Babatunde O’Sheg Oshinowo and Horst Friedsam
Published: Thursday, November 7, 2013 - 17:57 The NOvA (NuMI off-axis νe appearance) experiment is looking to answer fundamental questions about neutrinos and the role they play in the universe. NOvA will use two detectors, a near detector located underground at Fermilab in Batavia, Illinois, and a far detector located 810 km from Fermilab near the United States-Canada border in Ash River, Minnesota. The 14 kton far detector and the 0.3 kton near detector are composed of several cells of extruded PVC plastic in a cellular structure. This article discusses the survey of the NOvA detectors using a laser tracker and laser scanner. NOvA is a second-generation experiment on the NuMI (neutrino at main injector) beamline designed to search for oscillations of muon neutrinos to electron neutrinos (νμ → νe) by comparing electron neutrino rates at Fermilab with electron neutrino rates observed 810 km away from Fermilab at Ash River, Minnesota (figure 1).1 The NOvA experiment is an upgrade of the NuMI beam intensity from 400 kW to 700 kW. NOvA consists of a 0.3 kton near detector at Fermilab and a much larger 14-kton far detector in Ash River, Minnesota, just south of the U.S.-Canada border (figures 1 and 2). The NOvA detectors are sited 14.6 mrad off the center of the NuMI beam axis. The NOvA far detector is located in a new building that is 20.4 m wide by 113.8 m long, with the detector section sunk 16 m below the existing grade into granite rock (figure 3). Figure 4 shows the detector hall inside of the building. The NOvA near detector is located on the Fermilab site about 1 km from the NuMI target. It is similar to the far detector and is located 105 m underground in a new cavern adjacent to the MINOS access tunnel, downstream of the MINOS shaft (figure 5). The cavern is 6.0 m wide by 20.5 m long and 6.0 m high. Figure 6 shows the recently completed NOvA near-detector cavern. An earlier prototype of the near detector on surface (NDOS), shown in figure 7, exists as well at Fermilab.2 The 0.22 kton detector is located on the surface in the NOvA Near Detector Surface Building at Fermilab (figure 8). This prototype serves as a venue to test all the parts of the NOvA detectors together.3 The NDOS is about 107 mrad off-axis to the NuMI neutrino beam, and it has been operating since November 2010. NOvA uses liquid scintillator contained in rigid, highly reflective PVC (polyvinyl chloride) cells to detect neutrino interactions. The charged particles produced by the neutrino interaction inside the detector cause the liquid scintillator to produce light that is captured by optical fibers and carried to light-sensitive detectors. This article will focus on how the detectors are assembled using the PVC cells to construct planes and blocks used to make the detectors, and how the detectors are surveyed. Far detector Near detector NOvA PVC extrusions NOvA extrusion modules NOvA block B = h0v1h2v3h4v5h6v7h8v9h10v11h12v13h14v15… where v represents the vertical plane and h the horizontal plane. The number of planes is counted from 0 through 31. Similarly, the block configuration for the near detector is defined as: B = h0v1h2v3h4v5h6v7h8v9h10v11h12v13h14v15… where the number of planes is counted from 0 through 23. Far detector assembly FD → B0B1B2B3B4B5.........B22 B23B24 B25B26B27 where the number of blocks is counted from 0 through 27. The total number of planes or layers in the far detector is 896. Twelve extrusion modules are placed side by side on a flat Pivoter assembly table to form one plane of the far detector. The total number of modules per block is 384 and the total number of modules in the far detector is 10,752. The installation of the NOvA far detector is currently underway. The block assembly area houses the NOvA block pivoter machine as well as the pivoter assembly table that is used to build all the blocks (figure 12). Each of the 28 blocks in the detector is assembled on the assembly table while it is in its horizontal position. Each block is rotated to a vertical orientation after completion. Block assembly starts from plane-31 (v31) on the assembly table and ends with plane-0 (h0). The block assembly procedure starts at the glue machine (figure 13), where glue is applied to the extrusion modules. The modules are then transported to the assembly table by the vacuum-lifting fixture to be glued to the next modules to form planes or layers (figure 14). Alignment posts attached to the assembly table are used as guides for the module installation. The block is assembled in its horizontal position starting first with the downstream end. Once a block has been finished (figure 15), the block pivoter is used to move the block into place within the detector building, starting from the south wall (figure 16). The pivoter then pivots 90° to set the block upright to the ideal location. Near detector assembly ND → B0B1B2B3B4B5 