Focus Fusion?

[Eric L]

Hi Jean Tate,

It would be good to discuss issues in cosmology. But like everybody else, my time is limited, so we have to get this Focus Fusion discussion to a closing point first. Also, the Plasma cosmology thread looks almost as endless as the universe-- 91 pages, yikes!—so it would be probably be better to start a new thread about evidence for and against an expanding universe. But, as I say ,I can’t participate right now.

 

[ben m]

Again, this is name-calling instead of physics arguments. If the theoretical calculations in the Journal of Fusion Energy are wrong, what is wrong with them? Calling them a ghost or crack-pot is something you could say of any theory, without in any way actually criticizing it. The more detailed derivations in the old Particle Beams paper should be available on our site, so I just requested permission from the publisher to do that. I would assume I will get it after 32 years! But if you really know things that are wrong with the JFE derivations, here is your chance to say what they are.

That was not name calling. This is a serious question about whether your focus fusion work is attempting to be scientific or whether it's settling for the easier task of "let's convince ourselves, and then write it up so it sounds good".

Your 2012 paper states that it builds on a 30-year-long "quantitative elaboration" of Bostick's "theory" of plasmoids, which sounds more impressive than it turns out to be on clicking through. OK, y'know, I know you don't want to talk about signs of crackpottery, but "discussion of serious issues avoided by pretending it's resolved elsewhere" is a sign of crackpottery.

Let me dig through J Fusion Energ (2011) 30:367–376 (DOI 10.1007/s10894-011-9385-4) for details, for example.

The role of the plasmoids in producing the fusion neutrons
and the physical processes involved in their formation
and maintenance has been hotly debated among DPF
researchers for decades. The model that best fits all the
existing data makes the role of the plasmoids central to
neutron production. This model, initially developed by
Bostick and Nardi [4], and confirmed by observations of
several groups over three decades, was elaborated into a
more quantitative theory by the present author [20–24].

This paragraph is clearly intended to convey that, despite some past controversy, the Bostick "model" has now been firmly established---there is no need to worry your little heads about that, dear reader. But, seriously, has it? That "fact", perhaps the crucial one in this entire discussion, is the one you didn't bother citing! Is it indeed a fact? Are there people other than you who indeed believe that this is "the model that best fits all the existing data"? Given the antiquity of the Bostick paper, what model precisely is the one they're using, and is this the same one you're using? Did someone run COMSOL at some point and make predictions which you're confirming? It'd be nice to know. Your paper is no help.

Yes, I can find out more (and I will---wait a sec) but the "the reader should check for himself" thing is ... well, it's not supposed to be adversarial. I shouldn't have to do this. The norm in science is that the author walks the reader through it honestly.

The funny thing is, there do appear to be people doing modern-looking computational models of DPF plasmas. A nice review article by Soto (http://iopscience.iop.org/article/10.1088/0741-3335/47/5A/027/meta) certainly does NOT indicate that anything Bostick-related has been "confirmed by observations of several groups over three decades." Lee et. al. don't appear to be invoking Bostick or plasmoids. The Polish groups don't appear to be invoking Bostick or plasmoids. What am I supposed to conclude? Maybe the thing you describe as "hotly debated" is, in truth, basically you vs. everyone else? Maybe the thing you described as "confirmed by several groups" is, in truth, confirmed by you alone? Maybe you're talking about this guy Auluck, but didn't feel like citing him, just to make things harder for your readers?

In a later article Lee and Soto and others give a different version of the state of what you call a "raging controversy": (doi:10.1088/0741-3335/51/7/075006)

Mechanisms such as moving boiler, beam–target, gyrating particles [1–5] and others such as quasi-Maxwellian hot plasmoids [6] have been invoked to explain the high measured Yn. ... Again from such mechanisms come forth general results such as the temporal and spatial characteristics of the neutron pulses, and representative yield numbers put forward to illustrate the validity of the assumed mechanism. There do not appear to be any published results demonstrating non-thermonuclear modeling which may be applied to any particular machine to derive Yn in a manner where such modeled data may be compared with specific experiments

Translation: the experts do not find your model convincing or conclusive, Eric, and they said so in print. Two problems here, which again speak to the "does this make you sound serious or does it make you sound crackpot" issue:

a) You chose not to cite this, or anything like it. In fact you chose to polish it up (perhaps I should say "misrepresent it") when you summarized the "controversy" over these mechanisms in a way that made your version sound (to the non-reference-clicking reader) so nice and uncontroversial. That's not the sort of thing scientists do. If your approach is sui generis, you can say so.

