You are not logged in.
Pages: 1
i am not sure what the last poster means by a 60 % size reduction in a proton linac type particle accelerator. The problem with any other particle than an electron is as the speed changes during acceleration the length of the resonating structure must also change to keep all nuclii within 90 degrees phase of the acceleating voltage otherwise you cease to accelerate and instead decelerate since out of phase beam current losses both drop the impedance on the guide thus loading the rf and become waste heat, when you are talking about power levels that can vaporize any material in microseconds you must have design efficiency in the sense of capturing and accelerating the entire beam current "Fish" including inherant design features that will both bunch a lagard tail and disgcourage radial ion travel ( usually magnetics for this last) The only advantage that a non CV (Constant Velocity or greater than 99% C) accelorater has is that it makes it easier to eliminate in phase radial oscillations gaining in magnitude and thus contributing to parasitic power losses in the form of so called "Synchrotron Radiation". everywhere else it is a nightmare because the acceleration velocity gained by the beam at each stage must precisely match the both the overal mechanical length and the energy input to that stage. in practice it means the rf power must be ramped up as a fixed function of the beam current to maintain beam phase, assumming the extremly high "guide" impedance decrease vs beam current is virtually constant. If impedance has to vary agreat amount because of increased beam loading the guide will have negative stability and loss of phase synchronization thus the system will be unworkable. Also due to the extremely high (3 megaohm plus) overall impedance of the guide section the I squared R losses must be kept lower than even a pure copper surface can do. Therfore the resonating cavities and power coupling lines will have to be plated with superconducters
During the Eighties New England Nuclear tried to build an proton linac with a couple of milliamps beam current at 30 Mev and could not do it. Commercial medical and industrial linacs are all electron and the electrons get up to effective Constant velocity at under a hundred KV so actually the bunching cavity is nothing more than a half cavity on the beginning of the Guide . Bunching losses of a few micro amps at a hundred KV can be ignored Proton bunching losses of even 1 % are significant when you are talking 3 million volts at ten thousand amps, and consider you are going to need voltages like that to set up a strong enough electrostatic field, at a fairly wide accelerator drift tube appature neccesary to contain a proton beam of that high current. What it adds up to is about a half a liter/sec of H+ at STP and at say a 10 cavity machine using about a 25 megahertz frequency. The reasonant length would make the end cavity a toroid with a 1 meter gap and an overall thickness of about 1 meter with a half meter drift tube diameter and a 2and 1/2 meter inside diameter. that would still give beam densities about a hundred times more intense than the most powerful comercial proton cyclotron in use Today.(because of electrostatic forces within the beam this may be not possible.) Even with using only ten cavitieis the injector and resonant accelerator would still be about fifteen meters long. The overall "push" would be about twelve and a half metric tons utilizing only 43 kilograms of propellant fuel (H2) a day. If you could keep the overall weight down of the space craft down to 125 metric tons that would give you a constant one tenth g acceleration making the mars trip ( acceleration and deceleration) in less than twenty days at closest approach. That would mean two things. One you could send the supplies for a mars mission in a 1250 metric ton unmanned fregher with the same power plant about eight months ahead of time and since the life support in the crew pod would be minimal send the people in the smaller 120 ton machine. The interesting thing about this technology is that there does not have to be to much more developemental refinement on either the 50 megawatt reactor or the linac design except for the buncher .
