Fusion Engineering and Design 89 (2014) 2398–2403

Authors' copy - Article presented at the 11th International Symposium on Fusion Nuclear Technology (ISFNT-11), Sept. 2013, Barcelona, Spain Contents lists available at ScienceDirect

Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes

Progress in the design and R&D of the ITER In-Vessel Viewing and Metrology System (IVVS) Gregory Dubus a,∗ , Adrian Puiu a , Philip Bates a , Carlo Damiani a , Roger Reichle b , Jim Palmer b a b

Fusion for Energy, c/ Josep Pla, n◦ 2 – Torres Diagonal Litoral – Edificio B3, 08019 Barcelona, Spain ITER Organization, Route de Vinon sur Verdon, 13115 Saint Paul Lez Durance, France

a r t i c l e

i n f o

Article history: Received 6 September 2013 Received in revised form 21 December 2013 Accepted 8 January 2014 Available online 21 February 2014 Keywords: ITER Remote handling In-Vessel Viewing System Inspection Metrology Erosion monitoring

a b s t r a c t The In-Vessel Viewing and Metrology System (IVVS) is a fundamental tool for the ITER machine operations, aiming at performing inspections as well as providing information related to the erosion of in-vessel components, which in turn is related to the amount of mobilised dust present in the Vacuum Vessel. Periodically or on request, the IVVS scanning probes will be deployed into the Vacuum Vessel in order to acquire both visual and metrological data on plasma facing components (blanket, divertor, heating/diagnostic plugs, and test blanket modules). Recent design changes made to the six IVVS port extensions implied the need for a substantial redesign of the IVVS integrated concept – including the scanning probe and its deployment system – in order to bring it to the level of maturity suitable for the Conceptual Design Review. This paper gives an overview of the concept design for IVVS as well as of the various engineering analyses and R&D activities carried out in support to this design: neutronic, seismic and electromagnetic analyses, probe actuation validation under environmental conditions.

1. Introduction

2. Concept design for IVVS

Six In-Vessel Viewing and Metrology System (IVVS) units will be installed in ITER at lower port positions. Each IVVS unit will comprise a viewing and metrology probe capable of performing visual inspections as well as providing information related to the erosion of the in-vessel components. While the R&D supporting the concept of the scanning probe has been extensively documented over the past 15 years [1–3], the IVVS deployment system has remained relatively underdeveloped until a first activity laid the foundations of an integrated IVVS concept in late 2011 [4]. More recently a major piece of work by Oxford Technologies Ltd (OTL) has been achieved to bring this integrated design to a level of maturity suitable for the Conceptual Design Review. The outline of this paper is as follows. The concept design for IVVS is detailed in Section 2. Then, Section 3 gives an overview of the various engineering analyses and R&D activities supporting the design integration: neutronic, seismic and electromagnetic analyses, probe actuation validation under environmental conditions.

2.1. The IVVS port

∗ Corresponding author. Tel.: +34 93 320 11 33. E-mail address: [email protected] (G. Dubus).

Authors' copy © 2014 Authors. All rights reserved. http://dx.doi.org/10.1016/j.fusengdes.2014.01.012

The geometrical constraints on the IVVS units are mainly related to the shape and dimensions of the ports where they are mounted, still within the ITER primary vacuum (see Fig. 1). The IVVS ports extend beyond the bio-shield, in order to allow maintenance without the bio-shield removal. Each IVVS port is sealed by a circular closure plate mating directly with the vacuum vessel port extension. The IVVS concept also includes a vacuum feed-through to be located on the section of the port extension protruding inside the port-cell.

2.2. The IVVS cartridge The IVVS concept design contains components that will be permanently resident in the ITER ports (the so-called “passive assembly”) as well as a removable cartridge (the “active assembly”) that can be extracted from the port to allow for maintenance and/or replacement of the system. The IVVS cartridge (Fig. 2) is installed and removed by using a dedicated transfer cask including a specific handling unit.

