- Email: [email protected]
Manufacture and test of seismic bellows for ITER magnet feeder
G Model
ARTICLE IN PRESS
FUSION-8582; No. of Pages 6
Fusion Engineering and Design xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes
Manufacture and test of seismic bellows for ITER magnet feeder Chen Liu a,∗ , Kun Lu a , Liang Sheng b , Yuntao Song a , Jinjin Su a , Man Su c , Chenyu Gung c a
Institute of Plasma Physics Chinese Academy of Sciences, ShuShanhu Road No. 350, Hefei, China AEROSUN-TOLA Expansion Joint Co., Ltd., Jiangjun Road No. 199, Nanjing, China c ITER Organization, Route de Vinon-sur-Verdon, CS 90 046, 13067 St. Paul-lez-Durance, France b
h i g h l i g h t s • • • • •
The design of the double bellows was iterated with the results of analysis based on the Expansion Joint Manufacturers Association (EJMA) standard. The seismic bellows was tested with cyclic pressurization of the interlayer space to 2 bars absolute pressure for 5 cycles. 200 cycles of tensile fatigue test with 90 mm of stretching from the nominal design length. A full tensile test with 315 mm of stretching from its nominal length was conducted. The prototype bellows was qualified for its leak tightness (less than 1 × 10−9 Pam3 /s of helium) at all time during the qualification test.
a r t i c l e
i n f o
Article history: Received 29 July 2015 Received in revised form 2 February 2016 Accepted 19 February 2016 Available online xxx Keywords: Seismic bellows Pressure cycle test Fatigue test Leak test
a b s t r a c t This paper presents the key manufacturing and testing processes of the prototype ITER feeder seismic bellows. The design of the double bellows was iterated with the results of analysis based on the Expansion Joint Manufacturers Association (EJMA) standard. Each inner and outer bellows was supported in dedicated molds and formed by a hydraulic pressure machine rated at 800 tons. The double bellows were constructed by welding individual collars to the end flanges. The seismic bellows was tested with cyclic pressurization of the interlayer space to 2 bars absolute pressure for 5 cycles. This was followed by 200 cycles of tensile fatigue test with 90 mm of stretching from the nominal design length. After the mechanical fatigue test, a full tensile test with 315 mm of stretching from its nominal length was conducted. Helium leak tests, with the sensitivity of the helium leak detector set to 1 × 10−9 Pa m3 /s of helium, were performed at different stages of pressure and mechanical tests. The prototype bellows was qualified for its leak tightness at all time during the qualification test. © 2016 Elsevier B.V. All rights reserved.
1. Introduction The ITER magnet feeder system, which is the life-line to support and control the superconducting magnet system operated at liquid helium temperature inside the cryostat, supplies electrical currents and supercritical helium from the Tokamak gallery to the magnets, and transmits magnet diagnostic signals to the operators. The feeder cryostat feedthrough (CFT) as one of the major feeder subsystems, penetrates the Tokamak cryostat. The gap is sealed with an interface seismic bellows which is sandwiched by cryostat window flange and CFT terminal flange (Fig. 1) to maintain the main vacuum for the superconducting magnet system. The seismic bellows is constructed with a double bellows structure (Fig. 2) with the intermediate space charged with partial pressure of tracer gas such
as argon, neon, krypton, hydrogen, etc. The material of the bellows is stainless steel 304L which should follow special requirement of ITER. The bellows is designed to isolate the seismic acceleration of the cryostat, which is anchored to the Tokamak foundation, from that of the feeder CFT, which is supported by the floor embedment plates in the Tokamak gallery. The acceptance criteria for the seismic durability of the bellows is vacuum leak tightness before and after the dynamic mechanical loads simulating Seismic Level 2 accelerations of 78 m/s2 in the axial direction and 89 m/s2 in the lateral direction. The bellows also serves as a length compensator for the thermal contraction in the CFT vacuum duct in case of helium leak accident. The leak tightness requirement needs to be met after tensile fatigue cycles followed by an extensive stretching.
