Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
• 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) displacementsare 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
26th International Conference on Magnet Technology , September 22 - 27 , 2019 - Level 2 Posters 2I.D. number: Mon-Af-Po1.17-03 [55]
• Toroidal Field Coils Cases (TFCC) areSS316LN structures, which have towithstand strong magnetic fields (around12 T) in order to confine the hightemperature plasma (150M C°)
• Welding distortions can compromise thefinal shape of the assembly, generatingout-of-tolerances in the interface areasthat cannot be recovered by the extra-material foreseen.
• A Finite Elements Model has beendeveloped by Enginsoft S.p.A incollaboration with SIMIC and F4E.
• In July 2019 the welding phase of TFC09has been completed, monitoring thecase deformation by fiducial points alongthe 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-mechanicalanalysis.
• ‘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 andtemperatures have been used as calibration parameters (cut-off temperatures and coolinglaw, chord mesh size and coefficient of thermal expansion (CTE) of chord material). A blindtest 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 materialby laser tracking survey. The system accuracy depends on the distance from the target, the number offiducials and the number of position of the LT used. The uncertainty of the measurement evaluated isabout 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. Increasingthe 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 realwelding sequence, tack weld, buttweld, poloidal weld and splice platesweld.
• FEM boundary conditions allows all theDOF of the support with minimumreaction forces. The real supportconfiguration only allows BU case tomove, while AU remains fixed along theprocess. FEM deformations have beenfiltered with the average values of thedisplacements of the AU fiducialspoints in order to achieve the sameconfiguration of the DI.
4. Conclusions and future activities
3a. Butt weld phase
3b. Poloidal weld phaseAnalysis 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
FE
M t
ech
niq
ue
s
Su
rvey s
etu
p
Ste
ps c
om
pari
so
n
Fid
ucia
l
po
int
-9.00
-8.00
-7.00
-6.00
-5.00
-4.00
-3.00
-2.00
-1.00
0.00
1.00
TackWeld
Root pass 25mm 50mm 70mm 90mm
BU_70 X-displacements [mm]
FEM
DI
• The DI reported ~6mm of average closureafter the first 25mm of material deposition inthe chamfer, achieving at the end of the weldan average 7mm closure between parts.
• The cases misalignment along the verticaldirection has been verified by the DI data.The measurements show the same order ofmagnitude, but the real component present ahigher degree of misalignment localized inthe 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
fid
ucia
ls
XY
-pla
ne
dis
pla
cem
en
ts
[mm
]
Z
XY
-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
0.00Tack Weld Root pass 25mm 50mm 70mm 90mm
Average X-displacements [mm]
FEM
DI