b) You've been saying (as an attempt to insult/dissuade me) that only an incompetent or sloppy physicist could read your paper, follow the derivations, and do anything other than agree with your conclusions. That is manifestly not true. In fact, Lee and Soto's objections (your yield numbers get plugged into "assumed mechanisms" despite the lack of "non-thermonuclear modeling" of the mechanisms themselves) is almost exactly the same as my objection in post 119 ( http://www.internationalskeptics.com/forums/showpost.php?p=11108652&postcount=119) Huh. Are they "just name-calling" too?

 

[BenBurch]

Eric, you are really not dealing with this well. Talk it over with some of your academic friends and I believe they can tell you how to approach criticism like this. Some video published on youtube of shots being run would help greatly, for example. Some intermediate papers on what you've learned so far (after filing patent applications of course) would be awesome as well.

 

[Beelzebuddy]

Yes, I can find out more (and I will---wait a sec) but the "the reader should check for himself" thing is ... well, it's not supposed to be adversarial. I shouldn't have to do this. The norm in science is that the author walks the reader through it honestly.

Don't say things like that while I'm drinking my coffee, I nearly ruined my keyboard.

 

[ben m]

Don't say things like that while I'm drinking my coffee, I nearly ruined my keyboard.

Hmm, maybe that was unclear. I mean "adversarial" in the sort of courtroom sense, and I mean the relationship between author and reader. The relationship of an author to a reader should NOT mirror the relationship of two lawyers.

If you're (say) a lawyer defending an embezzler, you're perfectly welcome to think, "Hmm, it'd be good for my client if nobody ever asks whether he has huge gambling debts." You don't bring them up yourself; you tell the client not to mention them unless asked directly. If the prosecutor doesn't file the right discovery papers, doesn't ask the right pointed questions, and never finds out (or finds out that information exists)---fine. The jury didn't hear about the gambling debt, that helps you win, and you're happy. (On the other hand, the prosecutor is expecting this behavior, and knows they have to drill for information.) That's what I mean by "adversarial". The opposing parties are sitting on useful information, and the only way to get this information is ... well, adversarially.

Scientific papers aren't supposed to require adversarial discovery. The theory's defender should be volunteering pointers to information about its opponents. (For the purpose of criticizing/refuting them? Of course!) The theory's opponents should similarly be volunteering honest and reasonably thorough accounts (Critical, dismissive accounts? Sure!) of the existence of a case for the defense.

A mainstream paper would have said something like "Bostick's [1] model of plasmoids is not widely used [2-10] due to the trend towards smaller apparati [11,12,13] where the MHD approximation holds for longer; the success of MHD in those domains has deceived the community into neglecting plasmoids in the domain where we now know they occur. This author has consistently reported x-ray fluxes contradicting MHD models [13-20] and which we find to agree a Bostick-based model first outlined in [21] and debated over several years [22-25]. Several recent experimental counterclaims [22-30] are artifacts of contamination, which we will show in work currently in press." See the difference?

 

[Beelzebuddy]

Hmm, maybe that was unclear. I mean "adversarial" in the sort of courtroom sense, and I mean the relationship between author and reader. The relationship of an author to a reader should NOT mirror the relationship of two lawyers.

If you're (say) a lawyer defending an embezzler, you're perfectly welcome to think, "Hmm, it'd be good for my client if nobody ever asks whether he has huge gambling debts." You don't bring them up yourself; you tell the client not to mention them unless asked directly. If the prosecutor doesn't file the right discovery papers, doesn't ask the right pointed questions, and never finds out (or finds out that information exists)---fine. The jury didn't hear about the gambling debt, that helps you win, and you're happy. (On the other hand, the prosecutor is expecting this behavior, and knows they have to drill for information.) That's what I mean by "adversarial". The opposing parties are sitting on useful information, and the only way to get this information is ... well, adversarially.

Scientific papers aren't supposed to require adversarial discovery. The theory's defender should be volunteering pointers to information about its opponents. (For the purpose of criticizing/refuting them? Of course!) The theory's opponents should similarly be volunteering honest and reasonably thorough accounts (Critical, dismissive accounts? Sure!) of the existence of a case for the defense.

