Re:Has been around years but is still cool
on
Magnetic Fluids
·
· Score: 1
Their Feedthrough Catalog promises only 10^-9 mbar (I suppose you meant 10^-10 mbar, not 10^-10 bar). Anyway, do you really have a clean (hydrocarbon-free) mass spectrum after baking a chamber using these feedthroughs ? What about stray magnetic fields ? I know, I could ask the manufacturer but I'd like to hear the opinion of a user.
The principal factor affecting the loss in power with time is the Space radiation environment. For low radiation environments, such as low Earth orbiting, their effectiveness decreases around 1 to 2 percent a year. This means after a five year mission the solar panels will still be making more than 90% of what they made at the beginning of the mission (as long as they haven't gotten farther away from the Sun). In contrast, for missions in higher radiation environments, such as mid altitude Earth orbit (2000 to 10000 kilometers), arrays can lose half their power within 1 year. That is one reason few missions fly in this orbital range.
ever heard of rain ?
(not so good vor Nevada desert, I admit)
However, the radiation from the stars gets a lot lower the further away we get from them
isotropic radiation (i.e. from stars) falls of with 1/r^2 (r = distance to star). However, I would expect satellites to emit directed beams (think Laser), not isotropic - the losses would be much lower.
Why on earth do we not concentrate on energy saving instead of producing more and more energy?
It's stupid to "put all eggs into one basket". Better to proceed in both directions.
Refering to your last paragraph, I think cooling will require a larger portion of the total CPU costs in future. This is already true for todays CPU compared to older ones that only required passive cooling.
The main problems with the current cooling concept is that air has a very small heat capacity which means that you have to circulate a lot of air in order to remove a certain amount of heat, and second that only a small temperature difference T(CPU) - T(ambient) is allowed which makes cooling very inefficient. As a consequence, cooling by air will be replaced soon by other technologies that move the heat dump away from the processor (e.g. liquid cooling circuits, preferably liquids colder than room temperature). Cooling to below room temperature will be standard as this opens the possibility for even higher clock rates with only small additional hardware costs (running costs will be a lot higher of course).
The next step will be cooling facilities integrated into the chip (think of dedicated copper lines running from problematic 'hot spots' to the cooling interface).
Quantum entagnglement cannot be used to transmit information, because you cannot tell into which quantum state the "sending" particle collapses.
The classical analogon would be two "entangled" dice - say both dice always show an identical number. However, being dice you cannot predict which number they show (or force them to show a certain number) and thus you cannot send any information (only random numbers).
The connection between quantum entanglement and the dice is that collapsing a quantum particle into an eigenstate (e.g. forcing an electron into a spin-up/down state) is like throwing a die: you never know into which eigenstate it collapses (i.e. you cannot predict whether the electron goes into spin up or spin down state).
Yes, but you need a point to point connection to actually use quantum encryption. How do you maintain quantum correlation if you have routers, switches, firewalls in between (which also have to look at the data to do their job) ?
For instance a 1m tall person standing in that 5 m ship at 1G would have only 80% of the gravity at his feet acting on his head.
I think it's safe to assume that the average persion will be 1.8m tall *g*, i.e. experience only 64% of the feet-gravity at his head. Also things that fall "down" towards the outer cylinder wall would not do so in straight line in the rotating observer frame of reference but in a curved line, due to the Coriolis effect (of course in the non-rotating frame of reference it's a straight line). It would be a weird experience to pour yourself a cup of tea *g*
Okay, as I am a physicist I can tell you how we handle such cases: Every time I encounter a word/concept which I don't know I just replace it with "gizmo" and read over it. So, e.g. the sentence
According to quantum electrodynamics (QED), the relativistic quantum field theory of the interaction of charged particles and photons, an electron can emit virtual photons that can then emit virtual electron-positron pairs (e+, e-).
turns into
According to gizmo (gizmo), the relativistic gizmo theory of the interaction of charged particles and gizmos, an electron can emit virtual gizmos that can then emit virtual electron-gizmo pairs (e+, e-).
which is not that bad - you get the key message. Of course, if it reads
According to gizmo (gizmo), the gizmoic gizmo theory of the gizmoation of gizmoed particles and gizmos, an gizmo can emit gizmoal gizmos that can then gizmo gizmoal gizmo-gizmo pairs (e+, e-).
you have a problem. In that case, skip the sentence and carry on. If you are lucky it wasn't of key importance for understanding the concept anyway.
You should compare with gas leaks. Water leaks are obviously not considered critical - that's why they are so common (in you area). More care is put into gas pipes because gas leaks are pretty dangerous - a study on how often gas leaks occur should give a rough estimate (say upper limit) for reactor leaks.
