CLIC Post-Collision Line and Dump Edda Gschwendtner CERN

CLIC Post-Collision Line and Dump Edda Gschwendtner, CERN for the CLIC Post-Collision Line and Dumps WG

2 Outline • • • Introduction Background Calculations to the IP Magnet Lifetime Main Beam Dump Luminosity Monitoring Summary Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

3 Post-Collision Line Beam dump post-collision line 20 mrad beam 1 beam 2 detector Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

4 Some Numbers 50 Hz repetition rate 3. 7 E 9 e/bunch 14 MW beam power • • 156 ns bunch train length 312 bunches/pulse e+e- collision creates disrupted beam – Huge energy spread, large x, y div in outgoing beam total power of ~10 MW High power divergent beamstrahlung photons – 2. 2 photons/incoming e+e 2. 5 E 12 photons/bunch train total power of ~4 MW • Coherent e+e- pairs – 5 E 8 e+e- pairs/bunch. X 170 k. W opposite charge • Incoherent e+e- pairs – 4. 4 E 5 e+e- pairs/bunch. X 78 W Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

Purpose of the Post-Collision Line 1 - transport spent beam safely to dump 5 Transport of all charged particles (disrupted beam, e+ e- pairs) as well as beamstrahlung photons to dump Transport non-colliding particles to dump Controlled beam losses in magnetic elements and collimators, with a minimal flux of backscattered particles at the interaction point 2 – implement beam based diagnostics Monitor quality of collisions and luminosity for tuning purposes Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

6 Design Considerations Long drift section required to allow expansion of the beam before hitting the dump. Focusing of post-collision beam using quadrupoles not possible long low energy tail Based upon the characteristics of the disrupted/undisrupted beam and requirements from the exit window and dump overfocusing would lead to large energy losses Separate the charged particles from bremsstrahlung photons luminosity monitoring Vertical chicane Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

7 Baseline Design A. Ferrari, R. Appleby, M. D. Salt, V. Ziemann, PRST-AB 12, 021001 (2009) intermediate dump carbon based absorbers 1. 5 m side view ILC style water dump C-shape magnets 27. 5 m window-frame magnets 67 m 6 m 4 m 315 m • 5 window-frame dipoles and 4 C-shaped dipoles • Absorbers and an intermediate dump • To reduce beam losses in the magnets • Possible background sources: • Backscattered photons and neutrons from dump and along post-collision line Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

8 Background Calculations Lawrence Deacon Some changes/improvements done to the BDSIM model of the post-collision line • • • Thickness of elliptical beam pipes between the first 4 magnets An accurate model of the beam pipe after the intermediate dump. complicated shape. Main beam dump filled up with water Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

9 Photon Background at IP Lawrence Deacon Per bunch crossing, within a 12 x 12 m 2 square plane 3. 5 m from the IP Input particles Disrupted beam Beamstrahlung photons Flux [cm-2] [ke. V] 7. 5± 1. 1 360 <0. 5 Coherent pairs – wrong sign 8. 4± 0. 9 230 Coherent pairs – right sign 6. 1± 0. 6 300 All 22+3. 1 -2. 6 Per bunch crossing, within a 2 x 2 m 2 square plane 3. 5 m from the IP: All: 72 +46 -18 Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

10 Neutron Background Lawrence Deacon Per bunch crossing, within a 12 x 12 m 2 square plane 3. 5 m from the IP Input particles Disrupted beam Flux [cm-2] [ke. V] 1. 6± 0. 5 950 Beamstrahlung photons <0. 5 Coherent pairs – wrong sign Coherent pairs – right sign All 1. 6± 0. 4 350 0. 7± 0. 2 390 Edda Gschwendtner, CERN 3. 9+1. 6 -1. 1 LCWS 11, Granada, 26 -30 Sept 2011

Lawrence Deacon Origin of Backscattered Photons intermediate dump C-shape magnets 27. 5 m window-frame magnets 67 m 315 m Position of last scatter of backscattered photons 11 Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