B6B7 where the number of blocks is counted from 0 through 7. The blocks are currently being constructed at Fermilab CDF (figure 17). Block assembly starts from plane-23 (v23) on the assembly table and ends with plane-0 (h0). The total number of planes in the near detector is 192. Three extrusion modules get placed side by side on a flat assembly table to form one plane of the near detector. Total number of modules per block is 72. The block assembly procedure is similar to that of the far detector. The goal is to determine the relative positions of the modules within each detector block and the relative positions of the detector blocks with respect to each other. The relative position and orientation of the full detector with respect to the beamline also must be determined. Survey methodology Surface geodetic network Detector hall control network Tolerances of the NOvA module and block survey Overall tolerance: Near detector Far detector All the target holders are glued to the detector during module stacking in the assembly hall. For each location, the target is held in place by a 2.5 in. x 1.75 in. x 1.0 in. PVC block with a 0.625-in. threaded hole in the middle. These target mounting blocks are glued to the midpoint of a module (figure 22). Different types of targets can be screwed to the rectangular mounting block for use with the HDS6100 Laser Scanner paddle target, S6 Total Station prism reflector, or AT401 Tracker SMR retroreflector. Figure 23 shows the 6-in. black-and-white circular paddle target mounted on the module before scanning the layer. The length of the paddle target center to the base of the target base is 6.5 in. (16.5 cm). The length of the prism center to the base of the prism adapter is 18.3 cm (7.20 in.). A 1-in. diameter carbon-fiber extension rod is used to hold the target prism (figure 24) or SMR (figure 25). Three extension rod lengths are used: 25 cm (9.84 in.), 50 cm (19.68 in.) and 75 cm (29.53 in.). The eight targets marked by the heart-shaped symbols are fitted with laser scanner paddle targets in the assembly area so that the scanner can record their locations along with the module location during scanning in the horizontal position. These targets are later used for the block survey with the block in the upright position. Detector floor measurements Pivoter table measurements Far detector This graphic display is given to the construction supervisor to check if the plane meets the tolerance specified before construction of the next layer begins. Eight laser-scanner paddle targets with known coordinates on the walls of the assembly area are always part of each scan. Also, eight laser scanner targets are always fitted to the target mounting blocks on each block (Layer-2 and Layer-26) as part of each scanning procedure. Near detector The blocks are transported to and from the assembly hall to the detector hall in their upright positions one block after the other. Figure 35 shows Block12 of the far detector in the upright position Far detector A transformation is performed to transform all the horizontal position measurements [XHoriz , YHoriz , ZHoriz] into the upright position measurements [XUpright , YUpright , ZUpright] as follows: Where [XTrans , YTrans , ZTrans] is the vector containing the translation parameters in XYZ; R(εX,εY,εZ) is the rotation matrix; and S is the scale. For this survey the scale is fixed at S = 1.0. The measurements of the blocks in the detector hall require a survey crew to travel from Fermilab in Batavia, Illinois, to Ash River, Minnesota, for one week at a time. Two trips have been made so far, with more trips planned for the future as more blocks are installed. Near detector The NOvA far-detector construction is well underway. More than half of the modules have been installed (figure 39). First beam data is expected in the summer of 2013. NOvA will finish module assembly by early 2014. It is on schedule to be fully instrumented by August 2014. The NOvA far detector block survey is still in progress. As of May 2013, the following survey has been accomplished: Installation of the NOvA near detector started in July 2013 with the block assembly. The near detector cavern is now ready for the block installation. Acknowledgments References Dr. Babatunde O'Sheg Oshinowo is a Senior Geodesist at Fermilab where he is the survey and alignment task manager for the Fermilab Main Injector Accelerator, Recycler Antiproton Storage Ring, NOvA Detectors, NOvA ANU Recycler upgrades, Booster Accelerator, and other experiments. Dr. Oshinowo received a B.Sc. degree in Land Surveying Sciences from the University of East London, a M.Sc. degree in Geodesy from the University of Oxford, a M.S. degree in Geodetic Science from The Ohio State University, and a Ph.D degree in Surveying Engineering from the University of Washington. Horst Friedsam is AMG Department Head at the Fermi National Accelerator Laboratory.
