Injection Kickers

Vertical deflection, 0.85 mrad, integrated field 1.2 Tm. PFN voltage 54 kV. Four fast pulsed magnets per ring. Magnets between Q4 and Q5.

Safe injection in LHC Etienne Carlier Cham XII

High level requirements

• Visual confirmation of state - simple fixed display. Done.
• Applicaton: Appropriate high level commands to allow traversal of system state transition diagram. Done.
• [SCM: Definition of operational state diagram taking into account external conditions - interlocks/timing - not needed]
• OASIS/Application: Visualisation of kicker waveform (shall be recorded every cycle) - CMW. Done
• Verification of proper receipt of pre-pulses & slow timing.
• Beam permit windows. Signals from OASIS.
• Abort Gap keeper window. Signal from OASIS.
• BETs window. Application (comes from FESA). Done
• Settings: Shall provide settings management (including usual trim/archive functionality) - define parameter space (mrad to volts etc.) (remember possibility to change between injections) - use LSA!!!
• RBAC
• MCS: Safe settings management of kicker settings. - will be defined as safe settings in LSA.
• Inject and dump: For injection and dump, set delay in dump.
• Circulate and dump. - can do with timing table.
• XPOC: Not injection kicker specific - more operations - losses, lifetimes, tranverse oscillations (TFB response). Not direct feedback to injection kickers.
• Post Mortem - how do deal with erratics, dumps after pre-pulses etc. Will push everything to PM system when PM event arrives. Analysis we worry about later.
• Logging - standard logging of continuous varaibales - Marine
• Alarms - standard.
• Timing: define events & payloads - start putting together injection table - not directly concerned with injection kicker sw. One warning event required by kickers. Need destination ring. TEST. TEST. Check.
• Shot by shot logging. Digitization plus windows and then publication through OASIS - Stephane - thru CMW. Done
• Expert Settings. fine delays etc. windows - should be critical settings. Need to define in LSA (?). JAPC for the moment (not writtable)

Other issues

• Interface to beam permit loop - monitor
• Injection modes

Timing event per injection  (Etienne Carlier)

Do you have already an idea on how many set of settings (kick enable, kick strength, kicker delay and kick length) you would like to have for
the LHC injection kickers?

- One single set for all the injection (a la LEP). This option is possible as the last batch will be a long batch.

- One set for the first injection, one set for next injection and one set for last injection (first, next and last injections will have do be
managed through timing event)

- One set per injection (injection number performed will have to be identified by a timing event)...  how do we identify the different injections?

Following slide c/o Etienne

State Transition

Timing sequence

Settings

Architecture

Analog Acquisition

Flash Over (Bren)

After discussion with Volker and Laurent, a simple back-of-the-envelope calculation (based on measured failure rates) shows that we expect something of the order of one event over a 20 year LHC lifetime where the full injected beam will impact the primary collimator jaws, due to a  flashover in the MKI magnet cells. 

TDI position: 4.38 +/- 0.3mm (8.5 +/-0.58 sigma)

Primary collimator setting: 6 sigma

Total injected high intensity batches in LHC lifetime: 20 (years) x 200 (days) x 2 (rings) x 2 (fills) x 12 (batched) = 2 e5

4 MKI magnets per injection: 8 e5 magnet pulses in LHC lifetime

Measured MKI flashover rate is 2(+/-2) flashovers per 4 e5 pulses, so assume a rate of 1 per 1e5 pulses, i.e. 8 flashover events in LHC lifetime.

Average deflection per MKI is 18 sigma, for 33 cells, or 0.55 sigma/cell

A flashover at cell N produces x2 kick in first N cells, zero in remainder

So flashover needs to happen in cell 8 - 11 for the batch to get a dangerous kick (where the total is missing between 6 and 9 sigma) i.e. probability  is 4/33.

Therefore total number of dangerous events during LHC lifetime is 8 * 4/33 = 1.