Physics design of the CIADS 25 MeV demo facility

Nuclear Instruments and Methods in Physics Research A 843 (2017) 11–17

Contents lists available at ScienceDirect

Nuclear Instruments and Methods in Physics Research A journal homepage: www.elsevier.com/locate/nima

Physics design of the CIADS 25 MeV demo facility☆ a,⁎

a

a

a,⁎

a

a

Shu-Hui Liu , Zhi-Jun Wang , Huan Jia , Yuan He , Wei-Ping Dou , Yuan-Shuai Qin , Wei-Long Chena, Fang Yanb a b

crossmark

Institute of Modern Physics, the Chinese Academy of Sciences, Lanzhou 730000, China Institute of High Energy Physics, the Chinese Academy of Sciences, Beijing 100049, China

A R T I C L E I N F O

A BS T RAC T

Keywords: CIADS Beam dynamics Beam matching Frequency jump Error analysis

A superconducting linac has been proposed and under constructed to demonstrate the key technology and the feasibility for CIADS(China Initiative Accelerator Driven System)linac. This linac will accelerate the 10 mA proton beam to 25 MeV. There are some challenges in the physics design for the high power superconducting accelerator. In this paper, we focus on the matching between different cryomodules (CMs) and the frequency jump. This paper presents the physics design study together with the design principles and the simulation results with machine errors.

1. Introduction

2. Design philosophy and consideration

China Initiative Accelerator Driven System(CIADS) is a strategic plan to solve the nuclear waste problem and the resource problem for nuclear power plants in China. It aims to design and build an Accelerator Driven Subcritical System (ADS)demonstration facility in multiple phases lasting about 20 years. The driven linac will deliver a 1.5 GeV, 10 mA proton beam in CW operation mode. The general layout is shown in Fig. 1. The driver linac is composed of two major sections. One is the normal conducting section and the other is the superconducting (SC) section. The normal conducting section is composed of an electron cyclotron resonance (ECR) ion source with frequency of 2.45 GHz, a low energy beam transport (LEBT) line, a four-vane type copper structure radio frequency quadrupole (RFQ) with frequency of 162.5 MHz and a medium energy beam transport (MEBT) line. The normal conducting section will accelerate proton beam to 2.1 MeV. The SC section as the main accelerating section will accelerate the proton beam from 2.1 MeV up to 1.5 GeV. Then, the beam is transported to the beam dump going through the high energy beam transport (HEBT) line. For a high intensity proton superconducting linac running in CW mode, there are no operational experiences in the world so far. In order to overcome the challenges in technology and in physics, an experimental device with an energy of 25 MeV is proposed and constructed. Fig. 2 shows the schematic layout of the 25 MeV demonstration facility. In this paper, the beam dynamics design and simulation for LEBT, RFQ, MEBT, and SC section will be presented in detail.

Hands on maintenance and machine protection set strict limits, 1 W/m and 0.1 W/m respectively, on beam losses and have been a concern in high power linacs [1]. Therefore it is crucial to design a linac, which does not excite beam halo and keeps the emittance growth at a minimum level to avoid beam loss. Given the demands of stability and reliability, some guidelines are required to be considered in the design process. Although a lot of the design philosophy for the linac has been addressed in previous literature, we still consider some of them so important to be stated here, and the most important factors in designing our machine are the following: (1) Transverse period phase advances for zero current beams should be below 90° to avoid the structure resonance [2]. (2) Wave numbers of oscillations need change adiabatically along the linac, especially at the lattice transitions with different types of focusing structure and inter-cryostat spaces [3]. (3) Avoid strong space charge resonances through the judgment of Hofmann's Chart [4–7]. (4) Minimize the emittance growth and beam halo formation caused by mismatching in the lattice transition section. (5) Enough redundancy to avoid the beam loss along the linac. 3. Physics design of different section

☆ ⁎

The 25 MeV demo facility consists of ECR ion source with 2.45 GHz

Supported by the National Natural Science Foundation of China (Grant No. 91426303 and NO. 11525523) Corresponding authors. E-mail addresses: [email protected] (S.-H. Liu), [email protected] (Y. He).

http://dx.doi.org/10.1016/j.nima.2016.10.055 Received 7 June 2016; Received in revised form 27 October 2016; Accepted 28 October 2016 Available online 01 November 2016 0168-9002/ © 2016 Elsevier B.V. All rights reserved.

