RF power dissipation due to magnetic flux trapping

It is known for some time that trapped magnetic flux is one important contribution to the residual surface resistance of niobium and can easily dominate this.  Contrary to the theory of superconductivity, where any external magnetic flux is expelled (Meißner Effect), niobium often traps up to 100% of the flux in the form of normal conducting flux tubes [1]. In the RF fields these flux vortices oscillate and cause power dissipation.
To better understand flux trapping we have investigated a number of its aspects:

1. How do the Nb properties (crystallinity, purity, treatment history) impact the amount of trapped flux?

We have constructed a test stand using niobium disks to investigate the influence of niobium material properties and of the surface treatment history on the amount of flux that gets trapped inside the material. (Niobium sample results)

2. How do the cool-down conditions influence the expulsion of flux?

Though it has been shown that up to 100% of the ambient flux can get trapped we have seen that, in addition, the coolingrate influences the amount of flux that gets trapped. This observation brought the dynamics of flux tubes around the phase transition up for discussion and led us to perform a new set of experiments close to the transition temperature. (Cavity results)

3. Does the cool-down procedure of cavities affect the amount of trapped flux and can magnetic flux actually be generated by temperature gradients?

Past measurements have repeatedly demonstrate that a short thermal cycle of TESLA cavities from 1.8 K to a few 10 K and subsequent cooling back to 1.8 K can improve the Q0 dramatically (by up to a factor of 2).  A number of possible mechanisms, including the removal of adsorbates where considered.  Eventually we came to the conclusion that temperature gradients during the initial cool down are responsible for thermo-electric currents in the Nb-cavity/Ti-helium tank system whose magnetic field is subsequently trapped when the cavity transitions to the superconducting state.  A thermal cycle to T > 10 K and re-cooling effectively removes these thermal gradients so that the quality factor can improve.To study this effect in more detail we proceeded two fold.  In a “model” system consisting Nb and Ti rods connected in parallel we mimicked the cavity helium-tank system.  When applying temperature gradients during the cooldown through Tc we indeed found that additional magnetic flux was generated that then remained trapped in the Nb rod.  HIER LINK ZU NB SAMPLE RESULTSIn parallel to these measurements we equipped a TESLA cavity (including helium tank) with heaters at each end to intentionally generate temperature gradients as the cavity cools through the transition temperature following thermal cycles above Tc.  We found a clear correlation between the achieved Q factor and the temperature difference between the ends during the cooldown.  (Cavity results)

Citations:[1] S. Aull, O. Kugeler, and J. Knobloch, Phys. Rev. ST Accel. Beams 15 (2012).