Observatory Version of Misc Musings, Ravings, and Random Thoughts

[Dmytry]

I wonder which alpha-sensitive GM tube to substitute in place of the SBM-20?

Alpha particles are so 'big' you can detect them with few wires in air:

http://www.youtube.com/watch?v=HO8VDf_rCXo

also, ion chambers are pretty easy to make, i think it can count the alpha particles individually if the voltage is high enough:

viewtopic.php?f=26&t=1171162

 

[Dmytry]

Well i think it's the world that is being nice to physicists. In physics, something made of many atoms often follows simpler rules than individual atoms do.

In biology there was natural selection to try to make better use of the available complexity - something big and made of many cells can at start act simpler than a single cell, but then it will evolve and make use of the complexity it can attain with many cells. Basically, it's like, if physics was software, you could optimize it a ton by not calculating individual quarks n stuff, and still be close enough even at Newtonian physics level of complexity, and this allows us to, having limited computing power, make really cool technology nonetheless. For biology you can't do that so easily, there's minimum amount of calculations for finding what shape protein will fold into, and this minimum amount is not very small.

 

[Dmytry]

Something I've suspected for a while is that male pattern baldness serves a social function (like silver hair in gorrillas). It seems to neat, orderly, and consistent a pattern to not have some evolved purpose.

I've heard something along those lines. http://en.wikipedia.org/wiki/Baldness#E ... hypotheses

Few tens thousands years ago and back, anyone who was old enough to be bald, had badass genes, but also lower capacity to care for children, so one can imagine that there would need to be a honest-ish signal for that.

 

[Dmytry]

You can hide the high voltage wires behind metal mesh, or metal mesh + thin plastic wrap if you need it sealed from moisture. Regarding compact, you can make it in what ever size you need, with obvious ramifications (smaller size, fewer alpha tracks to detect).

 

[Dmytry]

So, something i was thinking about: inadequate application of scientific method when it comes to psychological conditions.

Many common medications used to deal with psychological conditions have only a very small effect over placebo; it seems that only a small sub population of those treated actually have a condition that is improved by medication, with the rest being helped equally well by half-placebo - the placebo effect you get when you are made aware there is 50% chance you are in the control group that is receiving placebo. That should not be surprising as it is known that the conditions in question often have non-chemical cause (i.e. some distressing information or imagery) and are conjectured to only sometimes have a chemical cause. Essentially, it is like taking anti-flu medicine for the cough that is usually caused by dust rather than flu. Two distinct disorders with same symptoms, and the medication only helping with one of the disorders.

Furthermore, in presence of side effects, the people experiencing side effects are more certain they are not in the control group, and subsequently are subject to stronger placebo effect, so one should expect a small above-placebo effect even in the drugs that have no effect on the actual condition, just as long as their side effects are sufficiently strong. (One could use toxins as placebo, but then there is the problem with placebo potentially worsening the original condition).

Through the entire history of medicine the most popular remedies have been complete bullshit; the improvement over bullshit came around, largely, from greater understanding of the disease allowing us to cull out the obvious bullshit even before the testing; we do not understand the workings of intelligence very well though, and subsequently some areas of medicine should be expected to stay well behind. If you look at history of medicine and go back as little as 100 years, or even 50 years, you find an enormous amount of what is now recognized as bullshit.

edit:formatting.

 

[Dmytry]

All the ground based electronics has to handle muons. The space based electronics has to 'handle more particles', but it is still single-particle events, just the tolerance on the glitch rate is stricter. Something that will glitch in orbit will glitch on the ground (and on airplane flights, just less often. The error correction codes and redundancy works for anything that has to be really really reliable, and just flying latest tech as is works fine for everything else like ipods, personal digital cameras, etc. A lot of the stuff that you get up high but not on the ground is not very high energy.

As of why stuff in space tends to be quite a bit older, legacy software for the most part, and such. There's very little advantage in upgrading, and many potential risks - and they need to test the stuff and deem it good for space, etc. The control systems aren't going to work better because they are running latest hardware, they're running on hardware that's already much more than fast enough (by a factor of thousands), it isn't the latest videogames they're wanting to run. The only point to upgrade would be saving power or weight (very little power and very little weight). Say hello to expensive redesign of electrical systems. Ain't going to be done every year.

