Development of srf spoke cavities for low and intermediate energy ion linacs

The fabrication of the first European spoke cavity prototype has been achieved in July 2002. A series of mechanical and RF tests at warm and cold temperature was performed this year. These tests have demonstrated, on the one hand, the good feasibility, stiffness and tunability of the spoke cavity and, on the other hand, its excellent RF performances with a maximum accelerating field of 12.5 MV/m reached at 4.2K. INTRODUCTION Spoke cavities have been studied for 3 years at IPN Orsay. Beam dynamics studies [1] dedicated to the EURISOL [2] and XADS [3] European projects have pointed out that this kind of cavity is particularly suited to be used from typically 20 MeV up to 100 or 150 MeV (for proton linacs). In this framework, IPN has studied a 2-gap, beta 0.35, 352 MHz spoke cavity. The optimization of the RF parameters (done with MAFIA [4]) and the structural analysis (done with ACORD-CP [5]) are presented in [6-8]. The prototype has been fabricated, from February to July 2002, by the French company Cerca [9] and a series of tests started since the delivery date. First of all, we will describe the measurements performed at room temperature to know in particular the mechanical properties of the cavity like its sensitivity versus displacement and load, its mechanical stiffness, the frequency shift due to fabrication and vacuum load, the accelerating field profile... Then, we will show the results obtained during the three tests, in January, March and July 2003, performed at 4.2K into our new vertical cryostat. Finally, the recent developments made on the new beta=0.15 spoke cavity will be presented. TESTS AT ROOM TEMPERATURE Dimensional Controls Before the tests, we studied the frequency variation (i.e. ∆f~+300 kHz) observed between the “theoretical” value calculated with MAFIA (fcalculated=358.55 MHz) and the frequency measured at the delivery (fmeasured=358.85 MHz). Note: calculations and measurements have been done for the cavity at atmospheric pressure. Thanks to the dimensional measurements of the main cavity pieces (e.g. the spoke bar, the cavity length...) done by Cerca during the fabrication process and the respective sensitivities calculated with MAFIA, we had estimated the frequency variation due to fabrication errors. As we can see in Table 1, the estimation (i.e. +455 kHz) is in good agreement with the measurement. On one hand, the little difference between the “theoretical” frequency and the “real” one and on the other hand, the very good accelerating field flatness due to strong magnetic coupling between both cells (see section above) means that spoke cavity doesn’t need to be tuned at room temperature. It’s a strong advantage as compared to elliptical cavities for instance. Table 1: Estimation of the frequency variation due to fabrication errors. Areas Sensitivity with MAFIA (kHz/mm) Errors (mm) Frequency variation (kHz) Cavity diameter 950 -0.1 +95 Spoke base diameter 650 -0.2 -130 Racetrack width 800 -0.2 +160 Racetrack thickness 600 +0.2 -120 Wall-to-wall cavity length 450 +1.0 +450 Total variation +455