SATURATION BEHAVIOUR OF THE LHC NEG COATED BEAM
SATURATION BEHAVIOUR OF THE LHC NEG COATED BEAM PIPES Tommaso Porcelli • Introduction • Experimental results • 2 m long NEG coated vacuum chamber • 28 m long NEG coated pilot sector • Applications to the LHC • Conclusions 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 1
Use of NEG in the LHC • • The LHC has a perimeter of about 27 km. Three different vacuum systems are present: • UHV beam vacuum at cryogenic temperatures; • Insulation vacuum; • 6 km of UHV beam vacuum at room temperature. • • • 29 th June 2012 Room-temperature beam vacuum sectors: Long Straight Sections (LSS). Experimental and utility insertions. Ti. Zr. V NEG coatings are extensively used on the inner walls of the LSS beam pipes, allowing to maintain the required UHV conditions. TE-VSC-LBV Tommaso Porcelli 2
NEG pumping characteristics • Ø NEG alloys are able to adsorb gas molecules (H 2, H 2 O, CO 2, N 2) on their surface by chemisorption. Noble gases and hydrocarbons are not pumped by NEG. Need for sputter-ion pumps associated to NEG. By adsorbing gas molecules (except for H 2), NEG surfaces progressively saturate and lose their pumping properties. A bakeout and activation process is necessary to restore the initial NEG pumping speed. 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 3
Saturation sources During operation with high-intensity proton beams: • Degassing in the room temperature LSS of the LHC from the not NEG-coated parts (cold-warm transitions, collimators, beam monitors). Ø Possible partial saturation of the NEG vacuum chambers. Need for a regular evaluation of the saturation level in the LHC NEG coated beam pipes: • To keep the designed vacuum performances; • To schedule technical interventions for NEG reactivation. 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 4
Experimental measurements • Experimental setup: • 2 m long NEG coated vacuum chamber; • 28 m NEG coated pilot sector. • Progressive NEG saturation by means of CO injections. • Calculation for H 2, N 2 and CO of: • Transmission; • Pumping speed; • Capture probability. vs NEG saturated length. • Benchmarking of results: • MOLFLOW+ (Test-Particle Monte Carlo); • VASCO (Multi-gas model). • Evaluation of NEG saturation in the LSS NEG coated vacuum sectors. 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 5
Experimental test bench NEG coated length: 2 m 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 6
Working procedure • NEG chamber activated • ΔP>1 order of magnitude • Progressive NEG saturation performed by means of a series of CO injections. After every CO injection, small injections of: • H 2 • N 2 All injection performed by increasing the injected gas flow step by step. Last CO injection NEG completely saturated Total quantity of CO injected: 0. 5 mbar∙l (2. 2∙ 1015 CO molecules/cm 2) PBA, ENTRANCE=PBA, END 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 7
Transmission measurements Sticking factors s of Ti. Zr. V coatings [1]: • ≈ 0. 7 for CO • ≈0. 1 for N 2 • ≈6∙ 10 -3 for H 2 Transmission strongly depends on NEG sticking factor. The higher is s, the higher is Tr. NEG saturated length Gas 17 cm 74 cm 145 cm 188 cm H 2 20% 46% 74% 90% N 2 10% 43% 83% 97% CO 8% 10% 26% 45% 29 th June 2012 • CO transmission decreases more slowly. • H 2 and N 2: Tr reduced of 1/2 when 1/3 of the NEG length is saturated. [1] P. Chiggiato, P. Costa Pinto, Thin Solids Films 515 (2006) 382 -388. TE-VSC-LBV Tommaso Porcelli 8
Saturation «seen» by gauges BA BA ENTR ANCE END Direction of saturation H 2 pumping speed as a function of the saturated length as seen at the entrance and at the end of the vacuum chamber 310 BAEND H 2 pumping speed [L/s] 260 210 BAENTRANCE Same pumping speed: Half chamber saturated 160 110 • 60 • BAENTRANCE sees variations immediately. BAEND: when about 40 cm are saturated. 10 0, 5 5, 0 500, 0 Saturated length [cm] 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 9
Transmission: MOLFLOW+ simulations • H 2 transmission as a function of the NEG sticking factor, for different saturated lengths. H 2 sticking factor • Most probable interval for H 2 sticking factor: 0. 001≤α≤ 0. 01. • As the NEG saturated length increases, H 2 transmission decreases. 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 10
H 2 transmission data Error bars: 20% (from accuracy of BA gauges) 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 11
Benchmarking data-MOLFLOW+ Error bars: 20% (from accuracy of BA gauges) Simulations in good agreement with the experimental data and with measurements made at CERNin the past [1] P. Chiggiato, P. Costa Pinto, Thin Solids Films 515 (2006) 382 -388. 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 12
