- Email: [email protected]
Construction and assembly of the neutron radiography and tomography facility ANTARES at FRM II
ARTICLE IN PRESS
Nuclear Instruments and Methods in Physics Research A 542 (2005) 38–44 www.elsevier.com/locate/nima
Construction and assembly of the neutron radiography and tomography facility ANTARES at FRM II Elbio Calzadaa,b,, Burkhard Schillingera,b, Florian Gru¨nauera a
Physik Department E21, Technische Universita¨t Muenchen, 85747 Garching, Germany b ZWE FRM II, TU Muenchen, Lichtenbergstr. 1, 85747 Garching, Germany Available online 30 January 2005
Abstract In 2003, the components for the neutron radiography and tomography facility ANTARES were constructed and assembled at the beam line SR4B in the northeast corner of the experimental hall at FRM II (J. Radiat. Isotopes (2002), accepted for publication). ANTARES consists of an external beam shutter, flight tube, radiation shielding and measurement blockhouse. The shielding elements are constructed as steel casings filled with heavy concrete. Many harsh constrains in space, accessibility and other conditions (such as a very limited area accessible by the hall crane and restricted space due to surrounding instruments) made it necessary to develop special technical solutions for every step of the construction of the facility. From filling up shielding elements with heavy concrete in winter conditions (temperature below 0 1C) to the development of an hydraulics driven external shutter and collimator positioning system, new convex-shaped nearly one meter diameter vacuum flanges and windows, the development of a movable 30-ton wall on rails as blockhouse door, upto the assembly and exact positioning of heavy components (upto 50 tons) using air cushions, ball casters and a special small truck-mounted telescopic crane. r 2005 Elsevier B.V. All rights reserved. Keywords: Neutron radioscopy; Dynamic neutron radiography; Image quality; Modulation transfer function
1. Introduction In the early planing stage of the reactor, the tomography facility was foreseen at a neutron guide leading out of the reactor hall to an external building. Corresponding author. ZWE FRM II, TU Muenchen, Lichtenbergstr. 1, 85747 Garching, Germany. Tel.: +49 0 89 289 14611. E-mail address: [email protected] (E. Calzada).
When it became clear that the beam geometry of a neutron guide was unsuited [1] for tomography, a classical flight tube design hat to be fitted into the remaining space in the reactor hall. The tomography facility at FRM II [2] shares the beam port (with two channels) with the UCN source and is surrounded by a spin-echo spectrometer and the platforms of the positron source at the inclined beam tube above its beams tube. Fig. 1 shows an overview of the facility, Fig. 2 shows a 3D view.
0168-9002/$ - see front matter r 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.nima.2005.01.009
ARTICLE IN PRESS E. Calzada et al. / Nuclear Instruments and Methods in Physics Research A 542 (2005) 38–44
39
Fig. 1. Complete layout of the facility: (1) Collimator inside the biological shielding, (2) external vertical beam shutter, (3) iris diaphragm, (4) pneumatic shutter, (5) flight tube, (6) variable beam size limiter, (7) measurement cabin, (8) wall and beam stop, (9) detector and camera system, (10) sample manipulator, (11) sliding door, (12) flight tube shielding.
Fig. 2. Tomography facility surrounded by other experiments. Fig. 3. Filled up of the shielding elements.
2. Construction of the wall elements The shielding elements and blockhouse walls were constructed as steel casings filled with heavy concrete [3]. With temperatures under 0 1C in January, to reach the maximum strength in concrete it was essential to have a closed hall conditioned with a heating system (Fig. 3). The complete shielding [3] first was mounted at the workshop without concrete, then for a second time to be filled up with concrete and to control the tolerances kipping the shape
during the filling and finally for a third time at FRM II (Fig. 4). The secondary shutter housing and the blockhouse elements were filled with concrete of density 4.5 tons/m3. As the collimated beam inside the flight tube does not touch the walls [3], a density of 3.5 tons/m3 was sufficient for the walls of the flight tube housing. Owing to the fact that the crane in the experimental hall has a nominal capacity of 10 tons only, all elements were designed for a maximum weight of about 9.5 tons. For some elements of the blockhouse wall near the beam
ARTICLE IN PRESS 40
E. Calzada et al. / Nuclear Instruments and Methods in Physics Research A 542 (2005) 38–44
per year), which was capable of handling all elements.
3. Mounting of the wall elements
Fig. 4. Antares shielding mounted and ready to be filled up with concrete.
