and the lower revolution frequency above the transition energy. Transition is the point where particles of all energy have the same revolution frequency. Thus there is no longitudinal phase stability at the transition point. One has to jump the accelerating phase of the RF at a precise moment of the transition. The transition gamma jump system makes the transition crossing easier. There are new lattice designs, which avoid transition or lattices with imaginary transition, however there is no such synchrotron constructed to date. Figure 2 shows the intensity history of the AGS, which is the highest intensity synchrotron operating at present. As can be seen in the figure, there are discrete steps which can be identified with higher injection energy or H^£­ charge exchange injection. In the 1980's, there were several proposal to build synchrotrons whose effective proton currents are ten to hundreds times higher than today's AGS. They were the AHF^[14] at Los Alamos, JHF^[15] in Japan, EHF^[16] in Europe and KAON^[17] at TRIUMF in Canada. None of these proposal have materialized, last of them was canceled in early 1990's. Post booster AGS intensities were adequate enough and expected to accomplish most of the physics objectives of these facilities and it was partially responsible for their demise. While preparing this manuscript the author has learned that redesigned new JHF^[21] has been approved in Japan. Figure 3 shows the intensity history of synchrotrons of the world presently operating and planned to be constructed in the future.

When the intensity records are combined with the additional capabilities of rapid cycling, a very high power pulsed proton accelerator is achievable. High intensity pulsed neutron sources are the subject of the next chapter.

Synchrotrons are not the only accelerators being developed to support higher intensity operations. There has recently been considerable development in proton linac technology. In addition to the development of the RFQ and higher frequency proton linacs, better understanding of beam dynamics of halo formation in a linac has led to the development of coupled cavity drift tube linac (CCDTL) to bridge between the low energy DTL and the high energy CCL. The development of superconducting RF cavity technology has led to a real possibility of constructing high current continuous wave proton linac. With these accelerator developments, the possibility of providing beams of many megawatts or hundreds of megawatts beam power is reachable.

PROTONS AS NEUTRON SOURCE

Traditionally the major source of epithermal to thermal neutrons has been nuclear reactors. They have been excellent continuous sources of neutrons. Although higher power reactors are needed, it has became next to impossible to construct any new reactor in today's political climate. The desire to have intense pulsed sources where the wavelength of the interacting neutrons is measured by time-of-flight has led neutron scientists to propose the use of a proton synchrotron as a spallation source. There are several pulsed neutron sources operating in the world, IPNS at Argonne, ISIS at Rutherford, KENS at KEK, and PSR at Los Alamos. The efficiency of the proton beam power producing neutron is shown in Figure 4. As can be seen, the efficiency is approximately 20 % between 500 MeV and 10 GeV. The proton energy range required for a spallation source is flexible. There have been many proposals to build accelerator based neutron sources. Virtually every accelerator laboratory in the world has proposed a machine over a wide range of proton energy. Only a few projects are promising at present. Table 2 lists the accelerator based neutron sources presently operating. Also included are the SNS^[19] and the ESS^[20] which are likely to be constructed in the near future. An artist's concept of the SNS is shown in Figure 5. Also included in the table is the AGS, which is the highest intensity per pulse synchrotron with very low repetition rate as comparison. The repetition rate is one of the important factors one has to consider for planning a high power synchrotron complex. The limitations of the power generated by a spallation sources arise from the beam current limitations of the injector linac and the rapid cycling synchrotron or the storage ring.

The enormous beam power one has to deal with brings new concerns. Even a minute amount of beam loss in the system causes insurmountable maintenance problem for the accelerators. For a beam power of 1 MW, even 0.1 % loss means 1 kW of beam power lost on the accelerator components.

In order to have hands on maintenance of the complex, one should not lose any more than 1 nanoampare of the beam current per meter of system, which translates for the SNS ring to less than 10^£­4 loss. This is a situation that has never been faced before.

Table 2. List of accelerator based neutron source.

Laboratory

Devise

Type

Energy (Ring/Linac)

Pulse width

Rep. Rate

Current

Power

Argonne, USA

IPNS

RCS

450/50 MeV

< Ռsec.

