Structure Design verification with Electrical Dynamic, Thermal Stress and Current Sharing Analysis for CRAFT Converter

doi:10.21203/rs.3.rs-1963213/v1

Download PDF

Research Article

Structure Design verification with Electrical Dynamic, Thermal Stress and Current Sharing Analysis for CRAFT Converter

https://doi.org/10.21203/rs.3.rs-1963213/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

During the design process of fusion power supplies, structure designs is a very important step for safety operation of fusion device. In this paper, the electric dynamic analysis based on the worst fault current and thermal stress analysis with maximum steady temperature are finished, respectively. In addition, a simple research method of current sharing characteristics of converter bridge arm by Q3D modeling with Simplorer is proposed, the results show that the current sharing characteristics are prosperous. Finally, the structural design of CRAFT converter meets the requirements, and the prototype will be manufactured later.

Electric Dynamic Analysis

Thermal Stress

Current Sharing Characteristics

CRAFT

Fusion energy has been widely studied all over the world because of its advantages such as clean and high output energy[1–3]. To promote fusion energy research, China has proposed a Comprehensive Research Facility for Fusion Technology (CRAFT) project, aiming to provide a good experimental environment for the fusion commercialization in China[4, 5]. The project needs a set of power supplies, which mainly composed of four sets of 0.25kV/25kA converter units, for the characteristic test of superconducting magnet[6]. To ensure that the CRAFT converter can meet experimental requirements, it is necessary to verify whether the structure meets the requirements on the basis of topology design and fault analysis[7, 8]. The fusion power supply requires that the converter must withstand the maximum fault current impact without obvious structural damage. The maximum transient impulse current in CRAFT converter is nearly 400kA as shown in Fig. 1[9]. So it is necessary to analyze whether the dynamic stability performance of CRAFT converter meets the standard.

The excessive thermal stress will increase the fatigue damage to the material, shorten the service life and lead to cracks in advance[10–12]. It is necessary to consider whether the thermal stress generated by the temperature during rated operation will damage the converter. When multiple thyristors are operated in parallel, the current sharing characteristics of the converter bridge arm must meet certain requirements. Otherwise some devices may overheat, which will affect the safety of the converter[13–15].

In this paper, firstly, the electric dynamic analysis of soft connector,AC busbar and converter are finished, respectively. Secondly, the thermal stress of single bridge and converter with and without support are verified. Thirdly, based on the inductance matrix, the current sharing characteristic is analyzed by co-simulation between Q3D and Simplorer in ANSYS. Finally, a conclusion is given in the last, and the structure design will be adopted in the future prototype manufacture.

The structural design requirement is that it is not allowed to be damaged at worst short circuit condition. Therefore, the short-circuit faults current at the front end of the converter reactor is taken as a reference. In order to ensure a certain margin, 400kA is selected as the fault current input. Based on this, the force analysis of soft connection, single bridge and whole converter with or without support is verified respectively. And compare whether the yield strength of the material is greater than the stress of the material to judge whether the structure can withstand the impact of fault current. For the convenience of evaluation, the material properties are given as shown in Table 1.

Table 1

Material parameters of each component in CRAFT converter
Material	Yield Strength	Poisson's Ratio	Specific Components
No.45 steel	355	0.3	bolt parts
6061 aluminium alloy	276	0.33	gasket clamp
6063 aluminium alloy	214	0.33	thyristor radiator
6106 aluminium alloy	193	0.33	busbars

2.1 Electric Dynamic Analysis of Soft Connector

To simulate the worst case, it is assumed that only one circuit of thyristor is connected during the fault, that is, 400kA current passes through the soft connector. The simulation results are shown in Fig. 2 and Table 2. The maximum physical density of the soft connector is 0.1N/mm³, the maximum deformation is nearly 0.02mm, and the maximum stress is 1.8MPa.

The deformation of long soft connector is larger than that of short one. The maximum deformation of the two kind of soft connectors occurs at the middle arc.

Table 2

Stress distribution of soft connector under 400kA
Position	Fault Current	Force Direction
Position	Fault Current	X/N	Y/N	Z/N	F_total/N
Long Soft Connector	+ 200kA	-1187.1	7002.7	769.4	7144.2
Short Soft Connector	-200kA	578.4	-7073.1	-2624.6	7566.5

It can be seen form the Fig. 2, force of two soft connectors are inconsistent in all directions, because their function is to connect the thyristor busbar and the fast fuse busbar, which is not completely symmetrical in space. The total force of long soft connector is slightly less than that of short soft connector, but its deformation is greater than that of the latter, because under the same vault mode, the former is longer and more prone to deformation.

