NOνA Project Status. Leon Mualem University of Minnesota. TAMU Seminar April 25, 2006

Transcription

1 NOνA Project Status Leon Mualem University of Minnesota TAMU Seminar April 25, 2006

2 Site for Far Detector: Ash River Ash River International Falls Duluth Voyageurs National Park (shaded brown) Minneapolis Ash River Fermilab This site is at 810 km from Fermilab, km off-axis Farthest available site in the U.S. along the NuMI beam US 53, St. Louis County Road all weather access April 25 TAMU Seminar Leon Mualem 2

3 Ash River details St. Louis County 129 Voyageurs National Park 1 mile Ash River 23.5 acre site Minnesota Department of Natural Resource land, adjacent to private land access via 3.6 mile logging road Elevation 1240 ft, on hilltop ~90 ft above Ash River 5-10 ft of soil over hard granite April 25 TAMU Seminar Leon Mualem 3

4 Far Detector Hall Excavated rock NOvA Far Detector soil granite 53 ft Beam View Main structure is 516 ft long by 73 ft wide Sunk into granite for secondary containment 3 m overburden of excavated rock (30-40% voids) Satisfies scientific performance ~ Equal to MiniBooNE Catwalk access to detector on both sides Access to detector top through truss system April 25 TAMU Seminar Leon Mualem 4

5 Near Detector at Fermilab MINOS Surface Building In the existing MINOS access tunnel 340 ft below grade Near Detector moves over a 75 m long part of tunnel Constraints on the Near Detector since access is through the MINOS shaft April 25 TAMU Seminar Leon Mualem 5

6 The Basic Detector element typical charged particle path To 1 APD pixel L W D Liquid Scintillator 5.5% pseudocumene as scintillant Mineral oil and waveshifters (PPO, bis-msb) PVC cell for primary containment Horizontals:3.87 cm x 6.0 cm x 15.7 m long Verticals: 3.76 cm x 5.7 cm x 15.7 m long Highly reflective, 15% titanium dioxide Diffuse reflection keeps light local to track along the cell length to ± 25 cm ~ cosine (angle to normal to wall) Looped wavelength shifting fiber to collect light 0.8 mm diameter, double clad, K27 waveshifter Almost perfect mirror, 3.6*light in 1 fiber Avalanche Photodiode 85% quantum nm Gain of 100, operate at -15 o C Low noise amplifier April 25 TAMU Seminar Leon Mualem 6

7 PVC cells in 32-cell extrusions Fiber manifold & closure at one end 64 fibers collected to match a 32 channel APD Closure plate at the other end Fiber manifolds Fiber manifolds 1.3 m 1.3 m 15.7 m All 32 cells in a common liquid volume 15.7 m 32 cells 32 cells 15.7 m = 51 ft. With ends, still fits on a standard 53 ft trailer truck Bottom Closure closure plates plates April 25 TAMU Seminar Leon Mualem 7

8 Assemble into planes 12 extrusions wide (= 15.7 m) So the beam view is a square detector Alternate cell orientations and readout from layer to layer Left, Top, Right, Top cm cm 6 cm 3.76 cm 5.7cm Plane of vertical cells Plane of horizontal cells Epoxy the planes together into blocks 31 planes thick So all horizontal extrusions are supported by two verticals April 25 TAMU Seminar Leon Mualem 8

9 Block Raiser lifts block of 31 empty planes (127 metric tons) m 15.7 m 15.7 m 54 blocks = 1654 planes form a structure 111 meters long Unique PVC Structure, Had an external engineering review of the structure in January. Each Block of 31 planes built horizontally, Block Raiser rotates it to vertical, positions it. Fill with liquid scintillator Filling trails assembly by ~ 4 of the 54 blocks April 25 TAMU Seminar Leon Mualem 9

10 And a more modest Near Detector 209 metric tons Similar R&D detector: Integration Prototype Serves to focus remaining R&D We need all the parts GOAL: Operate in MINOS surface building in calendar 2007 at 75 mrad off-axis see a beam from kaon decay This is our testbeam + Cross calibration Kept person at ~ the same scale as previous slide ν 3 GeV ν 2 GeV ~ 750 events in one year April 25 TAMU Seminar Leon Mualem 10