Nuclear Instruments and Methods in Physics Research A 843 (2017) 11–17

S.-H. Liu et al.

3.2. RFQ The CW RFQ plays a critical role in the CIADS driven linac. So, for this demonstration facility, a four vane RFQ has been designed in collaboration with the Lawrence Berkeley National Laboratory and fabricated at the workshop of the Institute of Modern Physics, Chinese Academy of Sciences [11]. This RFQ works at the frequency of 162.5 MHz and accelerates the proton beam of 15 mA from 35 keV to 2.1 MeV. The choice of the 162.5 MHz frequency is beneficial for the small surface power density because of its big size. Another advantage is the large aperture for lower frequency, that means larger acceptance. The main parameters of the RFQ are listed in Table 1. The total accelerator length of the RFQ is 4.208 m. It is composed of four coupled physical segments and each segment includes four technical modules connected together with flanges. Full model of the RFQ is shown in Fig. 5, and it has 16 pairs of pi-mode rods and 80 tuners [12]. The beam transport simulation along the RFQ was carried out using the PARMTEQM [13] code with a 4D-Waterbag input distribution, and 100,000 macroparticles were assumed for the initial distribution. Fig. 6 shows the output phase space distribution of the RFQ. The simulated RFQ output distribution has been used for the design of the SC linac.

Fig. 1. The general layout of the CADS linac.

Fig. 2. Schematic layout of the 25 MeV demo facility.

frequency, LEBT, RFQ, MEBT, SC section and beam dump, as shown in Fig. 2. The 35 keV proton beam from the ion source is bunched and accelerated to 2.1 MeV by a 162.5 MHz RFQ. The LEBT is used to match the beam between the ion source at the RFQ, and also to provide chopped beam for commissioning. The SC section accelerates proton beam from 2.1 to 25 MeV employing HWR cavities (162.5 MHz) and Spoke cavities (325 MHz). The beam dump line follows right after the SC segment.

3.3. MEBT The MEBT is used to match the beam between the RFQ and the SC sections with low emittance growth. There are seven quadrupoles and two bunching cavities to match the beam with the desired Twiss parameters in both transverse and longitudinal directions as shown in Fig. 7. As seen from Fig. 8, there are two beam waists formed in the middle and at the end of MEBT by three upstream quadrupoles and four downstream quadrupoles of the first bunching cavity [14]. MEBT is designed to contain diagnostic devices to measure the beam parameters after the RFQ cavity. There are four BPMs residing inside the quadrupoles and one BPM exists at the end of the MEBT. Two Alternating Coupled Current Transformers (ACCT) in the front and at the end of MEBT are placed to measure the beam current. Two sets of emittance scanning devices composed by slit wire scanner installed in the D-box (diagnostics box) to get the vertical and horizontal emittance in the middle of MEBT, One double direction wire scanner is installed between the sixth quadrupole and the seventh quadrupole. One Faraday cup (FC) is also placed inside the D-box.

3.1. Ion source and LEBT For the 25 MeV demo facility, an ECR ion source with 35 keV is chosen. A LEBT downstream the ECR ion source is used for matching between the ion source and the RFQ. The LEBT consists of two solenoid lenses, vacuum pumps, and beam diagnosis aperture. The length of the LEBT is 1670 mm. A cone has been installed at right before the RFQ entrance to substantially reduce the unwanted particles, such as 2H+ and 3H+, from getting into the RFQ [8,9]. The schematic figure of the LEBT and the chopper is shown in Fig. 3. This accelerator will be a CW machine, but the pulsed beam is an essential choice for the beam tuning stage. The front end needs to have the ability to provide adjustable beam. The particle trajectories along the LEBT and the phase space distribution with matched Twiss parameters at the RFQ entrance by the TRACK [10] code are illustrated in Fig. 4.

