Challenges—The Early Beam
These are solid accomplishments that put us closer to our ultimate goal of fusion energy. However, we took far longer than we expected to achieve these goals—a full year rather than the few months we thought it would take. This is largely due to the problem with the switches summarized in the next section, but it puts us a year behind our original schedule. (It is important to note that our earliest estimate for the total duration of the current experiment was three years, not two. By that earlier, and in hindsight, more realistic estimate, we are about on schedule.)
In addition, our goals are not fully achieved. We were able to achieve the fusion energy yields that we predicted only up to about 700 kA, but not up to our highest currents of over 1 MA. We have good evidence that this is because we have not been able to match the density of over 1021ions/cc that was achieved in the best shots in Texas and that we expected to match and exceed with FF-1. Instead we are about a factor of 10 below that, reducing our fusion yields by a similar amount.
In the course of our year’s work, we have discovered the probable explanation for this problem with density and yield, and therefore how to avoid it, but we do not yet have all the details we need. Our theory, like any theory, was incomplete and what was left out is a three-step process that occurs as the plasmoid contracts. This process produces three pulses of X-rays and a single neutron pulse. In our best shots, where the yield is as high as predicted, the third X-ray pulse coincides with the neutron pulse created by the fusion reactions. But in shots that fall short, the first one or two X-ray pulses are accompanied by an early production of the main ion and electron beams. These beams drain the plasmoid of energy before it has time to contract to the high density needed for the fastest fusion energy production. (We had previously referred to this problem as the pre-shock, but we now believe that there is no shock, but rather an early production of a beam.)
We don’t yet fully understand what is going on with the three steps, but we do know that only in the shots where the current is still rising fairly rapidly when the pinch begins, what we call short-pinch-time shots (or SPTs), is the early beam avoided and maximum fusion yield achieved. We have some, but no conclusive, evidence that these SPTs require the correct injection of angular momentum from the axial field coil (AFC). As a result of this phenomenon, we have not yet been able to get conclusive evidence that it is indeed because of the AFC that we are achieving the high efficiency of energy transfer we have sought.
Our next key goal is to thoroughly explore and understand this three-step process and the early beam that it produces so we know how to concentrate the beam production at the end of the process, when the plasmoid is densest. At the same time, we are going to shift our PMT detector to directly study the electron beam coming out of the plasmoid.
Once we have completed our shift to ruggedized switches, as described in the next section, we will be able to go to full bank power and higher current. This higher current will allow us to study the plasmoids over a greater range of conditions and more quickly achieve the optimal conditions for full fusion energy yield.
In terms of yield, achieving a factor of 10 increase in density will bring fusion yield up to around 1 J, which would be a record for any fusion device using pure deuterium with our input energy of only 50 kJ. With full power output at 45 kV, we would then expect to get fusion yield up to the area of 5-10 J. Further increases will occur when we go to shorter electrodes, higher densities, and heavier gas mixes, with still further increases when we transition to running with p-B11 later this year.
2. The Device
Obviously 2010 was the “Year of the Switch” at LPP. The switch problems led to very long delays in our project. Repeatedly we felt that we had the switches under control, only to be confronted with another failure, yet we have been able to make considerable progress. In September, we showed that increasing the distance to the adjacent switch electrode increased trigger voltage and led to simultaneous firing. At the end of December, a thorough redesign of the spark plug led to stable firing without pre-firing and breakage at 33 kV.
However, we are still not through with the switches. The tungsten rods are still eroding and have to be sanded routinely to avoid pre-firing. When we do have an occasional pre-fire, we can rapidly sand and eliminate the problem, but such continual maintenance is highly undesirable. In addition, while the insulator breakage rate has dropped sufficiently to put the lifetime of individual spark plugs in the area of hundreds of shots, there is still the need for occasional replacement. As explained in the December report, we are now in the final stage of a complete ruggedization re-design that should greatly reduce or eliminate the maintenance requirements of the switches. In addition, it will allow us to move up to full 45 kV operations. We expect to manufacture and test one example of the new switch in February and have full operation with 12 switches in March.
Unfortunately, we have had to undergo a crash course on switch engineering. We really did not have enough hard data on switch functioning until recently to confidently re-design the switch. Ideally, if we had received additional funding, we would have been able to start a parallel research project with at least one electrical engineer focused on the switch problem alone, but this was not possible with our current staff.
Having unreliable switches (until very recently) greatly restricted our ability to test the axial field coil under repeatable conditions. We have just now begun to do this.