∗ Corresponding author. Fax: +86 5515591310. E-mail address: [email protected] (C. Liu). http://dx.doi.org/10.1016/j.fusengdes.2016.02.065 0920-3796/© 2016 Elsevier B.V. All rights reserved.
Please cite this article in press as: C. Liu, et al., Manufacture and test of seismic bellows for ITER magnet feeder, Fusion Eng. Des. (2016), http://dx.doi.org/10.1016/j.fusengdes.2016.02.065
G Model
ARTICLE IN PRESS
FUSION-8582; No. of Pages 6
C. Liu et al. / Fusion Engineering and Design xxx (2016) xxx–xxx
2
Fig. 1. Connection configuration of seismic bellows.
Table 1 Working conditions of feeder seismic bellows.
Temperature Pressure Tensile displacement Compressive displacement Lateral displacement Angular displacement Maximum number of cycles
Condition 1
Condition 2
Condition 3
Condition 4
300 K External: 0.1 MPa Internal: vacuum 5 mm (Assembly) 5 mm (Assembly) 5 mm (Assembly) <1◦ (Assembly) 0 (Static)
300 K External: 0.1 MPa Internal: 0.11 MPa 5 mm (Assembly) 5 mm (Assembly) 5 mm (Assembly) <1◦ (Assembly) <5 (Static)
External: 300 KInternal 4–200 K External: 0.1 MPa Internal: 0.2 MPa <90 mm 5 mm (Assembly) 25 mm <1◦ (Assembly) <5
External: 300 KInternal 4–200 K External: 0.1 MPa Internal: 0.2 MPa <90 mm 5 mm (Assembly) 25 mm <1◦ (Assembly) <5
2. Seismic bellows shaping design 2.1. Working condition introduction Base on the technical specification from ITER, feeder seismic bellows have four different working conditions. Condition 1: normal operation; Condition 2: assembly/break cryostat vacuum; Condition 3: helium leaks in cryostat and in CFT; Condition 4: seismic event + helium leaks in cryostat and in CFT, which is an extremely unlikely rare event. The details are shown in Table 1 (refer to the technical specifications for feeder seismic bellows from ITER [1]). 2.2. Calculation analysis based on EJMA standard EJMA standard is an internationally recognized advanced standard, which is used for almost all bellows standard [2–4].
AEROSUN-TOLA Expansion Joint Co. Ltd., has their own proprietary software (the software interface as shown in Fig. 3) to do the quick design calculation, which follows EJMA 9th standard. The design of the double bellows was iterated with the results of analysis based on EJMA. The calculation results showed that under normal operation condition, the design fatigue life is nearly 40 million cycles. That means during normal operation period, the seismic bellows is impossible destroyed. 2.3. Safety assessment of the bellows under working condition 4 As above, working condition 4 is seismic event + helium leaks in cryostat and in CFT which is the worst condition for seismic bellows. In this condition, the lateral displacement is too big to make the wave shaping of the bellows will “yield” or “will be deformed plastically”, so the EJMA calculation cannot be used because it works
Fig. 2. Feeder seismic bellows.
Please cite this article in press as: C. Liu, et al., Manufacture and test of seismic bellows for ITER magnet feeder, Fusion Eng. Des. (2016), http://dx.doi.org/10.1016/j.fusengdes.2016.02.065
G Model FUSION-8582; No. of Pages 6
ARTICLE IN PRESS C. Liu et al. / Fusion Engineering and Design xxx (2016) xxx–xxx
3
Fig. 3. Design calculation system of bellows.
Fig. 4. Dedicated molds assembly and disassembly.
only in elastic regime. In this case, we just need to confirm the welding seams of the bellows will not be broken. The Eqs. (1)–(4) were used to do safety assessment for seismic bellows. = Fn =
Fn + Fg + Fz S 2 − 2 ) (Dm ×P 4
(1)
(2)
Fg = Fs × D
(3)
Fz = ma
(4)
In these equations, is the weld stress; Fn is internal pressure thrust (13275 N for outer bellows, 9977 N for inner bellows); Fg is spring force (20070 N for outer bellows, 14310 N for inner bellows); Fz is seismic loads (with axial and lateral derections); S is the weld force area (4710 mm2 for outer bellows, 4062 mm2 for inner bellows); Dm is average diameter of bellows; Ф is inside diameter of connection pipe; P is design pressure; Fs is spring rate; D is displacement; m is weight of bellows (60 kg for outer bellows, 55 kg for inner bellows); a is seismic acceleration (axial is 78 m/s2 , lateral is 89 m/s2 , that means should be calculated in two directions for outer and inner bellows).
