LPPFusion’s research team started our long-awaited new set of experiments August 4th, successfully firing FF-2B, our fusion experimental device, with its 16 new switches and newly-redesigned anode. While we are still adjusting and optimizing the new switches, we have already demonstrated a 6-fold decrease in erosion from the anode and an 8% increase in peak electrical current.
Figure 1. Our new polished anode (left) still shines mirror-bright after three shots. It is viewed through a window on our vacuum chamber. The anode is a bit more than two inches in diameter. The colors on the anode seen here are the result of lighting and camera response. The true color of the anode is still silver. In contrast, our first beryllium anode after one shot in 2019 (right) was covered with a dark ”snow” of beryllium oxide dust, which was vaporized and redeposited by FF-2B’s powerful electric currents. This is what we avoided by our new polishing procedure.
We devoted July to the tough task of reinforcing the insulation on the new switches so they could hold off 40 kV during the few seconds it takes us to fire the device, releasing the current into the central electrodes. The new switches are half the size of the old ones, (and twice as numerous) which enables them to carry more current. But the smaller size increases the electric fields, which makes arcing easier. After some trial and error, Research Scientist Syed Hassan devised a multi-layer defense in depth, using a combination of insulators to seal the metal parts on the bottom of the switch from contact with air. It is the contact of metal with air that leads to ionization – the stripping off of electrons – and in turn to arcing.
Once the switches were arc-proofed, we started firing the whole device on August 4. Each “shot” consists of charging a bank of energy-storing capacitors to 40 kV and then “firing” them in a microsecond-long surge of million-amp current. These initial set of shots are aimed at optimizing the performance of the new switches and getting them to fire synchronously and with minimum oscillations in current. We immediately noticed a major improvement over our first shots in 2019 with the first beryllium anode: greatly reduced erosion. Back in 2019, a thin oxide layer had vaporized, covering our electrodes with dark dust that interfered with functioning and took many shots to burn off. For this upgrade, we had hand-polished the anode – our central electrode – to remove the thin layer of oxides on the beryllium. The polishing worked well: after the first few shots, the anode remained mirror-bright. (Figure 1).
In subsequent shots – we have now fired 29 – erosion continued very low. While the anode has lost some of its mirror sheen, and there is roughening on the surface of its inner hole, measurements indicate that erosion is at least six time less than the best achieved with the old anode. We can estimate erosion by measuring the amount of material deposited on the vacuum chamber windows. Our spectrometer can do this with high sensitivity, since even a very thin layer of beryllium reduces short-wavelength (blue) light more than long wavelength (red) light. By this method we measured only about 10 nm (billionths of a meter) of beryllium is eroding from the inner anode with each shot. If erosion was the only concern, an anode would last 100,000 shots – a lot longer than we intend to experiment with this anode! Equally important, low erosion allows us to get high repeatability in our experiments, since the anode itself is not changing significantly.
Our first ICCD image, a photo taken with a 5ns exposure time, showed – unsurprisingly – a similar disruption of filaments as in 2020. We don’t expect this to improve until we optimize the functioning of the switches, among other steps we intend to take.
By August 13, we had achieved our first fusion shot – one with high enough temperature to produce fusion reactions. We also took a good image of the pinch region on film. (Fig.2) The image shows the relatively large (1 mm radius) plasmoid formed by the poorly-coordinated filaments. As we get better performance, we expect that plasmoid size will shrink dramatically. After that, we took a well-earned break for two weeks.
September got off to a rocky start with minor flooding of the lab by Hurricane Ida (a non-functional sump pump was at fault). No damage was done and the lab is on high-enough ground to avoid inundation by nearby Ambrose Brook (Figure 3). Within 2 weeks we resumed firing. We rapidly demonstrated that the new spark plugs with ceramic insulators were superior to the old ones with Lexan insulators. Importantly, we also showed that, as we adjusted the switches, we were able to increase the peak current by about 8%. This is not yet the full 20 % increase that we expect from the new switches, but is a significant advance. We were also able to rapidly optimize the switch gas mixture in the switches and the voltage of the trigger spark that fires the switches.
However, adjustment of the switches, still continuing, has been slowed by a number of factors. We found that too small O-rings (which seal with switches) allowed the spark plugs to move too much, so the rings were replaced. Resistors in the trigger headsalso had to be replaced to allow faster firing. In addition, significant time was consumed on seemingly small, but important tasks like improving the mounting of the doors on our oscilloscope racks. The doors are vital in sealing out electromagnetic pulses that often prevent our ICCD camera from working. The new mounting has reduced the noise entering the racks by a factor of 10. This now allows the ICCD to work on most shots, a big step for our most expensive instrument—one that gives us invaluable insights.
Details matter! This newly installed hinge on the doors of our oscilloscope racks helps to seal out intense radio waves from our FF-2B fusion device. These waves can interfere with the functioning of our instruments, like our ultra-fast ICCD camera. By connecting to the copper mesh on the doors and to the copper “fingers” on each side of the hinge, the new mounting forms a continuous electrical shield, protecting the instrument controls inside the rack.
The biggest hold-up has been the slow delivery of critical new parts. To observe the firing of all 16 light detectors, we planned to delay one signal for each pair of switches by sending it though a 100-meter coil of optical fiber. The delayed signal and a direct signal from the other switch in the pair would then be sent over a single optical fiber to our oscilloscopes, allowing us to easily distinguish when each switch fired. Unfortunately, no US supplier could sell us the correct Y-connector to join two fibers into one. The Chinese-supplied connectors, delayed by global shipping shortages, just arrived October 12. In the meantime, we could only observe 8 switches at a time, considerably slowing adjustment. We have now just started using the new connectors to monitor all 16 switches. Completing the adjustment of the switches will allow us to start increasing fusion yield.