Terahertz detection system and method for insulating layer of armored superconducting coil for nuclear fusion

Abstract

The invention discloses a terahertz detection system and method for an armored superconducting coil insulating layer for nuclear fusion, and belongs to the technical field of superconducting coil insulating structure detection, the defect depth d and dielectric constant change delta epsilon r are obtained through terahertz time-domain spectroscopy (THz-TDS), magnetic field disturbance delta B is measured in combination with a superconducting quantum interferometer (SQUID), and a quantitative relation model delta B = k * delta epsilon r / d2 is established. The system integrates a three-dimensional chromatography algorithm and a data fusion technology, and outputs a full-parameter detection report of defect position, size and dielectric loss. According to the method, synchronous detection of micron-grade defect positioning and micro Tesla-grade magnetic measurement is achieved, the depth measurement error is smaller than or equal to 0.02 mm, the detection efficiency is improved by 15 times compared with a traditional method, insulation structure defect evaluation of the superconducting coil can be achieved, and the problem that in Paschen discharge detection, high voltage breaks down a superconducting coil insulator structure is effectively avoided.

Description

Technical Field

[0001] This invention relates to the field of superconducting coil insulation structure testing technology, specifically to a terahertz testing system and method for the insulation layer of armored superconducting coils used in nuclear fusion, applicable to scenarios such as electrical performance testing of the insulation structure of superconducting magnet coils in nuclear fusion devices. Background Technology

[0002] As superconducting magnet systems in nuclear fusion devices (such as ITER) develop towards higher field strength and larger current, the cyanate ester-epoxy resin composite insulation layer (8-10 mm thick) prepared by vacuum pressure impregnation (VPI) faces the challenge of detecting defects such as micron-level air gaps and delamination. Traditional ultrasonic testing is limited by wavelength (≥100 μm) and signal-to-noise ratio, and is not suitable for 5×5 mm... 2 The following defects have an identification rate of less than 60% and cannot be quantified for changes in dielectric properties. In addition, the contradiction between the area array imaging speed and the area of ​​a single scan, as well as the aliasing interference of reflected signals from multilayer dielectric interfaces, make signal-to-noise ratio control more difficult in industrial environments. Summary of the Invention

[0003] To address the technical challenges of "difficult micro-defect identification, difficult dielectric property quantification, and low detection efficiency" in the detection of insulation defects in armored superconducting coils used in nuclear fusion, this invention provides a terahertz detection system and method for the insulation layer of armored superconducting coils used in nuclear fusion. This system establishes a quantitative mapping relationship between the geometric characteristics of insulation defects and their electromagnetic responses through the collaborative detection of a THz-TDS detection module, terahertz time-domain spectroscopy, and a superconducting quantum interference device (SQUID). It also achieves data fusion by using a depth-magnetic field perturbation model to integrate the detection results of two different physical quantities: dielectric measurement (THz) and magnetic measurement (SQUID).

[0004] Terahertz time-domain spectroscopy (THz-TDS) technology can penetrate nonpolar materials in the 0.1-3 THz band, providing sub-millimeter spatial resolution with wavelengths (100-300 μm) while avoiding ionizing damage. A center frequency of 0.8 THz achieves an optimized balance between penetration depth (10 mm) and resolution (20 μm) in cyanate ester-epoxy resins. Simultaneously, the defect depth d is calculated by measuring the reflected pulse delay Δt using the time-of-flight (ToF) method, with a measured error <0.1 mm. This invention enables the air gap to induce a local dielectric constant ε. r Decreases by 0.8 (3.2→2.4), reflection coefficient R∝Δε r This improves signal contrast by 3 times; simultaneously, the inverse Radon transform algorithm is used to reconstruct the time-domain signal into voxel data, achieving 20μm... 3Precision tomographic imaging. Therefore, this invention, through THz-TDS pre-positioning technology and the combination of broadband pulse scanning and three-dimensional tomographic imaging, provides the first solution for ITER-level superconducting coils that combines quantitative analysis and efficient detection.

