Unfortunately for ion drives you want high current but very low energy. The amount of electric power required by an ion drive increases as the exhaust velocity increases, and for present day space applications you are better off with less particle energy, not more.
Its a good question .
I don't understand astrophysical shocks, but see: http://www.slac.stanford.edu/e...
As far as I can tell the rely on magnetic fields bending the particles back into the shock.
When relativistic particle trajectories are bent by magnetic fields, they emit synchrotron radiation which increases rapidly with increasing particle energy.
Longitudinal fields don't do the same thing. There is a tiny amount of radiation, but it is not strongly dependent on particle energy. I believe this is because Lorentz contraction increases transverse, but not longitudinal electromagnetic fields: http://en.wikipedia.org/wiki/C...
Ideally the fields in the plasma accelerator are longitudinal on axis. If the particle enters slightly off axis it will get a transverse kick and will radiate synchrotron radiation, and we do see that. For very high energies that radiation might be large, but the effect would be to damp the transverse motion of the particle, but not affect the longitudinal acceleration.
I know that the plasma wakefield people are seriously thinking about TeV scale machines: https://accelconf.web.cern.ch/...
It is possible that the concept fails at some much higher energy.
That is a slightly different concept. This uses a medium-energy (20GeV), high current electron beam to drive the plasma, which then accelerates a high energy beam.
There is also a scheme to use a high energy proton beam to drive the plasma, and a scheme to use a high power (Peta-watt) laser to drive the plasma.
All are being seriously considered / developed by various laboratories.
This type of scheme probably doesn't apply well to a circular machine like LHC because the energy limit there is the magnets used to bend the beam into a ring
If you do everything right, you should get the same added energy from each section. So a 10 GeV input beam woudl go to 11.6, and a 1000 GeV beam would go to 1001.6. The beams are ultra-relativistic - for all practical purposes speed of light (off by only a part in a billion) and this acceleration mechanism doesn't depend on the beam energy .
Yes - in principal. You will need a separate bunch of 20 GeV drive electrons for each section, but that is not very difficult to do with a single accelerator. You need to separate the waste beam from the previous stage and the magnetic system to do that may be inconveniently long unless there is a beam-optics trick (which there may be).
Staging two sections together is on the list of things that they are going to try. The eventual goal is to put together a lot of stages to get to TeV scale energies.
If you do a Google search on
SLAC PUB plasma wakefield
you will find a lot of non-paywalled papers on this and related plasma accelerator experiments at SLAC.
I am peripherally involved in the SLAC plasma wakefield accelerator described in the article.
It provides a very high energy gain in a short distance, but needs to be driven by a high energy drive beam. The present design uses a 20GeV drive beam (using part of the old high energy physics accelerator).The required drive beam energy could be reduced to ~10GeV but probably not a whole lot lower. So this is a way to build a relatively short very high energy accelerator, but not a way to build a very short low-medium energy machine.
Other labs are working on laser driven plasma accelerators that do not need to start with a high energy beam, but do need an enormous laser system and are presently limited to much lower average beam powers .
Plasmas are very promising for future accelerators and there was some excellent work done at SLAC as well as laser / plasma accelerators at other labs. There is still a lot to do. There are issues with staging multiple plasma cells to get high energies, beam quality and stability issues etc.
We know how to put people in space and have been doing so for over half a century - spaceship2 is not doing anything new. The comparison with very early engines and engine driven vehicles is really not appropriate. The largest problem is that it has been very expensive to put people in space. Does spaceship2 address that problem? (there is also a very real safety issue, but presumably that would get better if the volume of space travel increased).
Spaceship2 uses air launch, and a conventional but low delta-v rocket to do a parabolic flight. Air launch has been used on small rockets before but is very difficult for manned missions because of the enormous fuel weight required to get a manned craft into orbit. Spaceship2 doesn't address this issue because it doesn't attempt orbital speeds and so can be very light weight.
Spaceship2 does not need to deal with significant reentry heating so its high drag mode doesn't really apply to a "real" reentry.
So I don't see any clear use of the spaceship2 technologies to "real" space vehicles.
That said, it is somewhat mysterious why space is so expensive. The fuel costs are 1% of the total launch costs, it is is not a fundamental "energy is expensive" problem. Presumably the cost is related to the combination of very low volume production and very difficult design issues. It is possible that the spaceship2 program would be able to address some of those issues, but I have not seen any discussion of specifics.
"Space planes" seem about as practical as "air ships" and "flying cars".
I know terrorists aren't the brightest bulbs in the chandelier but they know enough not to name their networks after a famous terrorist group. What exactly is the threat that caused the delay?
A "few hours" X 400 passengers IS a big deal - and delayed flights mean missed connections, The aircraft my not be available for its next schedule flight etc.
The compelling reason is that self-driving cars could free up tens of billions of man-hours a year in the US alone. People could use the time they spend in cars for entertainment or productivity. It would be one of the truly great labor saving inventions.
For that to work though, the car would need to be truly autonomous and that gets into tricky legal issues.
Agreed, and the problem with haggling is that it takes time. If I buy a $300 piece of electronics, it is not efficient for me to spend $100 of my time to reduce the price by $50. Unfortunately if it is possible to track individual purchasers habits, vendors can continue to increase the prices those consumers see until they are essentially forced to waste time price comparing.
Of course it won't be an "increase", Instead the "list" price will be $1000, and people will get varying discounts depending on how much price comparing they do. People who compare a lot, must not value their own time, and are more likely to only accept a low price. People who don't compare probably place a high value on their time and are willing to pay more.
This has the positive effect of effectively reducing wealth disparity, but the very large negative effect of providing an incentive for people to spend a large amount of time in a non-productive activity (price comparing). Its great for vendors, but bad for the overall efficiency of the economy.
Yes, by not seeing something under know conditions, you can rule out some possibilities. If the particles had properties withing a certain range they would have shown up in this experiment - since they did not, we know that the particles do not have properties in that range (assuming the experiment was done correctly).
This sort of null experiment is common in many types of science.
Why is matlab on a cluster stupid? It has some parallel tools that work well for certain types of jobs.
I use Matlab and Python at work. They are good for different things.
IMHO matlab has much more powerful graphics and debugging features. It is better set up for doing vector algebra problems. Python is better for some other forms of non-mathematical data operations. For most of my work Matlab is the better tool. On the windows platform it is much easier to install and manage .
We run the SLAC accelerators using both for analysis and non-realtime feedback, but Matlab is generally the preferred tool by most of the physicists.
Talking about which is better is like asking if a Semi-truck or a bus is a better vehicle. It all depends on what you are trying to move .
You can solve any problem in either of them (or in C, or fortran or COBOL if you really want), its just a case of which tool is better for the types of problems you are solving.
Seriously. By a large margin I am most likely to die due to an age related illness.Somewhere after that are non-age related illnesses. Then accidents.Then Suicide. Being killed by "bad people" is WAY down the list. Why on earth should I give up my rights to protect myself from a tiny chance of death?
Obviously people in power would like more control over me, but why should I agree to it?
Never test for an error condition you don't know how to handle. -- Steinbach