3.4. Superconducting section One of the most difficult problems is to efficiently accelerate the beam from the low energy at the RFQ exit to higher energy while maintaining beam quality at the same time [2]. For the 25 MeV superconducting linac, low-beta HWR cavities with β = 0.10 and β = 0.15 in 162.5 MHz and spoke cavities with β = 0.21 in 325 MHz were chosen to accelerate the beam from the 2.1 MeV at the RFQ to 25 MeV. The fabrications of the HWR cavities with β = 0.10 and β = 0.15 have been completed and the horizontal tests have successfully performed. In the rest part of this paper, we focus on the beam matching at the transitions and the frequency jump. 3.4.1. Beam matching at the transitions Beam matching at the transitions is very important for minimizing emittance growth and beam halo formation. Next, the longitudinal matching is discussed. In the absence of space charge, the rms envelope evolution meets the following equation:

R″ −

ϵ =0 R3

(1)

We assume ϵ remains constant, so beam size is proportional to β . The longitudinal β function in a drift obeys the relation:

Fig. 3. The layout of the LEBT.

12

Nuclear Instruments and Methods in Physics Research A 843 (2017) 11–17

S.-H. Liu et al.

Fig. 4. The envelope evolution in the X direction(left) and Simulated phase space distributions by TRACK code(right). Table 1 Main design parameters of the RFQ. Parameters

Value

Frequency(MHz) Input energy(MeV) output energy(MeV) Beam current(mA) Vane voltage(kV)

162.5 0.035 2.1 15 65 0.3

trans . norm . RMS ϵin (π mm mrad) Minimum apertures(cm) Kilpatrick factor(cm) Cavity length(cm) Beam transmission(%) . norm . RMS ϵtrans (π mm mrad) out

3.2 1.2 420.8 99.6 0.31 0.92

. norm . RMS ϵlong (keV.ns) out

Fig. 6. Output phase space distribution of the RFQ.

Fig. 5. Full model of the RFQ.

β (z ) = β * +

(z − z*)2 β*

(2)

where * is the waist position. Let z* = 0 , namely, the waist position is the original point, and L1, L2 denote distances from the upstream and the downstream end to the waist position. For a fixed value of L1, the minimum value of the β function at the upstream extremity occurs when

L2 d (β ) = 1 − 12 d (β*) β*

β* = L1, β = 2L1.

Fig. 7. The layout of MEBT.

minimum β for the drift region is obtained when L1 = L 2 = L . In that case,

(3)

β ( ± L) 2L = =2 β* L

The β function at the downstream extremity is similarly. So, a 13

(4)

Nuclear Instruments and Methods in Physics Research A 843 (2017) 11–17

S.-H. Liu et al.

Fig. 9. Schematic view of the lattice structures for the 25 MeV SC accelerating segment: (a) for HWR010, (b) for HWR015, (c) for spoke021.

k 02l =

2

(5)

In the presence of space charge, the rms envelope equation in a drift takes the form:

K ϵ R″ − − 3 = 0, R = R R

βϵ

KR2 + ϵ ϵ = R″ − 13 = 0 R3 R

(6)

ϵ1 > ϵ

(7)

From the above equation, we can get that the effect of space charge can induce an increase in emittance, but it does not change the optimal ratio, namely,

σ = σ*

2σ =

2 × σ* =

2 ×

β*ϵ

(8)

For the downstream cavity extremity,

σ=

β ϵ′

(9)

The ϵ′ presents the emittance of the downstream cavity extremity. And then,

β ϵ′ = σ 2 → β =

σ2 ϵ′

2fγf πqT sin ϕs mc 3βs3 γs3

(11)

3.4.3. Lattice design The SC segment of 25 MeV demo facility contains twelve HWR010 cavities, six HWR015 cavities and six spoke021 cavities housed in four CMs. The lattice structure is shown in Fig. 9. We have chosen two HWR cavities working at 162.5 MHz with β of 0.10 and 0.15, respectively, and a single-spoke cavities working at 325 MHz with β of 0.21, respectively. The main design parameters of the cavities are listed in Table 2. In the nominal design, only 75% of the maximum cavity voltage is used because of the element failure compensation requirements, and another 25% is reserved for the compensation, and this redundancy is also beneficial for the cavity reliability. The phase advance in the three planes are kept below 90° to avoid the parametric resonance [16–18]. The focusing fields in both the transverse and longitudinal directions are kept almost constant in each section to have almost constant envelope amplitude when the rms emittance is shrinking along the acceleration [19]. This also means constant phase advance in each section, but the absolute value of the synchronous phase decreases from the lower-energy section to the higher-energy section to obtain higher acceleration rate. Because of the limitation in the longitudinal phase advance per cell, a big synchronous phase (absolute value) has to be kept to ensure longitudinal acceptance, therefore, the cavity voltages at the beginning parts of the SC section may not be fully exploited. The synchronous phase evolution and the cavity field of the SC section of the 25 MeV demo facility are shown in Fig. 10. The SC segment of the 25 MeV demonstration facility consists of four CMs. Two problems need to be discussed here. The first is the matching between CMs. The distance between the two CMs has to be kept as short as possible to avoid emittance growth and beam halo formation caused by mismatching. In our case, 680 mm is required for the CMs warm to cold transition from the flange of the last cold