One should note that this is not eficeincy in any sense of the word the total energy expended to accelerate a mass of 1/2 gram per second to ~25% C should only be about 156 kilowatts if all energy was used with total efficiency. We are using 50 megawats of thermal to give us a little more than 30 megawatts electrical, Would any Varian engineers like to weigh in(annonamously or otherwise) at this point ---John
--------------------------------------------------------------------------------
Currently in use commercial and reasearch particle accelerators are very energy inefficient dc power is converted to rf power with about a 60% effeciecy no effort is made to use superconducting technology on either the accelerating cavities or the steering magnetics. In a space ship you would want to make the RF source for the accelerating wave guide operated by your fuel ions as well as their electrons Since vacuum is free you would use large vaccum full wave rectifier and capacitve ladders on your combo ion injection gun cum positive ion klystron rf generator you would shoot for multi megaohm rf mpedances on your guide, but it would stiil be very dificult to bring positive ions even H+ up to anything near a cv section on that accelerator. epecially since you would need tens of thousands of amps average beam current--- needless to say you would need a big nuclear reactor and waste cooling radiator to power all this. It would weigh a lot even if you used superconducting magnets and current carriers on the generators and kept the system as cold as possible(under 800c) to minimize waste heat, the turbine would have to be extremely largeto eliminate a monstrous condenser indeed you would most probably have to use the bulk of the ships hull surface for waste heat radiation esepcially during engine and reactor shutdown when the turbine is not turning the heat into power. you would need in space docks at both ends of the "run" just to utilize the cool down power and provide an ECCS and shielding for debarking and embarking passengers and for maintenance workers who would use hydraulic robots for the actual work. a nuclear submarine probably provides the best model for the type of power plant required. with the caveat that there is no nearby free source of coolant or umlimited thermal mass to get rid of unwanted heat, --One who spent 11 years doing this stuff
There needs to be a full scale design contract let on a low temp ( less than 1100 C )1200 megawatt interplanetary space reactor /turbo-generator coupled to a extreme high current both electron and proton operated combination klystron/linac with targeted exhaust velocities o 10e6 M/sec and exhaust masses of 10 g/sec. This would give a 50 metric ton vehicle an acceleration of about 1/5 G or about 2m/sec/sec, with a reaction propellant usage of 36 kilograms of H2 per hour. Note waste energy would be 100 to 200 megawatts and some of that would be used for internal spacecraft functions Nuclear plant thermal efficiencies would be increased to 90% in this model from the current less than 70%. If that is not possible increase the total reactor thermal output to 1500 megawatts and waste heat to 500 megawatts. That burn would make most Mars trips a twelve day cruise with just 60 hours combined acceleration and deceleration "burn time" and a reaction mass usage one way of 2.4 metric tons H2. Robot cargo ships would make the trip one way taking ten times as long with 500 ton total masses and then their reactors would be converted to surface power units. Best thing yet is that if we only achieve 50 percent of the power throughput goals the whole thing is still practical and very doable using technology that has been theoretically demonstrated today. The other thing is that the reactor radiation shielding only needs to be a 2 meter thick water tank between the crew section and the power plant and propulsion section and that the robot ships will make the flight first. you will note that this is about twice the thrust that you get per unit kilowatt than a current ion engine but the main accelerator klystron proton linac uses acceleration grids only in the preliminary low energy proton klystron stages and the target-less linac uses power only as a function of its own impedance, which using superconductors on the surface of its acceleration cavities would be quite high ( today?s commercial linacs use polished copper and have impedances in the 50 thousand ohm range. The throughput impedance on the DS1's ion drive is about 600 ohms. The high voltages and low ( comparative) frequency of this linac would enable the drift tube diameters to be quite large so that there would be no beam interaction with the accelerating current other than electrostatic. The proton klystron front end of the accelerator would be very long (50 M) because of the 1000 to one difference in the charge to mass ratio of a proton compared to an electron. and the frequency would be much lower because the initial electrostatic gun would be accelerating those protons to a much lower initial velocity, and because the system could accept virtually no out of phase beam losses both for efficiency reasons and power dissipation ones. The main source of power for the linac portion would come from the concurrent electron klystrons needed to neutralize the beam, and they of course would have no collectors just well designed Pulsed acceleration grids.and possibly quadrapole focusing magnets instead of annular windings. all current carrying surfaces would either be gold or superconductor plated and the reactor fluid waste heat and the hot side of the cryo-pumps would cool by the use of conventional space radiators, The design is doable.. all we need is for some Varian or Stanford engineer to crunch the numbers and come up with the exact specs --JJB
Pages: 1