1. ITER VV supports Cadarache 6 September 2007 A. Capriccioli

2. Outline 1. Pot bearings actual design: possible solution with a) new 40 MN downward force b) new 10 MN upward force c) toroidal restraint system 2. Flexible Plates, alternative solution 3. Conclusions

3. With reference to the new 40 MN downward force (total value: dead weight plus vertical and horizontal electromagnetic forces) and with reference to the actual design (net rubber diameter equal to 800 mm), the average pressure on the rubber (Neoprene) component results equal to 79.6 MPa. If the reference max value of 80 MPa, during downward transient VDE, can be assumed for the rubber component in the pot bearing pads, we obtain a safety margin SM = 1. Several solutions can be adopted to increase the SM and the easier, cheaper and feasible seems to be the upward translation of the pot bearing itself: 1.a) New electromagnetic downward vertical force: Port Bearing Pad Pedestal Ring

4. While in the toroidal direction seems feasible to increase the bearing pad size from 960 up to 1700 or more (see previous figure), in the radial direction the maximum neoprene diameter should be around 1200 mm. In every case, changing the neoprene diameter from 800 to 1200 mm induces an area increment of about 2.3 times and an average pressure less then 35 MPa (against the previous 80 MPa, with an increment of the SM from 1 to about 2.3).

5. 1.b) New electromagnetic upward vertical force: The max value of 10 MN upward per support was estimated The figure on the right side shows the long rods and ropes groups foreseen to prevent vertical detachment between VV and bearing pads. With reference to the previous meeting (28 June 2007), the use of not preloaded tie-rods is to avoid. The use of very stiff rods only can replace the previous one. Another solution is the use of vertical dampers (the Fig.1 next page shows an example of shock absorbers). In this case, two dampers 5000 kN each are necessary and diameters around 550 mm with minimum 1.7 m length are the standard dimensions.

6. Fig.1

7. The only way to reduce the dampers dimensions is the reduction of the axial force, through the amplification of the displacement Port  Damper. A proposal of alternative solution is shown in the scheme of Fig.2a. Only one damper is necessary (the horizontal device); the other elements are connections between Port and Ring. These last connections form an articulated structure attached to the Port/Ring side through Cardan or spherical joints (see Figures 2a and 2b). When the port tries to move vertically upward an axial force acts on the damper and its value is related to the slope of the stiff connections. Fig. 2b With INCONEL rods diameter of 200 mm and a vertical force of 5 MN, the max vertical displacement results about 0.35 mm per meter rod length. ~ 1400 mm  L  ____ mm Tvf/2 Tvf/2*0.4  20° 2*(cos    - cos   )  y b  x a  (sin    - sin   ) b y x a Tvf/2 = Total electromagnetic vert. upward force /2 PORT Fig. 2a

8. 1.c) Toroidal restraint system: The actual Toroidal restraint system seems to show potential seizing risks and when the vertical supports allow toroidal displacements, as in actual pot bearing (or spherical bearing) pads, the “pendulum” restraint system seems to be a really good choice. The figures on the left side show the W7-X Auto Centering System (see A.Cardella “ITER Vacuum Vessel Support System”, Working Group Meeting, Cadarache 28 June 2007). In the W7-X reactor no electromagnetic forces act on the Vacuum Vessel and the “pendulums” geometry is not a critical point. In the figure it is possible to note the relative high L/D ratio The system allows Vacuum Vessel vertical displacements and radial thermal expansion with small toroidal rotations of the whole VV itself.

9. The “pendulum” solution for ITER reactor has to take into account several basic points: the presence and the entity of the net horizontal force; the presence of the radial restraint system; during the disruption event, the opportunity to spread the reaction forces with different weights between radial and tangential ports (to minimize the stress level in the Ports-VV connection). At the moment no possibilities there are to perform a detailed design and analysis of a proper toroidal restraint style W7-X and only a scheme of “pendulum” with variable axial stiffness is shown in figures: It will be possible to change with continuity the axial stiffness at constant pendulum length, if the screws pitches are identical. This could be fixed on one end to the lower port and on the other end to the pedestal ring.