CMSC Survey of the NOvA Detectors at Fermilab
The NOvA experiment
Figure 1: Beam trajectory 810 km from Fermilab to Ash River, Minnesota
Figure 2: Ash River is located 810 km from Fermilab
Figure 3: Far Detector Building at Ash River
Figure 4: Far Detector Hall in Ash River
Figure 5: NOvA near detector in a cavern adjacent to the MINOS access tunnel
Figure 6: Recently completed NOvA near-detector cavern
Figure 7: The NOvA near detector on the surface (NDOS)
Figure 8: Near detector surface building at FermilabNOvA detectors
The 14 kton far detector is 15.6 m wide, 15.6 m tall and 60 m long (figure 9). This detector is composed of 344,064 cells of extruded PVC plastic in a cellular structure. Each cell is 3.9 cm wide by 6.6 cm deep and is 15.6 m long. The cells are filled with a total of 3.0 million gallons of liquid scintillator. The liquid scintillator comprises 63.5 percent of the total detector mass, making this a totally active tracking calorimeter detector optimized for identification of electron neutrino (νe) interactions. The detector is read out via 10,668 km of 0.7 mm diameter optical wave-shifting fiber into 344,064 avalanche photodiodes (APD) with associated electronics (figure 10). The far detector consists of 28 blocks of 32 planes (16 vertical, 16 horizontal) for a total of 896 planes4.
Figure 9: NOvA detectors
The 0.3 kton near detector is 4.2 m wide, 4.2 m tall and 14.3 m long (Figure 9). This detector is an identical copy of the far detector except that the length of the extrusion modules is 4.2 m. The near detector consists of 8 blocks of 24 planes (12 vertical, 12 horizontal) for a total of 192 planes followed by a muon catcher unit that consists of 10 pieces of 4 in. steel plates sandwiched with PVC modules.NOvA detector assembly
The rigid PVC extrusions are the basic building blocks of the NOvA detectors. They are the structural elements containing the liquid scintillator and light measuring fiber optics1. The total mass of PVC used to make the far detector is 5.0 kton, which is 36.5 percent of the total far detector mass. The extrusions have a cellular structure, with 16 isolated cells per extrusion. The cells have exterior PVC walls 4.8 mm thick with 3.3 mm thick interior webs between cells. The extrusion thickness is 6.6 cm. The length of the rigid PVC extrusions with 16 cells extruded together in a unit is 63.5 cm wide (figure 10). All far detector extrusions are 15.544 m long. Extrusions for the near detector are 3.988 m long. The width of a module is 1.27 m.
