Patent Grant
8878433
1. Technical Field
The present invention relates to magnetrons. More particularly, this invention pertains to a compact conventional (non-relativistic) magnetron suitable for numerous military and industrial applications.
2. Background Art
The magnetron is a crossed-field device in which electric fields (RF and d.c.) are perpendicular to a static magnetic field. It is recognized that numerous military and industrial applications exist for employing such a device capable of generating more than 1 Mega-watt of RF power near L-band with efficiencies of 85 percent or more.
Current commercially available magnetrons suitable for use in such critical military applications as providing countermeasures for improvised explosive devices are singularly inadequate for high microwave power (more than 1 Mega-watt RF output) at near L-band operation. For example, the 100L (L-3 Communications Electron Devices, California Tube Laboratory) is capable of only 100 kilowatt RF output near L-band at 88 percent efficiency. Another device, a 300L magnetron, is described by Wynn et al. in an article entitled “Development of a 300 kW CW L-band Industrial Heating Magnetron”. The 300L magnetron is capable of producing approximately 300 kilowatt RF near L-band (915 MHz).
Presently, no conventional magnetron (input voltage less than 100 kV) is capable of producing more than 1 megawatt of output power, near L-band (approximately 912 MHz) for diode voltages at or less than 45 kV with efficiencies exceeding 85 percent.
It is therefore an object of the present invention to provide a non-relativistic magnetron capable of generating greater than 1 Mega-watt of RF output power near L-band.
It is another object of the invention to accomplish the foregoing object at relatively low input voltage with a magnetron of compact design.
The present invention addresses shortcomings of the prior art by providing a magnetron for delivering mega-watt power at a non-relativistic diode voltage. Such magnetron includes an upstream coaxial waveguide comprising a central core and a surrounding exterior layer separated by an annular void defining an upstream chamber. The waveguide is electrically coupled to an adjoining downstream slow wave structure.
The slow wave structure includes a rod-shaped central supporting cylinder. A helical cathode comprises fourteen turns and is coaxial with the supporting cylinder. An anode slow wave structure is coaxial with and surrounds the support and the cylinder and is separated by a void defining a downstream interaction region.
The anode structure comprises fourteen interiorly-directed vanes separating and defining fourteen resonance cavities. Each of the cavities comprises a wedge portion adjoined to a neck portion.
The preceding and other features of the invention are described in the written description that follows. Such description is accompanied by a set of drawing figures. Numerals of the drawing figures, corresponding to those of the written description, point to the features of the invention. Like numerals refer to like features throughout both the written description and the drawings.
a) and 2(b) are cross-sectional views of the magnetron of
The present invention will be seen to provide a magnetron capable of providing megawatt power at conventional, non-relativistic diode voltages (less than 100 kV). Previously, the design of a magnetron capable of such performance was hindered by perceived difficulties. Concern about (breakdown) axial field stresses upon straps added to anode vanes for purposes of mode separation. An additional concern has centered about the issue of heating of the cathode. As described below, the inventors have designed a magnetron that offers architecture capable of sustaining the perceived stresses and heating to accomplish megawatt power outputs at conventional diode voltages. Simulations based upon such architecture are discussed to prove the efficacy of the inventors' magnetron design.
Turning now to the drawings,
A disk-shaped flange 20 surrounds a portion of the coaxial waveguide 16. Such flange 20 accommodates an interior reflector chamber 22 of like shape that is located and dimensioned to act as a reflector of r.f. energy, whereby energy is prevented from reflection out of the upstream end 12 of the magnetron 10, and guided to propagate toward a downstream slow wave structure 24. The downstream slow wave structure 24 includes chambers that communicate with the annular upstream chamber 19 including an upstream end cap chamber 26 and a downstream end cap chamber 28 separated by the interaction region 29.
a) and 2(b) are cross-sectional views of the magnetron 10 of
A virtual prototype of the magnetron 10 was configured for a performance simulation, discussed below, having the following parameters: cathode radii (major radius 2.2 cm; minor radius (i.e. cross-sectional radius) 0.386 cm); distance between cathode and center support cylinder (1.2 cm); and radius of center support cylinder (0.64 cm).
An anode slow wave structure 34 surrounds the helical cathode 30. The slow wave structure 34 consists of fourteen (14) vanes 36 through 62 that define fourteen (14) resonant cavities 64 through 90. The fourteen resonant cavities 64 through 90 serve to reduce the phase velocity of the RF modes in the cavity. Each cavity of the slow wave structure consists of two parts. Taking the cavity 64 as representative, a first part, a neck 92 has a generally square shape while a second part, comprises a wedge 94.