 

[Dmytry]

Probably just fine after the trip. Might have been glitches while it was in orbit, but the environment isn't going to produce any permanent damage (that takes quite a bit longer). Presumably iPods are designed well enough that they'll just crash and/or shut down rather than exploding due to random bits flipping or gates latching.

Edit: Assuming this isn't a design feature...

Most probably it got destroyed in space.
Electronics are really difficult to design in space. And the most troubling issue is thermal heating. Due to the lack of gravity no natural convection can build up what-so-ever. Electronics will simply sit in a bubble of heated air with no means of shedding heat. And the air temperature will increase ever more until the electronics are fried.

Radiation is most likely not an issue for an IPod. It doesn't matter if a few bits of music are flipped. And standard communication protocols have error correction already built in.

I would expect thermal shutdown, dunno though how well those things are engineered (i'd think the lithium battery has electronics that disconnects it when it is too hot). Also there's some forced air circulation there. The regular laptops are known to work fine in space.

 

[Dmytry]

Steve Pinker also believes in a zillion of cognitive modules that somehow evolved (very rapidly) to do very specific cognitive functions, in presence of training, that can demonstrably be done without those modules. I.e. non-falsifiable by design evo psych crap with zero predictive power.

 

[Dmytry]

Steve Pinker also believes in a zillion of cognitive modules that somehow evolved (very rapidly) to do very specific cognitive functions, in presence of training, that can demonstrably be done without those modules. I.e. non-falsifiable by design evo psych crap with zero predictive power.

Does this make his analysis of the trouble with Group Selection wrong?

I'm neither for nor against Steve Pinker. I'd never heard of him. The article linked in this thread is the only work of his I've ever read. However, it has its citations, and it seems accurate in its analysis and discussion. In this particular case, does it really matter if he is a crackpot about other things?

Well, basically, group selection undeniably has to happen to at least some extent (groups can die out), but the rate of group selection needs to be estimated, and may be very slow in all practical circumstances or drown in drift. Finding out how much it works is a task for applied mathematicians doing applied mathematics on many models (or stats on data), not for Pinker talking about what is necessary to explain what, greek philosophy style, bravely applying Occam's razor to cut off things you ought to calculate. (And this also is related to evolutionary psychology where real frigging rapidly whole new organs of cognition evolve despite the same function appearing if brain is damaged and other part is being used. This part is a job for neuroscience and again mathematics). Crackpottery is about not knowing what it takes to figure things out. In physics, people who try to verbally argument their way towards theory of everything (w/o mathematics) are that guy with i have theory of gravity thread. Biology is in a way much more complex topic and the method that fails physics fails biology even worse, but people who try to verbally argument their way can have credentials as biologists.

 

[Dmytry]

Well, basically, group selection undeniably has to happen to at least some extent (groups can die out), but the rate of group selection needs to be estimated, and may be very slow in all practical circumstances or drown in drift.

Groups are not coherent selectable units. They change and shift as members come and go, are born and die, as groups merge and separate. Do you count group selection as occurring if a group "dies out" by disbanding, even if no individuals die? How about if a group (as a selectable unit) "dies out" by being merged into a larger group?

Quit with strawmans

If the unit of selection is the group, how do you define the boundaries of that unit upon which selection is to act? Even with the difficulties of defining exactly what a gene is in gene selection, at least there is something workable with units of genetic material at a certain locus that moves through the generations as a coherent whole. Since groups don't have such replication and inheritance, what exactly is selection acting on in Group Selection?

Verbal irrelevance of precisely the kind that I do not like Pinker for. The group selection is e.g. along the lines of 'there are small tribes on that land, and they don't take in strangers. Tribes where people were assholes to each other performed worse and fared poorly in competition with other groups'.

Group selection doesn't explain anything that can't be explained by individual selection / gene selection, and those explains much more than group selection is capable of.

