1. Vertical Orbit-Excursion FFAGs (VFFAGs) and 3D Cyclotrons I.Principle & Magnetic Fields II.Proton Driver Study III.Isochronous 3D Cyclotrons June 2014 Stephen Brooks, IPAC’14 sbrooks@bnl.gov 1

2. I. Principle & Magnetic Fields June 2014Stephen Brooks, IPAC’142 In an FFAG (or cyclotron) the orbit moves across the magnet aperture during acceleration

3. Horizontal Aperture SC Magnet Getting vertical B field requires same-direction current windings on opposing sides B y proportional to x/(a 2 +x 2 ): cancels at x=0! June 2014Stephen Brooks, IPAC’143 a

4. Constructive Interference Getting horizontal B field requires opposite current windings and is easier B x proportional to a/(a 2 +x 2 ) June 2014Stephen Brooks, IPAC’144 a

5. Vertical Aperture SC Magnet But now the field is in the wrong direction! That's OK, rotate the magnet The dipole field is there But what sort of focussing does this magnet give? June 2014Stephen Brooks, IPAC’145

6. “Scaling” VFFAG Magnet Dipole field should increase moving up the magnet, so set B y = B 0 e ky on axis (x=0) Subtracting dipole component leaves the field of a skew quad: June 2014Stephen Brooks, IPAC’146 Exponential is good because moving upwards just scales the field and all gradients Thus closed orbits at different momenta are exactly the same shape, just translated upwards VFFAG = Vertical orbit excursion FFAG

7. Scaling VFFAG Field & Scaling Law June 2014Stephen Brooks, IPAC’147

8. FODO Scaling VFFAG Machine First VFFAG tracking simulation, for HB2010 – 2D, zero space charge, nonlinear magnets June 2014Stephen Brooks, IPAC’148 150mm.mrad  geom input beam Proton-driver-like but nasty circumference factor! (C=17)

9. June 2014Stephen Brooks, IPAC’149

10. Scaling VFFAG with Mismatch June 2014Stephen Brooks, IPAC’1410

11. VFFAG Acceleration June 2014Stephen Brooks, IPAC’1411

12. 2D Winding Model for Magnet June 2014Stephen Brooks, IPAC’1412

13. Application: Hadron Therapy? Low intensity but high rep-rate machines – Fixed field useful, space charge not too bad Small beams – The VFFAG magnet can be a narrow vertical slot – Less stored energy, smaller windings required Fixed tune allows slower acceleration, less RF Disadvantage: we still have the FFAG extraction-from-an-orbit-that-moves problem June 2014Stephen Brooks, IPAC’1413

14. Historical References “FFAG Electron Cyclotron” (Ohkawa, 1955) – T. Ohkawa, Physical Review 100 p.1247, abstract (1955) – Talk on isochronous electron VFFAG with exponential field, with and without edge focussing “Helicoidal FFAGs” (Leleux, 1959) – G. Leleux, J. Proy and M. Salvat, Rapport OC 70, Service de Physique Appliquee Section d’Optique Corpusculaire (1959) – Linear optics analysis of VFFAG “Accelerators with Vertically Increasing Field” (Teichmann, 1960-2) – J. Teichmann, translated from Atomnaya Énergiya, Vol.12, No.6, pp.475–482 (1962) – Isochronous, fixed tune electron VFFAG, exponential field, suggestion of curved orbit excursion June 2014Stephen Brooks, IPAC’1414

15. II. Proton Driver Study June 2014Stephen Brooks, IPAC’1415

16. Motivation: ISIS Energy Booster FFAG to accelerate the 50Hz pulsed ISIS beam Energy: 800MeV – 12GeV Superconducting magnets Ring radius 52m (2x ISIS) could do 2.5x,3x Mean dipole field in magnets 0.47 – 4.14T 30% RF packing, 20% magnets, 2-4m drifts Warm 6.2 – 7.3MHz RF Harmonic number 8 (10,12 in larger ring) June 2014Stephen Brooks, IPAC’1416

17. Scaling (V)FFAG disease Defocussing requires reverse bending, as in scaling FFAGs  large circumference Searched for “lopsided” scaling VFFAG lattices with good dynamic aperture [HB2010] – 10000 particles were tracked for 1km – Survival rate plotted on axes of lengths of “F” and “D” type magnets – This reveals both the lattice stability region and resonance stop-bands June 2014Stephen Brooks, IPAC’1417

