DTL status Laboratori Nazionali di Legnaro Italy A
DTL status Laboratori Nazionali di Legnaro (Italy) A. Pisent ESS Roma 2012
Group involved • The participation to IFMIF-EVEDA project (RFQ realization) is based on four INFN sites (LNL, Padova, Torino, Bologna) with competences in accelerator physiscs, RF, mechanical engineering, computer controls, vacuum, brazing. . . • Almost all of these competences (+Naples) will be gradually involved in ESS DTL, at present – – – – AP (coordination) M. Comunian (beam dynamics) F. Grespan (RF) R. De Prisco (RF) P. Mereu (Mechanical design, Torino) V. Vaccaro (RF stabilization, Napoli) C. Roncolato (vacuum) E. Fagotti (cooling system and acc. Phys. ) Laboratori Nazionali di Legnaro (Italy) IFMIF EVEDA RFQ organization In INFN LNL group A. Pisent ESS Roma 2012
Collaboration with CERN linac 4 • Collaboration on DTL since 2006, in design (F. Grespan 1 year visit at CERN for PHD thesis on DTL stabilization) • participation of INFN to the construction of DTL prototype. • INFN is in charge of the construction of movable tuners of linac 4. • 13 Sept 2011 we had a miniworkshop on main aspects of DTL realization. https: //indico. cern. ch/conference. Display. py? conf. Id=152766 • We are finalizing the paper work for full access to Catia drawings (possible starting point). Laboratori Nazionali di Legnaro (Italy)
DTL Physics design: general parameters RF frequency Design Current (10% margin included) Input energy Final energy (approx. ) Maximum power/Tank Maximum Surface field at 3 Me. V PMQ maximum integrated field Lattice PMQ law Trans. Input emittance Long. Input emittance Tank 1 2 3 4 5 Total Laboratori Nazionali di Legnaro (Italy) No modules/Cells 4/70 3/26 2/18 2/16 2/14 144 A. Pisent Length (m) 8. 19 5. 48 4. 67 4. 71 4. 52 27. 57 352. 21 55 3 81 2. 2 1. 5 1. 2 3. 8 FODO Equipart. 0. 24 0. 35 Wfinal (Me. V) 22. 08 39. 63 54. 64 69. 21 82. 52 ESS Roma 2012 MHz m. A Me. V MW Kilp kilp T mmmrad Power (MW) 2. 0 2. 2 2. 0. 2. 1 2. 0 10. 3
DTL Physics design: final energy 50 -80 Me. V shunt impedance higher than for any other structure (probably) RF frequency Design Current (10% margin included) Input energy Final energy (approx. ) Maximum power/Tank Maximum Surface field at 3 Me. V PMQ maximum integrated field Lattice PMQ law Trans. Input emittance Long. Input emittance Tank 1 2 3 4 5 Total No modules/Cells 4/70 3/26 2/18 2/16 2/14 144 Length (m) 8. 19 5. 48 4. 67 4. 71 4. 52 27. 57 Wfinal (Me. V) 22. 08 39. 63 54. 64 69. 21 82. 52 MHz m. A Me. V MW Kilp kilp T mmmrad Power (MW) 2. 0 2. 2 2. 0. 2. 1 2. 0 10. 3 2. 5 m module limit as for GSI galvanic Mohammad Eshraqi, Jim Stovall Laboratori Nazionali di Legnaro (Italy) 352. 21 55 3 81 2. 2 1. 5 1. 2 3. 8 FODO Equipart. 0. 24 0. 35 A. Pisent ESS Roma 2012
DTL Physics design: ramped field RF frequency Design Current (10% margin included) Input energy Final energy (approx. ) Maximum power/Tank Maximum Surface field at 3 Me. V PMQ maximum integrated field Lattice PMQ law Trans. Input emittance Long. Input emittance Tank 1 2 3 4 5 Total Laboratori Nazionali di Legnaro (Italy) No modules/Cells 4/70 3/26 2/18 2/16 2/14 144 A. Pisent Length (m) 8. 19 5. 48 4. 67 4. 71 4. 52 27. 57 352. 21 55 3 81 2. 2 1. 5 1. 2 3. 8 FODO Equipart. 0. 24 0. 35 Wfinal (Me. V) 22. 08 39. 63 54. 64 69. 21 82. 52 ESS Roma 2012 MHz m. A Me. V MW Kilp kilp T mmmrad Power (MW) 2. 0 2. 2 2. 0. 2. 1 2. 0 10. 3
