Choose timezone
Your profile timezone:
7th CLIC-ILC BDS+MDI meeting
→
Europe/Zurich
CERN
CERN
We change order of the talks to allow Andrei to attend Daniel's presentation which contains
new material.
Barbara presents the evolution of the CLIC FFS tuning:
-The first approach is Simplex+Knobs with a performance of 90% reaching 90% of the luminosity. This takes 17000 luminosity measurements
-The second approach consists of 5 steps: 1-to-1 correction, DFS, Multipole shunting, Knobs, Simplex. 95% probability of reaching 90% of the luminosity. The first 3 steps converge to a large horizontal beams size. The total number of luminosity measurements is 5000.
How fast can we measure luminosity?
Usually radiative bhabha are used but they will not be really visible for CLIC. Low angle bhabha takes 7-70 minutes for 1% resolution luminosity measurement. This clearly makes tuning catastrophically slow 70*5000=350000 minutes ~ 200 days. Need a solution.
Barbara has looked for a new signal. Hadronic events look promissing, in particular pions created from photons collisions. Measuring the multiplicity just with 3 trains would give 7% resolution. For 1% luminosity measurement resolution would need 100 trains. -> A good measurement of luminosity needs 2 seconds.
Using pions/hadrons signals does not seem to pose any problem. Of course a combination with many other
signals will improve the situatioin considerably. The 2 seconds imply 3 hours. At low luminosities beamstrhalung gives a good enough signal pulse-to-pulse. the stability of the luminosity has to be taken into account, since 10% will require averaging over few pulses to be improved.
The first method reaches the goal (110% lumi) for 70% of the machines.
The second does not make any machine in. This performance aspect should be reflected in the BDS chapter of the CDR. The detail description should go in the emittance presenvation chapter.
Daniel presents results of how fast we loose luminosity. Luminosity loss of 2% happens in 150s.
10% stability gives 1% precision averaging over 2s. In 150s we can perform 75 measurements.
The caveats are: -lower lumi might take longer with no margin in above numbers -Detailed full study is required (to be assessed if this is for CDR or not).
It has to be clearly mentioned that static errors are treated separately from dynamic errors.
Yves started working on DFS and ways to measure dispersion using energy jitter.
Jurgen looked into how well dispersion measurement could go with energy jitter.
Daniel reminds that magnet errors are not in there, two beam tuning is not yet used, etc.
new material.
Barbara presents the evolution of the CLIC FFS tuning:
-The first approach is Simplex+Knobs with a performance of 90% reaching 90% of the luminosity. This takes 17000 luminosity measurements
-The second approach consists of 5 steps: 1-to-1 correction, DFS, Multipole shunting, Knobs, Simplex. 95% probability of reaching 90% of the luminosity. The first 3 steps converge to a large horizontal beams size. The total number of luminosity measurements is 5000.
How fast can we measure luminosity?
Usually radiative bhabha are used but they will not be really visible for CLIC. Low angle bhabha takes 7-70 minutes for 1% resolution luminosity measurement. This clearly makes tuning catastrophically slow 70*5000=350000 minutes ~ 200 days. Need a solution.
Barbara has looked for a new signal. Hadronic events look promissing, in particular pions created from photons collisions. Measuring the multiplicity just with 3 trains would give 7% resolution. For 1% luminosity measurement resolution would need 100 trains. -> A good measurement of luminosity needs 2 seconds.
Using pions/hadrons signals does not seem to pose any problem. Of course a combination with many other
signals will improve the situatioin considerably. The 2 seconds imply 3 hours. At low luminosities beamstrhalung gives a good enough signal pulse-to-pulse. the stability of the luminosity has to be taken into account, since 10% will require averaging over few pulses to be improved.
The first method reaches the goal (110% lumi) for 70% of the machines.
The second does not make any machine in. This performance aspect should be reflected in the BDS chapter of the CDR. The detail description should go in the emittance presenvation chapter.
Daniel presents results of how fast we loose luminosity. Luminosity loss of 2% happens in 150s.
10% stability gives 1% precision averaging over 2s. In 150s we can perform 75 measurements.
The caveats are: -lower lumi might take longer with no margin in above numbers -Detailed full study is required (to be assessed if this is for CDR or not).
It has to be clearly mentioned that static errors are treated separately from dynamic errors.
Yves started working on DFS and ways to measure dispersion using energy jitter.
Jurgen looked into how well dispersion measurement could go with energy jitter.
Daniel reminds that magnet errors are not in there, two beam tuning is not yet used, etc.
There are minutes attached to this event.
Show them.