18. June 2014Stephen Brooks, IPAC’1418

19. June 2014Stephen Brooks, IPAC’1419

20. Lattices can't be very lopsided Unfortunately in all cases the region of dynamic stability sticks very close to the F=D diagonal line The 2 nd FDF stability region does not have enough dynamic aperture So basic scaling VFFAGs will always be big, with much reverse bending – Could edge focussing avoid reverse bends? June 2014Stephen Brooks, IPAC’1420

21. Vertical Edge Focussing [HB2012] June 2014Stephen Brooks, IPAC’1421 Superconducting magnets allow this to be smaller than synchrotron designs at lower energies

22. VFFAG with Edge Focussing June 2014Stephen Brooks, IPAC’1422 y z  =0  =L   = tan   = z –  y Scaling law:

23. Spiral Scaling VFFAG Magnet Field June 2014Stephen Brooks, IPAC’1423 5GeV 800MeV

24. June 201424Stephen Brooks, IPAC’14

25. Cell Beta Functions Doublet focussing nature – Visible in u,v planes FfD – Doublet controlled by  – Singlet controlled by k Ring tune sensitivity: June 2014Stephen Brooks, IPAC’1425

26. June 2014Stephen Brooks, IPAC’1426

27. June 2014Stephen Brooks, IPAC’1427

28. June 2014Stephen Brooks, IPAC’1428

29. 12GeV VFFAG RF Programme June 2014Stephen Brooks, IPAC’1429

30. Longitudinal Intensity Effects June 2014Stephen Brooks, IPAC’1430

31. June 2014Stephen Brooks, IPAC’1431

32. III. Isochronous 3D Cyclotrons June 2014Stephen Brooks, IPAC’1432 New: first successful tracking May 16 th, 32 days ago

33. Isochronous Cyclotron Disease Mean radius must satisfy r =  R – Where R = c/2  f rev is the limiting radius as v  c Mean B y = p/qr = m  c/q  R =  (mc/qR) =  B 0 This produces a quadrupole as radii bunch up: – dB y /dr = (B 0 /R)d  /d  = (B 0 /R)  3 Momentum only increases with  – Eventually quadrupole overfocusses the beam – Energy limit for any given planar cyclotron June 2014Stephen Brooks, IPAC’1433

34. Tilted Orbit Excursion June 2014Stephen Brooks, IPAC’1434  Teichmann (1962) also had idea

35. Extrapolation from Curved Surface Define C(x,y,z)=B(x,Y(x,z)+y,z) for a reference surface y=Y(x,z) so that C(x,0,z) is the initial condition. Transforming Maxwell's equations in free space to act on C gives: June 2014Stephen Brooks, IPAC’1435

36. 3D Cyclotron Field Model Spiral angular coordinate Isochronous sector field form: Must satisfy: Sectors constructed using June 2014Stephen Brooks, IPAC’1436

37. June 2014Stephen Brooks, IPAC’1437

38. June 2014Stephen Brooks, IPAC’1438

39. June 2014Stephen Brooks, IPAC’1439

40. Orbit Locations at Matching Plane June 2014Stephen Brooks, IPAC’1440

41. Deviation from Theoretical Orbit June 2014Stephen Brooks, IPAC’1441

42. Eigentunes Stable at High Energy June 2014Stephen Brooks, IPAC’1442

43. Focussing Eigenplanes in X-Y Space June 2014Stephen Brooks, IPAC’1443

44. Isochronism ±0.62% June 2014Stephen Brooks, IPAC’1444

45. Fields on Orbits <7T for R=3.1m June 2014Stephen Brooks, IPAC’1445

46. Comparison with Planar Cyclotron June 2014Stephen Brooks, IPAC’1446

47. Conclusion & Applications Non-isochronous proton VFFAGs provide rapid cycling, high fields and compact SC magnet designs – Neutron, neutrino and hadron therapy sources Isochronous proton 3D cyclotrons promise cyclotron- like CW acceleration above 1GeV – High cross sections for nuclear/spallation reactions – Nuclear waste transmutation, isotope production Electron VFFAGs are already almost isochronous – Multi-pass VFFAG-ERL-FELs with only one recirculating arc June 2014Stephen Brooks, IPAC’1447

48. Future Work Better alignment between the reference plane and the orbits, giving better convergence Field adjustments to improve isochronism from the current ±0.62% variation Vary  e with energy to improve tune control Make space for RF by using fewer sectors and/or more field-free space between sectors Find an example superconducting winding scheme Build electron model of the 3D cyclotron June 2014Stephen Brooks, IPAC’1448