Figure 10: NOvA PVC extrusions
The basic cell of the NOvA detector is a column or row of liquid scintillator encased in a PVC container. NOvA modules consist of two 16-cell PVC extrusions glued together to make a 32-cell extrusion assembly. Two 16 cell objects are attached with glue at the long sides, and the extrusion module is cut to an exact length. The module assembly shown in figure 11 defines a model plane. The extrusion modules are capped at one end by a simple PVC end cap to contain the liquid scintillator and are capped at the other end by a more complicated fiber manifold. The manifold routes the 64 fiber ends to 32 avalanche photodiode pixels. The assembled extrusion modules with fiber manifolds and end caps are 15.6 m long for the far detector. The end caps and fiber manifolds encase the entire 32 cells into a common sealed liquid volume. Thus the 1.27 m × 15.6 m extrusion module forms the primary containment vessel for the liquid scintillator. The length of the extrusion module for the near detector is 4.2 m. The assembly provides a leak-tight container for liquid scintillator.1
The NOvA detector is constructed from alternating planes (i.e., layers) of vertical and horizontal PVC extrusion modules, connected together by glue between planes. The basic structural unit of the NOvA detector is an assembly of 32 planes of PVC extrusion modules, called a block, for the far detector. The near detector consists of 24 planes. The modules in each block are glued together in a horizontal orientation on a block assembly table. The block configuration for the far detector is defined as:
...h16v17h18v19h20v21h22v23h24v25h26v27h28v29 h30 v31
...h16v17h18v19h20v21h22v23
Figure 11: PVC extrusion modules (two 16 cells). L = 15.6 m for far detector; L = 4.2 m for near detector.
The far detector (FD) consists of 28 blocks. Each block is 15.6 m wide by 15.6 m high by 2.141 m thick. The PVC in a 32-plane block weighs 177.7 metric tons and when filled with liquid scintillator, the weight of a 32-plane block is 487.5 metric tons4. The far detector block configuration starting from the upstream end is:
Figure 12: Pivoter assembly table in vertical (left) and horizontal (right) positions
Figure 13: The glue machine
Figure 14: NOvA far-detector assembly
Figure 15: Completed block on pivoter assembly table
Figure 16: Block transported to location by pivoter
The near detector consists of eight blocks of 24 planes and a 22-plane muon catcher. Each block is 4.2 m wide × 4.2 m high × 2.136 m thick. The PVC in a 24-plane block weighs 5 metric tons and when filled with liquid scintillator, the weight of a 24-plane block is 13.75 metric tons. The block configuration starting from the upstream end is:
Figure 17: The NOvA near detector assembly area at Fermilab CDFSurvey of the NOvA detectors
The survey instrumentation that is being used for the detector survey is as follows:
• An API T3 Laser Tracker and Spatial Analyzer software are used to establish control points in the entire NOvA far and near detector halls. It is used for the module plane measurements of the near detector and flatness measurements of the detector floors and block assembly tables. It is also used for the block survey of the near detector.
• A Leica AT401 Absolute Tracker and Spatial Analyzer software are used to establish additional control points in the NOvA far detector hall. It is also used for the block survey of the far detector.
• A Trimble S6 Total Station is used for the block survey of the far detector. It is also used for the flatness measurements of the far detector block assembly table.
• A Leica HDS6100 Laser Scanner system and its associated software are being used for mapping every far detector plane.
• A new Hexagon Romer Arm will be used for the near detector plane measurements.
• A Geodimeter Total Station is used for connecting surface control network to the detector hall network.
• A Leica DNA03 Digital Level is used for the vertical control network measurements.
• Trimble GPS receivers are used to establish a common control network at Ash River and Fermilab.
There already exists a well-established surface geodetic network on the Fermilab site. A GPS surface geodetic control network is established that connects points at Fermilab to Ash River. The network is tied to the National Geodetic Survey's Continuously Operating Reference Stations (CORS) precision geodetic network (figure 18). The network is based on NAD83 (North American datum 1983) for the horizontal datum and NAVD88 (North American vertical datum 1983) for the vertical datum.
A precision control network is established to bring horizontal and vertical controls into the near and far detector halls for positioning the detectors using the API Laser Tracker. This control network is extended to the higher levels of the west wall of the far detector hall using the Leica AT401 Tracker. The detector hall control network is tied to the surface network using the Geodimeter Total Station (figure 19).