A mainstream paper would have said something like "Bostick's [1] model of plasmoids is not widely used [2-10] due to the trend towards smaller apparati [11,12,13] where the MHD approximation holds for longer; the success of MHD in those domains has deceived the community into neglecting plasmoids in the domain where we now know they occur. This author has consistently reported x-ray fluxes contradicting MHD models [13-20] and which we find to agree a Bostick-based model first outlined in [21] and debated over several years [22-25]. Several recent experimental counterclaims [22-30] are artifacts of contamination, which we will show in work currently in press." See the difference?

Yes. That's why I was being facetious. Brutal honesty about one's own work is the ideal, but in practice it's rare to find a modern paper which does so. You're much more likely to find papers which guide the reader along a winding, biased path, carefully downplaying any unfortunate stumbling blocks, and it is up to the reader to become familiar enough with the field to at least know how myopic the presented story is. Otherwise it's just not sexy enough to get into a good journal.

Like it or not, whole fields are guilty of the same sins you ascribe to those papers, albeit not so flagrantly presented.

 

[ben m]

Yes. That's why I was being facetious. Brutal honesty about one's own work is the ideal, but in practice it's rare to find a modern paper which does so. You're much more likely to find papers which guide the reader along a winding, biased path, carefully downplaying any unfortunate stumbling blocks, and it is up to the reader to become familiar enough with the field to at least know how myopic the presented story is.

I disagree. In my own field, even in the midst of ugly controversies (barely-polite ArXiV shouting matches between disagreeing groups; an once-respected author whose mental health was a frequent conversation topic at conference coffee breaks; etc.) I can't recall a biased or self-serving introductory overview. (Biased, self-serving, or erroneous estimates of the importance/reliability of results? Sure.)

 

[Beelzebuddy]

I disagree. In my own field, even in the midst of ugly controversies (barely-polite ArXiV shouting matches between disagreeing groups; an once-respected author whose mental health was a frequent conversation topic at conference coffee breaks; etc.) I can't recall a biased or self-serving introductory overview. (Biased, self-serving, or erroneous estimates of the importance/reliability of results? Sure.)

Oh, I can give you some. Just yesterday I came across a review article which seemed to have conveniently forgotten about a good half-dozen esteemed scientists whose work disagreed with the subject matter. By all appearances it was a fine review, it just completely failed to cite them or otherwise acknowledge their existence.

More generally, technique papers often have very self-serving introductions. No one cares about a technique which offers minor but appreciable advantages in certain situations; it has to read like the second coming of Jesus would make out with the reincarnation of Niels Bohr, but for lack of this technique.

 

[Eric L]

plasmoid history

Ben M has finally raised some actual issues in physics, or at least in the history of ideas about the dense plasma focus. First, let’s look at the question of Bostick’s overall qualitative model, cited in my JOFE paper. This model says that the discharge create vortex filaments of plasma, which then , in the central pinch kink themselves up into a tiny plasmoid or ball of plasma, where the production of beams and the heating to fusion temperatures occurs.

We say in the JOFE paper has been plenty of debate over the role of plasmoids. Then we say the model that puts plasmoids as central is the best one. Do we say “most physicists in the field agree”? No we don’t actually. It is not implied either. We were clearly stating the views of the authors in saying this was the best model.

However, was the implication that Ben m derived, factually true at the time our paper was published? Did most DPF scientists agree that plasmoids were central to the mechanism? Yes it actually was. Here we are talking NOT about whether the qualitative model was valid, but the historical point of whether by 2011 (or for that matter 2016) most people in the field thought it was. Take a look, for example, at the most complete description of Bostick’s model which was published back in 1977 in an obscure journal, the International Journal of Fusion Energy—not related to JOFE. “The Pinch Effect Revisited” You can Google it and look up the citations. There are quite a few, but almost all of them after 2009. So why was a paper from 40 years ago suddenly being cited quite a bit, including by the researchers in the Polish group that runs the most powerful dense plasma focus device in the world?