where K > 0 , and it depends on the charge density. Rearranging the equation (6),

R″ −

=

In our case, to keep the same longitudinal acceptance, the synchronous phase is generally doubled after the frequency jump. To smooth the variation of the average phase advance across the frequency jump, the accelerating gradient of the first section downstream cavity is reduced. In our design, the scheme of keeping constant phase acceptance at the frequency jump, as well as phase advance per meter smoothing is valid.

Fig. 8. RMS envelope evolution along MEBT.

σz ( ± L ) = σz*

2πq sin ϕs mc 2βs3 γs3 λ

(10)

The longitudinal matching scheme is developed as the following steps: (1) first, drop the field next to last cavity; (2) second, iterate until the beam size in the last cavity is near the theoretical optimum; (3) third, adjust the field in the last cavity to get a waist in the center. The transverse matching is similar.

Table 2 The rf errors and misalignments amplitude for the injector error analysis.

3.4.2. Frequency jump To allow for a larger longitudinal acceptance at a low energy and also to ease the difficulty of the workload of component fabrication, a lower frequency is used at the front end of the ion accelerators [15]. With the rising energy, it is beneficial to increase the frequency to one of the higher harmonics of the bunch frequency to get a higher accelerating gradient and also a smaller size of the cavity. However, the frequency change causes an abrupt change in the average focusing forces in the longitudinal plane at the frequency jump according to Eq. (11), if handled incorrectly, it will cause redistribution of charged particles inside the bunch, emittance dilution and halo generation. 14

Error type

static (buncher/ cavity)

dynamic (buncher/cavity)

static Q/ solenoid

dynamic (Q/ solenoid)

δx (mm) δy (mm) Rx(mrad) Ry(mrad) Rz(mrad) δg (%) δ δφ (○)

0.1/1 0.1/1 2 2 × 0.5 0.5

0.002/0.01 0.002/0.01 0.02 0.02 × 0.25 0.05

0.1/1 0.1/1 2 2 2 0.5 ×

0.002/0.01 0.002/0.01 0.02 0.02 0.02 0.05 ×

δz (mm)

0

0

0

0

Nuclear Instruments and Methods in Physics Research A 843 (2017) 11–17

S.-H. Liu et al.

Fig. 10. Synchronous phase(left) and the Voltage(right) of SC segment cavities.

element of the first CM to the flange of the first cold element of the second CM. There are no warm BPMs between CMs because of the space limitation. The second problem is the frequency jump between the third CM and the fourth CM. The frequency jump may induce a discontinuity in the average longitudinal force per focusing period and shrink the longitudinal acceptance of the linac if this transition is not performed carefully. The phase and the field at frequency transitions should be tuned, including acceptance and phase advance per unit length issues [20]. The synchronous phase of the first cavity at the latter CM should be twice of the first cavity at the former CM. Meanwhile, the field of the corresponding cavities should be decreased to keep the phase advance per meter constant. 3.4.4. Beam dynamics Beam dynamics of the 25 MeV superconducting linac are carried out with TRACEWIN [21] code which adoptes 3D fields for cavities and solenoids. Fig. 14 shows the designed zero current phase advances per period. The tune depression evolution is shown in Fig. 11. The linac design is not equipartitioned and the working points are positioned in a relatively large region of stable area between the kl / kt = 1 and kl / kt = 2 stop bands on the Hofmann stability chart as shown in Fig. 12 [22]. The Hoffman stability chart is defined by initial rms emittance ratio, and the picture is plotted by kxy/kxy0 on ordinate axis as a function of kz/kxy on abscissa axis. The chart in color code indicates the theoretically expected growth rate of emittance transfer caused by resonant action of space charge. The blank area in between the resonances indicates absence of emittance coupling. Fig. 13 shows the RMS envelope evolution and the RMS emittance growths of the 25 MeV superconducting linac. The envelope is smooth at the transition of both CMs. A periodic lattice was designed to be a reference

