CCL

COMPLETION OF THE FIRST TF COIL STRUCTURE OF ITER

INTRODUCTION

M. Nakahira, M. Iguchi, T. Sakurai, H. Ozeki, E. Fujiwara, K. Takano, Y.S. Hong, M. Ino, M. Nishino, N. Koizumi, N. Sawa*, D. Hara*, T. Inagaki*, S.Y. Kim**, J.H. Choi**, S.S. Hwang** and C. Luongo***National Institutes for Quantum and Radiological Science and Technology, 801-1 Mukouyama, Naka-shi, Ibaraki 311-0193, Japan*Mitsubishi Heavy Industries, LTD, 1 Minamifutami, Futami-cho, Akashi 674-0093, Japan**Hyundai Heavy Industries co., LTD, 1000 Bangeojinsunhwan-doro, Dong-gu, Ulsan, 682-792, Korea***ITER Organization (IO), Route de Vinon-sur-Verdon, CS 90 046, 13067, St. Paul-lez-Durance, France

FIP1-1 / CN258

(i) Material

CCL

(ii) Welding Deformation

CP溶接準備中

CP溶接準備中

(iii) Partial Penetration Welding (PPW)

ConclusionITER Toroidal Field Coil

18m

TF Coils (18)

16m

Great Stupa (16m high, Sanchi, India)~300 ton Water in 25mx12m pool (~300ton)

TF Coil Structure (TFCS)• The biggest super conducting coil structures • The procurement responsibility :100% Japan Domestic Agency (JADA).

Winding Pack (WP)

AU

AP

BP

BU

TFCS

65 ton105 ton

8 ton

12 ton

A1

A2

A3B1

B2

B3

B4

Basic Segments (BS) Sub-Assembly (SA)

Challenges(i) Material:Control yield strength at 4KControl fracture toughness at 4K

(ii) Welding deformationControl welding deformationControl segments welding(iii) Partial Penetration Welding (PPW)PPW crack initiationPPW crack growth(iv) Ultrasonic testing (UT) Attenuation compensation methodAttenuation compensation factorUT for PPW(v) Fitting testFitting test for AU-AP and BU-BPFitting test for AU-BU

as-weld notch

AU/BU

AP/BP

Control yield strength at 4K

Special material is required with total amount about 5000 ton

Ensure Huge magnetic force(Static analysis)

y = 0.0106x2 - 6.4144x + 1336.3

y = 0.0074x2 - 4.4901x + 935.39

0

200

400

600

800

1000

1200

1400

0 50 100 150 200 250 300TEMPERATURE (K)

STR

EN

GTH

(MP

a)

��=　0.7

Measured value

Prediction curve (1)

Prediction curve by (C+N)

Figured out correlation between yield strength at 4K and C+N contentsHuge amount of 4K test is

neededActual materials

(total about 5000 ton)

In beginning of 2018, material procurement for TFCS was completed.

M14 R20

42 105

φ7±

0.01

C+N contentsYield strength at RT

SpecimenYield

strength at 4K

Predict

C

N

Control fracture toughness at 4KIn the additional work to improve control fracture toughness, JADA discovered the strong correlation between Md30 and fracture toughness at 4K.

Figured out correlation between fracture toughness at 4K and Md30

Found martensite at edge of cracked area

The Md30 has improved the quality of fracture toughness.

( )( )

( ) ( )308421068518713029182946282430

..Nb.Mo.Cr.CuNi.Mn.Si.NCMd

−ν−−−−+−−−+−=

y = αx+βLow fracture toughness

(ex. Glass: Hard but fragile)

High fracture toughness

(ex. Metals)

Chemical contents (9 elements)Grain size

Fracture toughness at 4K

Predict

Control weld deformationJADA performed 1) welding qualification using mock-ups, Mechanical properties of welding joints were confirmed, 2) Basic segment mock-ups. Control method of welding deformation was improved.

① Outer groove welding (1st)

② Inner groove welding (1st)

③ Outer groove welding (2nd)

④ Inner groove welding (2nd)

Welding with monitoring welding deformation (Balance welding)

B3 segment mock-up The deformation converges to 0mm.

Welding qualification

Improve by sequence

Control segments weldingSegments welding is the most difficult to control. Through trial, amount of deformation and tendency are figured out to implement to actual manufacturing.