The calculation results showed that the of outer bellows are 8.1 MPa in axial direction and 12.0 MPa in lateral direction; the of inner bellows are 7.0 MPa in axial direction and 8.0 MPa in lateral direction, which all were far less than 0.577 × 0.45[] = 30 MPa. In this formula, 0.577 is shear stress coefficient, 0.45 is weld joint efficiency (refer ASME-VIII-1), [] is allowable stress which is 115 MPa (refer ASME-II Part D). So welding seams of the bellows will not be broken under seismic condition, there is no leak risk for safety. 3. Key manufacture processes There two key manufacture processes of the seismic bellows, one is hydroforming another is welding. Firstly, the thin plates (1 mm) should be rolled and welded to a cylinder, automatic TIG (Tungsten Inert Gas) welding was used for this linear continuous butt weld. Then each inner and outer bellows was supported in dedicated molds (left picture in Fig. 4) and formed by a hydraulic pressure machine (Fig. 5) with 800 tons press capacity. After the forming process, the molds should be disassembled one by one (right picture in Fig. 4), then the formed outer and inner bellows should be welded together with the connecting collar by manual TIG welding (continuous fillet weld), finally the double bellows structure feeder seismic bellows was finished (Fig. 6).
Please cite this article in press as: C. Liu, et al., Manufacture and test of seismic bellows for ITER magnet feeder, Fusion Eng. Des. (2016), http://dx.doi.org/10.1016/j.fusengdes.2016.02.065
G Model FUSION-8582; No. of Pages 6
ARTICLE IN PRESS C. Liu et al. / Fusion Engineering and Design xxx (2016) xxx–xxx
4
Fig. 6. Completed feeder seismic bellows.
Table 2 Fatigue test requirement.
Fig. 5. The hydraulic pressure machine.
All welds quality on the bellows should refer to ISO 5817 level B, some nondestructive examination (NDE) like penetrant testing, radiographic testing were done for acceptance.
Displacement (mm)
Cycles
Temperature
Internal pressure
Remarks
90
200
Room
<1 Pa
Auto
(Refer to the technical specifications for FEEDER seismic bellows).
4. Qualification test The manufacturing and qualification of the prototype ITER feeder seismic bellows have been completed in October 2014. Base on the ITER feeder components qualification requirements, below qualification test were done.
Table 3 Leak test results for seismic bellows prototype. Test stages
Leak rate(Pa m3 /s)
Leak rate requirement(Pa m3 /s)
After final welding After pressure test After mechanical test
9.3 × 10-12 2.84 × 10-10 2.4 × 10-11
<1 × 10-9 <1 × 10-9 <1 × 10-9
4.1. Pressure test Refer to the working condition shown in Table 1, the biggest pressure change between the internal space of the double bellows and outside is about 0.1 MPa. So we use the gas pressure cycle from 0 to 0.14 MPa (relative pressure) for 5 times. The gas will filled in the internal space (shown in Fig. 7) of the seismic bellows. For safety considerations and observing the big leak point expediently (observe bubbles), the bellows was bathed in water. The configuration of test system as shown in Fig. 7. The pressure test must be done under the design displacement condition (5 mm axial displacement, 5 mm lateral displacement). Since the bellows material is assumed isotropic in EJMA, but actually the bellows material is not. So in order to avoid the experimental deviation, lateral displacement need to be converted into axial displacement in pressure test. The Eq. (5) was used for this calculation. ey =
3Dm y N(Lb + x)
(5)
In Eq. (5), ey is the axial displacement (for tensile) of each wave caused by lateral displacement; Dm is average diameter of the bellows; y is lateral displacement; x is axial displacement; N is the number of waves; Lb is the length of the bellows. Follow the calculation, the initial displacement is 42 mm in this pressure test. The test process onsite as shown in Fig. 8 . 4.2. Mechanical test The fatigue test requirements as shown in Table 2.