Well, because the debate had shifted. What was an extremely controversial stance, in 1977, hard to even publish, and hotly debated for the next 30 years, had now become –-pretty generally accepted. Bostick’s old paper is often the first or nearly the first citation in many of Kubes’ papers. (Kubes is one of the leading researcher in the Polish group). Take Kubes recent paper, “The evolution of the plasmoidal structure in the pinched column in plasma focus discharge”

Here’s the abstract in toto: “In this paper, a description is provided of the evolution of the dense spherical-like structures—plasmoids—formed in the pinched column of the dense plasma focus at the current of 1 MA at the final phase of implosion of the deuterium plasma sheath and at the phase of evolution of instabilities both at the time of HXR and neutron production. At the stratification of the plasma column, the plasma injected to the dense structures from the axially neighboring regions forms small turbulences which increase first the toroidal structures, and finally generates a non-chaotic current plasmoidal structure with central maximal density. This spontaneous evolution supports the hypothesis of the spheromak-like model of the plasmoid and its sub-millimeter analogy, high-energy spot. These spots, also called nodules formed in the filamentary structure of the current can be a source of the energy capable of accelerating the fast charged particles.”

Short paraphrase—plasmoids are the source of x-rays, high energy particles and neutrons(from fusion reactions). By the way, note that the plasmoids are referred to as plasmoids, nodules and hot-spots , all in the same paragraph. This lack of settled nomenclature does make our DPF literature confusing, but we are talking about the same objects, just different names.

What led to the shift that the plasmoid theory became uncontroversial? Well, I like to think our work played some role—I certainly have talked to Kubes and many others like Lee and Soto at conferences and by email. I think, however, it played a relatively small role. The biggest factor, IMHO, was the Grim Reaper removing all the original participants in the debate from the scene. The younger generation (and these are still very senior people right now in their 60’s-through 80’s) judged the experimental evidence more objectively and arrived at the conclusion that ,qualitatively, the Bostick/plasmoid model was right.

I’ll end by emphasizing this is the qualitative description of Bostick and Nardi, not my later quantitative theory based on their model.

This is long enough for one post.

 

[Eric L]

physics models

A second point Ben m is confused on is the basic one of what a quantitative model is. He seems to think that it is a simulation. He also seems to think that the commercial program COMSOL is a good way to simulate the DPF. Let’s sort this out.

My quantitative model is an analytical model, not a simulation. That is, I derived from basic plasma physics theory the formulae that describe how the plasma will evolve quantitatively (magnetic field, density, dimensions) given the Bostick-Nardi qualitative description of the evolution. This kind of analytical model, expressed as a set of equations, used to be the only kind of quantitative theory that existed in physics. People did do physics before there were computers, you know. Analytical models are still the best way to fully understand a phenomenon and connect it to fundamental theories like electromagnetism and quantum mechanics.

Of course, to get an analytical model with exactly solved equations you generally have to simplify things. If you try to introduce more physical phenomena—such as the fact that the filaments have to form, they are not exactly circular, and other stuff like that—a different kind of model, a simulation, becomes useful. In a simulation you put in more physical processes and do calculations on some sort of discrete grid. That is useful too, but it is a different type of model than an analytical model, not a substitute for an analytical model.

So why don’t we also have simulation? We are working on it, but it is very difficult and no one has done a full 3-D simulation of a DPF successfully –including the plasmoid formation process. COMSOL is totally inadequate for anything like this job. Ben m, you are again off by many factors of ten.

You need a program specifically developed for DPF. Why? Because the density of the plasma makes the simulation very hard. In a particle-in cell (PIC) simulation, you normally want to have a time step that can resolve what is called the plasma frequency. That means a time step of about 10^-14 sec for a process that lasts more than 10^-6 second, or 100 million times steps. Much worse you need to resolve a few microns or better in a device whose dimensions are cm. You end up with trillions of cells. If you take this brute force approach, you pretty much would need to hijack the biggest supercomputer in the world for a year for each run.

Why do you need 3-D? Because the formation of the plasmoid is inherently 3-D—you can’t see a kinking process in 2-D, it does not have axial or other symmetries. That’s why plasmoids can’t show up in a 2-D simulation.

Of course there are tricks to change the grid dynamically, etc., that may work to get around all this and get a good simulation, but so far neither we nor anyone else has succeeded in doing this. Right now, doing the experiments is probably faster than getting the simulations running.

But, final point, the analytical models give predictions that can be tested against experiment and they can guide how you build your machine. For example, my analytical model predicted that smaller electrodes would be better than big ones. That is the main reason our machine, although far from optimized, produces more fusion reactions than the more powerful one in Poland. Ours has electrodes that are 4 times smaller than theirs in all dimensions.

