Canted–Cosine–Theta Magnet(CCT)—A Concept for High Field Accelerator Magnets S.Caspi,F.Borgnolutti,L.Brouwer,D.Cheng,D.R.Dietderich,H.Felice,A.Godeke,R.Hafalia, M.Martchevskii,S.Prestemon,E.Rochepault,C.Swenson,and X.Wang
Abstract—Canted–Cosine–Theta(CCT)magnet is an acceler-ator magnet that superposfields of nested and tilted solenoids that are oppositely canted.The current distribution of any canted layer generates a pure harmonicfield as well as a solenoidfield that can be cancelled with a similar but oppositely canted layer. The concept places windings within mandrel’s ribs and spars that simultaneously intercept and guide Lorentz forces of each turn to prevent stress accumulation.With respect to other designs, the need for pre-stress in this concept is reduced by an order of magnitude making it highly compatible with the u of strain nsitive superconductors such as Nb3Sn or HTS.Intercepting large Lorentz forces is of particular interest in magnets with large bores and highfield accelerator magnets like the one foreen in the future high energy upgrade of the LHC.This paper describes the CCT concept and reports on the construction of CCT1a“proof of principle”dipole.
Index Terms—Accelerator magnets,Canted–Cosine–Theta magnet,CCT,highfield,superconducting dipole.
I.I NTRODUCTION
F OUR different types of superconducting accelerator mag-
nets have been built and tested over the years:Cosine-Theta,Common-coil,blocks withflared ends,and blocks with stress management.While each of the types attempted to ad-dress the incread complexity of brittle Nb3Sn superconductor, crossing10T and especially going beyond15T has proven difficult and unpredictable.For reasons,none more important than coil pre-stress,expected progress[1],[2]resulted in fewer successful attempts of magnets actually reaching their short-sample design limit.It became apparent that if we want to reachfields of20T yet another novel approach will have to be introduced that reduces conductor stress in magnets.With the major issue being the accumulative Lorentz forces between turns,intercepting such forces has been the primary goal of stresd managed coils[3].That approach us supporting beams to subdivide coils into blocks that intercept forces, however,within each block some degree of pre-stress is still using springs).Reducing the pre-stress to its bare minimum suggests blocks that are no larger than a single turn each.Doing so will cau the tangential Lorentz forces(at a
Manuscript received July15,2013;accepted September25,2013.Date of publication October4,2013;d
ate of current version October25,2013.This work was supported in part by the Director,Office of Science,High Energy Physics,U.S.Department of Energy under contract DE-AC02-05CH11231,and by the National Science Foundation under Grant DGE1106400.
The authors are with the Lawrence Berkeley National Laboratory,Berkeley, CA94720USA(e-mail:***************).强图片
Color versions of one or more of thefigures in this paper are available online at ieeexplore.ieee.
Digital Object Identifier
10.1109/TASC.2013.2284722Fig.1.Two nested layers with canted windings creating a dipole magnet. 20T level)not to exceed a conductor stress of25–30MPa. Tofirst order,conductor under such a low pre-stress may have reached a point where no pre-stress is needed at all. The intercepted Lorentz forces however do not disappear and will have to be carried by a strong internal structural support. Fusion magnets and very highfield solenoids have been using a similar approach by applying“cable-in-conduit”conductor to address stress interception(at a cost of a reduced current-density).Keeping a high current density in accelerator magnets is important and could be achieved if structural elements are combined with typicalfield-shape wedges.The CCT promis to accomplish that and in addition suggests the reduced func-tionality of an external supporting structure and replacing it with an internal structure that is part of the coils.
This paper describes advancements in the design of an inter-nal structure that intercepts individual turns by placing them into pre-machined channels around a mandrel(Fig.1).The magnetic concept wasfirst introduced by Meyer and Flasck in a1970publication[4],suggesting that the transver current density in canted solenoids cloly rembles that of a“perfect”cosine-theta magnet generating a“pure”dipolefield(Fig.2). The concept was sub-quentially considered and expanded in veral publications[5]–[11].Section II details a two-layer design of a CCT dipole using NbTi conductor.Section III discuss the construction status and future plans.