Also please consider, that said gas is dangerous too ("fire bomb"), as well as your cars' motor which actually has several thousand gas explosions per minute - but somehow it's not considered a danger at all. What I want to say: It depends on how much care and effort you take to make it safe.
Next, please note that they are talking about making the reactor "fail-safe". In this context "fail-safe" is a technical term that means "if it fails, it should by design fall into a safe state". It's not easy to make things fail-safe, maybe it cannot even be achieved - we have to wait and see. It would indeed be foolish to use a non fail-safe design for reactors without operator control.
Re:Man ya try to be a nice guy....
on
Super Hard Steel
·
· Score: 1
A good introduction into hardness can be found at http://www.calce.umd.edu/general/Facilities/Hardne ss_ad_.htm. It also contains a chapter on the relation between hardness and tensile strength (scroll down to section 7) which contains the following paragraph:
A correlation may be established between hardness and some other material property such as tensile strength. Then the other property (such as strength) may be estimated based on hardness test results, which are much simpler to obtain. This correlation depends upon specific test data and cannot be extrapolated to include other materials not tested.
I hope you don't want to suggest that no coherence is needed for interference effects. In the case of a double slit experiment one needs a transverse coherence length of at least the slit spacing to see interference effects. I guess you use a pinhole between lamp and slits to reduce the source size and put it far away from the slits. What you then have is a light source with a certain transverse coherence (although it's not a Laser).
I think our problem is that "regular light" was not defined exactly enough.
The reason why electrons don't plunge into the nucleus are:
1) The Pauli Principle: Electrons at lower energy states prevent the other electrons from going there. The reason why something like "states" exists is the wave-like nature of particles (electrons), which give also rise to:
2) The Heisenberg Uncertainty Principle: If the electron would be at rest (in the nucleus, or anywhere) it would have infinite uncertainty in momentum, i.e. in the next instant it would be somewhere else (and you wouldn't know where). So it IS possible for electrons to be at the position of the nucleus, but only for a very short time (this leads to something called "Fermi-Contact-Interaction" between s-electrons and nuclei).
Zewail didn't "resolve" the Heisenberg Uncertainty Principle (HUR) (what does "resolve" mean in this context anyway ?). He has just pointed out why the HUR was not an obstace in femtochemistry:
a) it's a matter of numbers
b) coherence (which is needed for spatial localization) is not as instable as initially thought
for a) it's important to remember that he is talking about positions of NUCLEI which are quite
heavy objects (compared to e.g. electrons). Heavy objects have large momentum (p=mv) therefore a small uncertainty is not as disturbing as for light objects.
Now some comments on Hemos' text/questions:
Femtosecond resolution apparently provides the localization needed to treat electrons as classical spheres in space, nearly following Newtonian physics.
No, he is talking about "atomic motion", i.e. about nuclei. See above.
However, femtosecond chemistry has been around for years, so why hasn't this worked yet?
It has worked. The article is just a summary/explanation of why it has worked.
Well, there is a great deal of error in gathering energy values, even when energies are collected at femtosecond intervals. This is due to freaky quantum physics i don't understand.
The error is given by the HUR: Delta(E) >= hbar/(2*Delta(t)). The shorter you look, the larger the uncertainty. However, it turned out that the timescale of several ten femtoseconds is still large enough to have sufficient energy resolution.
This is like the 'double slit experiment', in which regular light is shown through two slits, the waveforms either completely add or subtract, and what you see on the wall is a bunch of tiny spots of light at a defined point in space.
No. The double slit experiment doesn't work with regular light, you need *coherent* light (e.g. a Laser). That's exactly the point Zewail makes: if they had used incoherent states (or if the coherency was destroyed fast) they would not have observed localized atoms ( = your "tiny spots of light")
One has to keep in mind that magnetism esp. ferromagnetism is a very complicated phenomenon. In fact, ferromagnetism is a collective phenomenon and very hard to predict, esp. for unknown materials. However, the properties and laws that are used in the articles' prediction are very simple and basic and therefore we can put much more confidence in their predictions. OTOH the predictions result in so astronomical high numbers that we will not reach these barriers soon and much can happen meanwhile.
Neodym doped Yttrium-Aluminium-Garnet crystals are the lasing medium for this kind of lasers, properties of Nd:YAG see here, basically it's a high power (several Watt) infrared laser often used to pump other lasers. Also, it's not really small.
The dangerous life of plutonium is not determined by its half life, but by the speed of advancement (and persistance) of our technology.
