Polywell Fusion Design

For posts specifically relating to fusor design, construction, and operation.
Post Reply
Jeremy Adams
Posts: 26
Joined: Mon Aug 18, 2014 9:08 pm
Real name: Jeremy Adams
Location: Chicago Ill

Polywell Fusion Design

Post by Jeremy Adams »

Hi everyone,
I would like to know what you guys think of my fusion reactor plans. Here's what I got:

My next reactor is going to be a polywell fusion reactor. I'm very interested in the design and my friend and I who are building it are going to learn lots from it. Hopefully it won't be much more expensive than a regular Farnsworth-Hirsh reactor that fuses particles. We have already built a demo fusion reactor and understand a polywell is very complicated.
First the simple things: we're going to upgrade our chamber from our small glass and aluminum chamber in our demo model- the new one’s going to have less leaks and a viewport etc. We're getting a vacuum that's better, a used molecular turbo pump that produces an ultimate vacuum of 5x10^-8 torr. Because it is a very strong vacuum we'll use a shut off valve to prevent our deuterium from being taken out of the chamber. We already have compressed deuterium gas and will use it with a regulator and ionize it before it goes into the chamber.

Our Polywell electrode is going to be machined out of stainless steel or brass and we're getting this job done for almost free. Each ring is probably going to be machined in two separate parts that'll be welded together. The rings will be welded together as well. Each ring has the middle part "scooped out" (we have dimensions set for this part). Each ring has a cross sectional width and length of .25" and a radius of .5" (not including the cross sectional length). One corner of each ring is cut in half to make a triangle so the rings can be connected to make the polywell. I attached some pictures from a 3D modeling software. The electrode is small because our chamber most likely will not be large. The scooped out part of the electrode will be space for thin copper magnet wire to be wrapped around (wire with an acrylic coating) and produce magnetic fields to trap electrons in the center of the electrode. The electrode will have to be very symmetrical which creates some technical challenge. The reason 3D printing did not appeal was because plastics do poorly under a vacuum. Using polycarbonate would be an option but this would need to be cooked for a few hours to expel the hydrogen from the plastic and then put under a vacuum for many more hours. Ultimately using a plastic would be challenging under a vacuum and other 3D printed materials such as ceramic and some metals would do better but are expensive and hard to gain access to. We decided not to put space between the rings because this would probably hinder the symmetrical design and require more welding to be done, making the polywell less accurate. We also don’t know how to figure out the right measurements for the space between the rings and since our rings are not circular efficiency will be lost in recirculation attempts anyways. Therefore we are not trying to recirculate any electrons in our polywell, and we are building a model similar to the first WB designs.

For our polywell electrode we need 6 power sources, one for each ring, of high current and low voltage. Our current is going to be around 21 Amps and our voltage will be 1.6v. To get 6 separate outputs we're building two identical circuits. Both circuits will have a "modern" NST transformer outputting 6.5kv AC. This will be sent into a large step-down transformer about 1.8ft long. The wire will be thin copper magnet wire to allow the transformer to work properly (we have planned other dimensions for our transformer- Np = 4332, Ns = 2; Diameter of P and N winding column is 1”). The secondary winding of the transformer should output a current of approx. 65Amps and 3volts. This output is going to be full wave rectified with four diodes in a diamond shape. We'll have to get the right diodes and capacitor for our current & voltage levels. Since p-n junction diodes take .7 volts from a circuit and two will be crossed for both positive and negative outputs we will lose approx. 1.4v making our output voltage 1.6v (depending on the diodes we may have to change some aspects of the step down). The rectified output will be smoothed with a capacitor and then sent to 3 parallel circuits in which the current will be divided, in ideal conditions, equally into 3 outputs. Our circuit will use 4 types of wire. Thick tungsten, or 4 AWG wire will be used for the high current sections- the thickness will make it less resistive/durable and the tungsten will handle the heat from the high current. Two copper magnet wire types will be used- both single stranded but different thicknesses for the step down transformer and Polywell electrode. Pure copper wire with a larger diameter than the motor wire will be used in the parallel circuits to reduce resistance. We calculated the length of each wire type needed to create the right amount of resistance based of the resistivity of metals using R=p(L)/A and V=IR. We're certainly going to solder our circuits together to make good connections and use lots of insulation for safety.

To get electrons into our reactor we're using a simple tungsten filament producing thermionic emission from an outlet power source. We’ll also have a fuse just for safety.

We're going to need to figure out how to build an ion gun to ionize the deuterium since it won't naturally ionize like it would in a fusor with a voltage drop. I read the FAQ's but I didn't see much aside from the general theory on its purpose. I've read some on different design schemes, but we haven't planned much of our ion gun yet. Obviously it needs to produce positively charged deuterium ions which are attracted toward the electron well. We want to do something similar to a single wave device that could release large charges in a single moment and is not pulsed like other injectors. When we'd run our reactor we'd turn on the vacuum and get the electron trapping going, then send in the deuterons. We don't want to make this too complicated- not using any radio frequency etc. Any suggestions? Would using a high voltage electrical arc under a vacuum be a good way to ionize the deuterium gas? Should we direct the ions into the chamber with magnetic fields?

We'll figure out neutron detection equipment when we get there... And maybe we'll use a voltage bias probe to detect electron trapping too – just like Langmuir probe diagnostics. And taking it a step further we could build a simple venetian blind electrostatic direct converter and run the power gathered from the converter back to our ion gun or something… but if we did that’d we’d have to direct ions/electrons from our reactor toward the converter- we could change the current direction thus changing the magnetic field direction of one polywell ring while running the reactor and this would act similar to linear reactor by pushing the plasma toward the converter- what a cool thing that’d be!
Attachments
Part of our circuit design
Part of our circuit design
A 3D printed polywell design with our dimensions
A 3D printed polywell design with our dimensions
Polywell3.PNG
Polywell2.PNG
Polywell1.PNG
User avatar
Dennis P Brown
Posts: 3159
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: Polywell Fusion Design

Post by Dennis P Brown »

A deuterium gun can be simple or complex. They are shown in many posts including my gun for my accelerator. A cheap and small (battery operated 12 volt DC) 3 kV power supply can be had to ionize and accelerate the gas coming out of a small copper capillary tube. A trivially simple Einzel lens can then focus the ionized gas. See the photo for my deuterium gun (no power supply. Note deuterium gas connector and high voltage wire feed through on the left of the gun assembly.
Attachments
Picture of Deuterium ion gun in my accelerator tube. The gun assembly are the two metal parts in the center region of the glass tube and Einzel lens is an aluminum tube on the right
Picture of Deuterium ion gun in my accelerator tube. The gun assembly are the two metal parts in the center region of the glass tube and Einzel lens is an aluminum tube on the right
Post Reply

Return to “Fusor Construction & Operation (& FAQs)”