Minutes of March 29, 2007 LAL ILC-MDI/BDS/ATF2 group meeting (by Yves Rénier)
1) Optimisation of Guinea-Pig grid for ILC IP feedback simulations - Maria del Carmen Alabau
To keep a high luminosity, a feedback system is used to correct the offset produced by ground motion. The offset can be determined measuring the angle of deflection after interaction between the two bunches and simulations show it increases with time. The beam-beam interactions are simulated by the Guinea Pig software, but present computations are too long to see what would be the behaviour of the machine over long runs. So we need to adapt the computation grid as function of the offset to reduce computation time. It is specially needed for e- e- collision where the grid must be much larger due to repulsion of bunches.
Another suggestion was to not use Guinea Pig for large offsets, but instead a simpler model or just a look-up table.
2) Prospects to speed up the C++ version of the Guinea-Pig simulation - François Touze
When we look in Guinea Pig ++, there are two parts which are time-consuming: FFT algorithm and luminosity computation. Using FFT to solve the Poisson equation the computing time is longer with thinner mesh whereas luminosity computation is longer for a larger number of particles. Tests were done using the FFTW library version 2.1.5 and the computation time was divided by 5! To speed up the luminosity computation, the main idea is to distribute particles among the processes. The implementation is still under development.
3) Systematics from luminosity spectrum reconstruction - Filimon Gournaris
The ILC will allow high precision measurements, but this precision is often dependent on the knowledge of the luminosity spectrum. This spectrum is due to beam energy spread, beamstrahlung and initial state radiation (ISR). ISR can be calculated with high accuracy, beam spread is about 0.1% but beamstrahlung has to be simulated or parameterized (based on simulation results). A study of different systematics associated to beamstrahlung is being done to determine the ultimate precision of the current luminosity spectrum extraction technique.
There are minutes attached to this event.
Show them.
