by am_Unition

"So You Hate Flying Kites And Want To Hate Plasma Too"

Imagine you have a blank piece of paper. Imagine using a pen to draw the final, knotted-up lump that was the last kiteline you threw away. One or two-handed kite, one continuous line or two, don't care. Now, plop the ruined kiteline(s) squiggles onto an x-y plot, anywhere, and tell me how the points where the tangled mess of lines pass through relate to one another, using as basic of maths as possible. What a mess. You basically ruined the line's topological solvability (or linearity), because of losing control of the kite/stability/confinement. You dunno where to pull to get it untangled. On the drawing, one point along the X-axis is not guaranteed to correspond to a single point of the kiteline along the Y-axis.

Now, imagine, if the wind blew ever so slightly differently, maybe there would be no knot because you managed to react to the (now smaller) gust of wind in time to keep control. Now that the gust has passed, you and your kite are just cruisin' again. The kiteline's still not quite a straight line, it's hyperbolic, maybe parabolic-ish, but I can get good agreement with simple maths pretty easily.

The magnetic fields embedded in plasmas are your kitelines, and the winds of plasma do, in fact, blow very differently each time you go kiting/fusing. The onset of magnetic field-line reconnection/kiteline knotting happens when a gust/instability becomes unpredictable and the tokamak/you don't react in time to correct it. And yup, there are now automated, active confinement controls built into tokamak systems.

With plasmas, there's maaaaany qualities we gotta characterize if we are to accurately describe the physics compared to the fluids and particles we're used to thinking about, like the air around your kite. The math gets absolutely disgusting, and especially at these funding levels it'll still be a long time until we have simulations consistently matching e.g. specific loss of confinement event measurements. During loss of confinement, the plasma and magnetic flux should be ejected from resembling a fuzzy donut into the chamber wall in a somewhat unique way, every time. Surely the theory and modeling branch of the ITER team hopes to match specific event characteristics. I don't think that's been done very well, yet. Anywhere. Good simulations are still quite limited by computing power, even on state-of-the-art superclusters run by DoE. The technique requires GPU architecture (screw off crypto, GPUs go cheep now), btw. It could take longer than another decade, honestly. It be hard.

The same physical process that NASA designed the almost-$2 billion MMS mission for, magnetic reconnection, is the non-linear knot-tying that ruins stability and fusion. Unique and highly variable reconnection physics happens at the microscale, but is self-consistent with global topological reconfigurations. The two need each other. The magnetosphere and solar wind topologies MMS is flying through in space do another version of this dance, between reconnection and global-scale topology, but the magnetosphere is driven entirely differently than tokamaks, and the regimes of plasma conditions are quite different.

---

@somebody writes into our mailbag this week with an aritcle:

EVEN IF PLASMA STABILITY issues can be resolved, MCF systems still possess an inherent weakness: the fusioning plasma must interact with its physical environment, commonly referred to as the plasma-facing walls. Over the last 25 years, tokamak researchers have succeeded in making plasma pulses progressively longer—with no attempt at fusion production. This accomplishment amounts to little more than a heat-removal and plumbing exercise that has no impact on scientific feasibility. Any claim that these incremental improvements contribute to commercial viability is dubious, at best.

100% agree with the bit about plasma-chamber wall interactions. That's also how the heat is deposited for transfer into generating electricity... via boiling water from a tank near the inner chamber wall through turbines, lol. I don't even know if they've done it, or just calculated how much heat would have gone into heating a water tank. People are not allowed nearby in the lab during fusion "shots", btw. Which complicates things.

I kinda agree that chasing longer confinement/stability times without any fusion taking place is a weird thing to attempt. Like that blurb says, the heat produced from fusing nuclei should be a large portion of the overall energy budget, obviously, if you're trying to achieve a net positive yield of energy. Not accounting for it seems especially weird because of how non-linearly plasma behaves. Keeping stability's like trying not to set off a sensitive person, and, so far, instability seems like it's always eventually destined to find a way to ruin stability.

Seems unlikely and rare/difficult to find, but it could even be that several particular sets of non-fusion conditions that seem inherently unstable are actually stabilized by active fusion.

I imagine that the first thing they've gotta prove to make this seem like a good idea is to show that both their simulations and measurements can reproduce similar results to an actively-fusing confinement with similar conditions. I don't even know how anyone could do that... the plasma conditions in the core of the thing have to be drastically different for fusing vs. non-.

Also from @somebody's piece:

This essay compares the backgrounds and outcomes of the recent D-T campaigns at the JET and NIF facilities, shows why inertial confinement has established a clear lead over magnetic confinement in attaining reactor-relevant fusion conditions, and examines the future directions of both approaches. While there is a strong argument that the scientific feasibility of ICF has been demonstrated in recent experiments, the status and prospects for MCF are far less favorable.

LOL, this was wow. Shots fired. My feelfeels got hurt a li'l, 'cuz I have spent much, much more time thinking about tokamak or other magnetic confinement instead of inertial confinement. I'll have to spend more time deciding how believable the claims are. Better than break-even, OK, but how efficiently is the energy being harvested? I know that tokamaks' water-boiling scheme is funny, but this one baffles me even more, as of the moment. The complexity of the laser ignition facility is nuts, by the way. Not that tokamaks are simple or cheap.

Approaching inertial confinement first and foremost from a consideration of the magnetic field topology and reconnection is still the way to go. Simply (edit: lol) change the time and length scales and the regimes of plasma conditions. For inertial confinement, I think the game becomes: Design a laser and fuel geometry that produces a closed magnetic topology (or several) unconducive to widespread reconnection for as long as possible. Same overall game as magnetic confinement, different details. Maybe you could go the other way with it and intentionally design reconnection mimicking a coronal mass ejection to shoot some hot plasma at a target?

But I dunno, scaling down these huge labs, if they manage proof of concept, the National Ignition Facility, the ITER tokamak in France, etc., will be hella hard, in and of itself. And then interfacing with the grid en masse. I think the grid people are already too busy attending to integrating solar and wind, and rightfully so, I guess. Battery tech needs improvement no matter what. Would be nice to put a rare-Earth metals-rich asteroid in orbit, maybe around the moon, if you could prove it'd be less impactful to Earth than digging endlessly here.

I guess I'll learn more about inertial confinement. Whatever the form, we will eventually need a deep-space engine, and confinement fusion of some kind is our best shot.

The idea of combining both confinement methods into a single device is nightmare fuel. :D

link to Entry 2 of This Journal