G. Dubus et al. / Fusion Engineering and Design 89 (2014) 2398–2403

Fig. 1. IVVS port; vertical section. 1. Divertor 2. Blanket module #18 3. Vacuum vessel stub 4. VV side of the port 5. Bellows

Fig. 2. IVVS cartridge. 1. Cartridge frame 2. Push-chain magazine 3. B4 C movable block

6. Cryostat side of the port 7. Cryostat wall 8. Port extension 9. Closure plate

4. Cable loom 5. Counterweights 6. Leverages

2399

Fig. 4. Port section with IVVS cartridge.

the mobile assembly the chain is stored in a three layer magazine situated on the top of the cartridge. While deploying, the chain is guided by a profiled part of the guide tube. This guide also prevents the chain from jumping upwards during a potential seismic event. The mobile assembly remains identical to the one presented in [4]. It still includes a tilt motion mechanism actuated by the push chain. The generation of either a translation or a rotation of the scanning probe relies on the selective deployment of piezoactuated latches. The location and the operating conditions of IVVS make necessary a series of design choices. For instance, the need to operate with an active toroidal field limits the use of electromagnetic motors. For that reason, the main stepper motor is placed near the outer end of the port extension, where the magnitude of the magnetic field is low enough to permit its operation (circle in Fig. 4). Secondary actuators, located upfront, rely on the piezo-electric principle, inherently immune to magnetic fields. 2.4. The cable management system

Fig. 3. IVVS actuation system. 1. Push chain 2. Chain magazine 3. Stepper motor 4. Resolver

5. Gears 6. Push rod 7. The mobile assembly 8. Connector

The cables (power, signals, fibre optics) are all connected to the mobile assembly and must accommodate a 7.5 m translation. The individual cables are custom-manufactured using woven PEEK insulation implemented as screened twisted pairs. The ribbon cable is then assembled from these cables. The cable management system (Fig. 5) consists of two fixed pulleys and two travelling pulleys which can be moved forward and backward on a lead-screw. The motion is controlled by a series of synchronisation gears attached to the rigid chain drive, so that the pulleys move at the exact speed required by the cable deployment. The pulleys have a diameter of 250 mm to accommodate for the minimum bend radius of some of the fibre optics. One of the fixed pulleys is used to monitor the loom cable tension. This pulley is mounted on a spring loaded carriage which provides enough force on the loom to support its weight. The system controls both tension and bending applied on the cables. It also ensures that the umbilical cannot be trapped in the mechanism.

2.3. The main actuation system 2.5. The shielding blocks The main actuation of IVVS is achieved through a single chain which pushes or pulls the mobile assembly from the rear. It consists of interlocking links that behave like a chain and lock like a rigid bar, providing a safe, reliable and compact actuation mechanism that works in harsh environments and can be stored in a compact manner (see Fig. 3). This push chain is made of stainless steel and its links are vacuum coated. It is actuated through a special gearbox powered by a vacuum compatible stepper motor and position feedback is provided by a resolver. To allow for the 7.5 m travel of

A recent nuclear analysis showed that the direct neutron flux – resulting from the cut-out in the blanket – can reach significantly high values at the position where the scanning probe is stored during plasma operations (see Section 3.1). In order to protect the IVVS from neutron exposure while in stowed position, boron carbide (B4 C) shielding is provided. A moveable block, made of B4 C pellets encapsulated in a leaktight AISI 316LN enclosure, protects the probe by considerably

2400

G. Dubus et al. / Fusion Engineering and Design 89 (2014) 2398–2403

Fig. 7. IVVS B4C fixed blocks. 1. Cryostat side of the port 2. Fixed B4C blocks Fig. 5. IVVS cable management system. 1. Ribbon cable 2. Gears 3. Fixed pulleys

3. Movable B4C block

4. Lead screw 5. Travelling pulleys 6. Chain drive motor

reducing the direct neutron streaming (approx. 40 times). The block is lowered to allow the passage of the mobile assembly when an inspection is requested (see Fig. 6). The actuation mechanism has been integrated with the removable cartridge to permit remote maintenance. The failed state of this block is in the higher position, in order to maintain shielding for both scanning probe and closure flange in case of failure. The port also houses fixed shield blocks (Fig. 7) in order to protect IVVS from the scattered neutrons. These blocks are located around the resident guide tube, at the level of the scanning probe stowed position. 2.6. The scanning probe The concept of the scanning probe has not evolved since [4], although its design has been refined (see Fig. 8). The whole probe now weighs about 20 kg. The rotating head (Fig. 9) design has been revised so that it entirely fits within a 150 mm × 150 mm square section regardless of the tilt and pan positions. To that purpose, the prism dimensions have been reduced (from 141 mm × 55 mm to 139 mm × 52 mm) and its corners have been rounded. The tilt encoder has been reduced from Ø120 mm to Ø75 mm. The prism is now inserted in a Macor-made prism holder to allow easier mounting on the tilt rotator shaft. A Lambertian target