Fig. 12. Tune footprint of SC section for 25 MeV demo facility.

using truncated 4σand 5σ(transverse and longitudinal) Gaussian initial distribution. The normalized RMS emittance growths are 10.2%, 2.3% and 1% (x, y and z planes) respectively. During the design, the normalized RMS emittance growths are considered as one major criterion for determining the matching results between the two CMs. 3.4.5. Error analysis The errors from assembly alignment, RF jitters and variations are considered in error analysis. The beam orbit distortion may lead to direct beam loss in the downstream section. The beam off-centering can be corrected by a proper correction scheme that employs a number

Fig. 11. Phase advance at zero current(left) and the Tune depression(right) along the SC segment.

15

Nuclear Instruments and Methods in Physics Research A 843 (2017) 11–17

S.-H. Liu et al.

Fig. 13. RMS envelope evolution(left) and RMS emittance evolution(right) along the linac.

Fig. 14. The transverse particle density (graphs on the left and on the right) with errors and with corrections.

Fig. 15. The normalized RMS emittance growth and beam loss with errors left graph and corrections right graph.

tolerance of the basic design and estimate the beam properties evolution with or without corrections using the TraceWin program. All the errors are generated randomly and uniformly distributed between the maximum values. 1000 seeds are generated randomly for the error analysis. Fig. 14 shows the error analysis comparison with correction and without correction. The upper graphs show the transverse particle densities and the longitudinal power density with errors and corrections. As can be seen in the figure, the envelope evolutions

of correction magnets together with some beam position monitors (BPM). Radio frequency errors and misalignments are considered for all the optics in this section, including buncher/cavity errors, quadrupole/solenoid misalignments and field gradient fluctuations et al., as shown in Table 2. These numbers were determined according to the measured realistic results and experiences on FRIB ReA3 of MSU [23] and TRIUMF [24]. The error analysis is performed including MEBT and SC section. The error analysis was carried out to verify the

16

Nuclear Instruments and Methods in Physics Research A 843 (2017) 11–17

S.-H. Liu et al.