10. 2. Flexible plates, alternative solution Tab.1: an example with the Actual Allowable Space ( Width =960 mm and Length = 1200 mm ) 40 MN Vertical Force + Radial Displacement: Allowable 20 mm Thickness = 16 mmMembrane stress, σ m 43.4 MPa<141 MPa n° of plates = 60Membrane + Bending, σ m + σ b 149 MPa<212 MPa Width= 960 mmBuckling margin, m cr 2.5<3 Length = 1200 mmSM against collapse (Elas-Plas)3.3>3 Radial space, ΔR1430 mm If width = 1100 mm and n° of plates = 70 then Buckling margin, mcr = 3.3 >3 and ΔR = 1810 mm Tab.2: the same example with 2 plate thicknesses 40 MN Vertical Force + Radial Displacement: Allowable 20 mm Thicknesses = 14 / 20 mmMembrane stress, σ m 29 / 39.2 MPa<141 MPa n° of plates = 70Membrane + Bending, σ m + σ b 138.8 MPa<212 MPa Width= 1000 mmBuckling margin, m cr 3.3>3 Length = 1200 mmSM against collapse (Elas-Plas)3.5*>3 Radial space, ΔR2090 mm

11. A point to point out is the toroidal stiffness of the flexible plates system: in this case its value is basically high and cannot be easily changed (for example it is possible to divide each plate vertically in two or more parts). The values of the toroidal restraint and radial restraint stiffness are two basic characteristics to evaluate together with the tangential and radial stiffness of the Port / VV shell connection. Tab.3: Flexible plates 2.4 m height 25 MN Vertical Force + Radial Displacements: Allowable 20 mm30 mm40 mm Thickness = 43 mmMembrane stress, σ m 16.2 MPa <141 MPa n° of plates = 30Membrane + Bending, σ m + σ b 88.5 MPa131 MPa162 MPa <212 MPa Width= 1200 mmBuckling margin, m cr (*)12 (µ=0.5)>3 Length = 2400 mmSM against collapse (Elas-Plas)10.8 > 3 Radial space, ΔR1580 mm (*) see formula (3) from X.Wang, K.Ioki - ITER, August 7, 2007 "Preliminary Assessment of Multi Flexible Plates VV Support" Tab.2: Flexible plates 2.4 m height with 40 MN 40 MN Vertical Force + Radial Displacement: Allowable 20 mm30 mm Thickness = 43 mmMembrane stress, σ m 26 MPa <141 MPa n° of plates = 30Membrane + Bending, σ m + σ b 97.3 MPa 135 MPa <212 MPa Width= 1200 mmBuckling margin, m cr 7.4 (µ=0.5) >3 Length = 2400 mmSM against collapse (Elas-Plas)6.8*>3 Radial space, ΔR1580 mm

12. 3. Conclusions The two analyzed VV Support systems are: (1) Pot bearing + vertical upward restrain + toroidal system→ Vertical up/down + toroidal (2) Flexible plates→ Vertical up/down + toroidal -Both seem feasible (and for both other analyses are necessary). -The system (1) foresees common industrial use devices (pot bearings, shock absorbers) while the “pendulum“ (W7-X type) toroidal restraint system has to be analyzed. -The system (2) results easier, because in a single block are present all the restraints but it is not a commercial device (R&D), has a fixed toroidal stiffness and more vertical space would be necessary (see the buckling margin m cr ). -Common to both the systems (1, 2) is the radial restraint. -With the new radial forces during the downward vertical plasma disruption (73 MN against the previous 25 MN), the radial restraint system must be reviewed. - The Vacuum Vessel and Ports global model is essential to the evaluation of all the restraint systems.