binarycoffee

I am not sure I understand the objection (or maybe there wasn't one?) but since there seems to be some confusion in this thread on this point, the electrical efficiency of electric thrusters is very well defined, it is:

η = ½Q·(Isp·g₀)² / P

where Q is the mass flow rate (kg/s) and P is the input electrical power.

Hall thrusters have efficiencies typically ranging from 30% (small HTs) to 60% (large HTs). Gridded ion thrusters have slightly better efficiencies (add 5-10%), mainly because they operate at higher Isp (they would actually have worse efficiency than Hall thrusters when operating in the 1000-1500s range).

Why I object to the (admittedly common) characterization of Isp as an efficiency: it is always better to have a high efficiency, whereas Isp is always a trade-off between mass budget and power budget so optimal Isp is always mission-dependent.

Not sure what was meant in the article, but Isp is not the same as efficiency. Also, high Isp is not desirable for a power-limited spacecraft such as cubesats because the input power scales with the square of Isp.

The efficiency of an electric thruster is a measure of the losses related to ionization energy, beam divergence, beam temperature etc. In theory Isp and efficiency are therefore orthogonal, though Low-Isp devices indeed tend to have a lower efficiency due to higher ionization-to-kinetic energy ratio.

So this is not what the article is about.

It discusses a new method to generate integer values according to an arbitrary probability distribution, using as input a uniform random generator. Whether the input generator is truly random, pseudorandom, cryptographically secure etc, is irrelevant: the output will presumably only be as random/secure as the input RNG.

Admittedly, the article does a poor job at explaining what the FLDR method is about, and it looks as biased as the method itself (sorry for the easy pun). From my understanding of the paper [1], the method is better than the state-of-the-art Alias method [2] only when the entropy of the target distribution is low. When entropy is high, it performs similarly (or may even be a bit slower) but uses up to 8 times more memory space.

[1] https://arxiv.org/abs/2003.03830

[2] https://en.wikipedia.org/wiki/Alias_method

I fear GP's post has not been fully appreciated.

(I noted these words: [...] embrace [...] extend [...] extinguish [...])

Quite the opposite I would say: the harmonic mean can be seen as a very good justification for `1/0=inf` as the pragmatic choice.

If, say, `a` is mathematically very small (and `x` is accordingly expected very small`) but `a` become zero due to rounding behavior, then having `1/a=inf` results in `x=0` which is arguably closer to the expected result than e.g. `x=2b`.

As someone involved with numerical methods, I have very often relied on `1/0=inf` because it would ultimately warrant the proper behavior in case of rounding errors in the denominator.

Yes, operating a Hall or ion thruster on O2 and/or N2 is not new (and Busek was not the only one that did it either).

And it is definitely hard, but this is just part of the problem.

The really tricky thing is to get the whole system working. For instance it is not longer possible to use a "high pressure" (relatively speaking of course) propellant feed. Even with the passive compression stage discussed in the article, an off-the-shelf thruster wouldn't be able to operate due to the low inlet pressure.

Also, the intake/collector design is a problem of its own. AFAIK this was the first real-life test for the passive intake concept (the theory of operation and trade-off to be considered are discussed here [1]).

That being said there is still a long way to go before this can work in space...

[1] http://erps.spacegrant.org/uploads/images/2015Presentations/...

Yes, this is related (not the author of the cited article but I know him well).

There were also several precursor experimental studies funded by ESA, one of which [1] can be found in the same conf. proceeding.

[1] http://erps.spacegrant.org/uploads/images/images/iepc_articl...

Somewhat related is the E-sail [1] concept, a perhaps less ambitious but (probably) feasible idea to harness the momentum of solar wind particles with very long charged wires.

[1] https://en.wikipedia.org/wiki/Electric_sail

True, interesting idea.

[Long edit]

Thinking further about this idea, I realize this may even mitigate the catch 22 problem of very low orbits (<180km): the lower the orbit, the larger the drag and the required thrust power, meaning the solar arrays must be bigger, which in turn further increases the drag... Calculations suggests that with current solar array and thruster technology, flying lower than 150km with this concept is impossible.

But with an elliptic orbit, energy from the solar arrays can be stored on the low-drag portion of the orbit too and used during the perigee dip, thus decreasing the requirements in terms of solar arrays area.

Absolutely, this is a passive compressor.

It kind of works like an aerodynamic diode: it is much easier for the incoming particles to go through the intake tubes (because they are oriented along the spacecraft velocity vector) than to exit the collector. This is because after a they collide with surfaces inside the collector, their velocity vector is randomized and no longer aligned with the tubes.

[Edit] Mass is not such an issue when you contrast it with the mass of propellant that you save...

[Edit2] I forgot to precise that my description only pertained to the intake part: there is of course a plasma thruster at the back (no moving part either)

This is still very much sci-fi at the moment for anything flying above a 250 km earth orbit because atmospheric density decreases dramatically fast with altitude.

The missions targeted by this technology are GOCE-like spacecrafts which by design must fly low and need an insane amount of propellant to compensate for the high atmospheric drag at such altitude.

This is a customized Hall thruster.

Disclosure: I was involved in that project.

This is correct but incomplete: Vasimr may have a higher power density (be it power per thruster mass or per thruster volume) than ion thrusters but I am pretty sure it has lower power density than a Hall thruster, while being also much more complex.