12 Background Calculations to IP Preliminary results show that background particles at ‘outer edge’ of the LCD is low. The detector yoke and calorimeter will further shield the vertex and tracker against hits from background photons. Even if additional absorbers are needed, space is available in the forward region between the detector (6 m) and the first post-collision line magnet (27 m). Lumical BPM Spent beam N. Siegrist, H. Gerwig Edda Gschwendtner, CERN Beamcal Kicker QD 0 LCWS 11, Granada, 26 -30 Sept 2011

Lawrence Deacon Energy Deposition in Beam-line Components Masks 2 -4 Int. dump Magnets 2 -4 Magnets 5 -8 Magnets 1 a and 1 b Mask 1 • • Most magnets exceed 100 W/m, many exceed k. W/m As much energy lost in magnets as masks Edda Gschwendtner, CERN 13 LCWS 11, Granada, 26 -30 Sept 2011

14 Impact on Magnet Lifetime Lawrence Deacon, Michele Modena Magnet lifetime limited by radiation damage to the coil insulation material. • • Assume that conventional magnet coils (e. g. Cu insulated with epoxy impregnated fibre glass) can withstand 1 E 7 Gy. Worst case: Magnet 5 ~6 k. W. • • • Mass is 6. 9 E 4 kg, so it's dose rate is 8. 7 E-2 Gy/s. Assuming a homogenous magnet and uniform particle distribution, the magnet lifetime would be only ~4 years. changes to Post-Collision Line Simulation: • • Change material of absorbers for magnets 1 -4 from graphite to iron Lengthening the intermediate dump with 2 more metres of iron Assessment on specific CLIC radiation issues necessary • Robust coil fabrication technology for doses up to 1010 – 1011 Gy. Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

15 Results for Magnet Lifetime intermediate dump carbon based absorbers 5 side view 27. 5 m Lawrence Deacon 6 7 C-shape magnets 8 ILC style water dump 1 a 1 b window-frame magnets 2 3 4 Magnet coil Lifetime [year] - v 1 Lifetime [year] - v 2 1 a + 1 b 2500 2 5. 7 36 3 1. 2 7. 6 4 1. 2 8. 4 5 11 26 6 13 20 7 19 33 Preliminary results 8 170 Simulations ongoing Components downstream final dump were omitted Edda Gschwendtner, CERN 230 LCWS 11, Granada, 26 -30 Sept 2011

16 Main Beam Dump side view beamstrahlung photons 1. 5 Te. V 300 Ge. V disrupted beam + right sign coherent pairs 90 cm ILC type water dump 210 m to IP: 105 m Collided 1. 5 Te. V Beam at water dump 315 m from IP Right sign coherent beam Beamstrahlung photons Disrupted beam Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

17 Luminosity Monitoring Armen Apyan m+m- pair production Main dump as converter muons install detector behind dump perform simulations of expected muon signals at different beam collision parameters See Armen’s talk Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

18 Main Beam Dump Uncollided Beam Parameters e- beam energy [Ge. V] e- per bunch [109] bunches per bunch train bunch energy [MJ] frequency [Hz] beam power [MW] sx [mm] sy [mm] Area 1 s = psxsy [mm 2] Alessio Mereghetti CLIC 1500 3. 72 312 0. 28 50 13. 9 1. 79 3. 15 ILC 500 20 2820 4. 51 4 18. 0 2. 42 0. 27 TESLA 250 20 2820 2. 26 5 11. 3 0. 35 0. 5 17. 76 2. 05 0. 55 ILC 18 MW water dump – – 30. 0 cm Cylindrical vessel diameter Volume: 18 m 3, Length: 10 m window (Ti) 1. 0 mm thick Diameter of 1. 8 m Water pressure at 10 bar (boils at 180 C) – Ti-window, 1 mm thick, 30 cm diameter baseline for CLIC 2011 main dump Edda Gschwendtner, CERN 15. 0 mm thick Ti vessel Dumptype ILC axis water dump Diameter 1. 5 m Length 10. 0 m LCWS 11, Granada, 26 -30 Sept 2011

19 Alessio Mereghetti Energy Deposition in Main Beam Dump 230 J cm-3 per bunch train ILC: 240 J cm-3 /bunch train (6 cm of beam sweep); TESLA: ~20 k. J cm-3 / bunch train; Dx=0. 45 mm; Dy=0. 78 mm; Dz=20. 0 mm; Dr=2. 5 mm; Df=6. 0 deg; Dz=20. 0 mm; ~40 k. W cm-1 Dr=2. 5 mm; Df=6. 0 deg; Dz=20. 0 mm; ILC: 54 k. W cm-1; TESLA: 32 k. W cm-1; Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

