Here be Fusion – A Visit to General Fusion | Hacker News Reader

Here be Fusion – A Visit to General Fusion

wavewatching.net

81 pointscurtis12 comments

Would love to be pointed towards any kind of mathematical analysis of this at all. There's a really big problem -- you need high ion temperatures to achieve fusion, but it's electron pressure that's transfered by a shockwave, and therefore it should be electron temperature that will increase. It takes on the order of seconds for electrons to transfer their heat -- of a temperature of hundreds of millions of degrees kelvin to ions.

Can a spherically compressed high pressure concentration really maintain itself for the time required to heat the ions?

A good treatment of this is in the 1978 Nobel Prize Lecture of Soviet physicist Pyotr Kapitsa

http://www.nobelprize.org/nobel_prizes/physics/laureates/197...

The idea is to run just below the Bohm limit (see eg pg 13 in [1]), where the temperature and density required is much lower than that for Tokamak-style inertial confinement fusion. They adiabatically pre-heat and then inject a field-reversed configuration plasma immediately before implosion. So the plasma is at thermal equilibrium and the magnetic confinement minimizes conductivity loss during implosion.

Los Alamos has an MTF program that has demonstrated fusion under these conditions, and there are several other programs. See [2] and [3] for recent experimental and simulation results (different driver mechanism, but same ignition region).

[1] http://fire.pppl.gov/fpa04_wurden.pdf [2] http://fire.pppl.gov/IFE_NAS3_MTF_Wurden.pdf [3] http://sites.apam.columbia.edu/SMproceedings/11.ContributedP...

a bit of background about this tech (as understood by a non-physicist)

1. Take a big metal ball and surround it by a whole bunch of synchronized pistons

2. Fill said ball with molten metal

3. Fire a wave of plasma from both ends of the ball

4. Time the firing of the pistons so that the plasma and pressure wave from the pistons all meet in the center and create the conditions necessary for fusion to occur.

While it isn't yet proven, if this method works it would be quite a bit simpler than the other lines of research into fusion reactors which require massive arrays of magnets to suspend plasma and exotic materials that don't exist yet.

I have no idea if this is actually feasible, but it sure sounds cool!

This makes me wonder: what are the current constraints on fusion reactors? I mean, I had heard that the problem is that the current reactors consume more energy than they produce. Is that the case or is it just not possible to get fusion in the laboratory?

It's actually pretty straightforward to create fusion:

http://en.wikipedia.org/wiki/Fusor

The catch is doing it in a way that actually produces a surplus of energy that isn't a bomb. NB Even with bombs, most H-bomb designs actually get more energy from fission than fusion.

Edit: I wish there was some rule that said that a small percentage of mega-projects like ITER had to be used to fund competing approaches like this.

http://en.wikipedia.org/wiki/ITER

While we are completely off topic, for those who are interested: an H-bomb typically originates more energy from does get more energy from fission than fusion, but the fissile energy is converted into more destructive forms by the fusion reaction. Much of the destructive power of an H-bomb compared with a more conventional fission design is from the fusion reaction.

You are asking a few different questions here. The biggest problem for fusion in general is probably the limited theoretical understanding of plasmas, since simulations are intractable. The biggest practical problem faced by any commercial-scale deuterium-tritium fusion plant is a lack of neutron-resistant materials. Other concepts are built around aneutronic fuels such as pB11, but given the much higher temperatures required, my personal feeling is that this is shooting for Mars when one hasn't even made it to the moon.

As for whether breakeven is even possible (not counting H-bombs), we have to get a little technical. 'Breakeven' is defined as when the input power is equal to output power. However to generate net power, the fusion output power needs to be converted to electricity and fed into the reactor, which means there will be thermodynamic losses. A typical ratio quoted for net power generation, also termed 'ignition', is 1:5. ITER is shooting for 1:10.

The current record holder is the JT-60 tokamak in Japan. It has achieved slightly better than breakeven, with a caveat: these are simulated numbers. You see, for safety and convenience, JT-60 does not use tritium, just deuterium, so the numbers are extrapolated from the D-D case to the D-T case. This does not strike me as a major issue since other tokamaks (JET) have run with tritium, but I am not an expert.

Tokamaks are the furthest along in terms of achieving ignition, but there are still a host of practical issues that need to be solved before the technology can be commercialized.

http://en.wikipedia.org/wiki/JT-60 http://en.wikipedia.org/wiki/Lawson_criterion

You missed a step: 2.5: Spin the liquid metal really fast to create a vortex (like in a water bottle) such that the hole in the vortex reaches all the way to the bottom of the ball.

Also note that in step 3, "both ends" of the ball are the top and bottom. The plasma from the bottom is fired up through the hole in the middle of the vortex.

Ever since reading the original Popular Science article on this company it has amazed me how little attention it has received in the media. It seems like they are really on to something and their budget compared to other fusion projects is tiny. I hope they receive continued funding.

This sounds awesome; I wish there was more information available. I can't quite visualize how the vortex would work. It needs to reach all the way to the bottom of the chamber so the plasma can be injected through it, but it also needs to be small in diameter at the center of the chamber so the pressure wave can collapse it completely. Can those requirements be met simultaneously? How fast would you need to spin the metal? Is it really possible to make a stable vortex such that no liquid metal ever falls into the lower plasma injector (which would presumably destroy it), even as you send shockwaves through the chamber once every second?

You are over thinking it...the outlet port to circulate/drain the molten metal would be the same hole the lower plasma injector uses. By using a curved surface it can take advantage of the centrifugal force and surface tension of the liquid metal to prevent the metal from falling down lower into the injector. So basically there is a doughnut shaped orifice around the top of the lower plasma injector that's connected to a pipe which feeds into the liquid metal circulation system. Here is a cutaway drawing where you can see it: http://1.bp.blogspot.com/_VyTCyizqrHs/ScF7J2w-TOI/AAAAAAAADL...

Every picture in the article is just awesome. It's a dream lab: giant industrial machines mixed with lab tech and whatever-it-takes-to-work.

Every bit looks to have a story. On the image of the sensor there's a sharpie note "<-- STUD" pointing to where a stud was taken off (I assume). Or the aluminum foil wrapped around a pipe in the last image. This is what GTD engineering looks like; it can be polished later.

The entire setup just looks fun, like how I always imagined early rocket labs must have looked like. Serious, but utterly interesting and ever changing, tweaking, and upgrading.

Also the article is good. (But the eye candy made my day.)

Reply on news.ycombinator.com