You say Group Selection undeniably has to happen, and I say I'm not even convinced there is such a thing as group selection. :eyebrow:

No, you're just making irrelevant vibration of the air like Pinker does. Of course the group selection happens through selection of genes that are there, so what, 'genius', the proponents of group selection obviously don't know this, and you know, you're so smart¸reward activated, love Pinker by association. This is what I (and many others) don't like Pinker for. He's making those irrelevant vibrations, and people read those, and love to themselves make such vibrations, because it sounds like the biology is about such irrelevancies and about going on 'there was no definition i am allowed to make a strawman therefore its false', like frigging philosophy. People love that because it's something they can engage in very easily without knowing anything or doing anything hard. You can just sit around and feel the marvellous self which manages to easily reject, with ultra easy thinking, what many smart people don't find so easy to reject. *Spits*.

Which is a topic I want to raise: his sort of thing seem to happen a lot. Are there any consistent studies of why humans do this so much? I mean, loving the stuff which lets you 'solve' complex stuff by some sort of medieval scholastic argument.

 

[Dmytry]

WTF are you on about? Love Pinker by association because what I read of his for the first time ever happened to match what I already knew from evolutionary biologists for quite some time? Yeah, makes perfect sense. I don't really give a damn about your hatred for Pinker. Attack the ideas on their own merits or piss off. This is a science discussion, not Ad Homs R Us.

You want group selection to be in the mathematical models? I'm asking you to define your variables.

No, seriously. What exactly do you mean when you say "group selection", because based on your nonsense reply, I'm not even sure we are talking remotely about the same thing.

You are speaking of a strawman of some kind, so no we are not speaking of the same thing, and you are not speaking of the same thing as any group selection proponent. Group selection refers to selection of behaviours beneficial for joint survival of members of small groups of genetically related individuals in the conditions where groups compete with each other. It acts on traits controlling the joint behaviour of groups, and it is reducible to selection of genes (of course), like thermodynamics is reducible to atoms bouncing. Also if you want to discuss science, talk science, not argument over definitions where you just make some (blatantly obvious) strawman.

edit: actually, wikipedia is not too bad, you can just read this: http://en.wikipedia.org/wiki/Group_selection if you are confused about group selection after reading the strawman Pinker makes of it. And I'm not saying 'group selection is true'. It would be slow, and it's effect may easily be negligible in the circumstances in which it was asserted to happen. The issue is that the extent to which it happens is a mathematical problem. And the concept might not be very useful (certainly less useful than thermodynamics). But it can't be disposed of with the sort of arguments that Pinker makes. It was largely disposed of by models showing it to be negligible.

 

[Dmytry]

Ok... We are discussing two different interpretations of the phrase "Group Selection".

Your use is apparently the Multi-Level Selection idea, correct? (which, according to Stuart Smith, is mathematically equivalent to Inclusive Fitness Theory; link below) Or just a plain "competition between groups" meaning? Meanwhile my use was more the Wynne-Edwards version of selection for traits to benefit the group regardless of the cost to the individual. Since both concepts are equally referred to as "Group Selection", there confusion arises. My apparent "strawmen" are a direct result of this (as those "strawmen" are valid issues with Wynne-Edwards group selection).

If you have about 20 minutes to spare, Stuart Smith gives a great talk on this. Do you have any issues on his views of the lack of usefulness of Group Selection (in its many confusing uses)? (he doesn't refer to Pinker at all, so you can save any comments about Pinker love)

Or, if you don't have time for video, Jerry Coyne discussed the talk back in 2010 on "Why Evolution Is True".

As far as my understanding of MLS goes, as well as my understanding of Inclusive Fitness, barring further reading of the literature on my own, I agree with the viewpoints of Stuart Smith on this.

The confusion arises in the a: people whom don't understand reductionism, b: makers of strawmen, c: pop science book writers, d: pop science book readers. (some people fall into all 4 groups, some into first 3) . Regarding the usefulness of the concept, in general higher level concepts that summarize low level behaviours are useful, or we'd be modelling the airplanes on computer at the individual quarks level. (That being said, it is nowhere near as useful as approximations in physics, and perhaps we are better off modelling evolution at individual gene level, not needing to have any less detailed models. But I can see that if we e.g. are to model evolution of bacterial films and the like, it may be necessary to employ a more coarse model).

