LHC Injection Working Group

---

Summary notes of the meeting held on 28 February 2007

1. Injection Sequencer Requirements

R. Alemany Fernandez gave an overview of the on-going work of defining the sequences, sub-sequences and tasks to drive the injection into the LHC (slides). A Wiki page (http://wikis/display/LHCOP/LHC+Injection+Scenarios), currently maintained by R. Alemany Fernandez and V. Kain, was set-up to collect all available information and to allow everybody from inside CERN to add comments and feedback. For each of the injection related "scenarios" (nominal injection, inject & dump, commissioning,...), sub-sequences and tasks will be defined. Work is presently focused on the nominal injection scenario while trying to establish sub-sequences which can be re-used for other injection scenarios. Questions like "when does the injection equipment have to be switched on", "which timing tables are required", "what are the different modes and states the machine has to be driven through", "what are the different setting-up methods required", "when the LHC request which beams from the injectors", "which abort and recovery scenarios are there", "which machine protection aspects have to be treated within the sequencer", etc. have to be answered. A first draft for the nominal injection scenario was presented. The Wiki pages will ultimately be used as input to the LHC sequencer.

H. Burkhardt asked whether the sequencer drives directly the equipment. R. Alemany Fernandez and V. Kain replied that the software architecture chosen for the sequencer is a multi-tier architecture. There will be APIs interfacing with system specific applications to execute commands and drive hardware. V. Mertens remarked that people responsible for different systems (e.g. applications, equipment) should be involved during the preparation of the sequences to take account of possible limitations and constraints. B. Goddard suggested that the Wiki tables should allow to identify easily whether cells are left blank on purpose or because information was still missing. For the time being all tasks for the injection sequence have the same priority. B. Goddard remarked further that it could be useful to classify the tasks into "must not be skipped", "could be skipped" and to have the functionality in the sequencer GUI to skip tasks. Finally it was asked what of the sequencer functionality will be available on day 1 of the LHC start-up. B. Goddard commented that the database tables with the different tasks and sub-sequences should be available. R. Alemany Fernandez remarked that the sequencer should definitively be in place for the LHC start-up, but that the driving of the machine through the different states and modes will be manual in the beginning (step-by-step execution) and that more and more parts could be automated as experience would grow. .The sequencer itself will certainly also need commissioning time. The InjWG will follow the development of the injection sequences.

2. Inject-and-Dump Mode

B. Goddard summarised the latest understanding of how to handle the “inject & dump” mode within the LBDS, timing system and BIS (slides). The beam dump cannot be triggered directly via a timing event; instead opening the beam permit loop is required. Two inject & dump scenarios were identified. The first one needs an additional hardware box within the LBDS to allow 0 – 1000 turns of circulating beam before dumping. The second scenario will allow to have beam circulating for 0.1 – n seconds before dumping. In this case no hardware changes are required. The timing system sends an event to open the permit loops and the beam dump system is triggered in the “normal way”.  The 0 – 1000 turns injection & dump scenario will only be enabled for safe beam at injection energy. Screens can remain in the beam. This scenario will be required e.g. for aperture measurements in the injection and extraction channel. The additional hardware box in the LBDS system will have a turn counter which opens the beam permit loop of the chosen beam via the BIC in point 6 as soon as the required number of turns has been reached. Four additional timing events will be executed from the timing table during this scenario: “injection warning”, “suppress post mortem beam x” (x = 1, 2, for the requested beam), “unlink beams” and “dump beam x”. For inject & dump the permit loops are unlinked for the duration of the scenario and one beam can be dumped independently of the other. No post mortem event will be forwarded after the beam dump event. In order to carry out the XPOC of the LBDS, the acquisition of the relevant instrumentation is triggered by the event “beam x dumped” which is distributed as soon as the timing system detects the open permit loop. The timing event “dump beam x” is used as redundancy 1 ms after the requested number of turns. Hence the screens would have to stand at most another 10 turns after the requested turn number.

In the case of the other scenario, inject & dump after 0.1 – n seconds, the same four timing events have to be defined in the timing table. The main difference is that for this scenario the event “dump beam x” is used directly to open the permit loop.

B. Goddard went on to list the requirements for the different systems involved (see slides). After an inject & dump scenario the beam permit loops have to be recoupled via a timing event. This is crucial. A possibility would be to have the sequencer look after this. V. Mertens asked how one can make sure that this is not forgotten in case the machine is driven manually. B. Goddard replied that the MPWG will look into this.          

3. TDI news - outgassing tests, impedance issues, design changes

Y. Kadi reported on the latest news of the TDI production, the design changes to cope with the impedance issues (see last meeting) and the results of the outgassing tests (slides).