II.C ANTED–C OSINE–T HETA M AGNET—CCT1
A.CCT1Magnet
CCT1is a NbTi dipole magnet built to demonstrate CCT technology and determine its feasibility for highfield magnets using Nb3Sn conductor and HTS.The magnet is expected to reach a short-samplefield of2.6T within a50.8mm clear bore at a current of3660A.The cable,wrapped with braided s-glass insulation,is wound into pre-machined channels of Aluminum mandrels and then impregnated.Both layers are
U.S.Government work not protected by pyright.
Fig.2.Transver cut showing a “pure”cosine-theta current density distribu-tion with intercepting ribs and
spars.
Fig.3.CAD cross ction of a two-layer CCT1showing canted winding.
TABLE I
CCT1M AGNET P外交学院分数线
ARAMETERS
wound continuously and do not require an internal splice.A subasmbly of coils and Aluminum pads will be inrted into a cond subasmbly of an iron yoke and outer Aluminum shell (Fig.3)and the final asmbly will be completed using keys and bladders technology [12].The magnet parameters are listed in Table I.
A complete magnetic and structural analysis was carried out and the magnet is prently under
construction.
Fig.4.Varying size ribs attached to a spar guide and support the
求职信格式
conductor.
Fig.5.
喝雄黄酒Cross ction of a single lamination of coils ribs and spar.
B.Coil Winding Path and Mandrel
The eight strands Rutherford cable follows a winding path that intercts a plane and a cylinder at a
given angle (15deg.).The cable,simulated as a 4corners cross-ction plane,is placed normal to the winding direction and is oriented (with its narrow edge)tangentially to the cylinder (normal to its radius vector).Wound around the cylinder the cable is incrementally advanced ending with a final axial pitch of 7.60mm over one turn.The rib size between turns varies from 0.38mm at the mid-planes to 6.0mm at the poles (Fig.4).The mandrel channels were machined from a 6mm wall Aluminum cylinder at a depth of approximately 3mm leaving the rest as a supporting spar (Fig.12).C.Lamination
狮子座简谱
The entire winding mandrel can also be made from an asm-bly of identical laminations (a pitch thick)(Fig.5).Asmbling the laminations generates a continuous channel for the entire coil windings (Fig.6).Beside the impact on cost and eddy currents loss the laminations provide a major simplification in three-dimensional (3D)analysis.With proper symmetrical boundary conditions a 3D magnetic and structural model was made to determine the field and stress.(hardware laminations were not implemented in the CCT1).D.Magnetic Model
The CCT1TOSCA 3D model was created from 8node bricks simulating the entire two layers.Over the central straight-ction the field quality reflects the “purity”of the transver
CASPI et al.:CANTED–COSINE–THETA MAGNET (CCT)—A CONCEPT FOR HIGH FIELD ACCELERATOR MAGNETS
4001804
Fig.6.
Single turn fits within ribs of a single pitch
lamination.
Fig.7.Harmonics calculated by TOSCA (no-iron)at various bore radii.B1is the dipole and its value has been
truncated.
Fig.8.TOSCA field profiles along CCT1bore center (no-iron).
cosine-theta current density distribution in the windings.Fig.7shows the normal local harmonics in CCT1on a radius between 65%and 85%of the winding radius.Along the magnet “ends”the winding symmetry between any two layers naturally inte-grates all harmonics to zero [13].Fig.8shows the field profile along the magnet axis.
When iron is also considered a reduced model size is re-stricted to the volume of a meshed lamination (Fig.9).The con-ductor properties and the expected short-sample performance,with and without iron,are listed in Table II.In certain cas a two-dimensional cross-ction cut was made of the conductor and ud with the magnetic program Poisson (Fig.