 

[Dmytry]

Pinker is *not even making any right mistakes*. You need calculations and mathematics to rule this out, not sophistry along the lines of 'how useful it is', 'how necessary it is to explain whatever', etc. The mathematics-phobic handwaving-philes love that sort of stuff.

 

[Dmytry]

Pinker is *not even making any right mistakes*. You need calculations and mathematics to rule this out, not sophistry along the lines of 'how useful it is', 'how necessary it is to explain whatever', etc. The mathematics-phobic handwaving-philes love that sort of stuff.

You know, while mathematics is undoubtedly the most awesome tool humanity has ever devised for formal reasoning, it's not actually impossible to reason clearly without involving mathematical formalisms.

I can see where you're coming from with your “if it's not mathematics, then it's worthless” position, but it's wrong.

Yes, rigourous mathematical analysis of a suitable model of selection would be useful in identifying predictions which might be falsifiable. No, that's not the only possible mechanism of thought.

The sophistry along the lines of how useful a concept is, creation of strawmen and arguments over definitions (the sort of stuff that Pinker does which supposedly could constitute an argument about some factual matters), however, is NOT among the set of mechanisms of thought. edit: That is to say, the question is a quantitative matter. It's not so bad that you could reject it with handwave without ever manipulating any quantities.

On top of this, he's the evolutionary psychology guy, part of the modular mind crowd that would explain superior performance on Wason selection task with social situations by existence of some evolved 'social cognition' module.

 

[Dmytry]

Basically, you don't put the group selection into mathematical model, you put the elementary acts of survival and death of individuals with genes, and then you see how strong is the group selection effect in various conditions, and then you know where to look for the group selection in animals, and can see if it actually happen.

The question is whenever gene selection implies group selection in some circumstances, and the answer is quantitative, i.e. requires evaluation of quantities. If the 21th century mathematics actually didn't work there, the prehistoric grade mathematics of "one, two, a few, many, very many" which is used in language, complete with extra fallacies, would not work either.

 

[Dmytry]

This reminds me of a half baked idea I had: build very small mechanical hands (perhaps 1/5 or 1/10 the size), controlled from a glove, with tactile feedback (For the glove, I picture a glove with a lot of cables that work like bicycle brake line (steel cable in a shealth), connected to a box with reduction levers, with force-activated brakes to amplify the tiny feedback forces of the tiny arms). Would be awfully hard to build. Like a swiss watch. But such an awesome project to keep obsessive at bay for entire year. I might try making that when I settle down and get the tools.

 

[Dmytry]

Today I discovered http://what-if.xkcd.com/. It is awesome. You (yes, you) need to read both Relativistic Baseball and A Mole of Moles right now.

++

Both excellent articles. Because, really, what self-respecting nerd has never wondered what a mole of moles would be like. :bigdumbgrin:

Yeah, I really like the xkcd What If? stuff.

Hmm. I wonder what temperature it would have in the core. May actually end up quite hot, I mean, if you somehow started with a ball of loosely packed moles, there's a lot of potential energy released via just packing the moles. edit: also the water and organics are going to react under high pressure and temperature, probably releasing some hydrogen.

 

[Dmytry]

Hmm. I wonder what temperature it would have in the core. May actually end up quite hot, I mean, if you somehow started with a ball of loosely packed moles, there's a lot of potential energy released via just packing the moles. edit: also the water and organics are going to react under high pressure and temperature, probably releasing some hydrogen.

Mole oil! Moleoleum!

Would it be oil, though, with this much water? I need to calculate how hot it gets packing the moles but I am estimating well above 1000 Celsius.

 

[Dmytry]

Gravitational binding energy of uniform-density sphere is 3GM²/5r, so the binding energy per unit mass is 3GM/5r.

From http://what-if.xkcd.com/4/ M = 4.52x10²²kg and r = 2.21x10⁶m, so the binding energy per unit mass is about 819kJ/kg.