Figure 18: GPS Surface Geodetic Network
The NOvA module and block survey tolerances are specified as follows:5
• Relative 2 mm (horizontal) edge to edge
• Relative 0.75 mm or better (vertical) between adjacent module pieces
• Angular tolerance of ±2 mm/15.6 m = ±0.13 mrad
Figure 19: Far Detector Hall control networkBlock fiducialization
Several survey fiducials are mounted at suitable locations on the outside edges of each block for the near detector. Magnet rings are used as fiducials (figure 20). The magnet rings are glued to the points of interest with a 24-hour epoxy. The location of each fiducial is defined by the center of the SMR (spherically mounted retroreflector) as it precisely sits on the fiducial. A total of 24 fiducials are used for each block, nine on the module extrusions, and six on the blue-steel base plate. There are nine fiducials on the extrusions at about the first, middle, and last planes, and three on the base plate on the east side of the block. The same number of fiducials is used on the west side.
Figure 20: Fiducials on near detector module
There are 44 targets used for Block0 and Block1.6 There are two targets each on the top horizontal surfaces of Layer-2 and Layer-28. There are 10 targets each on the west and east vertical faces of Layer-1 and Layer-29. Figure 21 shows the upstream view of the target locations. The locations marked by red-colored heart symbols are fiducials for the block survey, and those marked with red or blue-colored stars are for the block shape monitoring. The targets on the top east corners, marked by blue stars on Layer-1 and Layer-29, are fitted permanently with prism reflectors in the assembly hall before blocks are rotated to the upright position. There are only eight targets mounted on Block2 through Block27, their locations marked by red heart symbols on Layer-2 and Layer-26 in figure 23.
Figure 21: Upstream view of targets on far-detector module for Block0 and Block1 in upright position. Click here for larger image.
Figure 22: Target mounting block on far detector module
Figure 23: Paddle targets on mounting block
Figure 24: Prism and extension rod on mounting block
Figure 25: SMR and extension rod on mounting block
Floor measurements are made with the Laser Tracker in the far and near detector halls to determine the flatness. These measurement results are used for shimming the pallets that the blocks sit on. Figure 26 shows the results of using a 720 grid point of 24 in. x 24 in. (61 cm x 61 cm) in the far detector hall. The maximum surface elevation difference is 1.1 in. (2.8 cm).
Figure 26: Far detector floor flatness. Click here for larger image.
Before each block is assembled, the block assembly table must be leveled. To determine the flatness, the surface of the near detector assembly table is measured with the API Tracker. The surface of the far-detector pivoter table is first measured with the S6 Total Station using 24 in. x 24 in. (61 cm x 61 cm) grids with the table in the vertical position (figure 27). The API Tracker is then used to measure the surface with the table in the horizontal position (figure 28). These measurement results are used for shimming the assembly table surface before the first plane is assembled. Figures 29 and 30 show the results for the far-detector assembly table surface. The surface elevation differences range from –0.8 in. (–2.0 cm) to 1.2 in. (3.0 cm).
Figure 27: Table surface measurements in vertical position
Figure 28: Table surface measurements in horizontal position
Figure 29: Far detector pivoter table flatness–API Laser Tracker. Click here for larger image.
Figure 30: Far detector pivoter table flatness—Trimble S6 Total Station. Click here for larger image.Block measurements in horizontal position
To measure the surface flatness, the upstream surface of each of the 32 planes of the NOvA far detector block is scanned with the HDS6100 Laser Scanner, which is located on the ceiling and inverted directly above the pivoter assembly table in the assembly area (figures 31 and 32). Prior to scanning the planes, the surface of the pivoter assembly table is also scanned. After each plane (or layer) is scanned, a program developed at Fermilab7 is used to reduce the large scanner data and display the surface shapes of the modules along their borders (figure 33).