Fig. 8. Scanning probe. 1. Rotating head 2. Optical box

3. Connection box

has been added on its back in order to internally calibrate the laser. Curved holes have been included through the prism support to group and guide the tilt optical fibres during rotation. Pan rotation is limited by using mechanical stops and rad-hard micro-switches. A support has been added in order to ease the integration of these end-of-run switches and of the pan collimator holder. It also guides both pan optical fibres and limit switch cables into the optical box. There, a polyimide-like flat guide protects the fibres and electrical cables through the mechanical structure. The optical box (Fig. 10) also includes several changes. In order to reduce its dimensions, the optical assembly has been rearranged by inserting a couple of mirrors on the transmission (TX) path and by having a direct reception (RX) path. To limit the impact of stray

Fig. 6. Movable B4C block.

G. Dubus et al. / Fusion Engineering and Design 89 (2014) 2398–2403

Fig. 9. Close-ups on the rotating head. 1. Prism and prism holder 2. Tilt actuator 3. Pan actuator 4. Tilt fibre guide

Fig. 10. Close-ups on the optical box. 1. TX fibre holder 2. TX focussing lens 3. TX micro-positioning actuator

2401

5. Tilt optical encoder 6. Pan optical encoder 7. Calibration target

4. TX path mirrors 5. RX focussing lens 6. RX macro-positioning actuator 7. FBG-based accelerometer

light, the TX laser beam is guided into an inner Ø10 mm TX tube while the captured RX signal passes through an outer Ø52 mm RX tube. Space has also been booked to integrate a fibre Bragg grating (FBG) accelerometer. Lastly, the connection box ensures the proper electrical and mechanical interfacing between the probe and the rest of the mobile assembly (see Fig. 11). The mechanical interface has been made compatible with RH operations, so that the probe can be mounted/dismantled remotely for maintenance in the Hot Cell. The coarse alignment is achieved thanks to a couple of Ø16 mm dowel pins, commonly used in RH applications. In order to accommodate for residual misalignment between male and female connectors, some compliance has been added to the interface by allowing the connector socket to move in both horizontal and vertical directions without rotation.

Since the mechanical interface ensures correct alignment and guidance before engagement of the connectors, standard Lemo FAG-5B/EEG-5B connectors have been employed. Their inserts allow for the electrical connection of the pan and tilt motor cables, the step focus motor cables, the thermocouples and the end-ofruns, and for the optical connection of the TX fibre, the RX bundle, 6 tilt fibres, 6 pan fibres and the fibres for the FBG accelerometer.

3. Engineering analyses and R&D activities This section reports on the engineering and R&D activities carried out since 2012 in order to support the concept design described in Section 2.

2402

G. Dubus et al. / Fusion Engineering and Design 89 (2014) 2398–2403

Fig. 11. Close-up on the probe/arm interface. 3. Compliant support plate 1. RH dowel pins 2. Lemo connectors

3.1. Neutronic analysis A series of Monte Carlo N-Particle (MCNP) transport calculations have been conducted with the aim of quantifying neutron fluxes and damage rates to various IVVS components and determining the adequacy of several port shielding configurations [5]. The neutron flux profile computed along the IVVS axis shows an effective decrease of the direct streaming beyond the movable shielding block. The total flux (summation of the global and streaming contributions) through the closest piezo rotary stage (probe tilt actuator) has been estimated in the order of 3.4 × 108 n/cm2 /s, for a total fluence of 5.8 × 1015 n/cm2 after 20 years of ITER operation. Even though the closest piezo rotary stage is located behind a boron carbide block, both neutron flux and damage rates in this component are still 85% dominated by the direct streaming contribution. The impact of a scenario where the movable shield block fails to return to its correct shielding position was assessed. In such a case, the neutron flux in the probe tilt actuator would be up to 1.4 × 1010 n/cm2 /s, for a fluence of 2.3 × 1017 n/cm2 . The shutdown dose rate around an isolated cartridge was calculated for this failure scenario, after full activation of the IVVS and of its surroundings. Contact dose rates were found to be in the order of 100–160 ␮Sv/h after 105 s of cooling. This is beyond the normal 100 × ␮Sv/h limit proposed in ITER for hands-on maintenance, but these computed values are conservative and could be lowered by using low activation materials. The presence of the IVVS penetration was found to increase damage rates and helium production in the divertor pipe region and vacuum vessel by a factor of 3 immediately behind the cutout in the blanket and triangular support; but the calculated values are still within ITER project limits. The effect on the magnetic coil nuclear loads is negligible. Lastly, this analysis demonstrated that active cooling was not required in IVVS or in the B4 C blocks as heating rates in these locations were generally very low (1 W in the movable shield block during a 500 MW pulse). 3.2. Seismic and electromagnetic analyses The material choice for the various parts of IVVS is mainly driven by the maximum stresses they will have to withstand during their service life. Consequently, the concept design has been checked for the various loads it will be subjected to. In particular, this task comprised a stress and deformation analysis for the critical parts of the system when subjected to a seismic