The configuration of the test system (for automatic 200 cycles) as shown in Fig. 9. The fatigue test machine PL-60/2000 was used for this test, the assembly process and test curve as shown in Fig. 10. Working condition 4 (seismic event + helium leaks in cryostat and in CFT) is extremely small probability and under this condition, the wave shaping will be severely deformed (in Section 2.3, we already confirm that the welding seams of the bellows will not leak), so the cycle just need one time. For the same reason with pressure test, the lateral displacement also need to be converted into axial displacement in fatigue test. The calculation results shows that the combined axial displacement is about 315 mm. So after the fatigue test, a full tensile test with 315 mm of stretching from its nominal length was conducted.
4.3. Leak test Helium leak tests, with the sensitivity of the helium leak detector set to 1 × 10−9 Pa m3 /s of helium (refer to “Guide to the Supply of Bellows for Use on ITER Vacuum System”, in ITER Vacuum Handbook Appendix 9 [5]), were performed at different stages, include after final welding process, after pressure test and mechanical test. The test results (Table 3) showed that the feeder seismic bellows prototype have an excellent sealing performance, even after waveform destruction mechanical test, the leak rate of the whole bellows was still less than 1 × 10−9 Pa m3 /s.
Please cite this article in press as: C. Liu, et al., Manufacture and test of seismic bellows for ITER magnet feeder, Fusion Eng. Des. (2016), http://dx.doi.org/10.1016/j.fusengdes.2016.02.065
G Model FUSION-8582; No. of Pages 6
ARTICLE IN PRESS C. Liu et al. / Fusion Engineering and Design xxx (2016) xxx–xxx
Fig. 7. Pressure test system configuration.
Fig. 8. Pressure test process.
Fig. 9. Automatic fatigue test system configuration.
Fig. 10. Test platform assembly and test curve.
Please cite this article in press as: C. Liu, et al., Manufacture and test of seismic bellows for ITER magnet feeder, Fusion Eng. Des. (2016), http://dx.doi.org/10.1016/j.fusengdes.2016.02.065
5
G Model FUSION-8582; No. of Pages 6
ARTICLE IN PRESS C. Liu et al. / Fusion Engineering and Design xxx (2016) xxx–xxx
6
5. Conclusions
References
It was shown by EJMA calculation and testing results that this ITER feeder seismic bellows can meet the normal and accidental conditions of ITER operation. The waveform of the bellows will be deformed when the seismic and helium leak accident happen, but even under this worst condition the leak tightness of the bellows is maintained, which satisfies the ITER requirements for Quality Class 1 component. All analysis and qualification test results can certify that this feeder seismic bellows can meet all technical requirements from ITER.
[1] Technical specifications for feeder seismic bellows. Technical documents from ITER IDM UID: A3XDTV. [2] J. Reich, A. Cardella, et al., Experimental verification of the axial and lateral stiffness of large W7-X rectangular bellows, Fusion Eng. Des. 82 (2007) 192–1924. [3] Isoharu Nishiguchi, Shinya Kashiwabara, On the pressure buckling of rectangular bellows for fusion reactors, Fusion Eng. Des. 41 (1998) 323–329. [4] E.N. Kashigin, K.Yu Zershchikov, et al., Design and manufacture of PTFE bellows, Chem. Petrol. Eng. 42 (2006) 7–8. [5] Guide to the Supply of Bellows for Use on ITER Vacuum System ITER Vacuum Handbook Appendix 9, IDM UID: 2E5LJA.
Acknowledgments This work was supported by the project of ‘ITER FEEDER’, and the authors would like to thank all their collaborators involved in the project.