Segment-to-segment welding (A1+A2)Example of welding deformation control

Welded Sub-assemblies (AU)

0

2

4

6

8

10

12

14

16

0

20

40

60

80

100

120

140

0 51 100 150 201 249 300 351 399 453

Defo

rmat

ion

(mm

)

Depo

sit t

hick

ness

(mm

)

Total deposit thickness of 3 plates (mm)

Outer plate (outside)

Outer plate (inside)

Side plate

Deformation

投影軸

ケガキ

＋

－

変形量

MarkLine

Controlleddeformation

ReferenceLine

4m

WeldingLine

A2

A1

~180mm~80mm

Side Plate

Outer Plate

Section at welding line

Welding trial (A1+A2(3m))

TargetPosition

The actual manufacturing, deformation is well controlled

Narrow work spaceFPW: Invisible weldingPPW: Visible welding

Plate shape attachmentsFPW: Impractical weld joint designPPW: Practical weld joint design

Full Penetration welding (FPW) is better. But…

Application of PPW is necessary

as-weld notch

or

S-N? LEFM?

Fusion line

Weld Metal

Fusion lineBase Metal

Base: 316LN (N~0.21%)Weld: JJ1 (12Cr-12Ni-10Mn-5Mo-0.13N)Method: TIG weld

Unwelded

WeldedWelded

As-weld notch

CT specimen (As-weld notch)

CT specimen (EDM notch)Test condition: Load control Frequency: 10HzStress ratio: 0.1

a)

b)

12,698

1,898

As-weld notch (no artificial treatment)

EDM notch (R0.1mmis remained on edge.)

=> Behavior like Crack

=> Propagation in weld

CT specimen (Weld metal/EDM notch)

d)FEM analysis and assessment allowable weld joint and maximum initial defect size

• FEM analysis (total 133 weld joints) • Allowable maximum initial defect size =>100mm2 of semi-elliptical at root

① Slow crack growth

③Application of common region to assessment

①

②③

c) Crack growth rate parameter

②Stress redistribution when sampling

Method of “Design by analysis” for PPW was successfully established!

CCL

(iv) Ultrasonic Testing (UT)

(v) Fitting Test

Principle of Ultrasonic Testing

CrackEcho

Back surfaceEcho

Initial pulse

Attenuation of weld metal was evaluated.

Transducer

Crack

Ultrasonicwave

Weld metal

Basemetal

Wavedecreasesin the weld

UT attenuation compensate method

DAC curves* were prepared by– Calibration block using base metal– Reference block including weld

metal

*: DAC curves (Distance Amplitude Characteristic curves)

1. Making DAC curve of the base metal using a calibration block

Plotted point

Distance

3. Calculating the sensitivities difference (A, A’, …) between each plotting point and DAC curve

Sensitivities difference

Calibration block

2. Plotting detected echo through the weld metal using a reference block with SDH

Reference block

The difference were quantified.

Ultrasonic attenuation (dB) vs. Weld metal path(mm)“y = ax” (a: 0.2 ~ 0.4)

Weld metal attenuation of TFCS was Properly corrected.

y = 0.313x

y = 0.395x

y = 0.216x

0

5

10

15

20

25

0 20 40 60 80 100

Sens

itivi

ty d

iffer

renc

e (d

B)

Weld Metal Path (mm)

Company ACompany BCompany C

UT for PPWEstablishment inspection method for PPW

Noise!

• High quality weld joint ← Inside defect inspection• Weld depth& initial crack size ← weld depth

Confirmation Ultrasonic Testing (UT) method

• Noise near root → Low accuracy on depth measuring?• Verification test by actual size PPW mock up.

UT procedure was defined.

+/-1mm accuracy for depth measuring.

1

AP/BPAU/BU

Gap

Mis-alignment

Gap: 0.5±0.25mmMisalignment:±0.3mm

Req. groove tolerances (inner)

Misalignment: ±1.3mm

Req. groove tolerances (outer)

AU/BU

AP/BP

Req. 2): AU-BU

Gap: 0.5±0.25mmMisalignment:±0.7mm

Gap: 0.5±0.25mmMisalignment:±0.3mm

Gap: 0.5±0.25mmMisalignment:±0.3mm

Back plate

Side plate

Side plate

Req. 1): AU-AP, BU-BP

Upper

Lower

AU and AP after fit-up

Nylon wedges

Stainless Steelwedges

Hydraulic Jack

Protection

Jig

Jigs used to fit up AP to AU

AU and BU after fitting up for the first EU products tested in horizontal position

AU and BU after fitting upfor the first JA coil tested in vertical position

Note: Md30 is defined as the temperature which 50% martensitic transformation occurs when 0.3 strain is applied. Originally, it’s tendency of martensitic transformation.

