Time: | 16:00 26 Oct 2005 |
Place: | 6/2-002 |
Participants: | W. Bruns, H. Burkhardt, A. Dabrowski, P. Eliasson, M. Korostelev, A. Latina, L. Neukermans, J. Resta Lopez, G. Rumolo, D. Schulte, R. Tomas |
Daniel started the meeting by a short report from the ILC meeting on 25 October. It had been discussed how to continue after 2007 when there will be no more EU money. Many institutes had problems hiring people and maybe it is possible to get an extension of the EU support. Daniel participated in a review of the UK work on linear accelerators. He was very impressed of what they had achieved in only two years.
Javier then presented the ongoing work on the non-linear collimation system and explained why he lost magnitudes of the luminosity compared to the linear lattice as mentioned last week.
Non-linear collimation presentation
The last slide showed that even with the sextupoles turned off, 50% of the luminosity is lost. The sextupole strength needed to get σrms=120μm decreased the luminosity with another 50%. During the simulations last week too strong sextupoles had been used.
It is not clear where the luminosity loss comes from. Rogelio thought it will be important to sort out where the luminosity loss with sextupoles turned off comes from. Daniel was wondering if the matching into the FFS has been done in the right way.
--> Investigate matching
--> Find out why 50% of the luminosity is lost even when sextupoles are switched off
Warner presented the results he had obtained with his new code. His plots showed charge density vs time for a fixed position at 0.5σ from centre of beam pipe. This was done both with and without the presence of a 1T dipole field. He also showed plots of number of emitted pairs and nr of particles hitting the wall vs time. Warner asked for comments about if the simulation results seemed realistic, but nobody could answer at this points, partly because it is necessary to know if it is necessary to know how many of the particles hitting the wall are ions (and how many are electrons).
The simulations (200000 timesteps) had taken 8-10 hours depending on how many particles were lost when hitting the walls. Daniel though this seems very timeconsuming. Giovanni wondered about the number of timesteps per bunch, which according to Warner was 10000. Should be possible to decrease this a bit to speed up simulations.
--> Giovanni should calculate if the there should be particle trapping or not for the parameters that Warner is using.
--> Warner should count the numbers of electrons and ions hitting the wall separately. He should also calculate the densities of the different particles in the beam
Maxim showed results of sensitivity studies for the BDS. The misalignments studied were quadrupole and sextupole misalignments and also quadrupole rotations. Maxim showed how the horizontal and vertical closed orbit were affected and also how sensitive the vertical dispersion was.
10μm quadrupole misalignments gave very large closed orbit offsets in the vertical plane.
--> Maxim should try to obtain the tolerances for the different misalignments. Daniel and Maxim will discuss this during the week.
Peder is still working on the paper on ILC tuning bumps. This is in the final stage.
Andrea said that he has started writing a tracking routine taking into account longitudinal motion. This piece of code will later be integrated in Placet. The basic structure of the code is finished and already working well. The code is also compatible with MPI.
Giovanni said that he is currently trying to get started by studying all new things. He was also wondering about collimator wakefield calculations in Placet. Daniel explained that for the moment only dipole wakefields work but that multipoles wakefields would be very fast to implement (already half done).
Rogelio reported that he has started trying Placet. He is now specified as developer in the Savannah CVS. He has also been working on transforming a MAD8 to a MAD-X lattice. This is not completely trivial. The two codes/lattices do not give the same results. He and Helmut are looking into this.
Helmut explained that he and Lionel have tried to investigate if the CTF3 can be used for benchmarking. Discussed this with Thibaut Lefevre.
--> Daniel though it might be a good idea if Helmut present a list of the codes to be used during the next meeting.
Anne said that she is still finishing her thesis in another group. She is working on the loss monitors in CTF3
Lionel said that he is translation some old fortran codes to C++ for beam-gas simulations. This should then be included in Placet.
Warner finally mentioned that anyone interested can have a look into the source code of his new program. The code is written in Fortran95. Daniel wondered if there are any open source compilers that can compile the code. No problem according to Warner.
Maintained by: Peder Eliasson
|
|
Last updated: 26 Oct 2005
|