Please cite this article in press as: C. Liu, et al., Manufacture and test of seismic bellows for ITER magnet feeder, Fusion Eng. Des. (2016), http://dx.doi.org/10.1016/j.fusengdes.2016.02.065
ARTICLE IN PRESS
FUSION-8582; No. of Pages 6
Fusion Engineering and Design xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes
Manufacture and test of seismic bellows for ITER magnet feeder Chen Liu a,∗ , Kun Lu a , Liang Sheng b , Yuntao Song a , Jinjin Su a , Man Su c , Chenyu Gung c a
Institute of Plasma Physics Chinese Academy of Sciences, ShuShanhu Road No. 350, Hefei, China AEROSUN-TOLA Expansion Joint Co., Ltd., Jiangjun Road No. 199, Nanjing, China c ITER Organization, Route de Vinon-sur-Verdon, CS 90 046, 13067 St. Paul-lez-Durance, France b
h i g h l i g h t s • • • • •
The design of the double bellows was iterated with the results of analysis based on the Expansion Joint Manufacturers Association (EJMA) standard. The seismic bellows was tested with cyclic pressurization of the interlayer space to 2 bars absolute pressure for 5 cycles. 200 cycles of tensile fatigue test with 90 mm of stretching from the nominal design length. A full tensile test with 315 mm of stretching from its nominal length was conducted. The prototype bellows was qualified for its leak tightness (less than 1 × 10−9 Pam3 /s of helium) at all time during the qualification test.
a r t i c l e
i n f o
Article history: Received 29 July 2015 Received in revised form 2 February 2016 Accepted 19 February 2016 Available online xxx Keywords: Seismic bellows Pressure cycle test Fatigue test Leak test
a b s t r a c t This paper presents the key manufacturing and testing processes of the prototype ITER feeder seismic bellows. The design of the double bellows was iterated with the results of analysis based on the Expansion Joint Manufacturers Association (EJMA) standard. Each inner and outer bellows was supported in dedicated molds and formed by a hydraulic pressure machine rated at 800 tons. The double bellows were constructed by welding individual collars to the end flanges. The seismic bellows was tested with cyclic pressurization of the interlayer space to 2 bars absolute pressure for 5 cycles. This was followed by 200 cycles of tensile fatigue test with 90 mm of stretching from the nominal design length. After the mechanical fatigue test, a full tensile test with 315 mm of stretching from its nominal length was conducted. Helium leak tests, with the sensitivity of the helium leak detector set to 1 × 10−9 Pa m3 /s of helium, were performed at different stages of pressure and mechanical tests. The prototype bellows was qualified for its leak tightness at all time during the qualification test. © 2016 Elsevier B.V. All rights reserved.
1. Introduction The ITER magnet feeder system, which is the life-line to support and control the superconducting magnet system operated at liquid helium temperature inside the cryostat, supplies electrical currents and supercritical helium from the Tokamak gallery to the magnets, and transmits magnet diagnostic signals to the operators. The feeder cryostat feedthrough (CFT) as one of the major feeder subsystems, penetrates the Tokamak cryostat. The gap is sealed with an interface seismic bellows which is sandwiched by cryostat window flange and CFT terminal flange (Fig. 1) to maintain the main vacuum for the superconducting magnet system. The seismic bellows is constructed with a double bellows structure (Fig. 2) with the intermediate space charged with partial pressure of tracer gas such
as argon, neon, krypton, hydrogen, etc. The material of the bellows is stainless steel 304L which should follow special requirement of ITER. The bellows is designed to isolate the seismic acceleration of the cryostat, which is anchored to the Tokamak foundation, from that of the feeder CFT, which is supported by the floor embedment plates in the Tokamak gallery. The acceptance criteria for the seismic durability of the bellows is vacuum leak tightness before and after the dynamic mechanical loads simulating Seismic Level 2 accelerations of 78 m/s2 in the axial direction and 89 m/s2 in the lateral direction. The bellows also serves as a length compensator for the thermal contraction in the CFT vacuum duct in case of helium leak accident. The leak tightness requirement needs to be met after tensile fatigue cycles followed by an extensive stretching.