• Solving the difficult challenges, TFCS became feasible.• Two TFCSs has been completed in 2018, and another

one will be completed soon. • The first Japan-manufacturing TF coil will be assembled

TFCS and WP from the fourth quarter of 2018 as the very first TF Coil of ITER

• All the TF coils will be delivered to ITER in 2021

1

COMPLETION OF THE FIRST TF COIL STRUCTURE OF ITER

2018.10.22 27th IAEA Fusion Energy Conference

Ahmedabad, India

M. NAKAHIRA, M. IGUCHI, T. SAKURAI, H. OZEKI, E. FUJIWARA, K. TAKANO, Y.S. HONG, M. INO, M. NISHINO, N. KOIZUMI (QST)N. SAWA, D. HARA, T. INAGAKI (MHI)S.Y. KIM, J.H. CHOI, S.S. HWANG (HHI)C. LUONGO (IO)

FIP1-1 / CN258

ITER Toroidal Field Coil

2

18m

TF Coils (18)

16m

Great Stupa (16m high, Sanchi, India)

~300 ton

Water in 25mx12m pool (~300ton)

TF Coil Structure (TFCS)• The biggest super conducting coil structures • The procurement responsibility :100% Japan Domestic Agency (JADA).

3

Winding Pack (WP)

AU

AP

BP

BU

TFCS

65 ton105 ton

8 ton

12 ton

A1

A2

A3B1

B2

B3

B4

Basic Segments (BS) Sub-Assembly (SA)

Challenges

4

(i) Material:Control yield strength at 4KControl fracture toughness at 4K

(ii) Welding deformationControl welding deformationControl segments welding

(iii) Partial Penetration Welding (PPW)PPW crack initiationPPW crack growth

(iv) Ultrasonic testing (UT) Attenuation compensation methodAttenuation compensation factorUT for PPW(v) Fitting testFitting test for AU-AP and BU-BPFitting test for AU-BU

as-weld notch

投影軸

ケガキ

＋

－

変形量

MarkLine

Controlleddeformation

ReferenceLine

Section at welding line

AU/BU

AP/BP

Challenge (i) Control yield strength at 4KSpecial material is required with total amount about 5000 ton

5

Ensure Huge magnetic force(Static analysis)

y = 0.0106x2 - 6.4144x + 1336.3

y = 0.0074x2 - 4.4901x + 935.39

0

200

400

600

800

1000

1200

1400

0 50 100 150 200 250 300TEMPERATURE (K)

STR

EN

GTH

(MP

a)

��=　0.7

Measured value

Prediction curve (1)

Prediction curve by (C+N)

Figured out correlation between yield strength at 4K and C+N contents

Huge amount of 4K test is neededActual materials(total about 5000 ton)

In beginning of 2018, material procurement for TFCS was completed.

M14 R20

42 105

φ7±

0.01

C+N contentsYield strength

at RT

Specimen

Yield strength

at 4K

PredictC

N

Challenge (i) Control fracture toughness at 4KIn the work to improve control fracture toughness, JADA discovered the strong correlation between Md30 and fracture toughness at 4K.

6Figured out correlation between fracture toughness at 4K and Md30

Found martensite at edge of cracked area

The Md30 has improved the quality of fracture toughness.

( )( )

( ) ( )308421068518713029182946282430

..Nb.Mo.Cr.CuNi.Mn.Si.NCMd

−ν−−−−+−−−+−=

y = αx+β

Low fracture toughness(ex. Glass: Hard but fragile)

High fracture toughness(ex. Metals)

Chemical contents (9 elements)Grain size

Fracture toughness

at 4K

Predict

Challenge (ii) Control weld deformationJADA performed 1) welding qualification using mock-ups, Mechanical properties of welding joints were confirmed, 2) Basic segment mock-ups. Control method of welding deformation was improved.

7

① Outer groove welding (1st)

② Inner groove welding (1st)

③ Outer groove welding (2nd)

④ Inner groove welding (2nd)

Welding with monitoring welding deformation (Balance welding)

B3 segment mock-up

The deformation converges to 0mm.

Welding qualification

Improve by sequence

Challenge (ii) Control segments weldingSegments welding is the most difficult to control. Through trial, amount of deformation and tendency are figured out to implement to actual manufacturing.

8

Segment-to-segment welding (A1+A2)

Example of welding deformation controlWelded Sub-assemblies (AU)

0246810121416

0

20

40

60

80

100

120

140

0 51 100

150

201

249

300

351

399

453

Defo

rmat

ion

(mm

)

Depo

sit t

hick

ness

(mm

)

Total deposit thickness of 3 plates (mm)

Outer plate (outside)Outer plate (inside)Side plateDeformation

投影軸

ケガキ

＋

－

変形量

MarkLine

Controlleddeformation

ReferenceLine

4m

WeldingLine

A2

A1

~180mm~80mm Side Plate

Outer Plate

Section at welding line

Welding trial (A1+A2(3m))

TargetPosition

The actual manufacturing, deformation is well controlled

Challenge (iii) Partial Penetration Welding (PPW)

9

Narrow work spaceFPW: Invisible weldingPPW: Visible welding

Plate shape attachmentsFPW: Impractical weld joint designPPW: Practical weld joint design

Full Penetration welding (FPW) is better. But…

Application of PPW is necessary

as-weld notchor

S-N? LEFM?

