Choose timezone
Your profile timezone:
PyHEADTAIL Meeting #19
Quick introduction to agenda by Kevin
- Jani
- History and background, motivation for circular convolution implemetation
- Jani explained the two different views how he thought of for the multi-bunch tracking.
- Jani explained the different reference frames for the different types of convolution.
- Jani then moved toward the diferent algorithms which are implemented today in PyHEADTAIL.
- Jani showed how we can priifot frm the cicular convolution spee-up if one removes slices ebtween bunches with adequuate padding. This still gives correct results at the location of actual yslices.
- Jani went through the different banchmarks that were done, against the original HEADTAIL code, among the different implementations of PyHEADTAIL and agains theory. Also a benchmark against NHTV was done and the results macth well.
- Today, simulating a full LHC beam is possible on the available HPC clusters of the departement.
- Jani presented some applications of the multi-bunch PyHeadtail codes. He shows two examples of studies, with the latter studying the intensity or impedance mragins where he seems to encounter phenomena beyong the simple coparson of instability growth rates and damping rates.
- Can we implement a test that compares the multi-bunch - single bunc reduction
- Noclar: the linear convolution and circular convolution sh0uold give exatply the msae results? Adrian mentioend that small difference can be expected.
There are minutes attached to this event.
Show them.
