Reactor Plasma Imaging

This area is for ADVANCED theory and discussions only. If you just joined, chances are this is NOT the area for you to be posting.
Post Reply
Posts: 14
Joined: Wed Feb 25, 2015 1:23 am
Real name: Daniel Christensen
Location: Sammamish, Washington

Reactor Plasma Imaging

Post by danielchristensen » Thu Mar 08, 2018 4:23 am

Hello folks,

It’s that time of year again- projects from Northwest Nuclear Labs have reached maturity and being prepared to be presented at the regional and state science fairs. I shared some about my project involving palladium last year and have more from my project this year. My team set out with a fairly bold goal of creating a complete system that is able to create maps of fusor plasmas this year and managed to do such. I will be sharing a quick brief on what has been done so far below.

The inception of our project was centered around an interesting piece of vacuum hardware (pictured below) that allows movement of a stick around the inside of a chamber that is at vacuum. One of my teammates came up with the idea of using a Langmuir probe integrated with that piece of hardware to undergo cross-sectional scans of our reactor’s plasma that could be compiled into maps of sorts that could visually represent properties of the plasma.
Reactor With Wobble Plate.jpg
We decided to develop our own plasma probe, which originally consisted of a fairly simple design. There were a series of tubes inside of one another that served various purposes. The outermost tube was made out of stainless steel and was used for structural purposes. Inside of that resided a double-bore ceramic tube that acted to reduce disturbances in the plasma that are caused by metal in the probe via the insulating properties. Inside of the bores in the ceramic were two tungsten wires that would act as electrodes. These layers were bonded together with epoxy to create a vacuum seal.

In testing our first prototype, we found out that the probe would gain a charge of a few hundred volts when it was in contact with the probe. We determined that in the time frame that we had to do our research for the science fairs, we would develop a more simple plasma probe which was not a true Langmuir probe but would get the job done. We pursued the creation of a voltage potential probe, which could be made with the materials we had on hand and completed in a favorable time frame. This variety of probe does not provide quantitative results about plasma density but would supply us relative information that we could use to create a map of the plasma.

In our second prototype the mechanical design remained similar, simply missing one of the two electrodes. To turn the new probe into a usable tool we designed, tested, and built a high impedance voltage divider and safety network with a ratio of 100/1 and an input impedance of 20 megaohms. The high impedance was necessary to avoid “loading down” the plasma.

To provide additional safety beyond the inherent measures in the reduction ratio, we implemented a stage fail-safe crowbar protection network made from a variety of gas discharge tubes (GDTs). When the input voltage goes too high, the gas discharge tube starts to conduct. This shunts the current from the probe input to ground, thus reducing the input voltage to the sub 50v range until the fault is cleared. If the main protection GDTs and voltage divider fail, there are additional GDTs on the output network. They are set at a lower voltage and are equipped with a thermal failsafe short device that will activate if a fault condition develops for sustained lengths of time. The entire circuit was tested with a 20 kVDC 10mA power supply and was found to effectively protect the operator and measurement device. The combination of these components enables the voltage potential of the plasma to be effectively measured while causing a minimal disturbance.

To reduce noise and increase safety, we attached the divider network directly to the end of the probe. The circuit itself was encased in a PVC housing filled with a thermally conductive potting compound that was rated for high voltage use. The entire circuit was grounded to the reactor ground with a copper braid to allow for an extremely low impedance ground.

The information we gained from this probe would be useless without a position monitoring system, which we also developed. The medium with which we developed this design was Lego technic, which allowed for rapid prototyping out of materials we already owned that had relatively good structural properties. To measure position, we decided on implementing an architecture that used two angular position sensors connected to mechanical linkages that through their geometric design enable position sensing by means of simple potentiometers. From there, various modular segments were constructed. We first constructed a module (pictured below) that addressed the need for gyroscopic motion of the pole. The module has the ability to swivel between mechanical linkages, hold the probe, and move freely enough that the probe’s motion is not inhibited. Another assembly (also pictured below) was constructed that is responsible for measuring the rotation of the linkages, which was conveniently based off of a gear. This allowed us to offset the position sensors from the part that is actually rotating through the use of another gear, reducing mechanical stress on the sensor and, in theory, lessening the chance of the sensor breaking.
Technic Gimbal 6.png
Gear Assembly.png
This design worked fairly well but had several inadequacies that had to be addressed. The center gimbal and mechanical linkages in the first prototype were found to allow the probe to give inaccurate position readings in situations where the probe was rotated. This was corrected by attaching the gimbal to the probe with tape and allowing the mechanical linkages to move more freely in multiple axes while remaining attached to the probe. Furthermore, we found backlash in the gear system to be a source of error and decided to reduce the backlash caused by the gears by eliminating them all together, driving the position sensors through the
movement of the linkage interface alone. These modules were combined into a single structure (pictured below) and attached to the reactor (also pictured below).
Rxr Mount 2 Complete.PNG
Reactor Technic 7.jpg
We integrated both of the systems onto our fusor and took data on a digital oscilloscope, 20 scans lasting 2 minutes each which ended up amounting to 72,000 data points. This data was not usable and needed some processing, which happened as follows. We first had to find the total range of the position reading from the two potentiometers, which was followed by calculations that found the center of the ranges from both potentiometers. We then used trigonometry to determine the total range of the potentiometers in degrees. The apparatus was not completely vertical, and therefore we had to determine the position of the potentiometers, which was done by simply rotating the coordinates around the origin by -35 degrees. Based off of this information, we were able to determine the slope of lines originating from the potentiometer to the point (0,0). We also had to determine a multiplier that would result in a circle with a radius of 1.

By multiplying the center of range by the total range multiplier and subsequently subtracting the slope to (0,0), we were able to account the offset in the center of range. This result was then used to correct the degree reading from the potentiometer to the true value. We took the tangent of this degree reading and found the slope of both potentiometers. These slopes were then fed into a system of equations that found the intersection of the two lines, giving us an x and y position. Our data and associated graph did not start out looking as planned. There were complications with the data that resulted in the graph appearing off. To fix this, we determined if the position data was outside of the possible range and deleted all data that fell out the range.

Following this processing, we plugged our information into a software called TeraPlot and created a really lovely cross-sectional image of our plasma (pictured below). The purple represents the densest areas and the red represents the least dense.
I would love to hear feedback and answer any questions that you may have.
Daniel Christensen

User avatar
Richard Hull
Posts: 11106
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

Re: Reactor Plasma Imaging

Post by Richard Hull » Thu Mar 08, 2018 6:42 am

Fabulous effort and report! In spite of a probe's ability to interfere with a complex plasma in a significant manner, you managed to show and, I think, prove that the greatest density of the plasma is at, more of less the center of the fusor. This is something that might be derived from common sense based on the physics of a regular shaped enclosure of same. The regions of greatest fusion however are known to reside elsewhere in the fusor, due to past research by the UofW. The latter discovery is less obvious at first application of common sense.

Your recent efforts have to command a win or prize for such an effort on a shoe string budget where noodling out a significantly successful measurement of plasma density has succeeded so brilliantly.

Again, another great effort by the group.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.

Post Reply