The blockhouse walls were mounted from five horizontal segments each, weighing more than 50 tons each. The assembled walls were fitted with mounting frames for upto six air cushions (Fig. 6). An external compressor was used to lift the walls and to move them into place, as the capacity of the house installations was not sufficient. Unfortunately, several cable channels in the hall floor were only roughly covered with uneven steel
Fig. 6. Assembly of the blockhouse walls. Fig. 5. Separately mounted concrete cylinder inside a wall to reduce weight for the crane.
entrance, this weight constraint could not be met; so additional removable concrete filled cylinders were constructed inside the wall volume (Fig. 5), which were mounted after the assembly of the wall elements. Additional problems arose because the concrete company guaranteed only for the minimum density of the concrete, but not for the nominal density. In reality, the density became higher than the calculated values, leading to a component weight of more than ten tons. Fortunately, the reactor hall crane has a special 10% overload range (of which the use is restricted to a few hours
Fig. 7. Moving the wall with air cushions.
ARTICLE IN PRESS E. Calzada et al. / Nuclear Instruments and Methods in Physics Research A 542 (2005) 38–44
plates, rendering an uneven surface and an escape path for the compressed air. They had to be covered with thin steel plates, but still, the walls could only transit the uneven regions with the aid of two steel cable tackles affixed to the reactor hall walls (Fig. 7).
41
predicted a material failure at the edge of the milled area, so the cover was constructed as a sandwich of a 3 mm aluminum plate, rubber seals and flange rings. The cover stones for the flight tube housing have retractable ball casters on which they can slide from the middle, crane-accessible position to their final position under the platforms.
4. The flight tube The flight tube consists of individual segments connected by rubber sealant rings, mounted on C-shaped brackets. The original design was taken from the vacuum housings for the neutron guides leading to the guide hall. No space was available to mount circular flanges, so the original square shape with rounded corners was altered by rounding the straight sections to a convex shape (Fig. 8). Each segment has an individual vacuum connector, so single segments can be removed for experimental installations, like a velocity selector, the remaining segments can be closed off by covers. Vacuum pump and vacuum control system were salvaged from the old reactor. For the final neutron window at the end of the flight tube, a massive aluminum cover with a thinmilled window was foreseen. Static calculations
Fig. 8. The flight tube with convex flanges.
5. Mounting of the blockhouse roof The area of the blockhouse is not accessible by the hall crane. To be able to mount the roof at all, it was constructed from small 2-ton bars. The original intention had been to use a balance bar and counterweight with the hall crane, which is able to telescope towards the corner of the hall just to the beginning of the blockhouse. Later on, a special compact truck-mounted crane was found which was just able to drive into the reactor hall in the time when the other experiments at SR-2 and 3 were not yet assembled, with 5 cm space left next to the already mounted measuring cabins. The roof elements were deposited on the platform for the positron experiment with the hall crane, the truck-mounted crane took over from there and deposited them in position on the blockhouse roof (Fig. 9).
Fig. 9. Mounting the blockhouse roof elements with a truck mounted crane.
ARTICLE IN PRESS 42
E. Calzada et al. / Nuclear Instruments and Methods in Physics Research A 542 (2005) 38–44
6. The external beam shutter The external beam shutter was constructed as a vertical concrete block containing two different collimators (Fig. 11) with collimation ratio L/ D ¼ 400 [1] und L/D ¼ 800 [1], driven by a hydraulic cylinder (Fig. 10). A servo system with position encoder and a control valve can position the shutter block to the accuracy of a tenth of millimeter to the collimator positions or to the closed position. When the position is reached, two hydraulic clamps hold the block in position and the hydraulic pump shuts off. A hydraulic pressure tank holds sufficient pressure to open the hydraulic clamps and close the shutter within one third of second in case of power failure or when the blockhouse door is opened (Fig. 11).
7. Sliding door As there is no space for a conventional door to swing open, a whole 30 ton wall section moves on rails parallel to the UCN experiment. The first step was to adjust accurately the blockhouse wall with the beam entrance hole, which defined the direction and position of the sliding door. Second and the most complicated was to get the permission to cut the floor (Fig. 12) in order to fix the precision rail guides in place.
Fig. 11. The picture shows the complete external shutter with the two collimators and the hydraulic clamps.
Fig. 12. Mounting and adjustment of the precision rail guides for the sliding door.
Fig. 10. The external beam shutter with hydraulic cylinder and clamps.
