• A. France, O. Delferrière, M. Desmons, Y. Le Noa, J. Novo, O. Piquet
  CEA, Gif-sur-Yvette

The RF de­sign of a RFQ should sat­is­fied sev­er­al con­di­tions, name­ly: volt­age pro­file re­quired by beam dy­nam­ics, a tun­able struc­ture, RF sta­bil­i­ty and rea­son­able sen­si­tiv­i­ty to pos­si­ble per­tur­ba­tions in­duced by power op­er­a­tion. Volt­age pro­file may be ob­tained ei­ther by a ded­i­cat­ed pro­fil­ing of 2D cross-sec­tion and/or slug tuner ad­just­ment. Tun­abil­i­ty is di­rect­ly re­lat­ed to spa­tial dis­tri­bu­tion of tuners. RF sta­bil­i­ty re­quires suf­fi­cient sep­a­ra­tion be­tween ac­cel­er­at­ing quadrupole mode and (i) ad­ja­cent quadrupole modes, or (ii) ad­ja­cent dipole modes. Quadrupole modes sep­a­ra­tion is di­rect­ly re­lat­ed to RFQ length, and can be in­creased if nec­es­sary via seg­men­ta­tion; po­si­tion of dipole modes spec­trum w.r.t. quadrupole spec­trum may be ad­just­ed using rod sta­bi­liz­ers in­sert­ed at RFQ ends and on ei­ther side of cou­pling cir­cuits. We pre­sent a thor­ough com­par­i­son of these two op­tions for a 6-me­ter long struc­ture at 352 MHz, and show they both lead to a tun­able struc­ture. The de­sign in­cludes 3D elec­tro­mag­net­ic sim­u­la­tion and ap­pli­ca­tion of trans­mis­sion line to tun­ing. The sen­si­tiv­i­ty of both de­signs to per­tur­ba­tions is also eval­u­at­ed.

• O. Piquet, O. Delferrière, M. Desmons, A. France
  CEA, Gif-sur-Yvette
• A.M. Lombardi, C. Rossi, M. Vretenar
  CERN, Geneva

In the Linac 4 and the SPL, a 3 MeV RFQ is re­quired to ac­cel­er­ate the H^- beam from the ion source to the DTL input en­er­gy. While the 6-me­ter long IPHI RFQ was ini­tial­ly cho­sen for this ap­pli­ca­tion, a CERN study* sug­gest­ed that a ded­i­cat­ed, short­er 3-me­ter RFQ might pre­sent sev­er­al ad­van­tages. The 2D cross-sec­tion is op­ti­mized for lower power dis­si­pa­tion, while fea­tur­ing sim­ple ge­o­met­ri­cal shape suit­able for easy ma­chin­ing. RF sta­bil­i­ty is eval­u­at­ed using a 4-wire trans­mis­sion model and 3D sim­u­la­tions, tak­ing elec­trode mod­u­la­tion into ac­count. The re­sult­ing RFQ is in­trin­si­cal­ly sta­ble and do not re­quire rod sta­bi­liz­ers. End cir­cuits are tuned with ded­i­cat­ed rods. RF power is fed via a ridged waveg­uide and a slot iris. Vac­u­um port as­sem­blies are po­si­tioned prior to braz­ing to min­i­mize RF per­tur­ba­tion. The 32 tun­ing slugs form a set of sta­ble sam­pling, able to tune 9 modes. Tuner pa­ram­e­ters are de­rived from bead-pull ac­cu­ra­cy spec­i­fi­ca­tion and fab­ri­ca­tion tol­er­ances. Sig­nals de­liv­ered by pick­up loops in­sert­ed in 16 of these tuners will be used to re­con­struct the volt­age pro­file under op­er­a­tion. Ther­mo-me­chan­i­cal sim­u­la­tions are used to de­sign tem­per­a­ture con­trol spec­i­fi­ca­tions.