11 Some numbers Massive, but unlike most HEP detectors, not many systems to build You just have to do everything 20,088 times 20,088 extrusion modules Each with 32 cells, each 1.3m x 15.7m total length 315 km 20,088 Avalanche Photodiodes Each 32 channels, total of 642,816 cells/channels 21,624 kilometers of wave-shifting fiber 13.8 tons ~ 67 tons of epoxy between the extrusion planes 25 kilotons total mass 6.9 kilotons of rigid PVC 18.5 kilotons of liquid scintillator, 73% active April 25 TAMU Seminar Leon Mualem 11

12 Performance of this design ~ 20 photoelectons / minimum ionizing particle from far end of cell Good ν e / ν µ separation Note 2 GeV ν e events are ~ 60 planes long 2 GeV ν e 2 GeV ν µ April 25 TAMU Seminar Leon Mualem 12

13 Performance of this design Electron energy resolution ~ 6% at 2 GeV Muon Quasi Elastic energy resolution ~ 3.5% at 2 GeV σ(e)/e Electrons in All electrons Quasi-elastic events 6 % Energy E in GeV σ(e)/e Muons in All muons Quasi-elastic events 3.5 % Energy E in GeV Figure of Merit = 32.2 >30 over a range of background cuts April 25 TAMU Seminar Leon Mualem 13

14 Design Parameter Compare to Scientific Performance Requirements Scientific Performance Requirement Distance off-axis 11.5 to 12.0 km km Performance of the NOνA Selected Design Distance from Fermilab Experimental Sensitivity As far from Fermilab as practically possible. Figure of merit greater than or equal to 30 (ν e signal events divided by the square root of the background.) 810 km, farthest possible site in the United States along the NuMI beamline 32.2 Energy resolution for ν e Charged Current events Less than 8% at 2 GeV 6% at 2 GeV Energy resolution for Quasi-Elastic?ν µ Charged Current events Far Detector overburden Near Detector Less than 4% at 2 GeV 2 meters of rock with density 2.5 g/cc At least a 20 ton fiducial volume located about 1 kilometer from the NuMI target with sufficient transverse and longitudinal size for neutrino event containment. 3.5% at 2 GeV 3 meters of excavated granite rock, rock density of 2.8 g/cc mixed with 40% voids ton ton fiducial volume km km from from NuMI NuMI target target Good containment Good containment All requirements satisfied April 25 TAMU Seminar Leon Mualem 14

15 Alternate Far Detector Sites Canadian Sites US Sites Baselines too short 1: Lake Superior 630 km 2: Cliffs-Erie, 712 km 3: Peyla, 735 km Viable 4: Orr-Buyck, 774 km (looked at 3 specific spots) 5: Ash River, 810 km (looked at 7 specific sites) In Canada 6: Mine Center, km power, road (OWR), no Canadian participation 7: Vermillion Bay, 950 km Beam is > 12 km above surface April 25 TAMU Seminar Leon Mualem 15

16 Choice between Ash River and Orr-Buyck Ash River more expensive access 3.3 mile access road Power upgrade ~ $ 3M more than Orr-Buyck Orr-Buyck has worse scientific performance Would take 10 kilotons more detector to equal Ash River ~ $ 60 M cost difficult for statistics to make up for shorter baseline matter effect is smaller, two mass hierarchies are less well separated In fact would just do an experiment with ~ 5% less reach in mass ordering Ash River is our selection Keep Orr-Buyck as a backup if some problem develops over land or access or April 25 TAMU Seminar Leon Mualem 16