level-2 event (SL2) in deployed configuration. The finite element analysis (FEA) software ANSYS v14.0 was used to perform both modal and spectrum analyses. The lowest natural frequency of the whole mobile assembly was found to be around 22 Hz. The spectrum analysis showed a maximal deflection of 1.3 mm at the tip of the probe as a result of an SL2 seismic excitation. It confirmed that all the mechanical components of IVVS will withstand the dynamic stresses with sufficient safety factors. In addition, the impact of the electromagnetic loads IVVS will be subject to has been analysed. This analysis addressed the forces and torques acting on IVVS in stowed position, as a result of a linear VDE cat-III plasma disruption [6]. While significantly high loads occur on the IVVS port structure, the maximum forces and torques exerted on the IVVS itself were found too small to have any significant effect on the system. 3.3. Probe actuation validation under magnetic field An experimental activity was conducted to evaluate the behaviour under an ITER-relevant magnetic field of the piezoelectric technology identified to actuate the probe prism. In particular, the test aimed at assessing how the motor speed is influenced by the field strength and by slow flux variations. It was performed using a 14 T superconducting facility hosted at ENEA Frascati. The motor speed was monitored in open-loop during magnetic field ramp-ups/-downs between 0 and 10 T. In all cases, the rotary stage was able to reach the maximal speed value of 11 rps, regardless of the field magnitude. Although speed variations were recorded during this test, there is no evident correlation with the field magnitude. Similar speed variations were observed outside of any magnetic field and should easily be solved by closed loop control. Full test results can be found in [7]. 3.4. Probe actuation validation under high vacuum and high temperature A second experimental activity was performed to evaluate the reliability of the probe actuation technology under high vacuum (5 × 10−4 Pa) and high temperature (120 ◦ C). For the purpose of this test, a dedicated vacuum facility was designed, assembled and equipped with a turbo-molecular pump able to achieve a base pressure in the order of 10−5 –10−6 Pa at ambient temperature. The task was divided in three phases respectively consisting in a functional test in vacuum, a functional test in vacuum/temperature and

G. Dubus et al. / Fusion Engineering and Design 89 (2014) 2398–2403

2403

a baking resistance test. In each case, the motor speed was characterised in open-loop. This experimental campaign demonstrated that, at ambient temperature, the motor performance at 10−4 Pa was equivalent as in air. It also showed that it is able to run up to 8 h at 120 ◦ C without any perceptible decrease of its performance. At last, it proved the ability of the motor to operate at 120 ◦ C after being baked at 200 ◦ C for 2 h. The strong dependence of the motor speed on the selected operating frequency was shown, suggesting that the piezo material resonant frequency may vary with temperature in the order of some kHz. By the end of 2013, this test will be complemented by a measurement of the outgassing rate of this piezo motor during baking at 200 ◦ C.

An experimental activity was set up in order to evaluate the capability of IVVS to detect erosion on the ITER first wall and divertor. A computational procedure was developed with the aim to calculate, from measurements acquired with the ENEA scanning probe prototype, the volume of eroded areas on a calibrated aluminium target representative of the first wall panel of a fusion device like ITER. This study confirmed that the current IVVS concept would be able to detect regions eroded in the order of some hundreds of ␮m, provided that ITER will comprise datum points which are not affected by dimensional changes over time. Full test procedure and results can be found in [9].

