Yongxiang Zhao Brookhaven National Laboratory Upton, NY, July 1998 CENTER FOR ACCELERATOR PHYSICS

Transcription

1 BNL CAP CCII, MUON-98C A NEW STRUCTURE OF LINEAR COLLIDER * Yongxiang Zhao Brookhaven Nationa Laboratory Upton, NY, RECEIVED AIK OSTI *This work was supported by the US Department of Energy under Contract No. DE-AC02-98CH Juy 1998 CENTER FOR ACCELERATOR PHYSICS BROOKHAVENNATIONAL LABORATORY BROOKHAVEN SCIENCE ASSOCIATES Under Contract No. DE-AC02-98CH10886 with the UNITED STATES DEPARTMENT OF ENERGY Submiued to Ihc 61hErrropuur Pam& Acceemor Conference, EPAC-98. sorktom. S~vedrn.June

2 DISCLAIMER This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their empoyees, makes any warranty, express or impied, or assumes any ega Iiabiiity or responsibiity for the accuracy, competeness, or usefuness of any information, apparatus, product, or process discosed, or represents that its use woud not infringe privatey owned rights. Reference herein to any spccific commercia product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessariy constitute or impy its endorsement, ncommendation. or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessariy state or refect those of the United States Government or any agency thereof.

3 DISCLAIMER Portions of this document may be iegibe eectronic image products. images are produced from the best avaiabe origina document.

4 . A NEW STRUCTURE OF LINEAR ACCELERATOR Yongxiang Zhao, Brookhaven Nationa Laboratory, NY,USA Abstract transverse dimension[]. Besides the beam transit time is R, consequenty the transit time factor is ow. A traveing wave structure can work at 2x13 ord2 mode has higher transit time factor, but the fied wi decay aong the structure and a damper at the end is needed. Not ony it compicates the structure, but aso wastes the power. One possibe soution is to recyce the remained power by connecting the outet to the inet to form a traveing wave resonator ring. This arrangement wi make use of the avaiabe power. However, the oop structure requires two ports on the acceerator, it is difficut from mechanica consideration, that there are soenoids surrounding the acceerator. Therefore, a new structure is proposed [2] and addressed as beow. A new scheme is proposed for the muon coider cooing RF, on which the cavity is amost a simpe pibox. The principe is to divide the acceerating ces into three separate channes with a phase difference of 2 d 3 between each other. A specia structure is designed in which each channe is ike a side-couped cavity structure operating at d2-mode, but an extra phase shift ofn is introduced so that the phase difference between two ces is 2x. Therefore, the beam sees a 2d3-mode impying a high interaction between cavities and beam, whie the RF system sees ad2-mode impying a high stabiity. In other words, it combines the advantages of both traveing wave and standing wave. The mechanica structure is aso simpe and compact because the side-couped cavity is 2 THE PRINCIPLE OF THE 2d3-MODE formed by a uniform rectanguar-ikewaveguide, which is INTERLEAVED STRUCTURE simpy an arc section attached on the main cavities. The Fig.1 shows the principe. At the input of the principe and the preiminary simuation are addressed in detai. This principe is aso appicabe for a 2 d 4 mode acceerator a 3-branches power divider is inserted so as to and other machine provided the couping due to aperture form three independent channes. There is a phase difference of 2 d 3 between each other. Each channe wi is weak enough. connect ony to one of every three ces of the acceerator structure, i.e. channe-1 ony coupe with those ces 1 INTRODUCTION numbered #1, #4, #7... channe-2 with #2, #5, #8... The muon coider is considered as an important fbture channe3 with #3, #6, #9... The ces are couped by the project for the high energy physics. One of the chaenges side cavities to form a side couping acceerator structure. is to coo the muon beam. It requires very high gradient As usua, it works in d2-mode; i.e. the adjacent cavity with reativey ow frequency, say 800 MHz. Since the has d 2 phase shift. tota RF power is very arge, minimising the RF power is of much concern. Meanwhie a high magnetic fied has to be appied for beam focusing. Consequenty, the acceeration structure is required to be compact as we. To meet above requirements, a few issues were considered. A norma cavity with a beam aperture, the ratio of maximum surface fied to the acceeration gradient is greater than 2 or more. If a pibox cavity is used, with the aperture for beam path being cosed by a beryium foi window, the ratio is one and thus enhances the acceeration gradient within the breakdown imit. Beryium is a ight meta causing itte oss when a muon beam passing through. Another measure is to appy iquid nitrogen to coo down the cavity so that significanty 4n 0 2x 4n 0 2x 4% o -2n3-4n3 0 2n reduces the RF power requirement. f f 3-37-T The acceerator structure is aso an important issue. A few structures were considered. The standing wave structure has highest impedance thus east power Fig. 1 The principe diagram requirement, but poor mode separation eads poor stabiity. Side-couped cavity structure (SCS) working at d2-mode can overcome this probem, but it is asymmetric making structure compex and has much arge If

