April 12 2014 demo run vid

For Farnsworth type designs.

April 12 2014 demo run vid

Postby Doug Coulter » Sat Apr 12, 2014 1:13 pm

Nothing special to those already in the know. I only hit around 2-3 million neutrons (twice that number of fusions) per second here, I was just trying to show the unitiated what we do with this. I limited the need for noisy fans by keeping the main power current limit lower than usual. As I type, there are 5 hours to go on the upload (just a little over 10 min of video), but I'm putting the link here now so I don't forget and lose it. More words later - a few times I was saying something important that got drowned out by the neutron detector audio. I'll have to watch again to fill that in here.
http://youtu.be/ZGiU_Ck192Q

During one of the times I got drowned out, what I was trying to explain is that our secondary grid, out in the larger tank, could light due to Paschen's law allowing a discharge to start at a lower voltage due to the higher distance. This is counter-intuitive at first, because most people have been taught that electricity always takes the shortest/straigtest path. Well, it doesn't always - seen a straight lightening bolt?
Here's a picture of the Paschen curve:
751px-Paschen_Curves.PNG
The law - no point trying to fight it, either.


From here, which explains why this is as well as I could myself. It boils down to electrons (there are always a few free from things like cosmic rays) either gaining enough energy from the field between collisions to ionize (knock an electron off a gas atom) or not. The X axis is therefore PxD (pressure times distance) and we are working on the left, steep side of this curve. Thus, even when 50kv can't "light off" over a 2.5" distance (outside of main grid to inside of 6" sidearm), it can light off over a ~14" distance at a lower voltage (5-10kv in our case) from our ion generator grid. For those into the actual numbers, we go out at around 2.2e-2 millibar, which is actually about half that if I believe the documentation on the PKR-251 pressure gage we use - it reads about factor of two high on hydrogen, according to them. For those really into numbers, there are 760 torr for one bar (atmosphere) so the conversion from millibar to torr is to multiply millibars by 0.76 to get torr. Going the other way, by 1.315789474.

Actually, for giggles (and learning) I just calculated. Our D is 6.35 cm. If I use that factor of two high, I can't make this work at all - looks like the Pfeiffer dox are just wrong when it comes to D, and the gage is reading right at 2.2e-2 mbar - then the Paschen's law calculation comes out as pxd = 0.106172, nope, oops, that's still off by factor 10 or so at the volts we have. Could it be some of the main grid field reaches out into the larger tank to get more D? Most good science doesn't come from "eureka" moments, it comes from "that's funny" types of things, this is obviously something I have to look into some more...We ARE pretty sure we don't have an even field distribution most of the time, certainly not when it's lit. But when it isn't - the place where Paschen's law works - we should have a fairly uniform one circumferentially, we work hard on making that true. And those lines are going pretty much straight up where the wiki chart stops bothering to plot...Could it be that despite our efforts to have real purity (our vacuum system base pressue hovers around 1e-8 mbar) we have a Penning effect? Now that one passes the R Feynman test - I can take it to the lab and test that prediction easily. And in fact, when we take the tank to STP so I can stick my hand in there to do something, guess what we use? Argon, dry nitrogen, or helium, depending on what's handy nearby at the time.
Right now, argon is what's hooked to the vent valve...

Why is the curve is U shaped? At high pressures, it takes a lot of field for an electron to gain much energy before it hits something, so it takes high voltages to "light off" a discharge. At very low pressures, an electron might go all the way across a tank and not hit anything but the other tank wall. Make sense? It also turns out that if electrons are going super-fast the chances of ionizing an atom they graze are less - they're just not around long enough to have much effect, which is partly why the left side of this curve is steeper than the right side.

We are constantly trying to run in lower pressure gas. Why? Actually it's the same idea. There's a concept called "mean free path" which is how far the average thing can move without colliding with something else. The lower the pressure, the longer the mean free path. One of those "obvious in hindsight" sort of things.

During the acceleration toward the grid, we don't want collisions - they reduce the energy we're trying to put into the ions, and deflect them. Some times there's even a charge exchange, so the old ion becomes neutral, and the new atom becomes an ion that's already partway down the field and therefore can't be accelerated to the full field voltage. The reality is a bit more complex - the electrons and counter moving ions mean the field we "think" we put on there isn't the field the particles themselves see - they make their own fields because they are charged. This is one of the current mysteries of the fusor, one we are trying to better understand, as it appears that even with a 50kv field, we're only getting about 1/10th of that onto our ions...and fusion is going up fast as we increase the voltage, as you might have noticed in the video. Any extra we have to put in is waste - if we could get to this level with 5kv instead of 50 - that would be a 10 times gain in efficiency all by itself.