DTL Physics design: FODO lattice Space available for dipole steerers and BPM (@SNS FF 0 DD 0 and @CERN FFDD) In operation Beam center can be tuned, not phase advance (should be sufficient). FODO allowes better shunt impedance if largesmall DT alternance is possible (em quads? ) RF frequency Design Current (10% margin included) Input energy Final energy (approx. ) Maximum power/Tank Maximum Surface field at 3 Me. V PMQ maximum integrated field Lattice PMQ law Trans. Input emittance Long. Input emittance Tank 1 2 3 4 5 Total No modules/Cells 4/70 3/26 2/18 2/16 2/14 144 352. 21 55 3 81 2. 2 1. 5 1. 2 3. 8 FODO Equipart. 0. 24 0. 35 Length (m) 8. 19 5. 48 4. 67 4. 71 4. 52 27. 57 Wfinal (Me. V) 22. 08 39. 63 54. 64 69. 21 82. 52 MHz m. A Me. V MW Kilp kilp T mmmrad Power (MW) 2. 0 2. 2 2. 0. 2. 1 2. 0 10. 3 2. 5 m module limit as for GSI galvanic Laboratori Nazionali di Legnaro (Italy) A. Pisent ESS Roma 2012
DTL Physics design: equipartitioning PMQ max=69 T/m PMQ min=28 T/m RF frequency Design Current (10% margin included) Input energy Final energy (approx. ) Maximum power/Tank Maximum Surface field at 3 Me. V PMQ maximum integrated field Lattice PMQ law Trans. Input emittance Long. Input emittance Tank 1 2 3 4 5 Total No modules/Cells 4/70 3/26 2/18 2/16 2/14 144 352. 21 55 3 81 2. 2 1. 5 1. 2 3. 8 FODO Equipart. 0. 24 0. 35 Length (m) 8. 19 5. 48 4. 67 4. 71 4. 52 27. 57 block design PMQ can be used. A. Pisent mmmrad Wfinal (Me. V) 22. 08 39. 63 54. 64 69. 21 82. 52 From the second Tank Laboratori Nazionali di Legnaro (Italy) MHz m. A Me. V MW Kilp kilp T ESS Roma 2012 Power (MW) 2. 0 2. 2 2. 0. 2. 1 2. 0 10. 3
Beam envelopes • • • Each tank begins with an empty Drift Tube (for BPM). Only a transverse matching is needed. Matching by using 2 PMQ at tank end and 2 PMQ at tank begin SNS Tank 1 -Tank 2: 1 bl Linac 4 DTL Tank 1 -Tank 2: 2 bl Laboratori Nazionali di Legnaro (Italy)
Intertank space: 1 bl F O D O bl bl F O bl From first to second tank=180 mm Laboratori Nazionali di Legnaro (Italy)
RMS emit Laboratori Nazionali di Legnaro (Italy)
Phase advance/m Laboratori Nazionali di Legnaro (Italy)
Output phase space using Gaussian dist. (6 s) Laboratori Nazionali di Legnaro (Italy)
DTL Physics design: error study • The error study is preliminary (not done with the last tank subdivision) Emit [75%] = 6. 9900 Pi. mm. mrad [ Norm. ] Emit [75%] = 8. 2433 Pi. deg. Me. V [ Norm. ] DTL 50 m. A Laboratori Nazionali di Legnaro (Italy) A. Pisent ESS Roma 2012
DTL Physics design: error study • The error study is preliminary (not done with the last tank subdivision) Errors are the double of the CERN PMQ tender specification • Magnetic center respect the geometrical center shake of +/- 0. 2 mm • Yaw/pitch/Roll of +/-1°=17 mrad • Gradient error of +/-1% • All errors apply together with a Gaussian (6 s) input beam distribution • 100 random DTL generated. • 10^6 particles i. e. 0. 32 W for particle at 100 m. A, 0. 16 Watts for particle at 50 m. A. • Separate X, Y Steerer used with max force of 1. 6 m. T*m. Laboratori Nazionali di Legnaro (Italy) A. Pisent ESS Roma 2012
Steerers on FODO Lattice • Using the empty space it has been put steerers X or Y almost at 90° phase advance apart for tank. • • 8 Steerers (2 couples for each plane) and 4 BPM for tank 1. 4 steerers and 2 BPM for the other tanks. Every tank starts with a BPM. Max steerer strength of 1. 6 m. T*m. SNS Steerer, max 1. 9 m. T*m DTL 100 m. A, Strength at max errors Laboratori Nazionali di Legnaro (Italy)
Errors results with correction steerers Step 1 Maximum Quad shake of X, Y ± 0. 2 mm; ± 1°; ± 1% DTL 50 m. A Laboratori Nazionali di Legnaro (Italy)
DTL 50 m. A Total=102 Watts Quad shake of X, Y ± 0. 2 mm; ± 1°; ± 1% Laboratori Nazionali di Legnaro (Italy)
The losses can be further decreased by increasing the DTL bore radius at high energy, above 50 Me. V. In the last tanks we can increase the bore radius with the same ZTT Laboratori Nazionali di Legnaro (Italy)