28 m long pilot sector • More complex vacuum system. • The pilot sector reproduces the typical structure of the LSS NEG coated vacuum sectors. • Three BA gauges and two RGAs to detect pressure variations. • Two sputter-ion pumps at the ends of the pilot sector. 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 13
H 2 transmission data vs simulations • Simulated transmission for a 7 m long NEG vacuum chamber: ≈2500 • Measured transmission for half of the pilot sector (14 m): ≈100 Transmission≈2500 Sticking=5∙ 10 -3 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 14
CH 4 influence on transmission • Measurements strongly affected by the CH 4 degassing of the Penning gauge in the injection line. Increase of NEG saturated length more H 2 molecules reach the ends of the pilot sector. ≈4% (Penning off) Centre: H 2 is always dominant during injections. ≈83% (Penning on) H 2 flow: Q=10 -6 mbar∙l/s 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 15
Influence of sputter-ion pumps Transmission variations for NEG saturated length comprised between 10 m and 20 m: • 30 -40 (pump on); • 2 -3 (pump off). If the pump is off: • CH 4 partial pressure at the ends of the pilot sector is very high; • significant variations of H 2 pressure at these positions are not visible. Ø It is much easier to recognise significant differences of transmission when the sputter-ion pumps are turned on. 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 16
CO transmission: calculated at the end of every CO saturation, before stopping the injection. • CO transmission remains constant until about 7 m of the NEG chamber are saturated and then begins to gradually decrease. • Independently of CH 4 influence, CO transmission is much higher with respect to H 2 ( sticking coefficients). 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 17
Application in the LSS Vacuum Pressure [mbar] ΔP [mbar] sector February 2009 March 2012 B 5 R 7. R 1. 9∙ 10 -10 1. 6∙ 10 -9 1. 4∙ 10 -9 B 5 L 7. R 1. 2∙ 10 -10 9. 6∙ 10 -10 8. 4∙ 10 -10 A 5 R 5. B 2. 3∙ 10 -11 1. 9∙ 10 -10 1. 7∙ 10 -10 A 5 L 5. R 2. 7∙ 10 -11 1. 6∙ 10 -10 1. 3∙ 10 -10 Pressures data: • 2009 (just after NEG activation) • 2012 (after the 2011 high intensity proton beam physics run). Data taken without beam circulating (static vacuum). ≈3. 5 m of NEG saturated on both sides of the collimator • • Thermal degassing in static vacuum conditions is constant. Determination of the loss of pumping speed seen by the BA gauge in the vacuum sector. Estimation of NEG saturated length. Results confirmed by VASCO code simulations of the vacuum sector. 2012: partial saturation ΔP=1. 7∙ 10 -10 mbar 2009: NEG activated 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 18
CMS: intervention area • Substitution of a non-conforming RF insert at 18 m on the right side of CMS interaction point (IP). • Need to operate under Ne flow to avoid NEG saturation. BA gauge at 13 m 29 th June 2012 BA gauge at 18 m TE-VSC-LBV Intervention area Tommaso Porcelli 19
CMS: H 2 transmission • H 2 source: NEG cartridge located at 18 m from the IP. • H 2 transmission measurements confirm NEG reactivation after partial bakeout. Before partial bakeout After partial bakeout Gauge Pressure Transmission position [mbar] ratio 18 m right side 6. 4∙ 10 -6 1 13 m right side 1. 2∙ 10 -7 13 m left side 18 m left side 29 th June 2012 Pressure Transmission [mbar] ratio 18 m right side 5. 6∙ 10 -7 1 53 13 m right side 8. 4∙ 10 -10 667 1. 4∙ 10 -8 457 13 m left side 2. 5∙ 10 -10 2240 1. 1∙ 10 -8 582 18 m left side 1. 3∙ 10 -10 4308 Gauge position TE-VSC-LBV Tommaso Porcelli 20
Conclusions Possible methods for assessment of NEG saturation: • CO transmission: § More precise than H 2; § Not feasible in the LHC because it causes NEG saturation. • N 2 transmission: § Causes NEG saturation. § Has a particular adsorption mechanism. Ø H 2 is chosen as reliable gas for transmission measurements in the LHC. § NEG cartridge installed along the LSS can be used for this purpose. § Strong influence of CH 4 degassing, especially on transmission, in the LSS room temperature vacuum sectors. § Need to operate with sputter-ion pumps switched on. 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 21
Thank you for your attention! 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 22
% of lost pumping speed abruptly increases only when even the last 20 cm of still vacuum activated NEG become saturated. Gas NEG saturated length 17 cm 74 cm 145 cm 188 cm H 2 17% 33% 52% 72% N 2 27% 48% 62% 75% CO 42% 58% 66% 73% 29 th June 2012 TE-VSC-LBV One tenth of the initial total length is sufficient to ensure about 25% of the initial pumping speed. Tommaso Porcelli 23
CH 4 influence RGACENTRE RGASALEVE • Negligible influence of CH 4 in the centre, where the gas flow is injected, independently of NEG saturation. • Extremities of the pilot sector: • CH 4 is dominant if NEG is totally or partially activated; • H 2 is dominant only if NEG is totally saturated. 29 th June 2012 TE-VSC-LBV Tommaso Porcelli 24
- Slides: 24