When the concrete hat hardened enough it was possible to bring the 30 to door with air cushions to the rail guides (Fig. 13).
ARTICLE IN PRESS E. Calzada et al. / Nuclear Instruments and Methods in Physics Research A 542 (2005) 38–44
43
Fig. 13. View of the sliding door already in end position.
Fig. 15. Injection pump, first radiography at Antares.
TARES was opened for the first time. With 50 kW, the reactor was running only at 1/400 of its nominal power (Fig. 15).
10. Outlook
Fig. 14. View inside of the blockhouse.
In the future, we want to add three components to make wider the possibilities of Antares: a filter wheel for phase contrast radiography will be mounted, a velocity selector and an X-ray tube.
8. Accssesories Acknowledgements In 2004, the walls were covered with B4C rubber mats to reduce the activation of the steel walls, the sample manipulator was transported to the blockhouse and the detector box was completed (Fig. 14). Also, the fast experimental pneumatic shutter and the beam limiter were mounted.
Thanks to the operation stuff from FRM II, specially, Mr. Stiegel, Mr. Weber, and Mr. C. Herzog, OSD people and Mr. D. Trentsch and Mr. F. Albert from Kinkele. References
9. First measurement On March 10, 2004, the shutter of the neutron radiography and tomography beamline AN-
[1] B. Schillinger, Estimation and measurement of L/D on a cold and thermal neutron guide, World Conference Neutron Radiography, Osaka 1999, in: Nondestructive Testing Evaluation, Vol. 16, pp. 141–150.
ARTICLE IN PRESS 44
E. Calzada et al. / Nuclear Instruments and Methods in Physics Research A 542 (2005) 38–44
[2] B. Schillinger, E. Calzada, F. Gru¨nauer, E. Steichele, The design of the neutron radiography and tomography facility at the new research reactor FRM-II at Technical University Munich, 4th International Topical Meeting on Neutron Radiography, Pennsylvania 2001, J. Radiat. Isotopes (2002). [3] F. Gru¨nauer, B. Schillinger, Optimization of the beam geometry and radiation shieldings for the neutron tomography facility at the new neutron source in Munich, 4th
International Topical Meeting on Neutron Radiography, Pennsylvania 2001, J. Radiat. Isotopes (2002), accepted for publication.
Further reading [4] www.ph.tum.de\antares
Nuclear Instruments and Methods in Physics Research A 542 (2005) 38–44 www.elsevier.com/locate/nima
Construction and assembly of the neutron radiography and tomography facility ANTARES at FRM II Elbio Calzadaa,b,, Burkhard Schillingera,b, Florian Gru¨nauera a
Physik Department E21, Technische Universita¨t Muenchen, 85747 Garching, Germany b ZWE FRM II, TU Muenchen, Lichtenbergstr. 1, 85747 Garching, Germany Available online 30 January 2005
Abstract In 2003, the components for the neutron radiography and tomography facility ANTARES were constructed and assembled at the beam line SR4B in the northeast corner of the experimental hall at FRM II (J. Radiat. Isotopes (2002), accepted for publication). ANTARES consists of an external beam shutter, flight tube, radiation shielding and measurement blockhouse. The shielding elements are constructed as steel casings filled with heavy concrete. Many harsh constrains in space, accessibility and other conditions (such as a very limited area accessible by the hall crane and restricted space due to surrounding instruments) made it necessary to develop special technical solutions for every step of the construction of the facility. From filling up shielding elements with heavy concrete in winter conditions (temperature below 0 1C) to the development of an hydraulics driven external shutter and collimator positioning system, new convex-shaped nearly one meter diameter vacuum flanges and windows, the development of a movable 30-ton wall on rails as blockhouse door, upto the assembly and exact positioning of heavy components (upto 50 tons) using air cushions, ball casters and a special small truck-mounted telescopic crane. r 2005 Elsevier B.V. All rights reserved. Keywords: Neutron radioscopy; Dynamic neutron radiography; Image quality; Modulation transfer function
1. Introduction In the early planing stage of the reactor, the tomography facility was foreseen at a neutron guide leading out of the reactor hall to an external building. Corresponding author. ZWE FRM II, TU Muenchen, Lichtenbergstr. 1, 85747 Garching, Germany. Tel.: +49 0 89 289 14611. E-mail address: [email protected] (E. Calzada).