Physics design conclusions • Defined the DTL design for 50 m. A, 5 tanks, 13 modules. • An alternative 4 tanks 16 modules is quite similar with pro and cons under evaluation. • Main differences of ESS design with respect to the Linac 4 design: – – – Use of FODO lattice instead of FFDD. Stronger PMQ in the first tank. Use of steerers and diagnostics (BPM) inside DT. Ramped Field E 0. Higher energy for the first inter tank transition. Shorter intertank space: 1 bl instead of 2 bl. • All these changes are in the direction of a smoother lattice. • Preliminary error studies show fair tolerances (looser than for linac 4 DTL) but the real evaluation has to be done on the integrated linac up to final energy. • The final ZTT optimization (with the design completely defined) will allow a further improvement of the power efficiency. Laboratori Nazionali di Legnaro (Italy)
Elements part of DTL WP (boundaries) LPRF Control system HPRF RF network RF window Local Control system Alignment references DTL INFN Vacuum pumps support Water cooling skid Heat exchanger General water distribution Laboratori Nazionali di Legnaro (Italy) A. Pisent ESS Roma 2012 Beam port flange MEBT RF coupler RF pick ups Spoke section
Girder (SLAVE) • Vacuum tank providing very high stiffness and ensuring structure stability (MASTER) • Girder with high machined bushing positioning (SLAVE) precision for DTs • No adjustment system (DTs alignment relies on mechanical ) tolerances Tank (MASTER) Laboratori Nazionali di Legnaro (Italy) A. Dallocchio 13. 09. 2011 22
~8 m • DTL made of 3 Tanks • Inter-Tanks ~8 m ~4 m • Each Tank independently supported on alignment jacks (isostatic configuration) • Each tank segment equipped with girder and DTs (~ 2 m long) • ESS lay out will be similar • Five tanks of 8. 2+5. 5+5+5+5 m, 4+3+2+2+2 modules, maximum module length 2. 4 m • Each tank with 2 couplers and 1 pumping station Laboratori Nazionali di Legnaro (Italy) A. Dallocchio 13. 09. 2011 23
Pick-up Movable Tuners RF Tuners Survey Target Post Couplers RF Wave-guide Coupler Alignment Jack Laboratori Nazionali di Legnaro (Italy) A. Dallocchio 13. 09. 2011 24
• Vacuum Tank: • Material → high quality Stainless Steel 304 L (UHV spec. ) for ESS perhaps soft iron, like for the prototype. • Very high stiffness → external diameter ~ 620 mm • Internal Copper plating on fine surface finishing (Ra 0. 8) to ensure low RF losses • Girders: • Material → Aluminum alloy EN AW 5083 -H 111 • Fretted Stainless Steel bushing to ensure DTs precise positioning • Helicoflex seals ensure vacuum tightness as well as RF continuity • Efficient cooling circuit for SPL 10% duty-cycle (precise deep-drilling of the vacuum tank) Laboratori Nazionali di Legnaro (Italy) A. Dallocchio 13. 09. 2011 25
Laboratori Nazionali di Legnaro (Italy) A. Dallocchio 13. 09. 2011 26
Bushing DT Assembly tool Helicoflex seal Drift Tube (including cooling circuit and precise PMQ housing) PMQ Laboratori Nazionali di Legnaro (Italy) Vacuum Barrier (Helicoflex seal) A. Dallocchio 13. 09. 2011 27
CERN-LNL DTL prototype • LNL contribution: maching of the tank and all components. • CERN for e-beam welding of the tubes, Cu plating of the tank, final assembly and RF high power tests. • High power tests concluded and published at Linac 08 Drift tubes, machined at Cinel srl (Italy) Laboratori Nazionali di Legnaro (Italy) Prototype Tank, 1 m long, machined at Cinel srl (Italy) A. Pisent ESS Roma 2009
First reduced prototype (1 m) built by CINEL (Italy) and successfully tested at CERN Laboratori Nazionali di Legnaro (Italy) A. Dallocchio 13. 09. 2011 29
DTL Tank design, copper plating • Copper plating at CERN will be almost impossible due to workload; industrial or GSI alternative. • Dimensions (2. 5 m long and 0. 6 m diameter), thickness 20 um, Ru 0. 8 um, are challenging. • R&D a full scale prototype of the tank (with design compatible with linac use) is necessary All flanges and the contact surface of the beam are protected by PVC covers, sealed with FPM o-rings Courtesy of G. Vandoni, TE-VSC Tank proto painted to avoid corrosion Laboratori Nazionali di Legnaro (Italy) courtesy M. Malabaila A. Pisent ESS Roma 2012
