Hacker News Reader

Comment by pontifier | Hacker News Reader

I feel a little like I'm on this road myself. I've got a design for a type of fusion reactor that is simple and elegant, but I'm having a devil of a time convincing anyone that it's worth working on.

It feels like the Idea is stuck in limbo and I can't get the resources together to test it.

I feel like I have a responsibility to the Idea itself. As long as there is a chance that it could work, dropping it feels criminal.

It needs to be killed completely, through testing, or it's going to keep haunting me.

View on HN

To convince someone to look at your fusion reactor plans, you have to first convince people that you'd be the kind of person capable of innovating in the field -- ie, publish papers, produce smaller results. Success follows success. Not knowing you or your ideas at all, I can already tell you that it's extraordinarily unlikely that you've come up with something new and effective, or you'd know how to get it built already.

I know HOW to build it, I just don't have a million dollars laying around to actually do it. The prototype needs to be built inside an MRI machine.

That's going to be one heck of a prototype or one heck of a bang. Either way the results will be spectacular.

I meant you would know how to get funding for it, since nuclear physics research is a well established field.

I've looked at funding. I've talked to people in the field. There does not appear to be any money available from any source to fund any fusion projects other than Tokamaks or similar devices.

Tokamaks by the way, were invented by an outsider with no previous domain knowledge...

People involved with certain “long term high capital expenditure” efforts have indicated to me that most venture capital funders and US DOE-like sponsors aren’t too interested in projects with potentially long time horizons. They told me they got the bulk of their capital from foreign sovereign wealth funds and university endowment investments.

Couldn't you produce exact diagrams of such a thing and model its properties to show that it would be able to effectively achieve a stable plasma?

I've done some simulations, but have taken them as far as I am able to without additional funding. The results were promising, but the cost of better software to MAYBE get better tests was insane... Over $50k for a package that might not even do what I need it to do.

I've contemplated writing my own software from scratch, but that just seems daunting. I'm quite burned out.

Don't need $1MM, a picoSpin 45 retails for ~$3k [1].

As an alternative, you could homebrew a simple NMR yourself with a cow magnet or a neodymium magnet.

[1] https://www.fishersci.ca/shop/products/thermo-scientific-pic...

Sorry, but that just won't work.

The device I've looked at that I believe will work is a Hitachi Airis II. It has just about the right field strength, working volume and field uniformity for the prototype to have a chance of working. It's an older model and I have seen some for sale for approximately $30k.

How serious are you about this? Why should anyone believe you when you blindly assert that you know how to build it? If you can convince someone in the field then there's a chance that the experiments can be done to see if it would work, but you'd need to convince someone that you know what you're talking about.

So where's your evidence? The usual method is to have a track record, another method is to be able to state very clearly what the new/novel idea is for this to be different and workable. In that case, what's your elevator pitch?

If you're outside the system it boils down to that question - if someone gives you three minutes, you need to be able to convince them to invest more time to understand more detail.

Do you have a description like that?

Why is your idea different? What's new about it?

Why should some invest the time to see how/whether it works?

I'm very serious, but very burnt out by many recent failures.

I would hope that nobody would blindly believe me, but would investigate the idea on it's merits.

It's basically an inside out cyclotron... instead of ions circling a central point, all ion trajectories intersect a focus point once per period. If they collide at that focus point without fusing, they will still be on a trajectory that brings them back to the focus.

There are no new physics, and only basic college level physics knowledge is required to understand the device.

Here is a simulation video I made using a program called simbuca that simulates ion motion in a penning trap. https://www.youtube.com/watch?v=RT6nvmN7GB0

As you can see, that simulation was made 3 years ago, but has very few views. I've tried, but don't know how to move forward without funding further research myself... I've basically stopped looking for outside funding.

It seems like you might be neglecting repulsive forces that would alter the trajectories as pressures increased far below those necessary for fusion. In short, it’s hard to see how your device would fuse anything; you’d need some way to constrict everything or it wouldn’t collapse so neatly in each cycle. That constriction would have to be magnetic, require a ton of energy and modeling to solve, and... then you’re in the same hole as every other fusion researcher, except you have a weird geometry that hasn’t been studied and isn’t efficient.

Sorry.

A uniform magnetic field is used to cause the trajectories in the simulation. Because the ions are allowed to follow large, circular (helical in 3D), cyclotron trajectories, self repulsion is minimal over a large portion of their motion.

At the focus, where repulsion is more of an issue, all ions are essentially flying straight at the focus (at the same time) with enough energy to pass through it easily unless they collide with another ion there.

There were some instabilities in some simulations that I did where I cranked repulsion to massive levels... It's beyond my expertise to interpret whether those instabilities are fatal or not, or would cause problems in an actual device. this simulation is here: https://youtu.be/NnbfgTP7v5M

One benefit of this weird geometry is that the walls of the device can be very far from any hot plasma. Energy losses should be minimal, and the self intersecting trajectory gives ions multiple chances to fuse once accelerated.

It may not work, but I feel that it deserves to be looked at.

I'm not much of a physicist, but I find the idea intriguing. Let me see if I understand. When you're talking about "the ions" you mean positively charged ions — nuclei, right? But there have to be electrons around somewhere so the whole thing is electrically neutral, right? Where are the electrons, and what are they doing?

The deuterium nuclei are positively charged and are basically in a penning trap. I don't think electrons would be stable there, and would be attracted to the positively charged electrodes at each end.

The deuterium would start out as a very diffuse gas and a high energy, narrow electron beam (pulsed at the cyclotron frequency) would be used to ionize them at the focus and add energy. The resulting deuterium plasma would be non-neutral. It would want to fly apart, but as it expands the self repulsion should get much weaker allowing the weak magnetic field to steer each ion back toward the focus again. (in 2d, cyclotron motion in a uniform magnetic field is circular)

There are many questions I can't answer about the technical details of how the plasma would evolve with repeated electron beam pulses, and what qualities the beam pulses should have to give the best result... I think it should be short duration, narrow, and high energy.