China, TUO3B02. [2] Z.H. Li, P. Cheng, et al., Physics design of an accelerator for an accelerator-driven subcritical system, Phys. Rev. ST Accel. Beams 16 (2013) 080101. [3] Petr Ostroumov, Fermilab Accelerator Advisory Committee, May 10th–12th, 2006. [4] I. Hofmann, G. Franchetti, J. Qiang, R. Ryne, F. Gerigk, D. Jeon, N.Pichoff, in: Proceedings of the 8th European Particle Accelerator Conference, Paris, 2002, edited by J. L. Laclare (EPS-IGA and CERN, Geneva, 2002), p. 74. [5] I. Hofmann, G. Franchetti, Phys. Rev. ST Accel. Beams 9 (2006) 054202. [6] I. Hofmann, in: Proceedings of the HB2012, Beijing, China, 2012, TUO3A01. [7] L. Groening, I. Hofmann, W. Barth, W. Bayer, G. Clemente, L. Dahl, P. Forck, P. Gerhard, M.S. Kaiser, M. Maier, S. Mickat, T. Milosic, S. Yaramyshev, Phys. Rev. Lett. 103 (2009) 224801. [8] Q. Wu, Z.M. Zhang, L.T. Sun, Y. Yang, H.Y. Ma, Y. Cao, X.Z. Zhang, H.W. Zhao, Rev. Sci. Instrum. 85 (2014) 02A703. [9] Y. Yang, Z.M. Zhang, Q. Wu, W.H. Zhang, H.Y. Ma, L.T. Sun, X.Z. Zhang, Z.W. Liu, Y. He, H.W. Zhao, D.Z. Xie, Rev. Sci. Instrum. 84 (2013) 033306. [10] V.N. Aseev, P.N. Ostroumov, E.S. Lessner, B.Mustapha, in: Proceedings of the 21st Particle Accelerator Conference, Knoxville, 2005, edited by C. Horak (IEEE, Piscataway, NJ, 2005), p. 2053. [11] Zhang Zhouli, He Yuan, Shi Aimin et al., in: Proceedings of the IPAC2014, Dresden, Germany, THPME027. [12] Chuan Zhang, GSI Helmholtz Center for Heavy Ion Research, Planckstr. 1, Darmstadt, Germany Chen Xiao, Institute for Applied Physics, Goethe-University, Frankfurt a. M., Proceedings of IPAC2013, Shanghai, China, THPWO022. [13] 〈http://www.laacg.lanl.gov/laacg/services/servcodes.phtml〉. [14] J.I.A. Huan, H.E. Yuan, Yuan You-Jin, Chin. Phys. C 39 (10) (2015) 107003. [15] M. Eshraqi, H. Danared, R. Miyamoto, in: Proceedings of the HB2012, Beijing, China, TUO3B02. [16] F. Gerigk, in: Proceedings of the 2002 Joint USPAS-CAS-Japan-Russia Accelerator School, 2002, pp. 257–288. [17] Li Zhi-hui, Tang Jing-Yu, Yan Fang, Geng Hui-Ping, Meng Cai, Sun Biao, Cheng Peng, Guo Zhen, Sun Ji-Lei, Chin. Phys. C. 37 (2013) 037005. [18] F. Gerigk, I.Hofmann, in: Proceedings of the 19th Particle Accelerator Conference, Chicago, Illinois, 2001 (Ref. [11]). [19] M. Reiser, N. Brown, Phys. Rev. Lett. 74 (1995) 1111. [20] M. Eshraqi, H. Danared, R. Miyamoto, Phys. Rev. Spec. Top. - Accel. Beams 10 (2007) 084201. [21] 〈http://irfu.cea.fr/Sacm/logiciels/index3.php〉. [22] I. Hofmann, Stability of anisotropic beams with space charge, Phys. Rev. E 57 (1998) 4713. [23] X. Wu, Accelerating Physics R & D for the Superconducting Linac Projects at Michigan State University, visiting report in IHEP, 2013. [24] R.E. Laxdal, et al., Cryogenic, magnetic and rf performance of the ISAC-II medium beta crymodule at TRIUMF, in: Proceedings of the 21st Particle Accelerator Conference, Knoxville, TN, 2005 (IEEE, Piscataway, NJ, 2005), iscataway, NJ, 2005).

are smooth, and the beam is under control longitudinally and there are relatively bigger margins on transverse planes. The horizontal direction, longitudinal direction and longitudinal normalized rms emittance growths are 6.5%, 6.2% and 13.7% respectively with errors and corrections for the basic design. If there is no correction, the corresponding horizontal, vertical and longitudinal emittance growths are 10.3%, 10.1% and 23.8%, respectively and beam losses appear as shown in Fig. 15. 4. Summary The physics design of the 25 MeV demonstration facility of CIADS is presented, including the ion source, LEBT, RFQ, MEBT and superconducting accelerating segment. For the SC linac, the beam dynamics was presented with emphasis on the method to handle the frequency and beam matching. The linac is designed respecting the rules of the thumb in high intensity ion linacs. Period phase advance at zero current less than 90° are considered to avoid resonance and to reduce the possible beam losses. The sensitivity of the linac to a set of defined errors was checked, and the results of the error study were presented in this paper. The simulation results show that the beam quality is preserved in the nominal case, and even when some errors are applied the linac can transport and accelerate the beam without any losses. Acknowledgements The authors would like to express their sincere acknowledgment to the colleagues in the CIADS accelerator team and especially the beam dynamics group for their comments, suggestions and discussions. Special thanks are expressed to Prof. Zhihui LI from Sichuan University for his kind helps and useful discussions. The work is supported by the CAS Strategic Priority Research Program-Future Advanced Nuclear Fission Energy (Accelerator-Driven Sub-critical System) (XDA03000000). References [1] M. Eshraqi, H. Danared, R. Miyamoto, in: Proceedings of the HB2012, Beijing,

17