DT for steerer dipoles SNS Steerer, max 1. 9 m. T*m • New DT with electrical and cooling water connection for steerers (and BPM) has to be developed. • Same mechanical interface if possible (perhaps looser positioning accuracy is possible). • R&D A prototype of such DT can be tested at CERN in the DTL prototype. PM drift tube (courtesy of Y. Cuvet) Laboratori Nazionali di Legnaro (Italy) A. Pisent ESS Roma 2012
Permanent magnets 16 silver-plated screws This is not the final quad!!!! 16 pole pieces 1/30/2022 Laboratori Nazionali di Legnaro (Italy) CERN DTL: prototypes with ASTER and European industry Elytt. We have to follow this production and possibly prototype our 70 T/m quads. It is a small serie (72 quads) other suppliers are possible. Cinel has experience with PM for example for ECR exapoles. (courtesy of A. Lombardi) A. Pisent ESS Roma 2012
Helico. Flex Sealing (w/ RF continuity ) on • Frontal Flange • Upper part of Stems PMQ(108 items) Os = 1 x 10 -9 mbar Lt/cm 2 -s As = 0. 1 m 2 Total Gas Load = 1. 08 x 10 -4 mbar Lt/s Laboratori Nazionali di Legnaro (Italy) Pumps Ion Pumps (rough estimation 10 units) Oil Free Primary Pumps and Turbo-Pumps NEG Pump
Power couplers • We have experience with design of high power coupling loops, but the coupler chosen by CERN worked very well in the tests of the DTL prototype. • The design has improved with the use of posts for tuning of the coupling constant b. • Is a 2 MW coupler possible? (R&D) Coupling to pi mode cavity TM (courtesy of Patricia Ugena Tiradof) Laboratori Nazionali di Legnaro (Italy) A. Pisent ESS Roma 2012
Tuning with movable tuners or water temperature • We have experience with both the approaches (TRASCO RFQ and linac 3 RFQ) • Movable tuners react faster, but with high duty cycle t regulation is more reliable, without sliding contacts. • We have not yet reliable thermal simulations to choose. Laboratori Nazionali di Legnaro (Italy) A. Pisent ESS Roma 2012
Accelerator assembly and installation • CERN DTL will be assembled (with DT aligned) in a dedicated hall in surface outside the tunnel. The second and third DTL tanks (about 8 m long) will be the largest pieces to be introduced in the tunnel. • If the same approach is implemented of ESS DTL, such installation has to be considered for building design Laboratori Nazionali di Legnaro (Italy) A. Pisent ESS Roma 2012
Conclusions • Plan for prototyping • Mechanical design based on tank design developed at CERN (accurate positioning and fixing of tube position, metallic gaskets). • Our design has FODO lattice, steerers or BPMs inside DT, field ramping short transition • We plan to develop DT for steerers or diagnostics, with electrical connections out of the stem. • Such DT will be tested at high power in the DTL prototype (developed by CERN and INFN) • PM with trapezoidal components should be prototyped • Construction and tests perspective – Tank and DT can be built in local industry (see again prototype). Brazing and e-beam welding are possible in 70 km distance. – We are looking for an industrial partner for copper electro-plating, the design is compatible with GSI galvanic. Full scale prototyping of a tank would be beneficial. – INFN (LNL and PD) has the structures for high precision milling, vacuum brazing, CMM, EDM, RF measurements, laser tracking, assembly of the tank and low power tests. – Moreover a power system could be installed in LNL for high power tests and conditioning of couplers and tanks (dedicated space under discussion with management). Laboratori Nazionali di Legnaro (Italy)
- Slides: 37