∗ Corresponding author. Fax: +86 5515591310. E-mail address: [email protected] (C. Liu). http://dx.doi.org/10.1016/j.fusengdes.2016.02.065 0920-3796/© 2016 Elsevier B.V. All rights reserved.
Please cite this article in press as: C. Liu, et al., Manufacture and test of seismic bellows for ITER magnet feeder, Fusion Eng. Des. (2016), http://dx.doi.org/10.1016/j.fusengdes.2016.02.065
G Model
ARTICLE IN PRESS
FUSION-8582; No. of Pages 6
C. Liu et al. / Fusion Engineering and Design xxx (2016) xxx–xxx
2
Fig. 1. Connection configuration of seismic bellows.
Table 1 Working conditions of feeder seismic bellows.
Temperature Pressure Tensile displacement Compressive displacement Lateral displacement Angular displacement Maximum number of cycles
Condition 1
Condition 2
Condition 3
Condition 4
300 K External: 0.1 MPa Internal: vacuum 5 mm (Assembly) 5 mm (Assembly) 5 mm (Assembly) <1◦ (Assembly) 0 (Static)
300 K External: 0.1 MPa Internal: 0.11 MPa 5 mm (Assembly) 5 mm (Assembly) 5 mm (Assembly) <1◦ (Assembly) <5 (Static)
External: 300 KInternal 4–200 K External: 0.1 MPa Internal: 0.2 MPa <90 mm 5 mm (Assembly) 25 mm <1◦ (Assembly) <5
External: 300 KInternal 4–200 K External: 0.1 MPa Internal: 0.2 MPa <90 mm 5 mm (Assembly) 25 mm <1◦ (Assembly) <5
2. Seismic bellows shaping design 2.1. Working condition introduction Base on the technical specification from ITER, feeder seismic bellows have four different working conditions. Condition 1: normal operation; Condition 2: assembly/break cryostat vacuum; Condition 3: helium leaks in cryostat and in CFT; Condition 4: seismic event + helium leaks in cryostat and in CFT, which is an extremely unlikely rare event. The details are shown in Table 1 (refer to the technical specifications for feeder seismic bellows from ITER [1]). 2.2. Calculation analysis based on EJMA standard EJMA standard is an internationally recognized advanced standard, which is used for almost all bellows standard [2–4].
AEROSUN-TOLA Expansion Joint Co. Ltd., has their own proprietary software (the software interface as shown in Fig. 3) to do the quick design calculation, which follows EJMA 9th standard. The design of the double bellows was iterated with the results of analysis based on EJMA. The calculation results showed that under normal operation condition, the design fatigue life is nearly 40 million cycles. That means during normal operation period, the seismic bellows is impossible destroyed. 2.3. Safety assessment of the bellows under working condition 4 As above, working condition 4 is seismic event + helium leaks in cryostat and in CFT which is the worst condition for seismic bellows. In this condition, the lateral displacement is too big to make the wave shaping of the bellows will “yield” or “will be deformed plastically”, so the EJMA calculation cannot be used because it works
Fig. 2. Feeder seismic bellows.
Please cite this article in press as: C. Liu, et al., Manufacture and test of seismic bellows for ITER magnet feeder, Fusion Eng. Des. (2016), http://dx.doi.org/10.1016/j.fusengdes.2016.02.065
G Model FUSION-8582; No. of Pages 6
ARTICLE IN PRESS C. Liu et al. / Fusion Engineering and Design xxx (2016) xxx–xxx
3
Fig. 3. Design calculation system of bellows.
Fig. 4. Dedicated molds assembly and disassembly.
only in elastic regime. In this case, we just need to confirm the welding seams of the bellows will not be broken. The Eqs. (1)–(4) were used to do safety assessment for seismic bellows. = Fn =
Fn + Fg + Fz S 2 − 2 ) (Dm ×P 4
(1)
(2)
Fg = Fs × D
(3)
Fz = ma
(4)
In these equations, is the weld stress; Fn is internal pressure thrust (13275 N for outer bellows, 9977 N for inner bellows); Fg is spring force (20070 N for outer bellows, 14310 N for inner bellows); Fz is seismic loads (with axial and lateral derections); S is the weld force area (4710 mm2 for outer bellows, 4062 mm2 for inner bellows); Dm is average diameter of bellows; Ф is inside diameter of connection pipe; P is design pressure; Fs is spring rate; D is displacement; m is weight of bellows (60 kg for outer bellows, 55 kg for inner bellows); a is seismic acceleration (axial is 78 m/s2 , lateral is 89 m/s2 , that means should be calculated in two directions for outer and inner bellows).