Right. The half life is kind of the upper limit. However, one has to remember that after one half life there is still half of the stuff there and after two half lifes a quarter and so on. So, actually it takes several half lifes for the stuff to decrease to non-dangerous levels.
But you can only have inductive losses when there is a changing magnetic field, i.e. for AC. With superconductors there is however no reason to use AC, so they probably use DC current and low voltages which eliminates inductive loss and leak currents due to high electric fields.
Slashdot Mirror
Their Feedthrough Catalog promises only 10^-9 mbar (I suppose you meant 10^-10 mbar, not 10^-10 bar). Anyway, do you really have a clean (hydrocarbon-free) mass spectrum after baking a chamber using these feedthroughs ? What about stray magnetic fields ? I know, I could ask the manufacturer but I'd like to hear the opinion of a user.
this site has the following information:
dust would cover the arrays
ever heard of rain ? (not so good vor Nevada desert, I admit)
However, the radiation from the stars gets a lot lower the further away we get from them
isotropic radiation (i.e. from stars) falls of with 1/r^2 (r = distance to star). However, I would expect satellites to emit directed beams (think Laser), not isotropic - the losses would be much lower.
Why on earth do we not concentrate on energy saving instead of producing more and more energy?
It's stupid to "put all eggs into one basket". Better to proceed in both directions.
Refering to your last paragraph, I think cooling will require a larger portion of the total CPU costs in future. This is already true for todays CPU compared to older ones that only required passive cooling.
The main problems with the current cooling concept is that air has a very small heat capacity which means that you have to circulate a lot of air in order to remove a certain amount of heat, and second that only a small temperature difference T(CPU) - T(ambient) is allowed which makes cooling very inefficient. As a consequence, cooling by air will be replaced soon by other technologies that move the heat dump away from the processor (e.g. liquid cooling circuits, preferably liquids colder than room temperature). Cooling to below room temperature will be standard as this opens the possibility for even higher clock rates with only small additional hardware costs (running costs will be a lot higher of course).
The next step will be cooling facilities integrated into the chip (think of dedicated copper lines running from problematic 'hot spots' to the cooling interface).
Quantum entagnglement cannot be used to transmit information, because you cannot tell into which quantum state the "sending" particle collapses.
The classical analogon would be two "entangled" dice - say both dice always show an identical number. However, being dice you cannot predict which number they show (or force them to show a certain number) and thus you cannot send any information (only random numbers).
The connection between quantum entanglement and the dice is that collapsing a quantum particle into an eigenstate (e.g. forcing an electron into a spin-up/down state) is like throwing a die: you never know into which eigenstate it collapses (i.e. you cannot predict whether the electron goes into spin up or spin down state).
Yes, but you need a point to point connection to actually use quantum encryption. How do you maintain quantum correlation if you have routers, switches, firewalls in between (which also have to look at the data to do their job) ?
For instance a 1m tall person standing in that 5 m ship at 1G would have only 80% of the gravity at his feet acting on his head.
I think it's safe to assume that the average persion will be 1.8m tall *g*, i.e. experience only 64% of the feet-gravity at his head. Also things that fall "down" towards the outer cylinder wall would not do so in straight line in the rotating observer frame of reference but in a curved line, due to the Coriolis effect (of course in the non-rotating frame of reference it's a straight line). It would be a weird experience to pour yourself a cup of tea *g*
The other option is to let out a tether with a countermass
...but it turns out to be much more efficient if you can just put something disposable on it
How do you accelerate such a construction ? How do you keep alignment for, say communication links back to earth or solar panels ?
But is it really efficient to carry more or less useless mass to mars ? Who did come to this conclusion (Zubrin?)?
turns into
which is not that bad - you get the key message. Of course, if it reads
you have a problem. In that case, skip the sentence and carry on. If you are lucky it wasn't of key importance for understanding the concept anyway.
You should compare with gas leaks. Water leaks are obviously not considered critical - that's why they are so common (in you area). More care is put into gas pipes because gas leaks are pretty dangerous - a study on how often gas leaks occur should give a rough estimate (say upper limit) for reactor leaks.
Also please consider, that said gas is dangerous too ("fire bomb"), as well as your cars' motor which actually has several thousand gas explosions per minute - but somehow it's not considered a danger at all. What I want to say: It depends on how much care and effort you take to make it safe.
Next, please note that they are talking about making the reactor "fail-safe". In this context "fail-safe" is a technical term that means "if it fails, it should by design fall into a safe state". It's not easy to make things fail-safe, maybe it cannot even be achieved - we have to wait and see. It would indeed be foolish to use a non fail-safe design for reactors without operator control.
