Start-to-end simulation with rare isotope beam for post accelerator of the RAON accelerator
SStart-to-end simulation with rare isotope beam for postaccelerator of the RAON accelerator
Hyunchang Jin ∗ and Ji-Ho Jang Rare Isotope Science Project, Institute for Basic Science, Daejeon 34047, Korea
Abstract
The RAON accelerator of the Rare Isotope Science Project (RISP) has been developed to createand accelerate various kinds of stable heavy ion beams and rare isotope beams for a wide range ofthe science applications. In the RAON accelerator, the rare isotope beams generated by the IsotopeSeparation On-Line (ISOL) system will be transported through the post accelerator, namely, fromthe post Low Energy Beam Transport (LEBT) system and the post Radio Frequency Quadrupole(RFQ) to the superconducting linac (SCL3). The accelerated beams will be put to use in the lowenergy experimental hall or accelerated again by the superconducting linac (SCL2) in order to beused in the high energy experimental hall. In this paper, we will describe the results of the start-to-end simulations with the rare isotope beams generated by the ISOL system in the post acceleratorof the RAON accelerator. In addition, the error analysis and correction at the superconductinglinac SCL3 will be presented.
PACS numbers: 41.85.Ja, 52.59.FnKeywords: start-to-end simulation, RAON accelerator ∗ Electronic address: [email protected] a r X i v : . [ phy s i c s . acc - ph ] J un . INTRODUCTION The Rare Isotope Science Project (RISP) was established in December 2011 to constructRAON (Rare isotope Accelerator Of Newness) accelerator for various science programs. Theproject of the RAON accelerator [1, 2] is in progress in order to generate and accelerate avariety of stable heavy ion beams and rare isotope beams to be used for a wide range ofbasic science researches and various applications. To produce various rare isotope beams,the RAON accelerator put to use the in-flight fragmentation (IF) system and the IsotopeSeparation On-Line (ISOL) system. At first, the IF system uses a diver linac, which consistsof a 28 GHz superconducting electron cyclotron resonance ion source (ECR-IS), a main lowenergy beam transport (LEBT) section [3], a radio-frequency quadrupole (RFQ) accelerator,a medium energy beam transport (MEBT) section, a low energy superconducting linac(SCL1), a charge stripper section (CSS), and a high energy superconducting linac (SCL2).The beams accelerated by the SCL2 collide with the IF target, and then the rare isotopebeams are created from the collision. Secondly, the ISOL system uses a 70 MeV cyclotronas the driver to deliver a 70 kW beam power up to the ISOL target. The rare isotope beamscreated by the ISOL system are accelerated again by a post accelerator, which consists of apost LEBT [4], a post RFQ, a post MEBT, and a low energy superconducting linac SCL3.The beams accelerated by the SCL3 will be delivered up to the low energy experimentalhall or to SCL2 after passing through the the post accelerator to the driver linac transport(P2DT) section [5]. A schematic layout of RAON accelerator is shown in Fig. 1.As above mentioned, the post accelerator consists of four sections; the post LEBT, thepost RFQ, the post MEBT, and the SCL3. First of all, the lattice of the post LEBT waspartly modified because that of the ISOL system was recently changed. The new latticedesign and beam dynamics simulations of the post LEBT were presented in [4]. The maingoal of the post LEBT is the stable transportation of the heavy ion beams and rare isotopebeams up to the post RFQ. To accomplish this goal, the electro-static quadrupoles are usedin the post LEBT. Secondly, the post RFQ of the post accelerator and the RFQ of the driverlinac are the same, which is about 5 m long and the number of cell is 245. The maximumpeak surface electric field is 17 MV/m and thus the post RFQ can accelerate 10 keV/u beamto 500 keV/u. Thirdly, the post MEBT includes 8 quadrupoles and 3 buncher cavities, whichare used to satisfy the beam requirements of the SCL3. Finally, the SCL3 is divided into1
IG. 1: Layout of the RAON accelerator.FIG. 2: Layout of the post accelerator in the RAON accelerator. three sections, SCL31, SCL32 and SCL33, depending on the cavity type and the numberof cavities in one cryomodule. The first section, the SCL31 is composed of 22 quarter-wave resonator (QWR) type cavities and each cavity is surrounded by one cryomodule.The second section, the SCL32 has 26 half-wave resonator (HWR) type cavities and onecryomodule surrounds two cavities. Third, the SCL33 has 76 HWR type cavities and onecryomodule surrounds 4 cavities. The schematic layout of the post accelerator is shown inFigure 2.In this paper, we will present the results of the start-to-end simulations for the postaccelerator with the reference beam Sn . In addition, the study of the error analysis2 ABLE I: Initial beam information.Parameter Value UnitBeam Sn -Energy 10.0 keV/uNormalized transverse emittance 0.1 mm · mrad β α and correction at the superconducting linac SCL3 will be described by using the graphicaluser interface (GUI) based on the MATLAB program and the DYNAC [6] code in RAONaccelerator [7]. II. START-TO-END SIMULATIONA. Beam information
The reference beam generated by the ISOL system was recently determined as a tin beam, Sn . At the end of the ISOL system, the beam energy is 10 keV/u and the normalizedtransverse emittance is 0.1 mm · mrad. The basic beam information at the entrance of thepost LEBT is listed in Table I. Based on this beam information, we carry out the start-to-end simulation with 100,000 macro-particles using the particle tracking code, TRACK [8].Figure 3 shows the beam distributions in the transverse and longitudinal phase spaces atthe entrance of the post LEBT. 3 IG. 4: Transverse rms beam size along the post accelerator.FIG. 5: Normalized transverse rms beam emittance along the post accelerator.
