Superconducting RF crab cavities are electromagnetic cavities used within particle accelerators, such as the Large Hadron Collider, to provide transverse deflection to particle bunches. Maximizing the luminosity is critical to achieve the maximum power of the accelerator system. In 1999 the A15 cavity was designed at Fermi National Laboratory (Fermilab) in Batavia, IL for the development of a separated K+ beam. It was designed as a deflecting mode cavity operating at 3.9 GHz. Early computational studies of the A15 cavity showed discrepancy with experimental studies. These discrepancies were worrying as it put into doubt the ability to use simulation tools for studies.
Dr. Leo Belantoni of Fermilab worked with Tech-X scientists using the VSim tool to uncover these discrepancies. The VSim embedded boundary method was used by Tech-X scientists to provide better accuracy than earlier simulations and the new frequency extraction method was used to extract the frequencies and the modes. Simulations were performed at very high resolutions on supercomputers at Department of Energy centers. These simulations allowed Dr. Belantoni and Tech-X scientists to understand that a discrepancy in measurement and machining caused the original differences in simulation and experimental results.
As a fully parallelized electromagnetic simulation software package that works on desktop systems as well as supercomputing systems with 10000s of processors, the VSim package was the ideal tool for solving this problem. Additionally, the very capable post-processing tools enabled efficient extraction of simulations with near degenerate modes. Lastly, the powerful visualization and flexible visualization program allowed high quality visualizations to include with the numerical simulation studies.
Previous simulation results for the A15 cavity showed a 2.8 MHz discrepancy with experimental measurements in the resonant frequency of the pi mode. Understanding this discrepancy was difficult but made much easier with the use of VSim [formerly Vorpal]. Working with Tech-X scientists, we discovered machining errors which, while within initially specified tolerances, were never accounted for in previous simulations. Once this was corrected in the simulation, we saw an accuracy of 1 part in 10^4 in the resonance frequency of the pi mode. Understanding this discrepancy improved our confidence in, and understanding the process of, designing these key components for the next generation of accelerators."
– Dr. Leo Belantoni, Fermilab