The calculation results showed that the of outer bellows are 8.1 MPa in axial direction and 12.0 MPa in lateral direction; the of inner bellows are 7.0 MPa in axial direction and 8.0 MPa in lateral direction, which all were far less than 0.577 × 0.45[] = 30 MPa. In this formula, 0.577 is shear stress coefficient, 0.45 is weld joint efficiency (refer ASME-VIII-1), [] is allowable stress which is 115 MPa (refer ASME-II Part D). So welding seams of the bellows will not be broken under seismic condition, there is no leak risk for safety. 3. Key manufacture processes There two key manufacture processes of the seismic bellows, one is hydroforming another is welding. Firstly, the thin plates (1 mm) should be rolled and welded to a cylinder, automatic TIG (Tungsten Inert Gas) welding was used for this linear continuous butt weld. Then each inner and outer bellows was supported in dedicated molds (left picture in Fig. 4) and formed by a hydraulic pressure machine (Fig. 5) with 800 tons press capacity. After the forming process, the molds should be disassembled one by one (right picture in Fig. 4), then the formed outer and inner bellows should be welded together with the connecting collar by manual TIG welding (continuous fillet weld), finally the double bellows structure feeder seismic bellows was finished (Fig. 6).
Please cite this article in press as: C. Liu, et al., Manufacture and test of seismic bellows for ITER magnet feeder, Fusion Eng. Des. (2016), http://dx.doi.org/10.1016/j.fusengdes.2016.02.065
G Model FUSION-8582; No. of Pages 6
ARTICLE IN PRESS C. Liu et al. / Fusion Engineering and Design xxx (2016) xxx–xxx
4
Fig. 6. Completed feeder seismic bellows.
Table 2 Fatigue test requirement.
Fig. 5. The hydraulic pressure machine.
All welds quality on the bellows should refer to ISO 5817 level B, some nondestructive examination (NDE) like penetrant testing, radiographic testing were done for acceptance.
Displacement (mm)
Cycles
Temperature
Internal pressure
Remarks
90
200
Room
<1 Pa
Auto
(Refer to the technical specifications for FEEDER seismic bellows).
4. Qualification test The manufacturing and qualification of the prototype ITER feeder seismic bellows have been completed in October 2014. Base on the ITER feeder components qualification requirements, below qualification test were done.
Table 3 Leak test results for seismic bellows prototype. Test stages
Leak rate(Pa m3 /s)
Leak rate requirement(Pa m3 /s)
After final welding After pressure test After mechanical test
9.3 × 10-12 2.84 × 10-10 2.4 × 10-11
<1 × 10-9 <1 × 10-9 <1 × 10-9
4.1. Pressure test Refer to the working condition shown in Table 1, the biggest pressure change between the internal space of the double bellows and outside is about 0.1 MPa. So we use the gas pressure cycle from 0 to 0.14 MPa (relative pressure) for 5 times. The gas will filled in the internal space (shown in Fig. 7) of the seismic bellows. For safety considerations and observing the big leak point expediently (observe bubbles), the bellows was bathed in water. The configuration of test system as shown in Fig. 7. The pressure test must be done under the design displacement condition (5 mm axial displacement, 5 mm lateral displacement). Since the bellows material is assumed isotropic in EJMA, but actually the bellows material is not. So in order to avoid the experimental deviation, lateral displacement need to be converted into axial displacement in pressure test. The Eq. (5) was used for this calculation. ey =
3Dm y N(Lb + x)
(5)
In Eq. (5), ey is the axial displacement (for tensile) of each wave caused by lateral displacement; Dm is average diameter of the bellows; y is lateral displacement; x is axial displacement; N is the number of waves; Lb is the length of the bellows. Follow the calculation, the initial displacement is 42 mm in this pressure test. The test process onsite as shown in Fig. 8 . 4.2. Mechanical test The fatigue test requirements as shown in Table 2.