30

15 ՌA

6.8 kW

Rutherford, UK

ISIS

RCS

800/70 MeV

< Ռsec.

50

200 ՌA

160 kW

KEK, Japan

KENS

RCS

500/40 MeV

< Ռsec.

20

4.6 ՌA

2.3 kW

Los Alamos, USA

PSR

SR

800/800 MeV

< Ռsec.

20

70 ՌA

56 kW

Los Alamos, USA

LANCE

Linac

800 MeV

1 msec.

60

1.25 mA

1 MW

PSI, Switzerland

SINQ

Cyclotron

590 MeV

CW

CW

1.5 mA

885 kW

KEK, Japan[21]

JHF Booster

RCS

3 GeV/200 MeV

< Ռsec.

25

200 ՌA

600 kW

Oak Ridge, USA

SNS

SR

1/1 GeV

< Ռsec.

60

1 mA

1 MW+

Europe

ESS

SR

1.334/1.334 GeV

< Ռsec.

50

3.75 mA

5 MW++

BNL, USA

AGS

SCS

28 GeV/200 MeV

2.7 Ռsec.

0.6

6 ՌA

140 kW

RCS; Rapid cycling synchrotron, SR; Storage ring, SCS; onsists of two rings. Slow cycling synchrotron. ⁺ Upgrade to 2 MW. ⁺⁺ consists of two rings.

Although H^£­ stripping injection provides great freedom for the phase space painting of the protons, it causes non-negligible losses at the injection point. Because of foil heating problem, the foil can not be arbitrarily thick to strip all the H^£­s. In the case of SNS injection, for carbon foils thicker than 400 Ռg/cm² the foils evaporate. Thus thinner foils are used. Those H^⁣ that do not strip to protons will emerge as H^o in various quantum states with varying life times in the magnetic field. This will create a significant amount of halo outside of the ring phase space. One has to create a scheme to minimize this halo This problem is common for both rapid cycling synchrotrons and accumulator rings except the loss occurs at a lower injection energy for rapid cycling synchrotrons.

The RF cavities will experience a new class of beam-loading problems. The circulating beam will interact with the vacuum wall by exchanging microwave energy causing the beam to become unstable. There have been many advances in this area over the last few decades, and one expects to cope with the new problems. However, injection and RF capture prior to acceleration without significant proton loss is not trivial. The synchrotron has to maintain full intensity beam for many milli-seconds during acceleration without loss, where as an accumulator storage ring only has to keep full beam intensity for the injection period.

Although a linac is much more expensive to build and less reliable than rapid cycling synchrotrons, the above mentioned problems leads one to choose a full energy linac and an accumulator storage ring for multi-megawatt spallation sources like the ESS and SNS. The debate will continue, however, over the choice of an accumulator storage ring or a rapid cycling synchrotron until one of them is built.

PROTONS FOR OTHER APPLICATIONS

Since very high power proton accelerators seem possible, other applications for intense protons beams have been proposed.^[22] All these proposals utilize the neutrons generated by accelerated protons. Typical of them are nuclear waste transmutation,^[23] energy amplifier^[24] and tritium production.^[25] For accelerator driven transmutation of waste, the neutrons are used to transmute long lived nuclear waste isotopes to shorter lived or stable elements. For the energy amplifier, a sub-critical reactor is supplemented by the neutrons generated by a spallation source. The reactor is safe because there is no possibility of a run away situation. The original energy amplifier proposal consisted of a cyclotron as a proton source. The consensus of a recent workshop^[26] is that a cyclotron is a reasonable choice for up to 2 MW of beam power but much R&D is needed beyond 2 MW. Linacs have become the choice for a high power spallation facility. Tritium has been produced very efficiently by reactors, however with present attitudes, one must consider a high power spallation source to produce tritium. The required beam power for this purpose is one to two orders of magnitude larger than any proposed pulsed spallation neutron source, and two to three orders of magnitude greater than any proton accelerator presently operating. Thus the beam loss requirement is many orders magnitude more severe. The accelerator must achieve reliability and availability greater than anything achieved before. Because of the stress on the target system one has to limit the number of sudden interruptions. This places sever requirements on reliability. According to the study done by the French TRISPAL project,^[26] the target system can accept no more than 10000 interruptions of 100 milliseconds or longer per year.