10).
Fig.9.3D TOSCA model of windings and a laminated meshed field.
TABLE II
CCT1M AGNETIC P
ARAMETERS
Fig.10.POISSON flux plot of CCT1showing the cosine-theta current density distribution.
III.L ORENTZ F ORCES AND S TRESS
The main advantage of the CCT is its capability to intercept Lorentz forces by ribs and spars.If we assume a rigid CCT mandrel structure without a turn-to-turn mechanical interaction,the Lorentz force will be confined tangentially by the channel walls and radial by an outer layer spar.The are the only force components needed to restrain the turn.Fig.11is a plot of the CCT1tangential and radial stress around a conductor turn.The tangential stress at the mid-planes (0and 180degrees)
4001804IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY,VOL.24,NO.3,JUNE
2014
Fig.11.Biot-Savart calculations of the tangential and radial Lorentz stress around a single turn of CCT1at 5
心疼的句子kA.
Fig.12.Photo of machined aluminum mandrills of CCT1layers 1and 2[15].
is 0,and below 5MPa at the poles (facing the axial direction).We note that the corresponding rib thickness at both locations,minimum at the mid-plane and maximum at the pole,reflect the structural need to constrain the force.Results of a complete 3D ANSYS stress analysis are reported in [14].
If the CCT1was designed as a “conventional”cosine-theta (CT)dipole a calculated pre-stress of 55MPa would have been required.Despite the low field and the low Lorentz force,the high-current density at that field in combination with a small NbTi cable still yields a substantial pre-stress.With stress interception,the pre-stress of the CCT1is less than 5MPa,an order of magnitude lower than a CT design.At such a low level the internal structure of ribs and spars may be sufficient to handle both azimuthal and radial forces without a supplemental external structure.Prently with a supplemental external struc-ture,the stress change in CCT1,between cool-down and 5kA (30%over the short-sample limit),is barely noticeable in the conductor with some affect on the spars in the form of bending.Accordingly,the plan for the CCT1is therefore to check its performance with and without an external structure followed by a cond model that replaces the prent NbTi conductor with Nb 3Sn.Replacing the conductor is expected to more than double the field and provide a good measure of comparison.In addition at 6T,where the pinning forces and therefore the Lorentz forces in Nb 3Sn,
are clo to their maximum,the impact of stress interception is expected to be significant.In
parallel,we are developing Bi-2212coil technology bad on the CCT layout,and intended as inrt for CCT1.
IV .C ONCLUSION
CCT1demonstrates the CCT concept as an accelerator mag-net.Pending the CCT12.6T magnet performance,the NbTi conductor will be replaced with Nb 3Sn to study the reaction process as well.Results on field quality and especially training will be a key factor in proving the concept’s future capabilities.The magnet test is planned to take place in early fall 2013.
R EFERENCES
[1]S.Caspi,“A perspective on LBNL 1m long Nb 3Sn superconducting
dipole magnets,”in LBNL Engineering Note 10674.Lawrence Berkeley Nat.Lab.,Jan.2012.
[2]D.Sutter and B.Strauss,“Next generation high energy physics collid-ers:Technical challenges,and prospects,”IEEE Trans.Appl.Supercond.,vol.10,no.1,pp.33–43,Mar.2000.
[3]P.McIntyre and A.Sattarov,“On the feasibility of a tripler upgrade
for LHC,”in Proc.Part.Accel.Conf.,Knoxville,TN,USA,2005,pp.634–636.
[4]D.I.Meyer and R.Flasck,“A new configuration for a dipole magnet for
u in high energy physics applications,”Nucl.Instrum.Method ,vol.80,no.2,pp.339–341,Apr.1970.狗咬主人
[5]C.L.Goodzeit,M.J.Ball,and R.B.Meinke,“The double-helix dipole—
A novel approach to accelerator magnet design,”IEEE Trans.Appl.Su-percond.,vol.13,no.2,pp.1365–1368,Jun.2003.