Again referring to a virtual prototype of the invention employed for the simulation to be discussed below, the following dimensions were employed to define the anode slow wave structure: radius of slow wave structure (3 cm); radial length of a neck (0.61 cm, extending from a radius of 3.04 cm to 3.65 cm); azimuthal length of a neck (0.35 cm); wedge radius (6.76 cm), angle (22.16 degrees).
A uniform axial magnetic field prevents charged particles from immediately accelerating across the interaction region 29. Rather, such particles undergo rotations about the cathode 30. If the particles' azimuthal velocity component is approximately to the phase velocity of a particular electromagnetic mode in the interaction region 29 then the possibility exists for energy exchange between the particle and the mode. Such resonance is known as the Buneman-Hartee resonance condition. As the particles rotate about the cathode 30 they gradually give up their potential energy to a mode or modes of the RF field as they migrate toward the anode slow wave structure 34. This is how RF oscillations are initiated in the magnetron 10.
b) in combination illustrate the RF extraction mechanism of the magnetron 10. Such extraction mechanism comprises six coaxial waveguides provided on the downstream end of the anode slow wave structure 34. The waveguides 98 through 108 are fixed to the downstream edges of (proceeding clockwise from the top of
b) in combination additionally illustrate straps 112 through 118 attached to the upstream and downstream ends of the slow wave anode structure 34 on alternating vanes. The straps are in place to create sufficient mode separation between the dominant π mode and the competing n−1 mode. Strap radii are measured from the center axis 110 of the magnetron 10. The inner strap 112 (referring to the upstream end of the slow wave anode structure 34) radius was modeled at 3.57 cm while the outer strap 110 radius was modeled at 4.0 cm. Radial strap thicknesses were modeled at 0.23 cm and strap separations at 0.19 cm.
Referring specifically to
The entire axial length of the magnetron 10 (from left boundary of the cylindrical core 14 to the inner surface 119 of the downstream wall 120 was modeled at 36.7 cm with the interior chamber 22 that serves as an RF energy reflector located at a distance of 3.27 cm from the upstream end of the magnetron 10 and a distance of 16.4 cm from the upstream end cap chamber 26.
A simulation was conducted of a magnetron 10 in accordance with the invention utilizing design parameters as set forth above. The simulation data discussed below were created using ICEPIC (Improved Concurrent Electromagnetic Particle-in-Cell) code, a numerical approach for evaluation of high power microwave tube designs. The ICEPIC algorithm solves Maxwell's equations and the relativistic Lorentz force law in the time domain on a fixed staggered grid.
For the simulations, a resolution of one grid cell length was used that was equal to 0.0508 cm on a uniform Cartesian grid. With the cathode radius at 2.2 cm and the vane tips at 3.0 cm, the interaction region was resolved at 16 grid cells. At this resolution, the grid volume was 590, 590 and 817 cells in the x, y and z directions respectively, yielding a total of approximately 284 million grid points. Simulations produce approximately 7 million charged micro-particles, requiring large high performance computing resources. The simulations of the magnetron described below were carried out on three different parallel computing platforms using 256, 2.6 GHz quad core Intel Hehalem CPU's, 256 2.7 GHz AMD Opteron CPU's, and 256 2.8 GHz Intel Xeon CPU's. Each simulation required approximately 2.5 days to reach 1.5 μs of simulation time at which time saturation was well established.
A pulsed power device was used to provide diode voltage. The circuit and switches that constitute the pulsed power device are not modeled here, rather, a time dependent voltage function was used to emulate the pulsed power source. The voltage function was continuous, consisting of a first part (a 50 ns linear ramp-up) followed by a second part that is a constant voltage amplitude lasting for the duration of the simulation. Voltages for the simulations ranged from 32 to 56 kV.
A uniform axial magnetic field existed for the duration of the simulations. Such field represented the insulating magnetic field that current carrying coils generate in the experiment. The coils produce a magnetic field that is uniform in the interaction region and throughout most of the magnetron.
Five sets of simulations were carried out. The magnetron was simulated at magnetic fields of 0.18 T, 0.2 T, 0.225 T, 0.25 T and 0.275 T. The voltage range examined at each magnetic field was approximately 6 kV in 500 V increments. A total of 65 simulations were carried out.