The results of impedance calculations for the TDI geometry led to the following design changes: ferrites on the front and end faces of the TDI, rounded edges (removing 2 mm of material) and RF fingers between the TDI segments in the form of undulated foils (0.5 mm thickness of CuBe foil, 2 - 3 mm retracted from the jaw surface). For delay and cost reasons a more standard design was chosen for the ferrites than originally proposed by the impedance experts, with only slight performance loss. As planned, the changes are already being incorporated for the TDI in preparation for LSS2L (installation foreseen in May/June 2007). The TDI installed in LSS8R will be replaced by a new version after the engineering run. Two spares will now be produced which will also include the latest changes.

Following the latest tests done with an hBN absorber segment the outgassing rates are acceptable for AT/VAC and are about a factor 10 better than for graphite. The much higher rate found earlier, which created some worries, was traced back to some air trapped in a badly mounted sub-assembly (info: F. Loprete). After some 3 weeks of bake-out also the presently installed TDI is ok from the vacuum performance. B. Goddard was concerned about the possible down time in case a TDI needed to be exchanged. Y. Kadi replied that the long bake-out was mainly due to a long exposure to air, which led to an enrichment of the absorber with water vapour. This exposure will be minimised in the future.

New FLUKA simulations have been made to take the new design features into account. All simulations were carried out for ultimate intensity. (4.9 x 10¹³ protons). The results are very similar to the ones obtained in earlier studies. B. Goddard asked whether the temperature increase for full impact in block 7 (Al) of 339 degrees is still acceptable or whether this was already beyond the elastic limit. Y. Kadi confirmed that this was still within the elastic limit. An additional study was carried out in which the original materials (hBN, Al, Cu) were replaced by different materials (C, TiAl, INCONEL). The outcome is that the original material choice leads to less energy deposition in the metallic blocks due to the increased attenuation of hBN compared to C, whereas the alternative material choice might be better for shock impact. The overall attenuation and hence energy deposition in the D1 would be slightly better for the alternative materials.

4. Status and issues concerning various equipment - TCCD/M, TCLIs, TCDIs

Y. Kadi went on to present the latest news on the status of TCCD/M, TCLIs and TCDIs (slides). For the moveable TCDD for IR2 big delays occurred in the design office. The design is very similar to the one of the TCLIA (2-in-1 design for the TCTV and the TCLI close to the D1). Thus the same motorisation, software and cooling can be used. The TCDDM masks for IR8  have been corrected after some initial alignment problems between the vacuum chamber ends and the absorber core; the tolerances are now within a few tenth of a mm. As the assembly was heated close to the weld seam during the re-alignment, the vacuum tightness had to be re-checked. The NEG coating can now be started. Installation should still well fit in the overall vacuum closure planning.

Concerning the TCLI it is not yet clear whether these collimators will be produced by CERCA or under the coordination of TS. The drawings for TCTV and TCLIA are being finalised and should be sent around for approval soon. Energy deposition studies for the TCLIA and TCLIB need to be made.

After some problems the collimator production seems now to resume speed. There are still 7 TCDI needed for TI 2 (plus spares), and the jaws of those installed in TI 8 need to be replaced. The present plan is to fabricate the TCDIs for TI 2 right after the production of the "minimum set of collimators" for the main ring, which would just fit the installation planning (June/July 2007). Due to lack of manpower, FLUKA energy deposition simulations are still missing for two critical locations in TI 2, where different masks will be needed for higher intensities. After some discussion the InjWG recommended to install the 6th TCDI (already at CERN) at the originally foreseen location (TCDIV87804) rather than sparing it for a position near the MSI in TI 2.

5. AOB

LHC V6.501: H. Burkhardt mentioned that the LHC optics team is planning an LHC optics version V6.501 which should represent the as-built machine more accurately. B. Goddard remarked that any strength change in the Q5, next to the MKI, might have a severe impact on the injected trajectory and hence on the MSI alignment. Any plans for changes should be presented and cross-checked because of the aperture restrictions and various other constraints (e.g. in phase advance, kicker voltage).

TPSG: Y. Kadi noted that the stress analysis of the TPSGs (for the new absorber layouts) in SPS LSS4 and LSS6 will no longer be carried out by CRS4. V. Mertens will discuss the situation with J. Lettry.

LHC commissioning procedures: V. Kain mentioned the work being done in the LHCCWG to define detailed commissioning procedures for the LHC with beam (http://lhccwg.web.cern.ch/lhccwg/overview_index.htm). The results for the first phase, A.1 – injection and first turn, will be summarised at one of the coming InjWG meetings. Documentation on the foreseen procedures is available in the form of web pages.

The list of open actions and future topics established some time ago has been updated and prioritised (coding: red = important/urgent, yellow = medium urgency, green = for later).

V. Kain + V. Mertens, 08/03/07.