Minutes PSB Upgrade WG Meeting 29th of November 2012

Present: E. Benedetto, J. Borburgh, A. Dallocchio, M. Delonca, A. Findlay, R. Froeschl, R. Garoby, K. Hanke, J. Hansen, D. Hay, S. Jensen, R. Losito, C. Maglioni, S. Mathot,B. Mikulec, A. Newborough, C. Pasquino, A. Perillo Marcone, V. Raginel, F. Roncarolo, A. Sarrio Martinez, J.Tan, W. Weterings, F. Zocca, L. Zuccalli.

Agenda:

1. Approval of Minutes

The minutes of the last meeting were approved.

2. Communications

K. Hanke asked all the WP Holders to send their update of the APT data before the 13th of December.

3. Status of the H0/H- Dumps

M. Delanco presented the status of the H0/H- dumps. The dump has been designed for 3 types of beams:

• H- injected or foil failure (¼ Linac 4 pulse, interlock after one pulse). The H- impact angle is assumed to be ~33 mrad.
• H0: unstripped particles (depend on foil efficiency)
  • 98% efficiency (operational case) -> Steady state, 2% of all H0.
  • 90% efficiency (degraded case) -> Steady state, 10% of all H0.

According to the Engineering Parameters document EDMS 1069240, the material of the dump should be completely non-magnetic, should induce little eddy current and should be at least slightly conductive (to not electrically being charged), hence ceramic material are good candidates. Out of 8 material candidates, Silicon Carbide (SiC) appears to be the most suitable candidate. With the actual design of the dump, only 3.5 mm are available between the dump and the vacuum chamber for the wires of dump monitoring and for the support (if needed).

The numerical simulations give with 1/4 Linac4 pulse a maximal ΔT of 171.34°C; at steady state with 2% of all H0 a maximal ΔT of 33.7°C and at steady state with 10% of all H0 a maximal ΔT of 80.6°C. It has to be confirmed that a temperature of 100°C would be acceptable for the vacuum.

The dump core will be placed in the vacuum chamber of the magnet KSW4. The actual baseline for the vacuum chamber is to use Inconel, instead of ceramic, which make it possible to use the vacuum chamber for supporting the dumps (development of the chamber in collaboration with EN-STI). Different solutions are proposed to support the dumps: support by vacuum chamber or/and brazing the dump on the flange and/or a mechanical solution with shrinking. It was said that in case of dump failure and need of replacement, brazing or shrinking were equivalent. Using the vacuum chamber to support the dump would release partly strength needed at the flange. Test and numerical simulations were done for brazing the dump in SiC, it appears that Molybdenum (Mo) is the most promising material to braze the SiC. A study is on going in collaboration with EN/MME.

The next steps are :

• Vacuum tests to be done (samples received)
• Brazing tests to be performed (SiC samples received, Mo samples to be delivered) ~ March 2013
• Cooling circuit integration
• Integration of the beam instrumentation

4. Status of the Head/Tail dumps.

C. Maglioni presented the status of the Head/Tails dumps. A series of dedicated meeting have been held with STI, MME, ABT, VSC, OP and RP to define responsibilities, needs, constraints and requirements.

• The detailed design of the dumps and of the support is on-going (EN/STI ↔ EN/MME):
  • The dumps will be made of graphite R4550 (dilution & activation)
  • The supports are defined, it will be in stainless steel (rigidity, fabrication, shielding)
  • The Dimensions of the dumps are fixed. The Head dump is fixed onto BI.SMV3 and the Tail dump onto SMV1. They both cover part of the SMV coils.
  • Cooling and bake-out integration are on going. The 2 dumps are cooled in series with copper pipe (9 mm outside, 6 mm inside).
  • Feed-throughs are defined.
• Vacuum analyses and calculations (TE/VSC) are completed
• Thermo-mechanical analyses (STI) are completed. Three loading cases have been studied (see EDMS document 963395):
  • Case 1 is accident scenario (1 pulse at maximum) : 373°C for the maximal ΔT.
  • Case 2 is nominal operation. The simulations show at steady state with active cooling an increase of temperature of ~1°C. With no water cooling, only thermal radiation the maximal ΔT is ~23°C.
  • Case 3, LL RF setting up happens for 1 week/year. The simulations show at steady state with water cooling an increase of temperature of ~18.2°C. With no water cooling, only thermal radiation the maximal ΔT is 183°C.
• Dump Eng. Spec. (STI) will be released soon.
• Activation analyses are on-going (RP)
• Integration and Interference are continuously checked (TE/ABT ↔ EN/MME)
  • Interface between the SMV and the dumps is frozen
• Assembly procedure of dump and magnets inside the tank (TE/ABT ↔ EN/STI ↔ EN/MME) is defined
• Maintenance scenarios (TE/ABT) have be studied.
• Monitoring / electronics / interlock are on going

The plan is to finalize the dump integration & design for March 2013 and the dump analyses & specification end December 2012. The production plan will be done in June 2013, the material procurement in September 2013 and the components production in December 2013.