The configuration of the test system (for automatic 200 cycles) as shown in Fig. 9. The fatigue test machine PL-60/2000 was used for this test, the assembly process and test curve as shown in Fig. 10. Working condition 4 (seismic event + helium leaks in cryostat and in CFT) is extremely small probability and under this condition, the wave shaping will be severely deformed (in Section 2.3, we already confirm that the welding seams of the bellows will not leak), so the cycle just need one time. For the same reason with pressure test, the lateral displacement also need to be converted into axial displacement in fatigue test. The calculation results shows that the combined axial displacement is about 315 mm. So after the fatigue test, a full tensile test with 315 mm of stretching from its nominal length was conducted.
4.3. Leak test Helium leak tests, with the sensitivity of the helium leak detector set to 1 × 10−9 Pa m3 /s of helium (refer to “Guide to the Supply of Bellows for Use on ITER Vacuum System”, in ITER Vacuum Handbook Appendix 9 [5]), were performed at different stages, include after final welding process, after pressure test and mechanical test. The test results (Table 3) showed that the feeder seismic bellows prototype have an excellent sealing performance, even after waveform destruction mechanical test, the leak rate of the whole bellows was still less than 1 × 10−9 Pa m3 /s.
Please cite this article in press as: C. Liu, et al., Manufacture and test of seismic bellows for ITER magnet feeder, Fusion Eng. Des. (2016), http://dx.doi.org/10.1016/j.fusengdes.2016.02.065
G Model FUSION-8582; No. of Pages 6
ARTICLE IN PRESS C. Liu et al. / Fusion Engineering and Design xxx (2016) xxx–xxx
Fig. 7. Pressure test system configuration.
Fig. 8. Pressure test process.
Fig. 9. Automatic fatigue test system configuration.
Fig. 10. Test platform assembly and test curve.
Please cite this article in press as: C. Liu, et al., Manufacture and test of seismic bellows for ITER magnet feeder, Fusion Eng. Des. (2016), http://dx.doi.org/10.1016/j.fusengdes.2016.02.065
5
G Model FUSION-8582; No. of Pages 6
ARTICLE IN PRESS C. Liu et al. / Fusion Engineering and Design xxx (2016) xxx–xxx
6
5. Conclusions
References
It was shown by EJMA calculation and testing results that this ITER feeder seismic bellows can meet the normal and accidental conditions of ITER operation. The waveform of the bellows will be deformed when the seismic and helium leak accident happen, but even under this worst condition the leak tightness of the bellows is maintained, which satisfies the ITER requirements for Quality Class 1 component. All analysis and qualification test results can certify that this feeder seismic bellows can meet all technical requirements from ITER.
[1] Technical specifications for feeder seismic bellows. Technical documents from ITER IDM UID: A3XDTV. [2] J. Reich, A. Cardella, et al., Experimental verification of the axial and lateral stiffness of large W7-X rectangular bellows, Fusion Eng. Des. 82 (2007) 192–1924. [3] Isoharu Nishiguchi, Shinya Kashiwabara, On the pressure buckling of rectangular bellows for fusion reactors, Fusion Eng. Des. 41 (1998) 323–329. [4] E.N. Kashigin, K.Yu Zershchikov, et al., Design and manufacture of PTFE bellows, Chem. Petrol. Eng. 42 (2006) 7–8. [5] Guide to the Supply of Bellows for Use on ITER Vacuum System ITER Vacuum Handbook Appendix 9, IDM UID: 2E5LJA.
Acknowledgments This work was supported by the project of ‘ITER FEEDER’, and the authors would like to thank all their collaborators involved in the project.
Please cite this article in press as: C. Liu, et al., Manufacture and test of seismic bellows for ITER magnet feeder, Fusion Eng. Des. (2016), http://dx.doi.org/10.1016/j.fusengdes.2016.02.065