5 a- The schematic structure is shown in Fig. 2. The sidecouped cavity is simpy a piece of rectanguar-ike 3 SIMULATION MAFIA simuation was done to verify the principe. The chosen parameters are isted beow: Acceeration ces: Ce Diameter 0.285m CH- 1 Ce ength 0.1Om. Period (3 ces) 0.312m Side cavity: Outer diameter m Inner diameter 0.296m Span ange 79.74deg Length 0.306m Couping sot: Spanange 45deg Width 0.01m CH-2 CH-3 Off edge 0.03m Fig.3 shows the resut of MAFIA simuation. The fixquency is 795 MHz. We chose one side couped cavity and two haf of acceeration cavities of one channe. The centra part refers to cavities of other channes. It was Fig. 2 Schematic structure excuded because they are independent without energy couping. It shows ceary that the fied in both waveguide with its wide side bent. Its ength is a haf of acceeration ces is in phase or 2x, whie the fied in the waveguide waveength. On its both sides there are two side cavity is pretty sma. This verifies the argument sots coupiig to two acceeration ces. The bottom of addressed in previous section. Fig. 2 shows the magnetic fied pattern in the sidecouped cavity, ceary, magnetic fied at both ends of a cavity has opposite direction, and it causes an extra phase shift of x. Thus between two acceeration ces there are 2x phase shift. Fig.2 shows the case of a traveing wave f h m eft to right with d2-mode. The arrows show the direction of energy fow. For a d2-mode, between two adjacent cavities, i.e. cross each sot, there s d 2 phase shift. Across two adjacent sots, there sx phase shift. As shown on the magnetic pattern, the directions of the magnetic fied on two sots are opposite each other, meaningx shift. However, the whoe pattern of two adjacent cavities is the same, meaning they are in phase or 2x shift. That is because an extra A is introduced as mentioned above. Fig.3 The eectric fied pattern from MAFIA Now, et the eft most and right most acceeration cavities serve as short end, then the wave wi refect and form a standing wave. In this case, the side couped 4 DESIGN APPROACH FOR THE BIcavities wi ocate at wave node, so that the fied inside PERIODIC STRUCTURE them are very weak. Therefore the side-couped cavity can The structure in question is bi-periodic. Besides, the be made with sma height without the risk of breakdown. unique character of an extrax phase shift cas for care That is a thin waveguide attaching to the acceerator, with the simuations. On Fig.3, from eft boundary to making the structure very compact. right boundary there is 2x or 0 phase difference, it impies In brief, it combines both advantages of standing wave 0 degree phase advance must be assigned when that 2x or structure and traveing wave structure. From the m n MAFIA. But, from the viewpoint of RF circuit, it has viewpoint of beam it is a TW with 2x/3-mode, that has ony x phase shift not 2%. In fact, between two adjacent higher transit time factor than that of a x-mode. cavities there is ony d 2 phase shift. This can aso be Meanwhie from the viewpoint of the RF source it ooks MAFIA running with periodic structure of 0 verified by as standing wave structure, with high impedance and no degree phase advance. The fied pattern of Ere and Eim, power wasted. Besides, it retains the advantage of d 2 i.e. the rea and image part of the fied or the fieds in 90 mode of the best mode stabiity. degree phase difference, show the eectric fied wi concentrated in either acceeration cavity or side couped cavity. Thus it is ad2-mode indeed and shoud have a it-xgn -4 I.

6 the d2-mode properties. The foowing formua shoud be kept in mind. +-=2++II where I+is the vaue assigned to MAFIA fie,+ is the phase refemng to the mode. Fig.4 shows the f+ curve, i.e. a-p diagram. The operation point is chosen at the midde,d2, (at MAFIA run it is 0 degree) where the group veocity is maximum and has maximum mode separation. It is aso a cross point of two modes, i.e. acceeration mode and side couping mode. In this case the frequency of 795MHz wi no ong be at d2-mode7 instead it is at n-mode, that this structure is regarded as a singe periodic, whie the side cavity is off resonance. Then the phase reationship shoud +-=I+-n In this case a the advantages ofd2-mode has ost. Therefore carefuy adjusting the iiequency is important for the design. By virtue of MAFIA, a resut of matched frequency has been obtained as shown in Figd. 5. SOME COMMENTS The above discussed structure has many advantages. It works on standingwave with high impedance and without damper in comparison with traditiona traveing wave structure. It works on d2-mode and avoid the shortcomings of simpe standing wave structure working at Ic-mode with zero group veocity making ong fiing time and poor stabiity due to poor mode separation. It has very simpe structure and compact size in comparison with a common side couped cavity structure, which usuay encounter difficuties. This principe can aso be appied to four channes with (I an (10 90 iao isn IW a d 2 phase difference between each channe that beam PHASE ADVANCE sees a 2 d 4 mode. Two channe intereave structure has Figd The dispersion curve been used ong time ago. However, present simpe side couped cavity is not appicabe because the ength of the waveguide-ike cavity must be onger thadg2, whie the Note that the two modes are in junction ony if the ength of beam trave of j3a.d is ess than that. resonant frequencies of acceerator and side cavities are The disadvantage of the structure is the requirement of equa or cose enough. Otherwise the two modes wi a three-channe high power spitter, which makes the separate. Fig.5 shows the case. input system being compex. 6 ACKNOWLEDGEMENT This work was supported by the US DOE under Contract No. DE-ACO2-76CHOOO16. The author was especiay indebted to R Pamer for his enightening comments. Thanks aso to many coaborators for beneficia discussions. REFERENCES t -180.a0 * m m PHASE ADVANCE tao a i T. G. Jurgens, H. W.Mier, A. Moretti and P. Zhou, "The determination of the 805 M H z side couped cavity dimensions for the Fermi Lab inac upgrade", Roc Linear Acceerator Conference, Sep. 1014, Fig.5 Two pass band not matched E21 Y. X Zhao, "The simuation of 2d3-mode Fig.5 is the case that the side-couped cavity has a itte intereaved structure", Muon FW workshop, Laurence Berkeey Laboratory, Feb , 1998 ower frequency. The ower pass band corresponds to side couping mode, whie the upper band acceeration mode.