RF output power is evaluated via the area integral of the outward Poynting flux. This integral covers all six coaxial RF outputs. This plane of integration is located at z=14.0 cm. RF output power at saturation was about 1.39 MW. Output power efficiency is defined as the ratio of radiated power to system input power. Input power is given by P=IV, where I is the input current supplied to the cathode and V is the upstream diode voltage. This current is calculated by performing a closed path line integral of the magnetic field around the area in which the current is flowing to determine the current and integrating the electric field radially to determine the voltage. For the simulation, RF power efficiency was 87.1 percent with an input current of 38.5 A and a measured voltage of 41.4 kV. There was no downstream leakage current. For a strapped magnetron operating at the MW power level there is concern about breakdown due to field stress. The Kilpatrick limit for breakdown in a magnetron using a copper anode operating at 912 MHz is approximately 284 kV/cm. A survey of electrical field data at saturation throughout the volume of the magnetron indicates that the critical location for breakdown is the downstream straps. Consequently a thorough examination of field stresses at this location was carried out. Field stress data at the straps was produced at every time step for two different times and extending over 50 oscillatory periods during saturation, one starting at 800 ns and the other at 850 ns. Results indicate that the axial electric field component peaks at 259 kV/cm and that the radial electric field component peaks at 346 kV/cm. (It should be noted that the axial magnetic field of B=0.2 T will act to insulate any charge flow along this direction, easily mitigating breakdown.)
The V=42 kV at B=0.2 T simulation ran at 87.1 percent efficiency, most of the remaining energy went into heating both the cathode and the anode slow wave structure. Simulations were equipped with diagnostic capability to record the kinetic energy of charged particles upon their impact with the cathode, anode slow wave structure or upstream/downstream structure. Impact data was taken during saturation at 850 ns at each time-step over one cycle to yield an approximate result for the character of energy deposition at saturation. Power loss to the cathode via back bombardment is approximately 3.4 percent of the total input power (i.e. approximately 53 kV). The heat burden may be mitigated by operating at a 5 percent duty cycle, in which case 2.65 kW heating results. Analysis indicates that a Tungsten cathode is capable of supporting this magnitude of heating. Power loss due to collisions with the anode slow wave structure is approximately 101 kW or 6.3 percent of input power. At a 5 percent duty cycle this yields approximately 5 kW, a rate of heating this is readily addressed by external cooling mechanisms.
The described simulation is representative of a battery of runs performed from B=0.18 T to B=0.275 T. A minor degree of mode competition present at startup, which may occur between 350 and 700 ns, is well on its way to decay once π mode saturation has been reached. The simulated magnetron operated in accordance with the Buneman-Hartee resonance condition. Operation was robust and predictable over the range of magnetic fields and voltages sampled. The voltage window for π mode oscillation for a given magnetic field was approximately 6 kV which advances performance stability.
Heating of the cathode and anode was measured at saturation for all simulations. For clarity, the simulation data for only the B=0.18 and B=0.275 T are illustrated in
The magnetron consistently oscillated, in simulation, in the π mode across a wide range of magnetic fields and voltages. It operated in a predictable fashion obeying the Buneman-Hartree resonance condition. The π mode resonance curve was used to successfully predict where in voltage/magnetic field space the magnetron would oscillate (i.e. oscillations tracked well with the curve).
Thus it is seen that the present invention provides a high efficiency, 87 percent conventional megawatt class magnetron. The magnetron has been demonstrated in simulation to be capable of supporting π mode oscillations over a 6 kV wide window absent any significant mode competition at output RF power levels that exceed 1 MW for voltages lower than 40 kV. No downstream current loss occurred and RF field amplitudes do not exceed the vacuum breakdown threshold. A magnetron capable of generating over a megawatt of RF output power near L-band at 87 percent efficiency for diode voltages below 45 kV, due to its high output power and minimum voltage requirements, allows the delivery of microwave induced effects over a much wider range of space than otherwise provided in the art.
While this invention has been described with reference to its presently preferred embodiment, it is not limited thereto. Rather, the invention is limited only insofar as it is defined by the following set of patent claims and includes within its scope all equivalents thereof.
The conditions under which this invention was made are such as to entitle the Government of the United States under paragraph 1(a) of Executive Order 10096, as represented by the Secretary of the Air Force, to the entire right, title and interest therein, including foreign rights.
| Number | Name | Date | Kind |
|---|---|---|---|
| 4169987 | Oguro et al. | Oct 1979 | A |
| 6643472 | Sakamoto et al. | Nov 2003 | B1 |
| 6653788 | Ogura et al. | Nov 2003 | B2 |
| 20020043937 | Ogura et al. | Apr 2002 | A1 |
| Entry |
|---|
| Wynn et al., “Development of a 300 kW CW L-Band Industrial Heating Magnetron”, IEEE Transactions (2004), pp. 164, 165. |
| Twisleton, “Twenty-kilowatt 890 Mc/s continuous-wave magnetron”, Proceedings of the IEEE, vol. 111, No. 1 (Jan. 1964), pp. 51-56. |
| Andreev et al., “Particle-in-Cell (PIC) Simulation if CW Industrial Heating Magnetron”, Journ. Microwave Power and Electromagnetic Energy, vol. 44, No. 2 (2010), pp. 114-124. |