So, in this beam collision device (the IECF moniker was a result of a huge misunderstanding by Farnsworth/Hirsch/Meeks but it stuck), we want no collisions except at the focus. We know we will never get there, as we have a "compression ratio" as things converge on the focus, but the hope is that by that time collisions don't matter as much - things are already going where we want at the speed we want, and it will then be hard for them to "miss" another ion and cause fusion.

One out of two DD fusions makes a neutron. That's why when we see ~2 million neutrons/second, we know that there were roughly twice as many actual fusions.
The reactions are D+D -> 3He + n + energy (mostly kinetic in the daughter particles), and D+D -> 3H + p + energy. Either of these (3He or 3H) can continue to be reacted in the fusor, but there aren't many at current production levels so it's on the rare side. We did once manage to measure the T coming out, after a 20 minute run, barely, at around 8x the background count from cosmic rays in a "heroic effort" detector. This could change as things get better, but for now, it's not a worry.

The energy comes from this - in D the proton and neutron are very loosely bound by atom standards. In the products, the binding is very much tighter. We get the difference out as energy, just as in letting a rubber band collapse. Nothing actually changes in identity here - protons all stay protons and so forth.

Actually, it's all D+D ->4He, but that would try to release ~17 MeV, and the He then breaks up into those other things, with less net release, on the order of 3.5 MeV. The reason appears to be various conservation laws we've known about since the '30s or thereabouts. Spin has to be conserved, for example, and the Pauli exclusion principle applies, among others. There seems to be a selection rule that says the He can't just give off a gamma ray (photon) since for fusion to happen at all, the D's have to have opposite and cancelling spins, and He has no spin. Photons do. It seems that it's quite rare to produce a pair of photons each at half the energy (to conserve momentum, another law we have to follow), like in e+/e- annihilation.

Other fuels will give different results. The main thing here is that the energy required to push the ions close enough for long enough for quantum tunneling into fusion is less with less charge, and everything else has more than hydrogen.
We are using DD instead of DT (t being tritium, very heavy hydrogen) for a rather large number of reasons. Deuterium is itself hard enough to get - it's a controlled substance for whatever reason (a stupid one, I'd guess, since DEA "owns" it). T is much more rare, and a VERY controlled substance. It's radioactive, bioactive, short half life (12.5 years more or less) and the DT reaction gives off...14+ MeV neutrons, which is enough to knock iron atoms out of their lattice and make dust out of the fusor, as ITER finally figured out and stopped to "solve" materials problems that are not soluble with any substance known (we do know the entire periodic table...). The big advantages of DT are this - around 16 Mev/fusion - over 4x what we get in DD fusion, and around 100 times the "cross section" for tunneling into fusion, since there are more ways to satisfy all those conservation laws given above in DT than there are in DD fusion. So a lot of fusion researchers, for good or ill, think "I only have to get within an order of magnitude of 2, and can switch to DT and be "there"". When in fact, those super high energy particles that come out of DT have their own very-very nasty set of issues - like taking your stuff apart.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: April 12 2014 demo run vid

Postby Doug Coulter » Sat Apr 12, 2014 4:07 pm

Why do we use silver activation? Because electronic neutron detectors can be fooled by a number of things - mainly EMI (radio noise) caused by other stuff happening nearby.
Or even dumber stuff - in the video above, my main logging detector failed as I kicked the plug out of the wall while filming. Life!

There is no way to "cheat" using silver - the only thing that will make it radioactive is neutrons absorbed by it. Why do we need a moderator? Well, the "cross section" of silver for fast neutrons (ours are around 2.4 MeV - pretty fast) is a lot smaller than for room temperature neutrons, which translates to around .025 eV. Actually there's a big absorbtion peak just above room temp, and we calculated our moderator first stage to try and hit that, thanks to Carl Willis' help. So, we force our fast ones to bounce around off H atoms (which weigh about the same as neutrons so it takes less bouncing to get the same slowing) to slow them down first, then let them hit the silver, making things more sensitive.

We can "get away" with using short half-life silver vs other harder to activate things, like indium (used by Fermi in the first nuclear pile work) or gold, both of which also have a longer half life, since we have a way, demo'd in the video above to know just what it would have been had we been able to measure the activity "instantly". This is possible because we have time on the X axis of the geiger plot, but the radiation counts are logged.

Since radioactive decay is exponential, a decay will produce a straight line on a semi-log plot like this, and it's trivial to interpolate back to what it would have been the instant the fusor shut off, so the short half life of silver doesn't hurt us or make things difficult. It's actually a big advantage as I can use the same piece over and over since it "cools down" pretty quick!
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA


Return to Fusors, Farnsworth type

Who is online

Users browsing this forum: No registered users and 2 guests

cron