[6]A.V .Gavrilin,M.D.Bird,V .E.Keilin,and A.V .Dudarev,“New concepts
in transver field magnet design,”IEEE Trans.Appl.Supercond.,vol.13,no.2,pp.1213–1216,Jun.2003.
[7]A.Devred, B.Baudouy, D. E.Baynham,T.Boutboul,S.Canfer,
M.Chorowski,P.Fabbricatore,S.Farinon,H.Félice,P.Fessia,J.Fydrych,V .Granata,M.Greco,J.Greenhalgh,D.Leroy,P.Loverige,M.Matkowski,G.Michalski,F.Michel,L.R.Oberli,A.den
Ouden,D.Pedrini,S.Pietrowicz,J.Polinski,V .Previtali,L.Quettier,D.Richter,J.M.Rifflet,J.Rochford, F.Rondeaux,S.Sanz, C.Scheuerlein,N.Schwerg,S.Sgobba,M.Sorbi,F.Toral-Fernandez,R.van Weelderen,P.Védrine,and G.V olpini,“Overview and status of the next european dipole joint rearch activity,”Supercond.Sci.Technol.,vol.19,no.3,pp.67–83,Mar.2006.
[8]S.Caspi,D.R.Dietderich,P.Ferracin,N.R.Finney,M.J.Fuery,
S.A.Gourlay,and A.R.Hafalia,“Design,fabrication,and test of a superconducting dipole magnet bad on tilted solenoids,”IEEE Trans.Appl.Supercond.,vol.17,no.2,pp.2266–2269,Jun.2007.
[9]S.Caspi,D.Arbelaez,H.Felice,R.Hafalia,D.Robin,C.Sun,W.Wan,
and M.Yoon,“Conceptual design of a 260mm bore 5T superconducting curved dipole magnet for a carbon beam therapy gantry,”IEEE Trans.Appl.Supercond.,vol.22,no.3,p.4401204,Jun.2012.
[10]S.Caspi,D.Arbelaez,L.Brouwer,D.R.Dietderich,H.Felice,R.Hafalia,
S.Prestemon,D.Robin,C.Sun,and W.Wan,“A superconducting magnet mandrel with minimum symmetry laminations for proton therapy,”Nucl.Inst.Methods ,vol.719,no.11,pp.44–49,Aug.2013.
[11]F.Bosi,P.Fabbricatore,S.Farinon,U.Gambardella,R.Munich,
R.Marabotto,and E.Paoloni,“Compact superconducting high gradi-ent quadrupole magnets for the interaction regions of high luminosity colliders,”IEEE Trans.Appl.Supercond.,vol.23,no.3,p.4001004,Jun.2013.
[12]S.Caspi,S.Gourlay,R.Hafalia,A.Lietzke,J.O’Neill,C.Taylor,and木兰诗的作者
A.Jackson,“The u of pressurized bladders for stress control of su-perconducting magnets,”IEEE Trans.Appl.Supercond.,vol.11,no.1,pp.2272–2275,Mar.2001.
[13]L.Jackson Laslett,S.Caspi,and M.Helm,“Configuration of coil ends
for multipole magnets,”in Particle Accelerators .New York,NY,USA:Gordon and Breach,1987,pp.22,1–14.
[14]L.Brouwer,S.Caspi,H.Felice,S.Prestemon,and E.Rochepault,“Struc-tural design and analysis of canted cosine theta dipoles,”IEEE Trans.Appl.Supercond.,vol.24,2014,to be published.
[15]A.Hafalia,S.Caspi,H.Felice,L.Brouwer,and S.Prestemon,“The
structural design for a ‘canted cosine-theta’superconducting dipole coil and magnet structure—CCT1,”IEEE Trans.Appl.Supercond.,vol.24,2014,to be published.