High Current in the Switches
Early last year, we were able to achieve the second of our major goals for the device, to function at 1 MA. We now routinely get this much peak current. In fact, at 33 keV, we are now getting 1.1-1.2 MA with shots that pinch, which are the vast majority of our shots. We are getting somewhat less current and a longer rise time than we expected in our design, but we understand why this is the case. We underestimated the inductance of the switches, leading to an overall reduction of current by about 25%. With our new understanding, we expect the peak current into FF-1 to be around 2.1 MA. This is above our minimum goal of 2 MA, but below the ideal design goal of 2.8 MA. At the moment, we don’t think that this will seriously impede our demonstration of scientific feasibility. It should reduce the fusion yield ratio (ratio of fusion energy yield to total energy input) by a factor of around 2. If we are still scaling upwards at that point, the achievability of net energy will still be demonstrated.
Higher current can be obtained with a device voltage higher than 45 kV. This could be achieved by putting a second set of 12 capacitors in series with the present one, lifting the maximum voltage to 90 kV. But it is not clear that this would be possible with the present device, although it certainly would be with a follow-on device, FF-2.
Elimination of Helium as a mixing gas
While we had originally planned to use both helium and nitrogen as heavier mixing gases to examine the effect of larger atomic masses, we are now going to proceed only with nitrogen. Dr. Subramanian pointed out that if we intend to use helium production as proof of p-B11 fusion, helium adsorbed by the vacuum chamber could be released and confuse our results. Indeed, we found that the amounts of helium trapped in this manner could well exceed what we expect to produce. To avoid this, we will not use helium. (Helium that was used in a few very early shots was out-gassed when the chamber was heated to high temperatures to anneal away its magnetization.)
At the beginning of 2010, we were observing our shots with just two instruments—the Main Rogowski Coil that measured the current, and the Far Time-of-Flight PMT which measured the neutrons and X-rays. As of now, we have 15 instruments: a silver activation detector, five bubble detectors, two Time-of-Flight PMTs looking at the neutrons, 3 X-ray PMTs, the upper and lower Rogowski coils studying the ion beam, the ICCD camera taking excellent high-resolution photographs, and a high voltage probe measuring the device voltage. In addition, we have up to 12 optical detectors monitoring the firing of the switches. We also have an X-ray pinhole camera installed, but it has not returned images yet as we don’t have enough X-rays. Two instrument remain to be installed, the X-ray spectrometer and an X-ray lens. Both of these instruments are intended to be used when we are producing more X-rays after we change to p-B11 fuel.
The present large set of instruments, perhaps the largest used to study a single DPF, has been the key to our ability to accurately measure and to understand the plasma produced in our shots.
We do not at present have the resources to have full time work on the simulation of the DPF that we are developing. However, with the help of LPP contractor Dr. John Guillory and FFS volunteers Henning Burdack and Warwick Dumas, we are approaching the point where a 1-D simulation of the formation of the filaments will be giving us useful information. We will also be working on ways to model the kinking of the central filament into the plasmoid.
At the moment, excluding the small simulation team, the project has the equivalent of three full-time technical personnel. Eric Lerner (the principal investigator) and Krupakar Murali Subramanian (the Senior Research Scientist) are both employed full time. Fred Van Roessel is our part-time electrical engineer, and Derek Shannon works part-time as a research assistant, and is also part-time as LPP’s Director of Business Development. This is in addition to Aaron Blake, our new full-time Chief Financial Officer, who will now have primary responsibly for conducting the capital drive.
The staff of three full-time equivalent technical personnel has proven to be just too small for optimal functioning of the project. Right now, the pace of progress is limited by the time available to the staff, no matter how productively we work. Mr. Lerner is wholly responsible for theory and data analysis, and has the primary role in producing reports and the published papers that are essential to LPP’s growing credibility, as well as directing the gathering of experimental data. Dr. Subramanian has had the primary role in the on-going re-design of the device, participates actively in the experimental shots, and has been very active in the maintenance of the instruments. Mr. Van Roessel has taken over a leading role in the development and maintence of the instruments, especially the ICCD, and is active in the work on the device itself. Mr. Shannon helps out wherever needed. The fact that all of us have multiple responsibilities means that, for example, we must frequently stop firing the machine to catch up on data analysis, or urgent redesigns, or writing papers.
To make maximum use of our facility, we ideally need a full-time technical staff of five. We need to hire an additional theoretically-trained plasma physicist to focus exclusively on working with Mr. Lerner on data analysis, and we need to hire an experienced electrical engineer to focus on the upgrading of the device and instrumentation. This would allow us to take data almost continuously and would almost double the amount of data we collect per month, which in turn would almost double our overall rate of progress.