Heat capacity of water at RTP is about 4.2kJ/kgK, so doing the division that's about 200K on average, not allowing for the heat capacity to vary with temperature/pressure and ignoring the moleish impurities or decay.

Assuming the incompressibility of moles the central pressure would be 2πGρ²R²/3 = 6.8x10⁸Pa, so about six times the pressure at the bottom of the Marianas Trench.

Edit: I should add that the last few moles to accrete onto Planet Mole would impact at about 1.6km/s.

Hmmm. That's way colder than I expected. Should start some interesting chemistry, though.

 

[Dmytry]

I just measured radon daughters in the rain, using an ion chamber... never did it before. If I didn't know what it was I would have so freaked out for the duration of a few half lives of that stuff. It's almost surreal - you just wipe the roof with a paper towel, then you get yourself a seriously radioactive paper towel, for a few hours (Mostly alpha and beta though). Much more radioactive than you'd think it'd be. Geiger counters also go nuts about that stuff (I seen a few youtube vids). You can do such an evil prank on someone who doesn't know about that. Get them to wipe fresh rain off something, then measure their hands.

edit: Actual science: I put it there - I have an ion chamber with a mesh to let alphas in. And I expect to get puny response on par with potassium chloride, I mean, the paper towel is small. I expect well less than 2fA, probably less than 1fA , something I'm not even sure it's there at all, something that I'd need long averaging time to detect . I get >5 fA , and it's very discernible from background radiation, at first I even thought it was humidity affecting the ion chamber. And that's with very low chamber bias voltage of 2.5 volts where a lot of ions may be recombining. 5fA is 31 000 electrons per second. edit: now, in terms of number of alpha particles in the chamber per second, of course that's not much, but keep in mind that most of the alpha particles from stuff on the paper towel are not going to even get out of the paper towel.

 

[Dmytry]

Alamout[/url]":1fwkb7kz]The HuffPo article is pretty terrible, but that "better news article" is even worse. The writer--apparently bursting at the seams with jealousy and feelings of inadequacy--insinuates a bunch of conspiracy shit that most people would rephrase as "very smart parents help their very smart daughter succeed in science." When an 18-year-old does something like this, there's no reason to assume her parents did all the work--they did the work of raising her, they didn't do her science project. That article reads like it's covered in a thin layer of slime.

Yup. Agreed. There's no reason why 18yo would be unable to do this (It's very impressive, of course, but it's not something with an age cut off). I almost ended up working as assistant at a laser optics lab at 15 or so but couldn't because of health issues and distance from where I lived (my family was very poor). I was also messing with ultra capacitors coz those are cool, but of course, lab less, there's not much interesting stuff you can do beyond experimenting with salt based electrolyte and activated carbon electrodes.

 

[Dmytry]

Adding a copper loop around the magnet might help some, by slowing the magnet down. It's a very weird device, though - I would of placed a hall effect sensor over the magnet, and not have it bang on any screws. With a screw nearby the magnetic flux through the hall effect sensor would change.

 

[Dmytry]

AFAIK rather than having each qubit interact with each other qubit (which is what makes large number of qubits problematic), in their circuit each qubit interacts with several neighbouring qubits in a lattice (grid?), and that's all. Subsequently it can't do the cool things that 512 qubit computer would do. Namely it doesn't do anything in the space of 2^512 .

 

[Dmytry]

ANSDAC[/url]":2847r89z]When oil runs up a paper towel, where does the energy come from?

Attraction of molecules. Basically, oil in touch with the towel has lower energy than oil in touch with the air, or oil in touch with itself. It's mostly van der Waals forces. Oil is non-polar so the forces between oil molecules are weak.

 

[Dmytry]

How durable are the implants? I was for some reason under impression that artificial teeth were - over time - less strong than the natural ones, as the natural ones repair themselves to some extent. edit: apparently it is better now:
http://en.wikipedia.org/wiki/Dental_implant#Failure
but still pretty bad

 

[Dmytry]

Yeah, those qualitative assessments of "evolution does this" are, frankly, ridiculous. There's a lot of evolutionary pressures working in different directions, and the combination, while it may look qualitative (monogamy or not) requires quantitative calculations, not "ohh there's a story why a is advantageous".