The development of superconducting RF cavity technology during the past decade seems to indicate that it is time to consider high intensity superconducting linac. Superconducting technology has certain advantages over conventional copper cavities. Since there is a negligible amount of cavity loss of RF power, considerably higher RF power efficiency is expected. The length of the linac can be reduced because much higher gradients are expected with superconducting cavities. However, there are many areas such as the construction of the cavity, cryogenic module, and RF coupler which need considerable development.

A larger beam aperture in a superconducting cavity can also be an advantage for reducing the beam loss. However, tolerance to instantaneous beam loss for a superconducting cavity is one or two orders of magnitude less than for a normal copper cavity. Because of the mechanical structure of cryogenic modules, room temperature quadrupoles in the machine are much further apart and the betatron amplitude tends to be larger than for a normal cavity. Also, only the halos which are bound (the halos created because of mismatch in focusing structure for example) can be contained in the aperture. Other halos like the protons that have slipped out of the longitudinal bucket are not bound and eventually reach the aperture. These beam halo loss processes are difficult to simulate and estimate.

Fig. 6. Schematic of KOMAC project in Korea.

Fig. 7. Schematic of APT project in USA.

Fig. 8. Schematic of TRISPAL project in France.

Figure 6, 7, 8 and 9 are the schematic diagrams of the Korean KOMAC,^[27] US APT,^[28] French TRISPAL^[29] and Japanese NSP^[30] projects.

Table 3. Ultra High Powered Accelerator Projects.

Inst.	Proj.	Power(MW)	Energy/Current	Ion Source	RFQ	DTL	CCDTL/SDTL	CCL
KAERI Korea	KOMAC	20	1.0 GeV 20 mA	50 keV Duo-plas'tron	3 MeV 350 MHz	―――	CCDTL 100 MeV 700 MHz	1.0 GeV SC 700 MHz
LANL USA	ATP	170	1.7 GeV 100 mA	75 keV ECR	6.7 MeV 350 MHz	―――	CCDTL 100 MeV 700 MHz	211 MeV NC 1.7 GeV SC 700 MHz
CEA France	TRISPAL	24	600 MeV 40 mA	95 keV ECR	5.0 MeV 352 MHz	29 MeV 352 MHz	SDTL 85 MeV 352 MHz	600 MeV NC 352 MHz
JAERI Japan	NSP	8	1.5 GeV 5.3 mA	70 keV Vol.	2.0 MeV 200 MHz	50 MeV 200 MHz	SDTL 100 MeV 200 MHz	1.5 GeV SC 600 MHz

summarizes these ultra high-powered proton accelerators. Since the KOMAC has been reported in detail,^[27] the following is a brief description of the other proposed facilities.

The US Accelerator Production of Tritium (APT) is the project to produce strategically needed tritium. Traditionally, needed tritium has been produced by specially designed nuclear reactors. These reactors are shut down and the choice has to be made either chose a reactor or develop an alternative method. In order to satisfy the strategically needed tritium in the US, it is estimated that 170 MW of proton power is needed. A 100 mA of 1.7 GeV specification has been chosen for this purpose. The accelerator consists of a 75 keV ECR source, a 6.7 MeV 350 MHz RFQ, a 100 MeV 700 MHz cavity coupled drift tube linac(CCDTL), a 217 MeV 700 MHz normal copper cavity coupled linac(CCL), followed by a 1.7 GeV superconducting CCL. There is a possibility to funnel two RFQs into one higher frequency CCDTL in order to achieve the required goal. The frequency chosen is to accommodate such a high proton current option. For example a 400 MHz RFQ may not efficiently accelerate 100 mA of proton current. At the moment, a demonstration accelerator(LEDA),^[31] 6.7 MeV RFQ is constructed and will be tested later this year. The decision whether to construct APT is expected soon.