When it became clear that the beam geometry of a neutron guide was unsuited [1] for tomography, a classical flight tube design hat to be fitted into the remaining space in the reactor hall. The tomography facility at FRM II [2] shares the beam port (with two channels) with the UCN source and is surrounded by a spin-echo spectrometer and the platforms of the positron source at the inclined beam tube above its beams tube. Fig. 1 shows an overview of the facility, Fig. 2 shows a 3D view.
0168-9002/$ - see front matter r 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.nima.2005.01.009
ARTICLE IN PRESS E. Calzada et al. / Nuclear Instruments and Methods in Physics Research A 542 (2005) 38–44
39
Fig. 1. Complete layout of the facility: (1) Collimator inside the biological shielding, (2) external vertical beam shutter, (3) iris diaphragm, (4) pneumatic shutter, (5) flight tube, (6) variable beam size limiter, (7) measurement cabin, (8) wall and beam stop, (9) detector and camera system, (10) sample manipulator, (11) sliding door, (12) flight tube shielding.
Fig. 2. Tomography facility surrounded by other experiments. Fig. 3. Filled up of the shielding elements.
2. Construction of the wall elements The shielding elements and blockhouse walls were constructed as steel casings filled with heavy concrete [3]. With temperatures under 0 1C in January, to reach the maximum strength in concrete it was essential to have a closed hall conditioned with a heating system (Fig. 3). The complete shielding [3] first was mounted at the workshop without concrete, then for a second time to be filled up with concrete and to control the tolerances kipping the shape
during the filling and finally for a third time at FRM II (Fig. 4). The secondary shutter housing and the blockhouse elements were filled with concrete of density 4.5 tons/m3. As the collimated beam inside the flight tube does not touch the walls [3], a density of 3.5 tons/m3 was sufficient for the walls of the flight tube housing. Owing to the fact that the crane in the experimental hall has a nominal capacity of 10 tons only, all elements were designed for a maximum weight of about 9.5 tons. For some elements of the blockhouse wall near the beam
ARTICLE IN PRESS 40
E. Calzada et al. / Nuclear Instruments and Methods in Physics Research A 542 (2005) 38–44
per year), which was capable of handling all elements.
3. Mounting of the wall elements
Fig. 4. Antares shielding mounted and ready to be filled up with concrete.
The blockhouse walls were mounted from five horizontal segments each, weighing more than 50 tons each. The assembled walls were fitted with mounting frames for upto six air cushions (Fig. 6). An external compressor was used to lift the walls and to move them into place, as the capacity of the house installations was not sufficient. Unfortunately, several cable channels in the hall floor were only roughly covered with uneven steel
Fig. 6. Assembly of the blockhouse walls. Fig. 5. Separately mounted concrete cylinder inside a wall to reduce weight for the crane.
entrance, this weight constraint could not be met; so additional removable concrete filled cylinders were constructed inside the wall volume (Fig. 5), which were mounted after the assembly of the wall elements. Additional problems arose because the concrete company guaranteed only for the minimum density of the concrete, but not for the nominal density. In reality, the density became higher than the calculated values, leading to a component weight of more than ten tons. Fortunately, the reactor hall crane has a special 10% overload range (of which the use is restricted to a few hours
Fig. 7. Moving the wall with air cushions.
ARTICLE IN PRESS E. Calzada et al. / Nuclear Instruments and Methods in Physics Research A 542 (2005) 38–44
plates, rendering an uneven surface and an escape path for the compressed air. They had to be covered with thin steel plates, but still, the walls could only transit the uneven regions with the aid of two steel cable tackles affixed to the reactor hall walls (Fig. 7).
41
predicted a material failure at the edge of the milled area, so the cover was constructed as a sandwich of a 3 mm aluminum plate, rubber seals and flange rings. The cover stones for the flight tube housing have retractable ball casters on which they can slide from the middle, crane-accessible position to their final position under the platforms.
4. The flight tube The flight tube consists of individual segments connected by rubber sealant rings, mounted on C-shaped brackets. The original design was taken from the vacuum housings for the neutron guides leading to the guide hall. No space was available to mount circular flanges, so the original square shape with rounded corners was altered by rounding the straight sections to a convex shape (Fig. 8). Each segment has an individual vacuum connector, so single segments can be removed for experimental installations, like a velocity selector, the remaining segments can be closed off by covers. Vacuum pump and vacuum control system were salvaged from the old reactor. For the final neutron window at the end of the flight tube, a massive aluminum cover with a thinmilled window was foreseen. Static calculations
Fig. 8. The flight tube with convex flanges.