3.5. Gamma irradiation of the probe key components

Acknowledgments

A gamma irradiation campaign was carried out in order to evaluate the reliability and lifetime expectation of key components of the probe: the piezo rotary stage actuating the prism, the optical encoder and small scale optical samples [8]. The test was conducted in the CALLIOPE gamma irradiation facility (ENEA Casaccia) between July and November 2012, up to a cumulated dose in the order of 4.8 MGy. The performance of the piezo actuator was characterised daily by assessing the minimum command values required to move the motor clockwise (CW) and counter clockwise (CCW), and to press limit switches representative of the resistive torque under nominal scanning operative conditions. After 2 months of irradiation, all measurements indicated that gamma radiation does not cause major damage to the piezoelectric motor. Throughout the campaign the motor was able to provide the torque necessary to rotate and to close the switches. It is to be noted that, for a given input value, the motor output speed increased over time. At the end of the test, it was rotating about three times faster CW and twice as fast CCW. To date, it has not been possible to interpret this observation with certitude.

The authors would like to thank all the contributors taking part to the activities touched upon in this paper. The views expressed in this publication are the sole responsibility of the authors and do not necessarily reflect those of Fusion for Energy (F4E) or of the ITER Organization (IO). Neither F4E nor any person acting on behalf of F4E is responsible for the use which might be made of the information in this publication.

3.6. Neutron irradiation of the probe main actuator A neutron irradiation campaign is currently being performed in order to verify if the probe piezo actuator is able to withstand a 1 MeV equivalent neutron fluence in the order of 2.3 × 1017 n/cm2 , which corresponds to the failure scenario referred to in Section 3.1. The test is being conducted in the TAPIRO reactor (ENEA Casaccia) and should be completed during Q4 2013. An intermediate characterisation of the motor was carried out after reaching a fluence of 1.3 × 1016 n/cm2 . No particular change in the motor performance was noticed. More detailed results can be found in [3].

3.7. Assessment of the erosion evaluation capabilities

References [1] L. Bartolini, A. Coletti, M. Ferri de Collibus, G. Fornetti, C. Neri, M. Riva, et al., Amplitude modulated laser in vessel viewing system (LIVVS) for ITER/JET, Fusion Technology 1 (1998) 685–688. [2] C. Neri, L. Bartolini, A. Coletti, M. Ferri de Collibus, G. Fornetti, F. Pollastrone, et al., The laser In-Vessel Viewing System IVVS for ITER: test results on first wall and divertor samples and new developments, Fusion Engineering and Design 82 (2007) 2021–2028. [3] C. Neri, U. Besi, M. Ferri De Collibus, G. Mugnaini, C. Monti, M. Pillon, et al., Status of the design refinement and the characterisation of the In-Vessel Viewing System for ITER, in: Proceeding of the IEEE 25th Symposium on Fusion Engineering SOFE, 2013, pp. 1–4. [4] G. Dubus, A. Puiu, C. Damiani, M. Van Uffelen, A. Lo Bue, J. Izquierdo, et al., Conceptual design finalization of the ITER In-Vessel Viewing and Metrology System (IVVS), Fusion Engineering and Design 88 (2013) 1973–1977. [5] A. Turner, R. Pampin, Z. Ghani, G. Hurst, A. Lo Bue, S. Mangham, et al., Nuclear analysis and shielding optimisation in support of the ITER In-Vessel Viewing System design, Fusion Engineering and Design (2014), http://dx.doi.org/10.1016/j. fusengdes.2014.01.066, in press. [6] G. Sannazzaro, C. Bachmann, D. Campbell, S. Chiocchio, J.-P. Girard, Y. Gribov, et al., Structural load specification for ITER tokamak components, in: Proceedings of the 23rd IEEE/NPSS Symposium on Fusion Engineering SOFE, 2009, pp. 1–4. [7] C. Monti, U. Besi, G. Mugnaini, C. Neri, P. Rossi, R. Viola, et al., Test of piezoceramic motor technology in ITER relevant high magnetic fields, in: Proceedings of ISFNT-11, 2014. [8] P. Rossi, M. Ferri de Collibus, M. Florean, C. Monti, G. Mugnaini, C. Neri, et al., IVVS actuating system compatibility test to ITER gamma radiation conditions, Fusion Engineering and Design 88 (2013) 2084–2087. [9] F. Pollastrone, M. Ferri de Collibus, M. Florean, M. Francucci, G. Mugnaini, C. Neri, et al., Erosion evaluation capability of the IVVS for ITER applications, in: Proceedings of ISFNT-11, 2014.