B. Simulation results
Along the post accelerator, the beam pipe radius of each section is different, which is 6.0cm at the post LEBT, 2.5 cm at the post MEBT, and 2.0 cm at the SCL3, respectively.Those beam pipe radii were determined by considering the beam energy and the optimizationof the cavities. Therefore, the beam size should be less than the beam pipe radius to avoidthe beam loss at each section. In view of such condition, the strength of each magnet wascalculated and the start-to-end simulation was conducted. Figure 4 shows the transverseroot-mean-square (rms) beam size along the post accelerator. Especially, the rms beam sizeis kept less than 0.5 cm at the SCL3 and the beam transmission after the post RFQ is about98 %, which is one of the advantages of the RAON accelerator.Figure 5 shows the normalized transverse rms beam emittance along the post accelera-tor. After passing through the RFQ, the energy spread becomes larger, which affects thetransverse emittance growth. Also, the vertical emittance increases at the SCL3 because ofthe effect of the electro-magnetic field of the QWR cavity.The normalized longitudinal rms beam emittance along the post accelerator is shown in4
IG. 6: Normalized longitudinal rms beam emittance along the post accelerator.TABLE II: Normalized transverse and longitudinal rms emittance at some positions of the postaccelerator.Parameter Horizontal [mm · mrad] Vertical [mm · mrad] Longitudinal [keV/u.ns]End of RFQ 0.113 0.111 0.750End of MEBT 0.113 0.116 0.758End of SCL31 0.114 0.119 0.742End of SCL32 0.117 0.120 0.856End of SCL33 0.119 0.125 0.889 Figure 6. After the beam is longitudinally bunched by the post RFQ, the energy spreadincreases because of the long tail of the beam and thus the longitudinal emittance increases.Table II shows the normalized transverse and longitudinal rms emittance at some position ofthe post accelerator. The growth of the horizontal, transverse and longitudinal emittancesfrom the end of the RFQ to the end of the SCL33 is about 5.3 %, 12.6 % and 18.5 %,respectively. The minimum requirement for the normalized transverse rms emittance at thelow energy experimental halls is about 2 mm · mrad which is much larger than the result ofsimulation.Figure 7 shows the transverse and longitudinal beam distributions at the end of eachsection. As mentioned above, the beam is longitudinally bunched by the post RFQ and itleads to the growth of the longitudinal beam emittance.5 IG. 7: Beam distribution in the phase space along the post accelerator: (a) post LEBT exit (b)post RFQ exit (c) post MEBT exit (d) SCL3 exit.
III. ERROR SIMULATIONSA. Error tolerance
In the SCL3, the beam orbit can be distorted by a variety of error sources like magnetmisalignment, cavity field error and so on. After the beam loss caused by the orbit distortionis checked, the effect of each error source can be verified. In this paper, the tolerance of eacherror source is determined when the beam loss caused by each error source becomes smallerthan 0.1 %. The TRACK code is used for the calculation of the tolerance and the resultsare listed in Table III. In the simulations, the errors of quadrupoles and cavities are givenby Gaussian distribution with the rms value and truncated at the 3 times of the rms value.Among the error sources, the tolerance of the quadrupole transverse rms misalignment isabout 0.018 cm and it is the most dominant one among all error sources.6
ABLE III: Tolerance of error sources at the SCL3.- - Value UnitInitial position x +0.68/-0.65 [cm]Initial position y +0.27/-0.25 [cm]Initial angle xp +8.0/-7.9 [mrad]Initial angle yp +4.0/-3.7 [mrad]Cavity rms misalignment x,y 0.12 [cm]Cavity rms misalignment z 0.08 [cm]Cavity rms field amplitude 2.1 [%]Cavity rms field phase 0.9 [deg]Quadrupole rms misalignment x,y 0.018 [cm]Quadrupole rms misalignment z 1.2 [cm]Quadrupole rms tilt 16.2 [mrad]FIG. 8: Layout of the SCL3 in RAON accelerator.