 

[Dmytry]

truth is life[/url]":232rfa5a]

Now, there are situations where radiation exposure can lead to "activation" of materials, making them radioactive themselves and hence dangerous to be around. But this occurs under special conditions of high neutron bombardment, and neutrons aren't a significant component of cosmic radiation fluxes, just the radiation produced by nuclear weapons (hence the so-called "neutron bomb").

Actually, any particle that's energetic enough can cause activation, even gamma rays. Neutrons are special because they can do it at low energies. The first time induced radioactivity was detected, it was with alpha particles and aluminium.

 

[Dmytry]

I was wondering the other day, to which extent various psychological trauma can be driven by expectation of said trauma or perceived uniqueness of the experience? Along the lines of nocebo effect. Are there any good studies of such effects?

 

[Dmytry]

NavyGothic[/url]":3pumoa4z]

Dmytry[/url]":3pumoa4z]I was wondering the other day, to which extent various psychological trauma can be driven by expectation of said trauma or perceived uniqueness of the experience? Along the lines of nocebo effect. Are there any good studies of such effects?

This is slightly different but (possibly?) related;

http://www.theguardian.com/science/2013 ... logy-worse
http://www.ncbi.nlm.nih.gov/pubmed/12076399 (the study referenced in the article above)

Single session individual debriefing did not reduce psychological distress nor prevent the onset of post traumatic stress disorder (PTSD). Those who received the intervention showed no significant short term (3-5 months) in the risk of PTSD (odds ratio 1.22 (95% ci 0.60 to 2.46 )). At one year one trial reported that there was a significantly increased risk of PTSD in those receiving debriefing (odds ratio 2.88 (1.11 to 7.53))odds ratio 95%). There was also no evidence that debriefing reduced general psychological morbidity, depression or anxiety.

The effect certainly exists for universal trauma counselling, so it wouldn't surprise me if the same were true for general awareness of PTSD and the like. Difficult to judge since "awareness" is something that's hard to control for, and it also wouldn't necessarily prove that said awareness is a bad thing; it's entirely possible that any increase in harm caused by over-awareness is less than the benefit gained from acknowledging PTSD sufferers.

It'd be better if sufficient data could be gathered on the suicide rate (to get rid of co-founding by the rate of reporting). The issue with 'debriefing' is that the psychologist is human, and humans, when there's no objective measure of how well they do their job, do a bad job or make things worse.

 

[Dmytry]

So, I came across a new anti-science anti-vaccination talking point... The argument is that at the confidence of 95% , i.e. 5% probability of a false positive, about 1 in 20 studies should find a link between vaccines and autism. But no studies listed in meta reviews as sound find such a link, in the pool of 40, therefore scientists must be covering something up.

My understanding is that while some simplistic set ups would provide exactly 95% confidence, in real studies all sorts of complications which can raise and raise the true confidence above 95% , which is perfectly okay as the standard is >95% anyway. Is that basically correct?

 

[Dmytry]

Alamout[/url]":sbck0w2y]A p-value cutoff is 0.05 is not the same thing as a false positive rate of 5%. p-values are tricky devils and are woefully misunderstood (and constantly abused).

In any case, let's pretend that we do expect 1 in 20 studies to find such a link, and furthermore that these studies are immediately published and don't need to further substantiate their claims (any such study is going to need much better evidence than "p < 0.05").

The probability of flipping a p = 0.05 coin 40 times and getting zero heads is...~12.9%. So we fail to reject the null hypothesis, which is that there is no cover-up. It's like a microcosm of their inability to understand science.

They just say that it is still 'substantial evidence' that scientists aren't publishing those results. The most annoying thing about the anti vaccination is that it's well off privileged people who peddle that crap.

 

[Dmytry]

Alamout[/url]":1zqgvd8x]Crazy news: a principal investigator behind the recent STAP/stem cell scandal appears to have committed suicide.