Note esp. the last sentence!
Can it have a soul?
Would a single cell know whether the whole thing has a soul or not?
I hope you don't want to suggest that no coherence is needed for interference effects. In the case of a double slit experiment one needs a transverse coherence length of at least the slit spacing to see interference effects. I guess you use a pinhole between lamp and slits to reduce the source size and put it far away from the slits. What you then have is a light source with a certain transverse coherence (although it's not a Laser).
I think our problem is that "regular light" was not defined exactly enough.
The reason why electrons don't plunge into the nucleus are:
1) The Pauli Principle: Electrons at lower energy states prevent the other electrons from going there. The reason why something like "states" exists is the wave-like nature of particles (electrons), which give also rise to:
2) The Heisenberg Uncertainty Principle: If the electron would be at rest (in the nucleus, or anywhere) it would have infinite uncertainty in momentum, i.e. in the next instant it would be somewhere else (and you wouldn't know where). So it IS possible for electrons to be at the position of the nucleus, but only for a very short time (this leads to something called "Fermi-Contact-Interaction" between s-electrons and nuclei).
Zewail didn't "resolve" the Heisenberg Uncertainty Principle (HUR) (what does "resolve" mean in this context anyway ?). He has just pointed out why the HUR was not an obstace in femtochemistry:
a) it's a matter of numbers
b) coherence (which is needed for spatial localization) is not as instable as initially thought
for a) it's important to remember that he is talking about positions of NUCLEI which are quite
heavy objects (compared to e.g. electrons). Heavy objects have large momentum (p=mv) therefore a small uncertainty is not as disturbing as for light objects.
Now some comments on Hemos' text/questions:
Femtosecond resolution apparently provides the localization needed to treat electrons as classical spheres in space, nearly following Newtonian physics.
No, he is talking about "atomic motion", i.e. about nuclei. See above.
However, femtosecond chemistry has been around for years, so why hasn't this worked yet?
It has worked. The article is just a summary/explanation of why it has worked.
Well, there is a great deal of error in gathering energy values, even when energies are collected at femtosecond intervals. This is due to freaky quantum physics i don't understand.
The error is given by the HUR: Delta(E) >= hbar/(2*Delta(t)). The shorter you look, the larger the uncertainty. However, it turned out that the timescale of several ten femtoseconds is still large enough to have sufficient energy resolution.
This is like the 'double slit experiment', in which regular light is shown through two slits, the waveforms either completely add or subtract, and what you see on the wall is a bunch of tiny spots of light at a defined point in space.
No. The double slit experiment doesn't work with regular light, you need *coherent* light (e.g. a Laser). That's exactly the point Zewail makes: if they had used incoherent states (or if the coherency was destroyed fast) they would not have observed localized atoms ( = your "tiny spots of light")
One has to keep in mind that magnetism esp. ferromagnetism is a very complicated phenomenon. In fact, ferromagnetism is a collective phenomenon and very hard to predict, esp. for unknown materials. However, the properties and laws that are used in the articles' prediction are very simple and basic and therefore we can put much more confidence in their predictions. OTOH the predictions result in so astronomical high numbers that we will not reach these barriers soon and much can happen meanwhile.
- Heisenbergs Uncertainty Principle
- Einsteins Energy-Mass relation E=mc^2
both are relativisticly correct (there is no Schrödinger equation involved!).Did you read the article ? I promise you will hear from it in the newspapers soon (in case you didn't already).
Neodym doped Yttrium-Aluminium-Garnet crystals are the lasing medium for this kind of lasers, properties of Nd:YAG see here, basically it's a high power (several Watt) infrared laser often used to pump other lasers. Also, it's not really small.
I wonder when someone will submit a story about the attosecond laser (Science 292, p. 1689) - far more ground braking than just another UV laser.
Yes, piezo drives. This one has a resolution of better than 0.01 nm. They are used e.g. for scanning tunnel microscopy.
The dangerous life of plutonium is not determined by its half life, but by the speed of advancement (and persistance) of our technology.
Right. The half life is kind of the upper limit. However, one has to remember that after one half life there is still half of the stuff there and after two half lifes a quarter and so on. So, actually it takes several half lifes for the stuff to decrease to non-dangerous levels.
But you can only have inductive losses when there is a changing magnetic field, i.e. for AC. With superconductors there is however no reason to use AC, so they probably use DC current and low voltages which eliminates inductive loss and leak currents due to high electric fields.
the Free Encyclopedia Project: http://www.gnu.org/encyclopedia
Better write them down in covariant form - even shorter and more elegant, and more geeky (will not be recognized by EE students however).