The main (and not yet verified) claim of Vasimr compared to existing and proven tech like Hall thrusters is the ability to greatly vary the specific impulse.

I guess you mean the power-to-thrust ratio, which indeed scales roughly with the specific impulse.

The thing is, pretty much every plasma propulsion device has the theoretical ability to throttle specific impulse. Gridded ion engines and Hall thrusters can for instance directly throttle specific impulse (Isp) by varying the discharge voltage. Other plasma thrusters can do this by other means like increasing RF heating, varying the magnetic field, etc.

Unfortunately, in real life the actual usable range of Isp always ends up much narrower that initially predicted because of a variety of factors that are difficult to predict theoretically (erosion and ionization efficiency in particular).

AFAIK, due to the shear size and complexity of Vasimr, its thrust has never been directly measured and we know very little about its actual performance. I would therefore wait a little and discard sensationalist articles until real, extensive testing has taken place.

fac = faculté, basically a French university.

I guess what the parents refer to is that Universities in France are not as elitist as in other countries; the most prestigious higher education public schools are often considered to be the so-called "grandes ecoles" (engineering schools) or some specific administration/economics schools.

Several comments have hinted at the possibility that this would be a over-unity device but as I couldn't find a specific demonstration, here is one that does not involve any fancy physics.

Let's assume we have a spacecraft powered by an energy source with total energy `E` and mass `m`. The EM-drive produces a constant thrust `F` at a constant power `P`.

Kinetic energy of the S/C after time `t`:

    Ek = 0.5 * F^2 * t^2 / m

If `t` is the time at which the energy in the source is exhausted, then the ratio of kinetic energy developed to the expended energy (E = t * P) is:

    Ek/E = 0.5 * F^2 / P^2 * (E/m)

So this basically says that we have a over-unity device whenever the energy source has a specific energy above:

    E/m > 2P/F ~ 1.6 10^6 J/kg

which is certainly not achievable in practice, but in theory is totally fine being still well below c^2 (which by the way justifies a-posteriori the neglect of relativistic corrections in the above).

easy: 1000000's of them rotating around a fixed point!

Interestingly, NASA used to have a group working on such projects.

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

And I believe they actually did a serious job, which is probably why they never found any warp-drive ;-)

I doubt Mr debunkit is from TU Dresden or he would probably have posted the link to the article first (though it could have been a ruse of course).

In any case, I am not from TU Dresden and have only a passing acquaintance with Tajmar, but I do agree that credit and recognition is not going where it should in this case.

Hall thrusters typically operate around 50mN/kW. Gridded ion thrusters a bit lower, like 30-40mN/kW. That being said, some exotic thrusters like the FEEP still manage to be useful operating below 10mN/kW for precise positioning.

[Edited erroneous data for FEEPs and needless rant]

True, that could have been brought up with greater diplomacy.

That being said, I can understand the over-reaction, it is indeed frustrating to see relatively weak articles making front page because they originate from NASA when earlier, more thorough research did not get any attention because their authors were more careful with their claims.

Mind you, I am not saying the Eagleworks paper should not have been published; any data point is good as long as the experiment is sufficiently described for it to be reproduced or disproved, and in that respect I think it is OK. But very good it is not.

Yes, sadly grand-parent appears to have been downvoted for posting one of the most insightful comment so far.

To be fair, the authors did enlist Tajmar's paper in the state-of-the-art but they do not discuss despite it being the most relevant prior art.

For anyone who knows a little bit Tajmar, if there was a tiny hint of a possibility that this effect could be real, he would definitely have pursued it; after all, his self-admitted goal is nothing else but to discover the warp-drive... That being said, he is no crackpot and the thoroughness of his article confirms it.

Incidentally, in a similar HN story a couple of days ago I shared my suspicion about spurious contributions from thermal drift, and unsurprisingly I see this was also the number 1 suspect for Tajmar. Beside using thermal barrier to mitigate the problem, he also went from electro-magnetic damping to oil damping and found completely different results. Meanwhile, the Eagleworks guys are still using an EM damper, working right next to a strong EM field source.

As an electric propulsion engineer used to measure thrust in the uN range, what disturbs me much more than thrust levels close to the sensitivity threshold is the fact that thrust is only detected some 10-20s after the RF is turned on (figs 9,8,13). I am rather confident that there is no physical time scale in the RF generation process that would account for a 10-20s ramp, and this would contradict their own report of the fast RF power stabilization. It definitely looks like thermal drift; in fact, thermal drift is _always_ the first thing to suspect when you get strange measurements with a low thrust balance.

It appears as well that the authors have conveniently prepared an answer for this objection with fig.5 which attempts to mislead the reader by suggesting that the very slow evolution of the balance displacement is to be expected as an artifact of the combined effect of thermal drift and RF pulse onset. But conveniently, the time units are arbitrary so that hopefully nobody will notice that RF onset was assumed in that figure to have exactly the same characteristic time as the thermal drift... When all measurements actually show that RF power stabilizes much faster.

I can recommend "Exploring BeagleBone" by Derek Malloy which offers a very gentle introduction to the PRU with some mildly advanced examples. Also, a shameless plug with more PRU code examples in case you are interested to use Rust rather than C for the glue code: https://github.com/sbarral/prusst