 

[Eric L]

Ben Burch, check out our website and the "about" section. You'll find we have published a lot since the patent. Also, just google focus fusion firing a shot video and you'll find one.

 

[BenBurch]

Ben Burch, check out our website and the "about" section. You'll find we have published a lot since the patent. Also, just google focus fusion firing a shot video and you'll find one.

Excellent.

 

[rwguinn]

An analytical model IS a simulation...

 

[JeanTate]

Hi Jean Tate,

It would be good to discuss issues in cosmology. But like everybody else, my time is limited, so we have to get this Focus Fusion discussion to a closing point first. Also, the Plasma cosmology thread looks almost as endless as the universe-- 91 pages, yikes!—so it would be probably be better to start a new thread about evidence for and against an expanding universe. But, as I say ,I can’t participate right now.

Hi Eric L,

No worries. FWIW, I went and read the entire 91 pages of that old thread (well, I did skip a fair bit; some of the stuff there is just, um, well not worth the electrons it's printed on).

Your papers - well, at least some of them - were cited, and quoted (sometimes without attribution, can you believe that?!?) perhaps a dozen or so times. As was your book. There are a lot of critiques of both, but in terms of Plasma Cosmology being woo, this quote (from WP?, but ultimate source is what?) seems to have been regarded as pretty convincing (link to post by Zeuzzz in which it is quoted; my bold):

3. Since every part of the universe we observe is evolving, it assumes that the universe itself is evolving as well, though a scalar expansion as predicted from the FRW metric is not accepted as part of this evolution (see static universe).

In terms of the match between observation and theory (i.e. yours, based on your papers and book), the killer - in the eyes of the critics - seems to have been the inability of your model(s) to match the highly detailed observations of the CMB and of high-z radio sources.

 

[JeanTate]

Hi Jean Tate,

It would be good to discuss issues in cosmology. But like everybody else, my time is limited, so we have to get this Focus Fusion discussion to a closing point first. Also, the Plasma cosmology thread looks almost as endless as the universe-- 91 pages, yikes!—so it would be probably be better to start a new thread about evidence for and against an expanding universe. But, as I say ,I can’t participate right now.

I'll probably sit out most of the discussion on Focus Fusion, but I'd like to give you a bit more of context, so that you can understand at least part of your audience better.

I'll start with a brief look at the Zooniverse, perhaps the largest online citizen science activity^. Initially, almost a decade ago now, it was just Galaxy Zoo, but has now expanded to over 40 projects. Several studies have been published on who the volunteers/citizen scientists/zooites are. To take a somewhat old one, Raddick+ (2013) found that over half had a university degree, and nearly 10% a PhD (or equivalent), which is a substantially higher fraction than in the general adult population (of the US and UK, the two countries they used as reference). Others have found a substantial over-representation of present or past (many zooites are retired) employment in STEM (US term) fields, and a strong interest in science, and learning a big motivator.

I don't know if similar studies have been done of ISF members, but my - thoroughly subjective - impression is that it's likely similar.

Why is this important?

Well, at least part of your (potential) audience knows about journals, papers, paywalls, etc. And quite a big subset is not put off by a technical discussion. And at some level they do not take kindly to being 'talked down to'; sure, a very simplified overview is expected, but so is a set of easily-accessed explanations/descriptions/resources which go deeper, right down to the level of published papers.

There's also "the Electric Universe" ("the EU").

This pseudo-science/anti-science set of ideas can be found all over the internet, particularly anywhere astronomy or plasmas are discussed (as I noted above). My - thoroughly subjective - impression is that almost everyone who is interested in astronomy (etc), and who has a university degree, has encountered (shall we say) some zealous EU fan or twenty ... and concluded that the idea is pure crackpot nonsense. At some level, your website seems to have similarities with the EU (also as I noted above), which will likely turn away many a potential supporter.

The same is true of 'non-standard' approaches to controlled fusion, albeit to a much lesser extent.

So what?

Of course, it's your project, your website, etc. However, if you are aiming to get more people interested (and not turned off), more investors, etc, I would recommend that you think long and hard about how you position your research (etc); in particular, how you distinguish it from "the EU" and the many other non-standard efforts to develop 'table-top' controlled fusion devices.

Now back to your regular program ...