The French TRISPAL is a project for TRItium production using SPALlation neutrons. The project is similar to the APT project to satisfy French strategic tritium need. This project proposes to build a 600 MeV, 40 mA CW, 24 MW proton linac. The linac consists of a 95 keV ECR source, a 5 MeV 352 MHz RFQ, a 30 MeV 352 MHz DTL, a 85 MeV Separated Drift Tube Linac SDTL, followed by a 600 MeV 352 MHz normal conducting copper CCL. They are planning to stay with the same 352 MHz structure through out in order to avoid troublesome bunch compression problem when the frequency of later structures are doubled. The project is only a conceptual study and there is no firm plan for construction.

The Japanese project NSP is Neutron Science Project developed at JAERI. It is an offshoot of their OMEGA(Option Making Extra Gain from Actinides and fission products) project. A two-stage development is planned. The first stage is to build a pulsed linac for a spallation neutron source, and the second stage is to upgrade to CW mode. The pulsed mode is to operate at 1.5 GeV, 50 Hz with an average current of 1 mA. The linac will then be upgraded to an average current <5.33 mA CW operation or pulsed operation with peak current of 30 mA. The accelerator system consists of a 70 keV volume source to operate either H⁺ or H^⁣, two 2 MeV 200 MHz RFQ one each for pulsed and CW, a 50 MeV 200 MHz DTL, a 100 MeV 200 MHz SDTL, followed by a 1.5 GeV 600 MHz superconducting CCL.

The frequency of the CCL section is three times greater than that of the lower energy sections. The addition of accumulator storage rings for pulsed mode to operate a short pulsed spallation source is also proposed.

Whatever the scheme or project proposed, beam loss is the major problem. A lot of work has been done to improve accelerator efficiencies over the past decades,^[32-34] however, proof of practical applicability is yet to come. It is the task of accelerator physicists and engineers to solve these problems, because one of the challenges facing the world is the energy problem. We have to face and solve the problems related to nuclear power since like it or not, nuclear power is one of the major components of energy supply for the coming days.

The author would like to thank Ms. M. Campbell, Mr. J. Tuozzolo, and Dr. D. Lowenstein for their help in preparing this manuscript.

REFERENCES

[1] R. J. Van de Graaf, 'A 1,500,000 volt electrostatic generator,'Phys. Rev. 387, 1990-20.

[2] J. D. Cockcroft and E. T. S. Walton, 'Experiments with high velocity ions', Proc. Royal Soc., Series A 136 (1932), p. 619.

[3] G. Ising, Arkiv for Matermatik. Astronomi och Fysik 18, 1 (1924).

[4] R. Wideroe, Arch. Fur Elektrotechnik 21, 386 (1928).

[5] E. O. Lowrence and N. E. Edlefsen, Science 72, 376 (1930).

[6] See, for example, P. Waloschek, 'The Infancy of Particle Accelerators,' DESY Report 94-039 (1994).

[7] D. W. Kerst, Phys. Rev. 60, 47 (1941).

[8] D. W. Kerst and R. Serber, Phys. Rev. 60, 53 (1941).

[9] E. M. McMillan, Phys. Rev. Lett. 68, 1434 (1945).

[10] V. Veksler, J. of Phys. USSR 9, 153 (1945).

[11] E. D. Courant, M. S. Livingston and H. S. Snyder, Phys. Rev. 88, 1190 (1952).

[12] E. D. Courant and H. S. Snyder, 'Theory of alternating-gradient synchrotron,' Annals of Physics, No. 3 (1958), p. 1.

[13] N. C. Christofilos, Unpublished Report (1950).

[14] Plan for 45GeV Facility, LA-10720-MS (1986).

[15] K. Imai et al., Proc. 5th Symposium on Accelerator Science and Technology, KEK (1984), p. 396.

[16] Proposal for a European Hadron Facility, edited by J. F. Crawford, EHF-87-18 (1987).

[17] KAON Factory Proposal, TRIUMF (1985).

[18] A. Boudard et al., Nuclear Physics News 8(1) (1998).

[19] SNS Conceptual Design Report (1997). The SNS is approved to be constructed from oct., 1998.

[20] ESS Conceptual Design Report (1997).

