1
• The FEM model foresees the AU-BU gap to reduce uniformly with time of ~5mm along the radial (X in the TGCS reference frame) direction, and a global misalignment between the two sub- components of 1.4mm along the vertical direction. No out of horizontal plane (Y axis) displacements are predicted. -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 BU_16 BU_18 BU_20 BU_21 BU_23 BU_24 BU_25 BU_27 BU_28 BU_29 BU_32 BU_33 BU_34 BU_35 BU_38 BU_39 BU_41 BU_43 BU_45 BU_46 BU_47 BU_49 BU_51 BU_53 BU_55 BU_57 BU_58 BU_59 BU_60 BU_62 BU_63 BU_64 BU_65 BU_67 BU_68 BU_69 BU_70 BU_73 BU_74 BU_75 BU_76 BU_78 BU_79 BU_80 BU_81 BU_82 BU_83 All fiducials Z-displacements [mm] FEM Root pass FEM 25mm FEM 90mm DI Root pass DI 25mm DI 90mm Comparison of FEM Predicted and Measured values of the TF coil closure welding distortion Edoardo Pompa, SETIS & University of Rome Tor Vergata; Marc Jimenez, Gabriele D'Amico, Boris Bellesia, Alessandro Bonito-Oliva, Fusion for Energy 26 th International Conference on Magnet Technology , September 22 - 27 , 2019 - Level 2 Posters 2 I.D. number: Mon-Af-Po1.17-03 [55] • Toroidal Field Coils Cases (TFCC) are SS316LN structures, which have to withstand strong magnetic fields (around 12 T) in order to confine the high temperature plasma (150M C°) • Welding distortions can compromise the final shape of the assembly, generating out-of-tolerances in the interface areas that cannot be recovered by the extra- material foreseen. • A Finite Elements Model has been developed by Enginsoft S.p.A in collaboration with SIMIC and F4E. • In July 2019 the welding phase of TFC09 has been completed, monitoring the case deformation by fiducial points along the whole process. 1. Introduction 2. FEM model definition and survey setup • Theoretical and technical methods applied to the model: • Quasi-steady state analytical solution (Rosenthal) to de-couple the thermo-mechanical analysis. • ‘Birth’ and ‘death’ technique to simulate material deposition during weld. • Clustering technique of the weld passes to achieve mesh reduction maintaining as much as possible the real sequence in the material deposition. • Tuning of the model through experimental campaign: • First stage: welding of six qualification coupons. Measured displacements and temperatures have been used as calibration parameters (cut-off temperatures and cooling law, chord mesh size and coefficient of thermal expansion (CTE) of chord material). A blind test on a coupon has been carried out to final validate the model. • Second stage: more refined tuning of the parameters on three TFCC full scale mockups, reproducing three zones of the case. • Deformation of the TFCC structure has been monitored approximately every 25mm of deposited material by laser tracking survey. The system accuracy depends on the distance from the target, the number of fiducials and the number of position of the LT used. The uncertainty of the measurement evaluated is about 0.2 mm. Quick surveys on 45 accessible fiducial points on the case with high repeatability. Globally, the model reproduces the with good fidelity the TFCC deformation, in terms of behavior of deformations and order of magnitude of the displacements. Further work is foreseen in order to improve the prediction of deformation in specific areas (IOIS), while maintaining the global coherence. Increasing the structure stiffness, fine-tuning of material properties are foreseen as possible solutions to improve the model prediction capabilities. 3. Results comparison • The TFCC full model (190k elements) includes all the steps of the real welding sequence, tack weld, butt weld, poloidal weld and splice plates weld. • FEM boundary conditions allows all the DOF of the support with minimum reaction forces. The real support configuration only allows BU case to move, while AU remains fixed along the process. FEM deformations have been filtered with the average values of the displacements of the AU fiducials points in order to achieve the same configuration of the DI. 4. Conclusions and future activities 3a. Butt weld phase 3b. Poloidal weld phase Analysis step Bevel fill [mm] Tack Weld 1 - Butt Weld Root Pass 8 6 25% 12 25 50% 15 50 75% 18 70 End 21 90 Poloidal Weld Root Pass 31 6 50% 73 55 End 54 110 Splice Plates Weld Root Pass 57 6 25% 58 25 50% 59 50 End 61 90 FEM techniques Survey setup Steps comparison Fiducial point • The DI reported ~6mm of average closure after the first 25mm of material deposition in the chamfer, achieving at the end of the weld an average 7mm closure between parts. • The cases misalignment along the vertical direction has been verified by the DI data. The measurements show the same order of magnitude, but the real component present a higher degree of misalignment localized in the termination area. • Only AP and BP plates have been foreseen to deform under the welding loads. The FEM model predicts negligible deformations on the rest of the case. • The FEM prediction is localized in the bevel zone, while the DI demonstrated to be extended to the case sides. This caused the fiducials on the IOIS interfaces to rotate according to the cases sides angular deformation. • By taking into account only the fiducials not affected by the angular deformation, the average X displacements reported from the DI are in line with the model. FEM DI IOIS fiducials XY-plane displacements [mm] Z X Y -9.00 -8.00 -7.00 -6.00 -5.00 -4.00 -3.00 -2.00 -1.00 0.00 Root pass 55mm 110mm Average X-displacements [mm] FEM DI -8.00 -7.00 -6.00 -5.00 -4.00 -3.00 -2.00 -1.00 0.00 Tack Weld Root pass 25mm 50mm 70mm 90mm Average X-displacements [mm] FEM DI