5. Vacuum systems

C. Pasquino presented the BISMV & H0-H- dump vacuum requirements.

The vacuum system of the BI.SMV is composed of 3 Sputter Ion pumps Starcell 500 and 2 Sputter Ion pumps Starcell 300. Outgassing tests were carried out on graphite, the results can be found in the EDMS document 1254422. The results show that the graphite is compatible with Ultra High Vacuum (UHV) systems only if:

• A Vacuum firing at 950 ◦C is done for 2 hours.
• A In Situ bake out at 200 ◦C is done for 24 hours.
• The working temperature of the graphite block is under 50 ◦C.

Outgassing rate measurements of the SiC are on going but from the very first results, it looks very promising for UHV applications. More detailed result can be shown beginning of 2013.

6. H0/H- Instrumentation

F. Roncarolo presented the status of the H0 - H- current monitor. The H0/ H- current monitor is needed in front of the dump to allow efficient setting up of the injection and to monitor the efficiency of the stripping foil (detect degradation and failure). The H0/H- current monitor is composed of a plate intercepting the H0 and H- ions acting as a Faraday cup for the stripped electrons (stripping & collection) and of polarization rings which are needed to avoid emission of secondary electrons that would affect the measurement. The polarization rings have a voltage of 1 kV, so 1 kV cables have to be feed through. The effect of back scattered electrons still need to be checked. A consistent fraction of the H0/ H- stripped electrons could be back scattered and hence not being collected (20-50%, depending on the plate material). This effect is under study for the SEM grids and WS in LINAC 4, it was already checked with an electron beam and it needs to be tested with the H- beam.

To select the material of the plates, different properties have to be considered. The electron yield is only 2-3% of the primary particles for all considered materials. Positron yield is very low for all considered materials. The plate ageing is proportional to the neutron yield so a low Z material is preferable. Aluminum or titanium seems to be a good choices, with their good electrical conductivity, but it need to be checked if it could influence the magnetic field.

The next steps H0/ H- current monitor are:

• To establish a mechanical design including fixation to the dump & integration.
• To dimension the electronics for both H0 and H- and the desired dynamic range. It was said that it should be checked if the electronics could be located far from the injection region to avoid radiation damages.
• To establish the possible tests during Linac4 commissioning

It has been underlined not to forget the connection to the interlock system, as this instrument's aim is also to protect the H-/H0 dump from continuous high load.

7. Radio Protection

R. Froeschl presented the RP Status for the PSB Injection Dumps.

The Head Dump RP study is based on the load assumptions from the EDMS document 963395. In which the operational year has 2 phases, the LLRF setup (max. 1 week at the beginning of the operational year) with 278e10 pps and the normal operation with 17e10 pps. Simulations allow to conclude that:

• Graphite is a well suited material for the dump core from an RP point of view.
• After 3 hours, the dose rates is dominated by the magnet and not by the dump. The expected dose rate at 1 m is less than 700 μSv/h after LLRF week and less than 100 μSv/h after normal normal operation). It was said that it needs to be clarified whether the LLRF setup has to be done with the full current as stated in EDMS document 963395.
• The production of tritium in the cooling water is very low so the connection to the (then) underground PSB water cooling circuit should be OK.

For the H-/H0 dumps, the load assumptions come from the same EDMS document 963395. Two types of load are considered: a continuous load (9.61e18 H-/y at 14.2 W) and the load after a foil degradation (max. 8 hours for about 10 times per operational year, 7.99e17 H-/y at 71 W). The simulations show that:

• The dump core is the dominant contributor to the dose rate, even after 1 day cool down as Na-22 (half-life 2.6 y) is the dominant radionuclide.
• The dose rate goes up to a couple of mSv/h at 1 meter distance if the dump is unshielded. During a follow up meeting on the 4th December it was clarified that this radiation levels would be without shielding, while in the actual set-up of the dumps will be shielded by the iron of the magnets. R. Froeschl estimates that taking into account this shielding values, dose rates of 0.2 – 0.3 mSv/h at 1 m are obtained. These estimated dose rate values have to confirmed by FLUKA simulations.
• Water cooling have to be studied.
• Intervention scenario have to be developed for the injection area.

A. Newborough has asked what would be the activation of the main dipole behind the H-/H0 dumps. It may be useful to also implement the geometry of the downstream magnet.

The next LIU-PSB meeting will be held on Thursday 13th of November at 10h30 in 865-1-D17.