2.2 electric dynamic Analysis of AC Busbar

Assuming that the fault current of 400kA is taken as the total input by phase A, the AC busbar of phase B and phase C bears half of it, respectively. When taking this as the input condition for analysis, only the busbar part is considered, and the other front and rear stages influence is ignored to analyze the stress of AC busbar at this time. As shown in Fig. 4, the identification method here is the same as the previous text, which is called A1, A2, B1, B2, C1 and C2 phase AC busbar from left to right. It can be seen form Fig. 3, the maximum physical density is 0.003N/mm³, the maximum deformation is about 0.01mm, and the maximum stress is 1.34MPa.

Table 3

Stress of three-phase AC busbar at 400kA
No	Fault Current	X/N	Y/N	F_total/N
A1	+ 200kA	30343.7	402.9	30346.4
A2	-200kA	29954.8	370.7	29957.1
B1	+ 100kA	7449.6	44.5	7449.8
B2	-100kA	8069.3	29.3	8069.4
C1	+ 100kA	7367.0	3.3	7367.0
C2	-100kA	7430.6	68.7	7430.9

It can be seen from Table 3 that in A, B and C three-phases, the global stress of each adjacent busbar is basically same. Among them, phase A bears the largest force because its passing current is twice as large as that of the other two phases. The force of phase B is slightly larger than that of phase C, because it is closer to phase a, so it is more affected by phase A. At the same time, the force on each phase busbar in the Z direction is 0. This is because the six busbars are placed in parallel in space, and the current direction are parallel, so only the XY direction is stressed.

Of course, the above analysis only assumes that phase A is the total fault current input. In fact, phase A, phase B and phase C may be the total fault current input, so this assumption is applicable to the maximum stress condition of three-phase AC busbar, that is, each busbar must have the maximum stress condition that can withstand this time.

After analysis, when the converter is short circuited at the front end of the reactor, the stress of AC busbar is less than the material yield strength, which meets the design requirements.

2.3 electric dynamic Analysis of Converter

The simulation results of the converter with support are shown in Fig. 4. The maximum deformation is nearly 0.15mm. The maximum stress is 12.2MPa. The maximum force is at phase A support of three phase AC busbar. The maximum force of the converter is nearly 5.0MPa

After analysis, when the converter is short circuited at the front end of the reactor, the stress of converter is less than the material yield strength, which meets the structure design requirements.

Although the converter is cooled by deionized water in practice, it will still have a certain temperature rise. The temperature variation will lead the change of material properties, and then cause certain thermal stress. When there are no external constraints, the material will expand freely, while when there are constraints, it will lead to the increase of thermal stress. The thermal stress changes of a single bridge arm and the whole converter with or without support are analyzed below. For the convenience of analysis, the overall temperature of the converter is set to 40℃, and the temperature of all soft connections is set to 70℃. The analysis results are as follows.

3.1 thermal stress analysis of single bridge

As shown in Fig. 5, the maximum equivalent stress of a single bridge arm without support is 67.2MPa, and the maximum equivalent stress with support is 88MPa. The maximum equivalent stress of a single bridge arm is 1.3mm without support and 0.83mm with support. The maximum equivalent stress of a single bridge arm without support is 36.3MPa, and the maximum equivalent thermal stress with support is 50.6MPa.

3.2 General Thermal Stress Analysis of Converter

Similarly, the thermal stress of converter with or without support is analyzed in this part. When there is no support, only the AC incoming line is fixed. When the device is support, the whole converter is supported by epoxy resin. As shown in Fig. 6, the maximum equivalent stress of the converter is 67.2MPa without support and 91.5MPa with support. The maximum equivalent stress of the converter without support is 1.3mm, and the maximum equivalent stress with support is 0.83 mm. The maximum equivalent stress of the converter without support is 36.3MPa, and the effect is equal when there is supported. The maximum force is 91.5MPa. The thermal stress of the whole craft converter are less than the yield strength of aluminum, which meets the design requirements.

Through analysis, it is found that the thermal stress of a single bridge arm increases when there is supported, because the support limits the free expansion of the bridge arm at this time. The overall thermal stress is less than the yield strength of the material, meeting the design requirements.

The bridge arm of the converter is composed of multiple thyristors in parallel, and each thyristor is connected in series with a fast fuse through soft connector, and then converged to the fast fuse busbar. The bridge arm of the converter adopts the AC side to flow current from the lower end, and outputs current through the upper end of the DC side after rectification. During steady-state operation, because the current paths of each thyristor branch are different, their respective currents are not exactly same, so it is necessary to verify the current sharing effect of all branches. The current sharing effect is determined by the current sharing coefficient (k_unb), which is defined as:

\({K_{unb}}=\frac{{{I_{\hbox{max} }} \times {N_p}}}{{{I_{total}}}}\)

Where I_MAX is the maximum branch current, N_P is the number of branches in parallel, and I_total is the total output current. When k_unb is less than 1.2, it meets the engineering application conditions.