20 Activation of Materials Alessio Mereghetti Rate of production of radio-nuclides for un-collided beam (stable nuclides are not shown – collided beam: ~3% lower values): In the TESLA report, the production rates for 3 H and 7 Be are expected to be roughly a half of the present values. Edda Gschwendtner, CERN Main species in H 2 O: 3 H 7 Be 11 C 13 N 15 O t½ 12. 3 y 53. 6 d 20 m 10 m 2 m R [s-1] 3. 12 1014 3. 39 1014 5. 51 1013 1. 19 1015 LCWS 11, Granada, 26 -30 Sept 2011

21 Water - Thermodynamic Transient Cesare Maglioni Instantaneous p rise (k. Pa due to 1 bunch train) – whole dump, fast-transient analysis › › Beam energy : 1. 5 Te. V 3. 72 E 9 e-/bunch Edda Gschwendtner, CERN dynamic pressure wave, estimated at 5 -7 bar at walls (static bath, pin ~ 2 -2. 5 bar) LCWS 11, Granada, 26 -30 Sept 2011

22 Window Thermo-Mechanics › Cesare Maglioni With such a simplified preliminary design, the window cannot withstand the hydrostatic pressure (10 bar) + dynamic wave : only energy deposition analysis 1 bunch-train Temperature rise Steady-State Temperature rise (multi bunch-trains) Temperature rise per 1 pulse: DT=1. 44°C Stable temperature after 3 -4 sec with DT=24°C Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

23 Considerations & Next Steps Cesare Maglioni • • Min Water flow = 30 l/s, nominal is 60 l/s, @ v =1. 5 m/s, Tin = 22°C • • If no water circulation is active (failure), water boils after few pulses • • TH 2 O, boil = 180°C @ 10 bar drive the choice of the hydraulics of the primary water loop, to be designed for high turbulence, mainly radial water flow and heat evacuation (orthogonal heat flux = 14 MW/ps 2) Energy density peak is at E=230 J/cm 3/bunch train. The instantaneous T rise is 55°C. The interlock system must be as fast as 3 bunch trains. The dynamic pressure wave inside the bath, estimated at 5 -7 bar at walls (in the case of a static bath) has to be taken into account for dimensioning the water tank, beam window, etc. . Hydrodynamic effects due to water flow have to be further analysed. Of help : stiffeners on tank and window, sweeping system, shock absorber / diluter (gas bubbles? this requires a detailed analysis and dedicated CFD simulations). A steel / copper plate internal to the water bath, near to the end of it, and cooled by water directly, helps absorbing the tail of the beam. Dump window needs special care, detailed study and the design of a maintenance system Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

24 Summary • Many new results achieved since last year – Background calculations to the IP – Magnet life-time estimates – Thermo-dynamical results on the beam dump and dump window – Muon monitoring for luminosity measurement • Next steps – Continue work on these issues to come up with a more detailled design Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

25 SPARE SLIDES Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

26 Post-Collision Line Magnets Designed considerations: • average current density in copper conductor < 5 A/mm 2. • magnetic flux density in magnet core is < 1. 5 T. • temperature rise of cooling water < 20° K. Magnetic length Full magnet aperture horiz. / vert. [m] Full magnet dimensions horiz. / vert. [m] Power consumption Mag 1 a 1 b 2 m 0. 22 / 0. 57 1. 0 / 1. 48 65 k. W Mag 2 4 m 0. 30 / 0. 84 1. 12 / 1. 85 162. 2 k. W Mag 3 4 m 0. 37 / 1. 16 1. 15 / 2. 26 211 k. W Mag 4 4 m 0. 44 / 1. 53 1. 34 / 2. 84 271 k. W Mag. C-type 4 m 0. 45 / 0. 75 1. 92 / 1. 85 254 k. W Dipole names All magnets strength of 0. 8 T In total 18 magnets of 5 different types Total consumption is 3. 3 MW Edda Gschwendtner, CERN M. Modena, A. Vorozhtsov, TE-MSC LCWS 11, Granada, 26 -30 Sept 2011