Figure 31: Laser scanner on the detector hall ceiling above the block
Figure 32: Block directly below laser scanner
Figure 33: Result of the surface analysis at a module boundary for first layer of Block0.8
After all 32 planes have been assembled, several API Tracker measurements will be made to the 24 block fiducials while the block is in the horizontal position in the assembly area. To determine the module positions, groove measurements will be made at specified locations by placing the SMR where the horizontal module grooves intersect the extreme end of the vertical plane (figure 20). The SMR will be placed on every third groove from the left (–13) and right (+13) edges of the module, and on the middle groove (0) as shown in figure 34. Measurement will also be made on top of the block along the groove of the vertical modules.
Figure 34: Groove locations for near detector blocksBlock measurements in upright position
With the blocks in the upright position, measurements are made to targets on the east and west sides of the blocks (figures 36 and 37) and on the top of the blocks (figure 38). As of May 2013, the following measurements have been completed for Block0 through Block12:
• The top of Block0 through Block12 is measured with the 75-cm rods and SMR retroreflectors using the AT401Tracker on a wall bracket mounted up near the ceiling of the detector hall. In the process, one of the MAP platforms is used as a work platform. The same measurements are repeated for Block0 and Block1 using 50-cm rods. The two targets with different lengths are used as vector targets at each point. This will allow any desired location to be computed along the vector.
• Measurements on the east (beam-right) sides of Block0 through Block12 are made with prism reflectors without any rod extensions, using the S6 Total Station.
• Measurements on the west (beam-left) side of Block0 through Block12 are made with prism reflectors and the 25-cm rods using the S6 Total Station. The same measurements are repeated for Block0 and Block1 with the prisms, without extension rods.
Blocks in the upright positions will be measured with the API Tracker inside the cavern. Similar transformation is performed as above to transform all the horizontal position measurements into the upright position measurements.
Figure 35: Block12 in the upright position
Figure 36: Block measurements on the lower (left) and upper (right) east side of the far detector
Figure 37: Block measurements on the west side of the far detector
Figure 38: Block measurements on the top of the detectorStatus of the NOvA detectors
• Block0 and Block1 surveyed with and without liquid scintillator
• Block2 through Block6 surveyed only with filled liquid scintillator
• Block7 through Block12 surveyed without liquid scintillator
Figure 39: Status of the NOvA far detector. Updated weekly here.
I would like to thank the Alignment and Metrology Department members who participated in the NOvA setector survey, especially Gary Crutcher and Chuck Wilson, who many times made the survey trip with me from Illinois to Minnesota. I would also like to thank Pat Lukens and Ting Miao of the NOvA Collaboration for all the information and assistance.
1. NOvA Technical Design Report. Fermilab, Oct. 8, 2007.
2. B. O. Oshinowo. "Survey of the NOvA Near Detector at Fermilab." 11th International Workshop on Accelerator Alignment, Sept. 13–17, 2010. Hamburg, Germany.
3. G. Feldman. "Physics of the NOvA Experiment." Fermilab NOVA Document 7733-v1, July 2012.
4. P. Lukens. "Nova-FD-Parameters." Fermilab NOvA Document 2066-v2, January 27, 2012.
5. P. Lukens. Fermilab NOvA alignment meeting. March 11, 2010.
6. T. Miao. Fermilab NOvA alignment meeting. Aug. 9, 2012.
7. B. Mercurio. "Quality Control of the NOvA Far Detector." 12th International Workshops on Accelerator Alignment, Sept. 10-14, 2012, Fermilab.
8. H. Friedsam. "Laser Scan Analysis of the NOvA Far Detector Layer Surfaces." 12th International Workshops on Accelerator Alignment, Sept. 10-14, 2012. Fermilab.
About The Author
Babatunde O’Sheg Oshinowo and Horst Friedsam
© 2023 Quality Digest. Copyright on content held by Quality Digest or by individual authors. Contact Quality Digest for reprint information.
“Quality Digest" is a trademark owned by Quality Circle Institute, Inc.