10

1: Confirmation of crack initiation behavior

Fusion line

Weld Metal

Fusion lineBase Metal

Base: 316LN (N~0.21%)Weld: JJ1 (12Cr-12Ni-10Mn-5Mo-0.13N)Method: TIG weld

Unwelded

WeldedWelded

As-weld notch

CT specimen (As-weld notch) CT specimen (EDM notch)Test condition: Load control Frequency: 10HzStress ratio: 0.1

a) b)

Challenge (iii) PPW crack initiation

12,698

1,898

As-weld notch (no artificial treatment)

EDM notch (R0.1mmis remained on edge.)

=> Behavior like Crack => Propagation in weld

11

Challenge (iii) PPW crack growth

CT specimen (Weld metal/EDM notch)

d)FEM analysis and assessment allowable weld joint and maximum initial defect size

• FEM analysis (total 133 weld joints) • Allowable maximum initial defect size =>100mm2 of semi-elliptical at root

① Slow crack growth ③Application of common region to assessment

①

②③

c) Crack growth rate parameter

②Stress redistribution when sampling

Method of “Design by analysis” for PPW was successfully established! 11

Challenge (iv) Ultrasonic Testing (UT)• Principle of Ultrasonic Testing

12

Transducer

Crack

CrackEcho

Back surfaceEcho

Initial pulse

Attenuation of weld metal was evaluated.

Ultrasonicwave

Weld metal

Basemetal

Wavedecreasesin the weld

Challenge (iv) UT attenuation compensate method• DAC curves* were prepared by

– Calibration block using base metal– Reference block including weld metal

13*: DAC curves (Distance Amplitude Characteristic curves)

1. Making DAC curve of the base metal using a calibration block

Plotted point

Distance

3. Calculating the sensitivities difference (A, A’, …) between each plotting point and DAC curve

Sensitivities difference

Calibration block

2. Plotting detected echo through the weld metal using a reference block with SDH

Reference block The difference were quantified.

Challenge (iv) UT attenuation compensation factor• Ultrasonic attenuation (dB) vs. Weld metal path(mm)

“y = ax” (a: 0.2 ~ 0.4)

14

Weld metal attenuation of TFCS was Properly corrected.

y = 0.313x

y = 0.395x

y = 0.216x

0

5

10

15

20

25

0 20 40 60 80 100

Sens

itivi

ty d

iffer

ence

Weld Metal Path (mm)

Company ACompany BCompany C

15

Challenge (iv) UT for PPWEstablishment inspection method for PPW

Noise!

• High quality weld joint ← Inside defect inspection• Weld depth& initial crack size ← weld depth Confirmation

Ultrasonic Testing (UT) method

• Noise near root → Low accuracy on depth measuring?• Verification test by actual size PPW mock up.

UT procedure was defined.

+/-1mm accuracy for depth measuring.

Challenge (v) Fitting Test

16

AP/BPAU/BU

Gap

Mis-alignment

Gap: 0.5±0.25mmMisalignment:±0.3mm

Req. groove tolerances (inner)

Misalignment: ±1.3mmReq. groove tolerances (outer)

AU/BU

AP/BP

Req. 2): AU-BU

Gap: 0.5±0.25mmMisalignment:±0.7mm

Gap: 0.5±0.25mmMisalignment:±0.3mm

Gap: 0.5±0.25mmMisalignment:±0.3mm

Back plate

Side plate

Side plate

Req. 1): AU-AP, BU-BP

Upper

Lower

- Strict alignment accuracy requested on welding groove to assure welding quality- Actual AU, AP, BU, BP were tested

Challenge (v) Fitting Test for AU-AP/BU-BP

17

Nylon wedges

Stainless Steelwedges

Hydraulic Jack

ProtectionJig

Jigs used to fit up AP to AUAU and AP after fit-up

• Difficulty: To control the precise position of AP and BP.To control the flexibility of AP and BP shape.

• Solution: Several types of guide jigs to adjust their position.Lever hoist to control their axial flexibility

• Result: Tests were successfully completed.

Challenge (v) Fitting Test for AU-BU

18

Aligned grooves

AU and BU after fitting up for the first EU products tested in horizontal position

AU and BU after fitting upfor the first JA coil tested in vertical position

• Difficulty: To find the exact position of actual AU and BU to achieve the target criteria• Solution: Virtual fitting based on the dimension survey data by laser tracker

Find the optimum position of AU and BU• Result: Tests were successfully completed utilizing the above

CONCLUSION

• Solving the difficult challenges, TFCS became feasible.• Two TFCSs has been completed in 2018, and another one

will be completed soon. • The first Japan-manufacturing TF coil will be assembled

TFCS and WP from the fourth quarter of 2018 as the very first TF Coil of ITER

• All the TF coils will be delivered to ITER in 2021

19

आपका ध्यान के�लए धन्यवाद

20