27 Magnets C-type Mag 1 a (window frame type) incoming beam 5. 1 cm 11. 1 cm Mag 1 a Mag 1 b 27. 5 m 30. 5 m Edda Gschwendtner, CERN 20. 1 cm Mag 2 38 m 34. 3 cm Mag 3 46 m 40. 8 cm Mag 4 54 m 49 cm Intermediate dump 67 m 54. 1 cm spent beam Ctype 75 m LCWS 11, Granada, 26 -30 Sept 2011

28 Main Dump • 1966: SLAC beam dump – 2. 2 MW average beam power capacity – Power absorption medium is water • 2000: TESLA – 12 MW beam power capacity – Water dump Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

29 › › More Detailed View of Main Beam Dump An internal (closed) water loop => heat dilution › Safety confinement Vortexes, turbulence, 30 An external water cooling system => heat removal Dump window Dump shielding Water bath Vessel Internal primary water loop External secondary Water cooling system beam 10 bar HE Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

30 Alessio Mereghetti Collided CLIC Beam at Main Dump Distribution of secondary particles generated by the GUINEA-PIG code and transported up to the dump entrance via the DIMAD code (M. Salt). Collided electrons Coherent electrons Window Photons Dump IP-dump distance: 315 m; Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

31 Energy Deposition in CLIC Beam Dump Window Titanium window: 1 mm thick, 30 cm diameter, cooling on internal surface by dump water at 180°C Total deposited Power: ~6. 3 W uncollided beam max ≈ 4. 3 J/cm 3 max ≈ 0. 13 J/cm 3 collided beam A. Mereghetti, EN-STI Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

32 Assumptions › › › Cesare Maglioni Thermodynamic estimates – some easy numbers to get the feeling : › › Input punctual values from FLUKA analysis Symmetric un-disrupted beam adsorption (worst case) (Static) water bath Thermodynamic Transient – preliminary simulations › › › Energy Density map from FLUKA analysis Symmetric un-disrupted beam adsorption (worst case) Static water bath : meaningful of what may happen in case of major issues with the cooling / circulating system Thermo Mechanics of the (simplified) dump window › › › Simplified geometry (circular ᴓ 30 cm, flat, Ti-6 Al-4 V), perfect convection Only Energy Deposition analysis. Energy Density map from FLUKA Symmetric un-disrupted beam adsorption (worst case ? ) Edda Gschwendtner, CERN LCWS 11, Granada, 26 -30 Sept 2011

33 Thermodynamic Estimates › Average Power deposited P = 14 MW mass flow : = 60 l/s to keep reasonable DT (50°C) = 25 l/s not to reach boiling point (180°C) › Energy Density Peak E* = 230 J/cm 3/bunch train Peak Instantaneous T rise ~ 55 °C › › interlock system as fast as 3 bunch trains ( DT < 180°C) rapid intervention of safety system and valves Peak Instantaneous p wave~ 560 bar › › › Edda Gschwendtner, CERN Proportional to initial pressure pi = 10 bar : With exponential (radial) decay 6 – 7 bar at walls Risk of cavitation LCWS 11, Granada, 26 -30 Sept 2011

Considerations & Next Steps 34 1 Component Window design Issues Stresses due to : A. Dynamic pressure wave B. hydrostatic pressure C. beam heating Possible Solution A. Choice of material B. Sweeping system C. Stiffeners D. Multiple window design E. Additional water-jet cooling F. Low pressure water and free surface 2 Tank integrity Stresses due to : A. Dynamic pressure wave A. B. C. D. 3 5 Water circuit Maintenance . . . A. Periodical window A. Remote handling System replacement B. Automated window exchange B. Management of radionuclides system and radiolysis products C. Local, separated storage of water 6 Safety A. Containment of activated water in case of accident C. Maglioni A. B. C. D. Sweeping system Shock Absorbers / gas bubbles Stiffeners Low pressure water and free surface Water tight dump enclosure Safety enclosure + basin Pre-window on beam pipe Interlock and safety system for nuclear pressurized vessel CLIC Water Dump Thermo-Mechanical Analysis - 18 March 2018 34