For those not following along: they published a high-profile paper in Nature, claiming to have created adult stem cells using an extremely simple procedure. No one could reproduce it, and an investigation concluded that the work was fraudulent and they eventually agreed to retract the paper.

Apparently they were considering shutting down the whole research institute. While suicide is always an extreme decision, it sounds like this guy's career in Japan was completely ruined. What I am totally flabbergasted by is that they thought they could publish the fake research in the first place.

This reminds me of a conversation I had recently with some anti scientific nutters. Basically they point at the fairly high "hypothesis confirmed" rate among published research, and their favourite explanation is that research must be fraudulent and if someone found that vaccines caused autism blah blah we wouldn't hear of it. But the truth as I see it, fake research with an impact gets overturned quite well. There may be a few problem areas (apparently there's very few replications in some branches of psychology, and when replications are attempted, a lot of research is not replicated), but ironically this research is the sort of thing said nutters almost religiously trust. (Not to mention that this whole 'set a hypothesis, then tell if its confirmed' is a gross oversimplification, and it is very common that rather than testing a hypothesis, a measurement of a value is performed and reported, or a working technique for doing something is reported, or the like).

 

[Dmytry]

I were looking into the ball bearing motor (prompted by the recent article about classical physics), and apparently it is still not settled what exactly makes it spin. There's the explanation from contact heating, which seems like it would be very sensitive to how tight is the bearing, and thus seem very unsatisfactory.

There's an explanation where the conductive rotating balls in the circular magnetic field produce a radial component which then works on the shaft as in homopolar motor, which does seem sensible but would require careful analysis to make sure that the opposing torque on the balls is not braking the rotation more than the forces in question. Then there's the problem that it only seem to have been tested with ferromagnetic bearings, which adds hysteresis to the equation, making everything even more complicated. It's like one of those puzzles my physics teacher would present the class with, except hellishly more evil.

 

[Dmytry]

crazydee[/url]":3502b0ri]The thermal people seem to have a very strong point if it does, as claimed, run equally on AC or DC.

Well, if it has no permanent magnets, and runs electromagnetically, we should expect it to run on either AC or DC (the same way an universal motor does), because all magnetic fields will reverse on AC.

edit: I think it could be useful to try making one with a pipe instead of the central rod, and another rod in the middle, so that the magnetic field of the rod can be cancelled out, or the bearing motor could be spun while the current is supplied to the middle rod, to see if that works as a generator.

 

[Dmytry]

crazydee[/url]":19cke1xx]If the AC is a symmetrical wave form then presumably is could only run at a harmonic of supply frequency - so it should be easy to prove this by changing frequency, but maintaining power.

I dunno, everything in the motor is round so it's not clear what should move at the harmonic.

The thermal explanation seem plausible. Each ball bearing is acting as a spring releasing the tension caused by thermal expansion; except the spring in constantly being "re-compressed" by the thermal expansion of the next part of the ball.

The reason it seems unsatisfactory to me is that it should depend a lot to the pressure on balls in the bearing. Also, this thing reaches thousands RPM on the shaft, tens thousands on the balls, so the hotspots got to be very small for them to cool down in time. But if they're very small (micrometres) the size of resulting thermal expansion would be in the nanometres range.

 

[Dmytry]

crazydee[/url]":20z2eb0h]
Looking at the experimental set-ups online with the races in vices I can believe you have to crush them till every nanometre counts.

If it is electromagnetic then getting one going on DC them reversing the polarity should be trivial and either stop or reverse.

I don't think that's true. If I make a homopolar motor using a coil instead of the permanent magnet, it will not stop or reverse if polarity is reversed, because both the stator field and the current in the rotor will reverse. If a contraption doesn't rely on permanent magnetization and the only source of the fields is the electric current, and everything is linear (and ignoring mass of the electrons), it shouldn't be possible to identify direction of electric current with it, because if you change direction of the current, all fields reverse and all forces double reverse (i.e. stay the same).

Speaking of shape change of the balls, there's also magnetostriction to consider...

 

[Dmytry]

(Sits rotating left hand around different axes) I think I see that; if the field in the stator is induced with DC then the field has to be reversed when the current is reversed. A "standard" battery cylindrical homopolar motor reversing the current and the field preserves direction of rotation; to reverse rotation you need to change only one.