5. Mounting of the blockhouse roof The area of the blockhouse is not accessible by the hall crane. To be able to mount the roof at all, it was constructed from small 2-ton bars. The original intention had been to use a balance bar and counterweight with the hall crane, which is able to telescope towards the corner of the hall just to the beginning of the blockhouse. Later on, a special compact truck-mounted crane was found which was just able to drive into the reactor hall in the time when the other experiments at SR-2 and 3 were not yet assembled, with 5 cm space left next to the already mounted measuring cabins. The roof elements were deposited on the platform for the positron experiment with the hall crane, the truck-mounted crane took over from there and deposited them in position on the blockhouse roof (Fig. 9).
Fig. 9. Mounting the blockhouse roof elements with a truck mounted crane.
ARTICLE IN PRESS 42
E. Calzada et al. / Nuclear Instruments and Methods in Physics Research A 542 (2005) 38–44
6. The external beam shutter The external beam shutter was constructed as a vertical concrete block containing two different collimators (Fig. 11) with collimation ratio L/ D ¼ 400 [1] und L/D ¼ 800 [1], driven by a hydraulic cylinder (Fig. 10). A servo system with position encoder and a control valve can position the shutter block to the accuracy of a tenth of millimeter to the collimator positions or to the closed position. When the position is reached, two hydraulic clamps hold the block in position and the hydraulic pump shuts off. A hydraulic pressure tank holds sufficient pressure to open the hydraulic clamps and close the shutter within one third of second in case of power failure or when the blockhouse door is opened (Fig. 11).
7. Sliding door As there is no space for a conventional door to swing open, a whole 30 ton wall section moves on rails parallel to the UCN experiment. The first step was to adjust accurately the blockhouse wall with the beam entrance hole, which defined the direction and position of the sliding door. Second and the most complicated was to get the permission to cut the floor (Fig. 12) in order to fix the precision rail guides in place.
Fig. 11. The picture shows the complete external shutter with the two collimators and the hydraulic clamps.
Fig. 12. Mounting and adjustment of the precision rail guides for the sliding door.
Fig. 10. The external beam shutter with hydraulic cylinder and clamps.
When the concrete hat hardened enough it was possible to bring the 30 to door with air cushions to the rail guides (Fig. 13).
ARTICLE IN PRESS E. Calzada et al. / Nuclear Instruments and Methods in Physics Research A 542 (2005) 38–44
43
Fig. 13. View of the sliding door already in end position.
Fig. 15. Injection pump, first radiography at Antares.
TARES was opened for the first time. With 50 kW, the reactor was running only at 1/400 of its nominal power (Fig. 15).
10. Outlook
Fig. 14. View inside of the blockhouse.
In the future, we want to add three components to make wider the possibilities of Antares: a filter wheel for phase contrast radiography will be mounted, a velocity selector and an X-ray tube.
8. Accssesories Acknowledgements In 2004, the walls were covered with B4C rubber mats to reduce the activation of the steel walls, the sample manipulator was transported to the blockhouse and the detector box was completed (Fig. 14). Also, the fast experimental pneumatic shutter and the beam limiter were mounted.
Thanks to the operation stuff from FRM II, specially, Mr. Stiegel, Mr. Weber, and Mr. C. Herzog, OSD people and Mr. D. Trentsch and Mr. F. Albert from Kinkele. References
9. First measurement On March 10, 2004, the shutter of the neutron radiography and tomography beamline AN-
[1] B. Schillinger, Estimation and measurement of L/D on a cold and thermal neutron guide, World Conference Neutron Radiography, Osaka 1999, in: Nondestructive Testing Evaluation, Vol. 16, pp. 141–150.
ARTICLE IN PRESS 44
E. Calzada et al. / Nuclear Instruments and Methods in Physics Research A 542 (2005) 38–44
[2] B. Schillinger, E. Calzada, F. Gru¨nauer, E. Steichele, The design of the neutron radiography and tomography facility at the new research reactor FRM-II at Technical University Munich, 4th International Topical Meeting on Neutron Radiography, Pennsylvania 2001, J. Radiat. Isotopes (2002). [3] F. Gru¨nauer, B. Schillinger, Optimization of the beam geometry and radiation shieldings for the neutron tomography facility at the new neutron source in Munich, 4th
International Topical Meeting on Neutron Radiography, Pennsylvania 2001, J. Radiat. Isotopes (2002), accepted for publication.
Further reading [4] www.ph.tum.de\antares