[21] Y. Mori, Proc. European Accelerator Conf. (1998).

[22] See preceding papers and their references.

[23] H. Takahashi et al., 'Use of linear accelerator for incinerating the fission product of Cs137 and Sr90,' Conf. on Nuclear Waste Transmutation, Austin TX, USA (1980).

[24] C. Rubbia et al., 'Conceptual Design of a Fast Neutron Operated High Power Energy Amplifier,' CERN/AT/95-44, (1995).

[25] Los Alamos National Laboratory APT Conceptual Design Report, LA-UR-97-1329 (1997).

[26] J-M Lagniel, 'A Review of Linacs and Beam' (1998).

[27] C. K. Park et al., 'The KOMAC project,' Proc. of APAC98 (1998).

[28] APT Conceptual Design Report, Los Alamos Report (1997).

[29] Rapport de faisabilite TRISPAL, CEA-DAM, Juin (1998).

[30] M. Mizumoto et al., 'High intensity proton accelerator for neutron science project at JAERI', Proc. EPAC (1998).

[31] J. D. Sherman et al., 'A DC proton injector for use in high-current CW linacs', Proc. of EPAC (1998).

[32] J-M. Lagniel, 'Halos and chaos on space charge dominated beams,' Proc. of EPAC96 (1996), p. 163.

[33] S. Y. Lee et al., 'Envelope hamiltonian of an intense charged particle beam in periodic solenoidal fields,' Phys. Rev. E 51 (1995).

[34] M. Pabst, 'Progress on intense proton beam dynamics and halo formation,' Proc. of EPAC98 (1998).

Key Words

1) ¾ç¼ºÀÚ °¡¼Ó±â (proton accelerator)

¼ö¼ÒÀÇ ¿øÀÚÇÙÀÎ ¾ç¼ºÀÚ¸¦ ³ôÀº ¼Óµµ ¶Ç´Â ¿¡³ÊÁö¸¦ °®µµ·Ï °¡¼ÓÇÏ´Â ÀåÄ¡ÀÌ´Ù. À̸¦ °¡¼ÓÇÒ ¼ö ÀÖ´Â ÀåÄ¡·Î´Â cyclotron, synchrotron µî ¿©·¯ Á¾·ù°¡ ÀÖÀ¸³ª, mAÀÌ»óÀÇ ´ëÀü·ù¸¦ ³ôÀº ¿¡³ÊÁö·Î °¡¼ÓÇÏ´Â ÀåÄ¡·Î´Â ¼±Çü°¡¼Ó±â°¡ ÀûÇÕÇÏ´Ù. ¼±Çü ¾ç¼ºÀÚ°¡¼Ó±â´Â Á÷¼±ÇüÀÇ ±¸Á¶¸¦ °¡Áö¸ç, À̿¿ø¿¡¼­ ¸¸µé¾îÁø ¾ç¼ºÀÚºöÀ» °íÁÖÆĸ¦ ÀÌ¿ëÇÏ¿© RFQ, DTL, CCL ¶Ç´Â ÃÊÀüµµ ¼±Çü°¡¼Ó±â µî ¼Óµµ¿¡ µû¶ó ¿©·¯ Á¾·ùÀÇ ¼±Çü °¡¼Ó±â·Î ¼øÂ÷ÀûÀ¸·Î °¡¼ÓÇÑ´Ù. ³ôÀº ¿¡³ÊÁöÀÇ ¾ç¼ºÀÚ´Â ÁÖ·Î ¹°¸®ÇкоßÀÇ ¿¬±¸¿¡ »ç¿ëµÇ¾úÀ¸³ª, ÃÖ±Ù¿¡´Â µ¿À§¿ø¼Ò »ý»ê, Áß¼ºÀÚ ¹ß»ý, ¾ç¼ºÀÚ Ä¡·á µî »ê¾÷ ÀÇ·áºÐ¾ß¿¡ »ç¿ëÇÏ´Â µî È°¿ëºÐ¾ß°¡ ¸¹ÀÌ È®´ë º¸ÆíÈ­µÇ°í ÀÖ´Ù.

2) ÇÙÆļâ (spallation)