When multiple devices and branches are connected in parallel, each branch contains not only its own resistance and inductance, but also the resistance and inductance parameters caused by the coupling of different branches. Therefore, when the bridge arm of the converter is equivalent to a circuit by structure, it will be very complex. In this paper, for convenience, Co-simulation between Q3D and Simplorer in ANSYS is used for current sharing analysis of the converter.

Because the CRAFT converter is a three-phase six pulse bridge structure, one of the phases is selected for analysis. The bridge arm connection mode is in-phase inverse parallel structure, and each bridge arm is divided into upper and lower parts. In each work cycle, only half of the thyristors will work. For simplicity, half of the bridge arm is chosen for calculation. It is divided into three parts as shown in Fig. 7 for equivalence. Part 1: AC incoming line to thyristor contact surface; Part 2: soft connection part; Part 3: contact surface of quick fuse to DC busbar. And all thyristors are installed in front and back sides. Here, for illustration convenience, the imprinted surface of front devices and busbar is set as a red solid line, and the imprinted surface of back devices and busbar is set as a black dotted line. It should be pointed out that this is just for the convenience of identification and illustration. In practice and simulation, the marking surface of the black dotted line identification is taken on the other side of each branch.

The inductance matrix of each equivalent branch obtained after Q3D analysis is shown in Table 4, Table 5 and Table 6. Among these tables, Th 1 to Th 6 represents thyristor 1–6, Fuse 1 to Fuse 6 represents fast fuse 1–6, Sh 1 to Sh 3 represents the black identification branch in Fig. 7 (c), and Lo 4 to Lo 6 represents the red identification branch in Fig. 7 (c).

Table 4

Inductance matrix of each thyristor branch of Part 1 (uH)
	Th 1	Th 2	Th 3	Th 4	Th 5	Th 6
Th 1	0.24889	0.25104	0.25208	0.1397	0.14195	0.14294
Th 2	0.25104	0.2786	0.28097	0.14185	0.14674	0.149
Th 3	0.25208	0.28097	0.31024	0.1429	0.14902	0.15404
Th 4	0.1397	0.14185	0.1429	0.24884	0.2511	0.25208
Th 5	0.14195	0.14674	0.14902	0.2511	0.27865	0.28118
Th 6	0.14294	0.149	0.15404	0.25208	0.28118	0.31011

Table 5

Inductance matrix of each thyristor branch of Part 2 (uH)
	Sh 1	Sh 2	Sh 3	Lo 4	Lo 5	Lo 6
Sh 1	70.84209	14.24596	2.71798	0.75111	1.20906	1.65073
Sh 2	14.24596	127.8696	14.00963	2.62073	2.95455	2.09168
Sh 3	2.71798	14.00963	71.09683	2.05487	1.18888	0.59476
Lo 4	0.75111	2.62073	2.05487	79.21759	10.11542	2.32921
Lo 5	1.20906	2.95455	1.18888	10.11542	77.76124	8.72899
Lo 6	1.65073	2.09168	0.59476	2.32921	8.72899	68.72007

Table 6

Inductance matrix of each thyristor branch of Part 3 (uH)
	Fuse 1	Fuse 2	Fuse 3	Fuse 4	Fuse 5	Fuse 6
Fuse 1	744.3998	711.6131	675.5905	376.3765	367.8866	355.9346
Fuse 2	711.6131	704.8263	670.2677	367.8052	361.7508	350.9767
Fuse 3	675.5905	670.2677	661.2155	356.3332	351.3928	343.0532
Fuse 4	376.3765	367.8052	356.3332	745.7004	712.9019	675.3547
Fuse 5	367.8866	361.7508	351.3928	712.9019	706.0777	670.062
Fuse 6	355.9346	350.9767	343.0532	675.3547	670.062	660.5837

Based on the stray inductance parameters of each divided branch extracted by Q3D, a single converter bridge arm model is established in Simplorer in ANSYS. Each partition module can be automatically encapsulated into a sub circuit with input and output. The model is used to analyze the current sharing effect of the parallel branch in the bridge arm, and the results are shown in Fig. 8,where, B1-B6 means the branch 1 to branch 6.

The simulation results show that the current sharing effect of the six branches of the bridge arm is very good (k_unb=1.05 < 1.2), which meets the design requirements. However, the triggering time and closing time of different branches are not exactly the same, which is due to the different stray parameters of each branch. The maximum error time is less than 200us, which is acceptable in practical application.

