Patent Grant
7345435
This invention relates to linear accelerators and specifically to an improved accelerating structure for accelerating a high current beam of charged particles.
High power Free Electron Lasers (FEL) with power levels of 10 kW of infrared laser light have recently been demonstrated at the Thomas Jefferson National Accelerator Facility (TJNAF) in Newport News, Va. Although 10 kW of laser light is a substantial achievement, even higher levels of power would support advanced studies of biology, chemistry, and physics and enhance manufacturing technologies.
A linear accelerator (linac) supplies the FEL with electrons at relativistic speeds. For higher levels of FEL power, such as 1 MW, improvements are required in the accelerating structures in the linac. To achieve a power level of 1 MW in the FEL, electron beams in the range of 500 to 1,000 mA have to be accelerated in the linac supplying the FEL.
Successful operation of a linac at 0.5 to 1 amperes of current will require the accelerators to be based on superconducting technology. Additionally, Higher Order Modes (HOM) excited by the beams must be effectively damped to allow stable operation of the linac.
Therefore, one of the requirements for increasing the power level of an FEL to the 1 MW range is an accelerating structure capable of accelerating an electron beam in the linac to the level of 0.5 to 1 ampere and having sufficient damping to suppress the HOMs excited by the beams.
The invention is a superstructure for accelerating charged particles at relativistic speeds. The superstructure is made of two weakly coupled multi-cell subunits and equipped with HOM couplers. A beam pipe connects the subunits and an HOM damper is included at the entrance and the exit of each of the subunits. A coupling device feeds rf (radio frequency) power into the subunits. The subunits are constructed of niobium and maintained at cryogenic temperatures. The length of the beam pipe between the subunits is selected to provide synchronism between particles and rf fields in both subunits.
The superstructure beam of the present invention enables acceleration of electrons to achieve an electron beam at 0.5 to 1 A. Use of a superstructure over conventional accelerating structures advantageously increases the active cavity length as a percentage of the total length of the linac and also significantly reduces the amount of microwave components. The superstructure of the present invention reduces the number of cells per structure and therefore reduces the amount of trapped HOMs. The reduction in number of cells and microwave components leads to a significant reduction in the cost of the accelerating structures. The superstructure is very compact and will provide excellent HOM damping. The compactness of the superstructure and the reduction in the number of power feeds for the subunits reduces the number of required rf components and therefore significantly improves the economics of the linac.
The following is a listing of part numbers used in the drawings along with a brief description:
With reference to
Referring to
The RF properties are summarized in Table 2.
The superstructure design features a rather large beam hole for good coupling of HOMs. The large beam hole compromises the R/Q to some extent, but still results in a reasonable ratio of Epeak/Eacc. With the R/Q value of ˜89 Ω/cell, the power dissipated in the superstructure at a Q-value of 8.109 is approximately 68 W.
Table 3 lists the first 20 monopole modes. Mode No. 14 (TM020) is the mode with the highest impedance of 36.5Ω, which is however a factor of 10 smaller than the fundamental mode (R/Q) value (mode No. 3). The field distribution of the mode No. 14 shows sufficient field strength at the locations of the HOM couplers thereby indicating appropriate damping. The total impedance for the first 16 monopole modes is approximately 140Ω. None of the first 16 modes falls on a machine line (N*1500 MHz).
The dipole modes up to a frequency of 1950 MHz have been calculated with MAFIA and are listed in Table 4 below. MAFIA, an acronym for Maxwell's Equations by the Finite Integration Algorithm, is a computer program for solving problems in the simulation of electromagnetic fields. As shown in Table 4, the shunt impedances (R/Q) are favorably small.
The choice of the frequency of 750 MHz as opposed to 500 MHz was dictated by the following considerations:
a). Under the assumption that the beam alignment will be the same for a 1 A beam as it is for a 100 mA beam, which was the threshold current for the 1500 MHz superstructure and the achieved damping results for dipole modes, the threshold current at this frequency would be 2× higher, since the impedance (R/Q) scales with 1/r2 where r=iris diameter, which scales with frequency. The threshold current is proportional to 1/v(R/Q). A threshold current of 1 A is therefore achievable with a reduction of the number of cells from 5 to 2 and an appropriate opening of the iris diameter.
b). The second consideration for proposing a 750 MHz cavity is the existing infrastructure for cavity treatments. Both the cabinets for chemical polishing and high pressure rinsing are size limited to a cavity length of ˜130 cm. The 750 MHz superstructure as shown in
The superstructure will be fabricated of niobium. The niobium cavities will be operated at cryogenic temperatures, or a temperature below 4.2 K, so that they are superconducting.
Table 5 includes a detailed list of proposed assumptions and parameters for a 1 MW FEL based on the superstructure disclosed herein.
The present invention as described herein is a superstructure for a 1 Amp beam, resonating at 750 MHz and consisting of two 2-cell subunits coupled weakly by a beam pipe. This superstructure features a total of 6 coaxial type HOM couplers. Two HOM couplers are located at the end of each subunit and two at the interconnecting beam pipe. The superstructure also includes a high power coaxial input coupler of the KIK (SNS) type, as described in “Superconducting Cavities for HERA”, B. Dwersteg et al., Proceedings of the 3. Workshop on RF Superconductivity, Report ANL-PHY-88-1, p. 81ff, ANL, 1987.
Simulation calculations indicate that the ratio of peak surface fields and accelerating gradients are reasonable, whereas the shunt impedance suffers somewhat from the large iris diameter. The HOM spectrum of this structure is quite favorable as the highest parasitic shunt impedance is only 10% of the fundamental mode shunt impedance indicating that the HOMs would be damped to Qext values of less than 1000. In this case the HOM power generated by the beam would be only 60 W, distributed over 6 HOM couplers.
As described herein, the present invention describes a method of developing a superstructure for accelerating charged particles in a high energy particle beam. The method for developing a superstructure includes 1) selecting a resonant frequency, 2) providing a cell including a cell length, an equator diameter, and an iris diameter, 3) connecting a plurality of the cells into a multi-cell subunit, 4) connecting the subunits with a beam tube, 5) weakly coupling the subunits, 6) providing a plurality of couplers for feeding rf power into the subunits to define a superstructure, 7) identifying the monopole modes for the superstructure, 8) adjusting the field strength of each monopole mode to achieve appropriate damping, 9) determining the dipole modes at a range of frequencies, and 10) verifying that the impedances for each dipole mode are small.
Although the specific embodiment described herein is comprised of 2-cell subunits, the superstructure for high current applications can be constructed of subunits having between one and nine cells per subunit.
Having thus described the invention with reference to a preferred embodiment, it is to be understood that the invention is not so limited by the description herein but is defined as follows by the appended claims.
The United States of America may have certain rights to this invention under Management and Operating contract No. DE-AC05-84ER40150 from the Department of Energy.
| Number | Name | Date | Kind |
|---|---|---|---|
| 4629938 | Whitham | Dec 1986 | A |
| 4988919 | Tanabe et al. | Jan 1991 | A |
| 6366641 | Whitham | Apr 2002 | B1 |
| 6376990 | Allen et al. | Apr 2002 | B1 |
| 6407505 | Bertsche | Jun 2002 | B1 |
| 6642678 | Allen et al. | Nov 2003 | B1 |
| 6646383 | Bertsche et al. | Nov 2003 | B2 |