Likewise, I don't know how high density can get before it starts to cause synchronization problems.

> The resulting deuterium plasma would be non-neutral

Meaning, it would be negatively charged because of the added electrons?

My guess is, you want the plasma to be neutral. (The term I've seen is "div-free" — you want the divergence of the E field to be everywhere zero.) But maybe removing the excess electrons wouldn't be too hard.

Anyway, if you want to chat about this more, feel free to email me — address in profile.

The nuclei would be positive ions... I don't believe free electrons would be stable in a penning trap set up for positive ions. The strongly positive endcaps (my guess is 60kv) would make it unlikely that they would stay in the same region as the positive ions, though I'm not sure.

The goal is for the ions to act like individual ballistic deuterium nuclei flying along a cyclotron trajectory. The density should be very low most of the time, but spike dramatically very briefly.

That line between high density and acting like individual particles is one of the reasons this device is difficult for me to model.

Hi. It might be good if you email me, as I'm not sure how much I can put into this post.

I've followed a similar path to the one you want to take. I got funding and built a fusion reactor design with some similarities (see US patent 5818891 for some details). What you are describing is more similar to a Migma design (https://en.wikipedia.org/wiki/Migma). If you are not already aware of Migma, I recommend researching it to see what the problems were. See if figure 3a looks familiar here http://www.rexresearch.com/maglich/migma.htm.

This type of design will have a few problems. First, to work the way you simulate, it will have to remain at low density. This is because of space charge (all those repulsions add up as you add more particles). Also, the particle velocities will smear out over time as the particles adopt a Maxwellian distribution (see https://dspace.mit.edu/handle/1721.1/11412). You can get around the space charge issue by putting electrons into the mix, but then you have a large source of energy losses (bouncing electrons emit photons that leave the area).

I'd love to see someone crack the self-sustaining fusion problem. If you would like to talk fusion, feel free to email me.

Wow, I like your idea... It's very similar in concept. Set up a system which naturally resonates at a specific frequency, then add energy at that frequency until something has to give...

I am familiar with Migma, which has self-intersecting orbits. I believe the main difference is that my ions will have a thermal distribution from the start, yet still be able to self intersect. (Particle velocity does not change cyclotron orbit timing until much higher energies. So Rider's arguments are not as applicable)

My hope is to exceed the Brillouin limit only very briefly, and that most of the time, the plasma will be diffuse enough that space charge won't be an issue...

The size of the focus should be the diameter of the electron beam, and electron beams can be very narrow. If the beam diameter is made very very small (electron microscopes prove this is possible), density at the focus could be very high for very short amounts of time... higher than even the Z machine, yet would quickly drop to levels low enough that space charge is not an issue.

The possibility of this periodic density spike is predicted in "Beyond the Brillouin limit with the penning fusion experiment" which implies that local density can exceed the limit in either space or time. They chose space, I've chosen time.

I'd love to email you about this, but fear I'm falling well into crank territory as it is. I'm pontifier everywhere, including gmail.

If you look at that second simulation, about midway through you can see the problem I was talking about, and FiatLuxDave goes into in more detail. The repulsive force is keeping the density of the plasma very low, and you end up with a cloud of weakly interacting, orbiting ions that never fuse. The problem really is that the repulsive force increases enormously at very close ranges, so it may seem like the ions are colliding, but they’re not going to regularly have the energy required to overcome Coulomb repulsion, which remember increases with the inverse of the square of the distance.

To overcome that you’d need to increase the strength of the magnetic containment, and shrink your container, but that will raise the density and change the behavior of your plasma while heating the walls of the chamber.

Sounds a little reminiscent of General Fusions piston design: http://generalfusion.com/subsystems-fusion-energy-technology...

The essence: A sphere with an array of pistons compresses a central point.

Yes, periodic compressions are a feature of both.

I've looked a lot at what they are doing and hope to replicate their story. If I can generate some "marketing neutrons" like they did, maybe funding will follow.

That's funny, I had never seen that! It reminds me of a couple of things, water hammers and sonoluminescence:

https://en.wikipedia.org/wiki/Water_hammer

https://en.wikipedia.org/wiki/Sonoluminescence

The important equations are probably under:

https://en.wikipedia.org/wiki/Water_hammer#Slow_valve_closur...

So for high flow velocity and small time delta, the pressure can be pretty astronomical. If you could somehow create a little bubble in the center of the Earth and let it collapse, you might get a little fusion, statistically speaking, before it flattens out to steady state conditions and obviously a low probably of fusion since we aren't standing on a star!

Edit: I realized one other connection here - I think this is what happens when red supergiant stars run out of nuclear fuel and can no longer produce enough heat to hold up their gas against gravity. The gas cools, collapsing into a small volume near the limit of a black hole's density, and under such tremendous pressure, much of the star's mass undergoes fusion simultaneously and releases 10 billions years worth of energy in an instant, blowing the remaining material out into space. That's where the heavy elements in our bodies like iron (and even heavier elements that require added energy to fuse) come from:

http://cse.ssl.berkeley.edu/bmendez/ay10/2000/cycle/snII.htm...

I really like situations like these... places where you get singularities.. divide by zero situations that break things. It's where really interesting stuff happens that can be unexpected and lead to new ideas and understanding.

This reminds me of certain aspects of the Triple Alpha design. I think they’re still hiring.

https://en.m.wikipedia.org/wiki/TAE_Technologies

Reply on news.ycombinator.com