To verify the structure design reliability of CRAFT converter, detailed electric dynamic impact analysis and thermal stress analysis are carried out in this paper. It is verified that the converter can withstand the maximum fault current and the worst steady-state temperature without causing structural damage. At the same time, the inductance of the bridge arm of the converter is extracted by Q3D and combined simulation with Simplorer in ANSYS. The results show that the current sharing coefficient of the bridge arm is 1.05, which meet the design requirements.

Acknowledgments

The authors will thank for the help of Institute of Plasma Physics Chinese Academy of Sciences (ASIPP) in the fusion power converter design process. This work was supported by the University Natural Sciences Research Project of Anhui Province, No. KJ2021A0981 and No. KJ2021A0979, Talent Research Fund of Hefei University No. 21-22RC09, and Excellent Young Talents in Universities of Anhui province, No. gxyq2021229.

R. Kembleton, A.W. Morris, G. Federici, A.J.H. Donné, Design Issues for Fusion Commercialization[J]. IEEE Trans. Plasma Sci. 48(6), 1703–1707 (2020)
A. Taherigaravand, A. Nassiri, S. Gharibzahedi, Electromagnetic and mechanical characterisation of ITER CS-MC conductors affected by transverse cyclic loading. III. Mechanical properties[J]. IEEE Trans. Appl. Supercond. 9(2), 1069–1072 (2017)
J. Chen, D. Brower, W. Ding et al., Correlation polarimeter-interferometer in the DIII-D tokamak[J]. Rev. Sci. Instrum. 92(4), 043502 (2021)
J. He, J. Wu, W. Wen et al., Analysis and Verification of CRAFT TF Coil Turn Insulation Wrapping System[J]. IEEE Trans. Appl. Supercond. 31(7), 1–5 (2021)
J. Xue, J. Wu, W. Wen, Z. Yan et al., Subwinding Pack Embedding Experiment of CRAFT TF Coil[J]. IEEE Trans. Appl. Supercond. 32(7), 1–4 (2022)
R. Huang, Y. Zhu, P. Fu et al., R esearch on Short Circuit Operation Control Strategy of a Superconducting Magnet Power Supply for CRAFT[J]. J. Fusion Energy 40(21), 1–9 (2021)
P. Wang, Z. Song, C. Li, el at. Prototype Design of External Bypass for ITER Poloidal Field Converter[J]. IEEE Transactions on Plasma Science, 2016, 44(9): 1525–1528
Z. C.li, P.Fu Song et al., Dynamic stability analysis of high -power DC reactor for ITER poloidal field converter[J]. J. Fusion Energy 34(2), 202–206 (2015)
Z. Wang; Fu Peng; Hang Liansheng; et al. Prelim inary design of high power magnet converter for CRAFT[J]. Plasma Science & Technology, 2020, 22(4): 1–17
O.B. Shagniev, V.S. Modestov, A.A. Listopad et al., Modelling of the stress-strain state of ITER upper ports ##02 & 08 constructions' components under thermal, electromagnetic and seismic loads[J]. Fusion Eng. Des. 146(Sep.), 2421–2425 (2019)
Y. Zhang, T. Lu, Unsteady-state thermal stress and thermal deformation analysis for a pressurizer surge line subjected to thermal stratification based on a coupled CFD-FEM method[J]. Ann. Nucl. Energy 108(Oct.), 253–267 (2017)
S.L. Cha, Y. Lee, J.K. Kim, Thermal stress analysis of silo in radioactive waste repository considering construction conditions[J]. Nucl. Eng. Des. 322, 412–426 (2017)
X. Zhang, L. Huang, X. Chen et al., Fault analysis and solution in current sharing test of ITER PF converter bridge[J]. Fusion Eng. Des. 167(4), 112328 (2021)
P. Wang, Z. Song, Thermal design of external bypass for ITER poloidal field converter[J]. Fusion Eng. Des. 133(Aug.), 130–134 (2018)
Y. Xiao, Y. Wu, Y. Wu et al., High-capacity MVDC Interruption Based on IGCT Current Sharing Integrated with Freewheeling[J]. IEEE Trans. Industr. Electron. PP(99), 1–1 (2020)

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Structure Design verification with Electrical Dynamic, Thermal Stress and Current Sharing Analysis for CRAFT Converter

Status:

Version 1

Abstract

Figures

1 Introduction

2 Converter Electric Dynamic Analysis

2.1 Electric Dynamic Analysis of Soft Connector

2.2 electric dynamic Analysis of AC Busbar

2.3 electric dynamic Analysis of Converter

3 Thermal Stress Analysis Of Converter

3.1 thermal stress analysis of single bridge

3.2 General Thermal Stress Analysis of Converter

4 Current Sharing Analysis Of Converter

5 Conclusion

Declarations

Acknowledgments

References

Additional Declarations

Status:

Version 1