Seems fine visualising the shaft as a magnet and the balls as current paths perpendicular to that field. Though that suggests that bearings with exposed races should work better than ones with ferromagnetic side plates.

Wait, that makes sideway forces... the shaft's field is in circles around it, and the forces on the balls should push balls along the axis, away from the inner part of the motor... What I were thinking is that the shaft's magnetic field through the balls, as they rotate, acquires a radial component. The radial component would interact with currents in the inner raceway, applying rotational force.

This paper, behind the paywall, claims to have solved everything exactly:

http://iopscience.iop.org/0305-4470/23/ ... 14_017.pdf

but I don't have access and don't want to pay for it.

Also, a tangentially related question: suppose I pass direct current through an iron rod. How will the magnetic field (enormously strengthened by ferromagnetism) interact with the current? Will it concentrate the current into the middle of the rod?

cdclndc[/url]":euw2d88m]There seems to be a very plausible explanation at the end of this video

It's scary to look at this video. He left the HV winding on the microwave oven transformer, and he's waving his hands next to it.

The bulges explanation is what we were discussing... the reason I'm not persuaded is that the thing can spin very fast (thousands RPM), so the hotspots are necessarily quite small (they have to cool down in time), and if they are small, even at the high temperature, it seems the 'bulging' due to the thermal expansion of those small regions would be very small and smaller than surface roughness. Keep in mind also that the bulge is the original explanation by a guy who came up with the bearing motor and then went all sorts of crazy about this motor, claiming free energy on the basis of a calorimetric study.

This guy:
http://www.mtaonline.net/~hheffner/HullMotorA.pdf
claims that it doesn't work with non ferromagnetic (austenitic stainless steel) bearings.

 

[Dmytry]

blargh[/url]":1nxbidui]

Dmytry[/url]":1nxbidui]

cdclndc[/url]":1nxbidui]There seems to be a very plausible explanation at the end of this video

It's scary to look at this video. He left the HV winding on the microwave oven transformer, and he's waving his hands next to it.

From his description, he's using the transformer in reverse, so the output voltage is actually quite low, and high amperage.

At 11 seconds, you can see it from the above... on the left is the mains winding, connected to mains, in the middle there's filament winding (would normally supply current to the filament in the magnetron) which he connected to the ball bearing motor, and on the right there's HV winding, disconnected from everything. At 2:51 you can clearly see the disconnected HV tab on the right, and he proceeds to wave his hand around it.

These experiments all seem to have a lot of unknowns related to poor component material controls, and poor instrumentation. I'd say studying this would make a pretty decent gradudate, or even undergrad, research project.

Yeah. Spot welding is definitely an issue.

 

[Dmytry]

.劉煒[/url]":3ldal8d9]Partially evacuated tunnel, muzzle on top of a mountain? Say, the everest ridge?

I had an idea a while back. Suspend the gun from a lot of hydrogen filled balloons to it's sides*. You can picture what it would look like with two balloons . The amount of hydrogen required would be quite small compared to what you burn in a rocket - hydrogen is about 14 times lighter than air, and thus lifts 13 times it's own weight.

It would seem to be only a matter of money. A lot of money. A gun can also be very long and suspended all along it's length.

(*yes yes I know Hindenburg, in WW1 they had real trouble lighting the blimps on fire even with incendiary rounds. We have far better tech now as far as dealing with electrostatics and lightning goes, and we have gas sensors to detect leaks before mixes enter flammable range).

 

[Dmytry]

Well, the idea is that it is not a real gun, it's an electromagnetic launcher, with a track, so if it is a little bent it is just a slightly curved track. It's not a bullet getting stuck in a barrel narrowed by bending. If many projectiles are launched in sequence, the forces will act to straighten out the track. Likewise, as for the recoil, it's a matter of the ratio of the mass of the gun to the mass of the projectile - divide the projectile velocity (say, 10km/s) by the mass ratio (say, 5000) and you get the gun velocity after the recoil (2m/s).