17 Aternate Detector Technologies Detector Technology Water Cherenkov Liquid Argon TPC Sampling Calorimeter with Solid Scintillator Sampling Calorimeter with RPCs Sampling Calorimeter with Liquid Scintillator Comments Not well suited to 2 GeV with multiple tracks (multiple rings) Requires R&D beyond our timescale Just too expensive, almost twice the price of liquid scintillator Not cheaper than liquid scintillator, risks: gas contamination, glass,, worse performance since digital, energy resolution is worse than totally active designs (2.5x) and would not meet performance requirement Only few % cheaper than totally active, energy resolution is worse than totally active designs (1.5x) and would not meet performance requirement Totally Active Liquid Scintillator Our judgment is that it has best performance to cost ratio, Only solution with adequate energy resolution for scientific performance requirements Opportunities to improve event selection algorithms April 25 TAMU Seminar Leon Mualem 17

18 Alternate Liquid Scintillator Structures 15.7 m m 15.7 m 4 ft 9 ft Incoming beam direction 53 ft Selected alternate 19.2 psi at bottom of vertical cells held in check by PVC Vee Rotate by 45 degrees Reduces hydrostatic pressures but doesn t eliminate the problem Bathtub Zero hydrostatic pressure But must go to shallow stereo to do this Loss of active volume at edges is substantial Shipping Containers Low pressures since only 9 ft tall No adhesive But cracks when stack up 6 of them 5-10% worse performance ~10% additional cost April 25 TAMU Seminar Leon Mualem 18

19 Additional Alternates (CDR has chapters 7-16 organized by Level 2 subproject) Will not discuss these today in any talks but we have thought about them discuss tomorrow in breakouts if you like Far Detector Hall Above ground, halfway in the ground, 10 m underground Scintillator Biodegradable Fiber Other wave-shifting dyes, spliced fibers PVC extrusions Different plastics, aluminum Extrusion Modules Other schemes for closure plate and manifolds, spiders to hold fiber away from PVC walls Photodetectors & Electronics Multi-anode PMTs, VLPCs, HPDs, silicon PMTs different amplifiers with switched capacitor arrays, different cooling schemes DAQ Optical links, wireless Near Detector New tunnel, faster electronics for a few planes Far Detector Assembly 8-plane blocks, multiple assembly tables April 25 TAMU Seminar Leon Mualem 19

20 Optimization of the Selected Design April 25 TAMU Seminar Leon Mualem 20

21 Optimized Detector Segmentation To 1 APD pixel FoM vs. W, D (suggests might go wider fewer cells, cheaper) Cell Width (W) 3.9 cm 5.0 cm 8.0 cm 12.0 cm L Cell Depth (D) 4.5 cm cm cm cm typical charged particle path Light vs. W,D (but then less light so performance suffers) Cell Width (W) 3.9 cm 5.0 cm 8.0 cm 12.0 cm W D Cell Depth (D) 4.5 cm cm cm cm April 25 TAMU Seminar Leon Mualem 21

22 APD Photodetector Manufacturer Pixel Dimensions Pixel Pitch Array Size Die Size Quantum Efficiency (>525nm) Pixel Capacitance Bulk Dark Current (I b ) at 25C Bulk Dark Current (I b ) at 15C Peak Sensitivity Operating Voltage Gain at Operating Voltage Operating Temperature (with Thermo-Electric Cooler) Array per plane APD Channels per plane Total Number of planes APD pixels total Hamamatsu 1.95mm x 1.0mm 2.65,1.7mm 32 pixels 15mm x 15mm 85% 10pF 10pA 0.15pA 600nm 400±50V C ,816 April 25 TAMU Seminar Leon Mualem 22

23 Detected Spectra QE and Intensity m 1m 2m 4m 8m 12m 17m PMT QE APD QE Wavelength (nm) April 25 TAMU Seminar Leon Mualem 23

24 CMS Experience CMS has 124,000 5mm x 5mm APDs in ECAL 3 square meters Consistent quality Delivered at predicted cost No installed failures NO A will use 22,000 devices, 32x1.95mm*1mm 1.4 square meters active 4.5 square meters total April 25 TAMU Seminar Leon Mualem 24

25 Proposed APD Layout Optimized for 2 fiber mounting Industrial Si mounting (Flip-Chip) April 25 TAMU Seminar Leon Mualem 25