B. Orbit correction
Figure 8 shows the schematic layout of the SCL3. For the orbit correction, a horizontalcorrector and a beam position monitor (BPM) are located at first quadrupole and a verticalcorrector is located at second quadrupole at each warm section. With the correctors andBPMs, the orbit correction for the distorted orbit is carried out at the SCL3.The errors used in the orbit correction simulations are listed in Table IV. The beam orbitis distorted by these errors and then the orbit correction is carried out by using the correc-tors and BPMs with the singular value decomposition (SVD) method. The rms transversemisalignment of the quadrupole, which is the most dominant error source, changes from 100 µ m to 500 µ m. 7 ABLE IV: Errors used in the orbit correction simulations.Parameter Value UnitInitial position x,y 10 µ mInitial angle xp,yp 10 µ radCavity rms misalignment x,y 10 µ mQuadrupole rms misalignment x,y 100 – 500 µ mFIG. 9: GUI for the orbit correction in RAON accelerator with the quadrupole misalignment 300 µ m and 200 random seeds.TABLE V: Summary of the orbit correction for the quadrupole misalignment 300 µ m and 200random seeds.Parameter Value Unit < x > rms,BP M before correction 48.9 µ m < y > rms,BP M before correction 36.3 µ m < x > rms,BP M after correction 2.9 µ m < y > rms,BP M after correction 2.1 µ mAverage horizontal corrector strength 0.30 mradAverage vertical corrector strength 0.27 mrad IG. 10: Rms orbit size at BPMs before and after the orbit correction.FIG. 11: Maximum and average corrector strength for the orbit correction.
For the orbit correction in RAON accelerator, the GUI based on the MATLAB programand the DYNAC code has been developed. Figure 9 shows the GUI screen after the orbitcorrection with quadrupole misalignment 300 µ m and 200 random seeds. As listed in Ta-ble V, the horizontal (vertical) rms beam size decreases from 48.9 (36.3) µ m to 2.9 (2.1) µ m after the orbit correction. The average corrector strength is less than the mechanicalmaximum value.Figure 10 shows the rms orbit size at the BPMs before and after the orbit correctionfor 100–500 µ m rms quadrupole misalignment with 200 random seeds. For the quadrupolerms misalignment from 100 µ m 500 µ m, the rms orbit size decreases about -90 %. Themaximum and average corrector strength for the orbit correction is shown in Figure 11.For some random seeds, the corrector kick angle becomes larger, but the average value isless than 0.6 mrad which is much less than the mechanical maximum kick angle, about 2.7mrad. The research for the proper number and position of the correctors and BPMs will be9ontinued. IV. CONCLUSIONS
We presented the results of the start-to-end simulations with the rare isotope beam in thepost accelerator of the RAON accelerator. For the post accelerator, the new reference beam, Sn , from the ISOL system was tracked from the post LEBT to the superconductinglinac SCL3. At the end of the SCL3, the beam energy reached at about 27.7 MeV/u and therms beam size along the post accelerator was kept much less than the beam pipe radii. Theerror analysis and correction were also performed at the SCL3. The tolerance of each errorsource was calculated and the beam orbit distorted by the errors was corrected with theSVD method. Additionally, for various error sources, the orbit distortion was checked andcorrected with the correctors and BPMs. As a result, the average kick angle of correctors forthe orbit correction was much less than the mechanical maximum value. The research forthe proper number and position of the correctors and BPMs will be carried out continuously. Acknowledgments
This work was supported by the Rare Isotope Science Project of Institute for Basic Sciencefunded by Ministry of Science, ICT and Future Planning and National Research Foundationof Korea (2013M7A1A1075764). [1] S. Kim et al. , Baseline Design Summary, http://risp.ibs.re.kr/orginfo/info_blds.do (2012).[2] D. Jeon et al. , J. Korean Phys. Soc. , 7 (2014).[3] H. Jin et al. , J. Korean Phys. Soc. , 8 (2015).[4] H. Jin et al. , Rev. Sci. Instrum. , 2 (2016).[5] H. Jin et al. , Nucl. Instr. and Meth. A , 65 (2015).[6] E. Tanke, TH429, LINAC02 (2002).[7] H. Jin et al. , MOPJE017, IPAC15 (2015).[8] V. Aseev et al. , TPAT028, PAC05 (2005)., TPAT028, PAC05 (2005).