^I think they now claim to have nearly 1.5 million members

 

[JeanTate]

Take Kubes recent paper, “The evolution of the plasmoidal structure in the pinched column in plasma focus discharge”

Here’s the abstract in toto: “In this paper, a description is provided of the evolution of the dense spherical-like structures—plasmoids—formed in the pinched column of the dense plasma focus at the current of 1 MA at the final phase of implosion of the deuterium plasma sheath and at the phase of evolution of instabilities both at the time of HXR and neutron production. At the stratification of the plasma column, the plasma injected to the dense structures from the axially neighboring regions forms small turbulences which increase first the toroidal structures, and finally generates a non-chaotic current plasmoidal structure with central maximal density. This spontaneous evolution supports the hypothesis of the spheromak-like model of the plasmoid and its sub-millimeter analogy, high-energy spot. These spots, also called nodules formed in the filamentary structure of the current can be a source of the energy capable of accelerating the fast charged particles."

Kubes+ (2016)?

If not, what?

Eric L, you can now post links (the threshold is 15 posts, I think); may I ask, why didn't you?

 

[JeanTate]

<snip>

My quantitative model is an analytical model, not a simulation. That is, I derived from basic plasma physics theory the formulae that describe how the plasma will evolve quantitatively (magnetic field, density, dimensions) given the Bostick-Nardi qualitative description of the evolution. This kind of analytical model, expressed as a set of equations, used to be the only kind of quantitative theory that existed in physics. People did do physics before there were computers, you know. Analytical models are still the best way to fully understand a phenomenon and connect it to fundamental theories like electromagnetism and quantum mechanics.

May I ask, what critiques of your analytical model have been published?

<snip>

So why don’t we also have simulation? We are working on it, but it is very difficult and no one has done a full 3-D simulation of a DPF successfully –including the plasmoid formation process.

May I ask, are you the only group working on this kind of simulation?

COMSOL is totally inadequate for anything like this job. Ben m, you are again off by many factors of ten.

Can you provide some details? Particularly as to why COMSOL "is totally inadequate"?

 

[BenBurch]

There is some virtue in building a physical system to test an idea, especially a relatively attainable system like this one. Simulations rarely have perfect fidelity.

I guess I am by nature a tinkerer and a constructor, and have sympathy for people whose approach to a problem is to actually start bending metal.

Worked pretty well for Edison.

 

[ben m]

An analytical model IS a simulation...

It's worth clarifying: no, it's not quite. An analytic model includes the phenomena which you thought to include. If you think you're got a spherically-symmetric implosion, like if it were 1970 and you were daydreaming about building NIF---you can write down an analytic description of the temperature/pressure/fusion rate as a function of time in a spherically-symmetric implosion.

But if real explosions are not spherically symmetric, then your analytic model is wrong---and quite likely "silently" wrong, not accompanied by any error term or crosscheck that flags the result as more or less reliable.

A computational model contains the laws of physics which you thought to include. (It's much easier to get this list right than it is to build an analytical model.) The simulation output then shows you all of the phenomena that will happen given the time- and granularity- limits. The nice thing is, the effect of those limitations is frequently testable---convergence and so on.

Tokamaks, for example. It's easy to build an analytic model of a tokamak that makes them look easy. "Hey, the plasma is going around like X, and you scale this parameter Y until the breakeven equation says Z=1. Done." But the plasma isn't going around like X.

 

[rwguinn]

It's worth clarifying: no, it's not quite. An analytic model includes the phenomena which you thought to include. If you think you're got a spherically-symmetric implosion, like if it were 1970 and you were daydreaming about building NIF---you can write down an analytic description of the temperature/pressure/fusion rate as a function of time in a spherically-symmetric implosion.

But if real explosions are not spherically symmetric, then your analytic model is wrong---and quite likely "silently" wrong, not accompanied by any error term or crosscheck that flags the result as more or less reliable.

A computational model contains the laws of physics which you thought to include. (It's much easier to get this list right than it is to build an analytical model.) The simulation output then shows you all of the phenomena that will happen given the time- and granularity- limits. The nice thing is, the effect of those limitations is frequently testable---convergence and so on.

Tokamaks, for example. It's easy to build an analytic model of a tokamak that makes them look easy. "Hey, the plasma is going around like X, and you scale this parameter Y until the breakeven equation says Z=1. Done." But the plasma isn't going around like X.

Let me qualify:
An analytical model is a simulation of the conditions you know or think you know for the event of interest.
Nothing says it has to be a CORRECT simulation or analysis.
.
There is more to technical analysis than just transient events