26 Optimized Cell Light Threshold FoM vs. threshold for different light levels Not so sensitive since 20 pe at FAR end, 45 pe typical Calorimeter, many pieces of information Even as a tracker, OK to have some layers of your tracker with inefficiency This is not like setting a threshold on a single counter Signal ( pe ) Threshold ( pe ) Gaussian amp noise, high tail from APD excess noise Example, for 20 pe far end (red is integrated) Hit probability from cosmic rays 25% of cosmic rate April 25 TAMU Seminar Leon Mualem 26

27 d (m) pe W Optimized Horizontal Cell D L Readout Orientation Top, Right, Top, Right Integrated over all planes, range of response is pe Top, Left, Top, Right Integrated over all planes, range of response is pe April 25 TAMU Seminar Leon Mualem 27

28 Continuous Digitization 32 channel front end ASIC T f T f T f T r T r T r 8:1 8:1 Mux Mux 8:1 8:1 Mux Mux 8:1 8:1 Mux Mux (4x) 8:1 Analog multiplexers Each runs at 16 MHz 62.5 ns settling time per channel 500 ns equivalent sampling time Digitized by quad 40MHz ADC AD41240 used in several systems in CMS 8 T f T r 8:1 8:1 Mux Mux April 25 TAMU Seminar Leon Mualem 28

29 Optimizations Optimization of these sections include Pixel size, APD mounting Match 2:1 aspect ratio of fibers Industrialize manufacture of APD modules Optimize gain to minimize threshold Optimize temperature to balance noise and cooling costs and complexity Design to minimize heat load April 25 TAMU Seminar Leon Mualem 29

30 Gain Optimization Amplifier noise is a constant number of electrons. Increasing gain reduces the effective noise contribution of the amplifier. Increasing gain increases dark current contribution to the noise. The optimum gain is at the minimum of this curve. April 25 TAMU Seminar Leon Mualem 30

31 Operating Temp Optimization Reducing the operating temperature reduces the noise by reducing the dark current. Lower temperatures increase cooling power required, reduce efficiency of TE cooler Little to gain by operation below 15C. Noise (e-) APD267 odd APD268 Odd 1000 APD287 Odd 900 APD288 Odd 800 APD267 Even 700 APD268 Even APD287 Even 600 APD288 Even Temp (C) April 25 TAMU Seminar Leon Mualem 31

32 Organizational Breakdown Structure NOvA Project John Cooper Ron Ray Dave Pushka John Oliver Anna Pla-Dalmau Suzanne Pasek Bill Freeman Harry Ferguson Keith Schuh Project Manager Deputy Manager Project Engineer, Mechanical Project Engineer, Electronics (Harvard) Project Chemist Budget Officer Project Scheduler Documentation & Schedule, Hazards ES&H Site & Building Scintillator Wavelength Shifting Fiber PVC Extrusions PVC Modules Electronics, Trigger, DAQ Detector Assembly Project Management Steve Dixon (Fermilab FESS) Stuart Mufson ( Indiana) Carl Bromberg (Michigan St) Rich Talaga (ANL) Ken Heller (Minnesota) Leon Mualem (Minnesota) Dave Ayres (ANL) John Cooper & & April 25 TAMU Seminar Leon Mualem 37

33 DOE Fermilab Director P. Oddone Higher Level Structure, complication of Cooperative Agreement Cooperative Agreement (bid process) Associate Director for Research H. Montgomery NOvA Project Manager J. Cooper Particle Physics Division Head J. Strait NOvA Spokespersons line management advisory Deputy Manager R. Ray NOvA Technical Board G. Feldman M. Messier Some entity selected via a DOE bid process reporting 2.1 Site and Building S. Dixon, FESS Other Level 2s NOvA Executive Committee NOvA Institutional Board Design-Build Contractor April 25 TAMU Seminar Leon Mualem 38

34 Cost All slides are in FY06 $ These are burdened : 16% on $500,000 M&S, 1.5% pass-thru M&S, 31% Fermilab SWF, Univ. include their burden. The cost of the building and of the commodities (scintillator, fiber, PVC) dominate this detector. Labor is not as large a fraction as we are all used to for other HEP detectors Worked very hard during the last 5 months to extract real costs for the cost driver items Particularly since several were linked to the cost of crude oil needed to learn HOW they were linked We went after bids on Integration Prototype quantities with a unilateral option for the whole detector to be exercised before November 1,2007 (FY08) We now KNOW many of these costs, could buy today and therefore the contingency is low April 25 TAMU Seminar Leon Mualem 39

35 Cost Table: FY06 $, burdened costs NOvA 's Cost Estimate FY06 $M Estimated Cost (with indirects) Contingency Estimate Contingency % Total WBS Items M&S Labor 1 Total M&S Labor 1 Total M&S Labor 1 Total Cost 2.1 Far Detector Site and Buildings $ 28.8 $ 1.1 $ 29.9 $ 6.6 $ 0.4 $ % 33% 23% $ Liquid Scintillator $ 28.1 $ 0.4 $ 28.6 $ 8.2 $ 0.4 $ % 92% 30% $ Wave-Length-Shifting Fiber $ 18.8 $ 0.0 $ 18.8 $ 5.3 $ 0.0 $ % 50% 28% $ PVC Extrusions $ 30.9 $ 0.5 $ 31.4 $ 12.3 $ 0.2 $ % 37% 40% $ PVC Modules $ 5.7 $ 3.1 $ 8.8 $ 1.9 $ 3.0 $ % 96% 55% $ Electronics Production $ 13.4 $ 0.4 $ 13.8 $ 6.4 $ 0.1 $ % 31% 47% $ 20.3 TEC 2.7 Data Acquisition System $ 1.2 $ 1.0 $ 2.2 $ 0.3 $ 0.2 $ % 19% 24% $ Near Detector Assembly $ 0.2 $ 0.5 $ 0.7 $ 0.2 $ 0.5 $ % 100% 98% $ Far Detector Assembly $ 5.4 $ 5.8 $ 11.1 $ 3.9 $ 5.8 $ % 100% 87% $ Project Management $ 0.4 $ 2.6 $ 3.0 $ 0.1 $ - $ % 0% 4% $ 3.1 Subtotal Construction $ $ 15.4 $ $ 45.1 $ 10.5 $ % 68% 38% $ PED $ 7.9 $ 0.3 $ 8.2 $ 1.7 $ 0.1 $ % 26% 22% $ 10.0 Total TEC: $ $ 15.8 $ $ 46.9 $ 10.6 $ % 67% 37% $ R&D $ 5.4 $ 6.0 $ 11.5 $ 0.1 $ 0.1 $ 0.2 2% 1% 2% $ 11.7 OPC Total OPC: $ 5.4 $ 6.0 $ 11.5 $ 0.1 $ 0.1 $ 0.2 2% 1% 2% $ 11.7 TPC: $ $ 21.8 $ $ 47.0 $ 10.7 $ % 49% 34% $ Notes: 1. Labor costs presented here include all project labor from Fermilab, other DOE facilities, and Universities. PED funds in FY07 for design -build at Ash River Cost Drivers Building is 19 % Commodities (scintillator, fiber, PVC) are 52% April 25 TAMU Seminar Leon Mualem 40

36 Schedule: List of Milestones April 25 TAMU Seminar Leon Mualem 41

37 Schedule: Graphical Form Number of Modules (4,000 = 5 kilotons) 24,000 20,000 16,000 12,000 8,000 4,000 FY08 FY09 FY10 FY11 FY12 Data, Near & Far Extrusions Fiber Modules Assembly Scintillator Setup equipment in Far Hall Builiding under construction Month from Start of Project April 25 TAMU Seminar Leon Mualem 42

38 NOνA s Recent Past/Near Future Reviews, reviews Director s CD1 review March 06 Lehman CD1 review April 06 More reviews Director s CD2 review August 06 Lehman CD2 review September 06 Just in time delivery to get into FY08 budget. Actual Operating Near detector in MINOS surface building at FNAL Collecting off-axis beam neutrinos; electron and muon Running by end of 07! April 25 TAMU Seminar Leon Mualem 43