[0005] This invention obtains the relationship between defect depth d and dielectric constant change Δε using terahertz time-domain spectroscopy (THz-TDS). r By combining the measurement of magnetic field perturbation ΔB using a superconducting quantum interference device (SQUID), a quantitative relationship model was established:

[0006] ΔB=k·Δε r / d 2 The system integrates 3D tomography algorithms and data fusion technology to output a full-parameter detection report on defect location, size, and dielectric loss. This invention achieves simultaneous detection of micron-level defect location and micro-Tesla-level magnetic measurement, with a depth measurement error ≤0.02mm. It improves detection efficiency by 15 times compared to traditional methods, enabling the assessment of insulation structure defects in superconducting coils and effectively avoiding the problem of high-voltage breakdown of the superconducting coil insulation structure in Paschen discharge detection.

[0007] The principles involved in this invention include:

[0008] 1. Principle of THz-TDS detection layer:

[0009] When a 0.8 THz pulse (wavelength 375 μm) propagates in an insulating material, it will produce Fresnel reflections and group velocity changes (time delays) when encountering defects such as air gaps / delamination. By measuring the amplitude attenuation ΔA and time delay Δt of the reflected pulse, the local dielectric constant change Δε can be calculated. r .

[0010] An improved time-domain back projection tomographic inversion algorithm is employed to convert the time-domain signal into a spatial distribution, achieving a 20μm resolution. 3 Precision defects in 3D imaging.

[0011] 2. SQUID detection layer principle:

[0012] Defects in the insulation layer of a superconducting coil can cause distortion in the current distribution, generating a magnetic field disturbance ΔB. This disturbance is converted into a measurable current signal through the phase difference Δφ of the Josephson junction and is eventually captured by the SQUID.

[0013] DC-SQUIDs were fabricated using YBa2Cu3O7 superconducting thin films, with each unit containing a double Josephson junction (critical current I0). c =20μA), flux transformer (coupling coefficient k=0.85) and gradient meter structure (common mode rejection ratio>60dB), with an operating temperature of 77K.

[0014] 3. Data fusion model:

[0015] Based on Maxwell's equations and London's equations, a quantitative relationship between defect depth d and magnetic field disturbance ΔB is derived:

[0016] ΔB=k·Δε r / d 2 ;

[0017] Where k is the electromagnetic coupling coefficient of the material, in T·m 2 The conductivity and permeability of the insulating material, as well as the geometry of the coil, are determined by these properties.

[0018] The SIFT feature matching algorithm was used to achieve spatial alignment between THz and SQUID data, with a registration error of <5μm.

[0019] 4. A three-level linkage mechanism of triggering, data acquisition, and registration is adopted.

[0020] To achieve the above objectives, the present invention adopts the following technical solution:

[0021] A terahertz detection system for the insulating layer of an armored superconducting coil used in nuclear fusion includes:

[0022] THz-TDS detection module, where THz stands for terahertz and TDS stands for time-domain spectrum;

[0023] Superconducting quantum interference device detection array;

[0024] A synchronization controller is used to trigger the THz-TDS detection module and the superconducting quantum interference device detection array to achieve timing alignment.

[0025] The data processing unit is used to receive signals from the THz-TDS detection module and the superconducting quantum interference device detection array, perform three-dimensional tomographic reconstruction and data fusion, and output a defect detection report.

[0026] Furthermore, it also includes:

[0027] The base has an electric rotary support platform fixedly installed on its top. A rotary support column is fixedly installed on the electric rotary support platform. An electric rotary gear sleeve is sleeved on the outside of the rotary support column. The electric rotary gear sleeve is connected to the superconducting coil support base through a square support bushing. The superconducting coil support base and the superconducting coil intermediate support frame are reinforced and connected by a first reinforcing rib, a second reinforcing rib, and a third reinforcing rib.

[0028] The first superconducting coil and the second superconducting coil are respectively mounted on the superconducting coil support base and the superconducting coil intermediate support frame;

[0029] An embedded movable electric drive slider is installed on the inner wall of the vacuum chamber. A lifting rod is fixedly installed on the embedded movable electric drive slider, and the top of the lifting rod is connected to the terahertz time-domain spectroscopy detection module.

[0030] A cylinder is fixed to one side of the superconducting coil support base, and the piston rod of the cylinder is connected to the side wall of the vacuum chamber.

[0031] Furthermore, a vacuum port and a gas source port are provided on the side wall of the vacuum chamber. The vacuum port is connected to a vacuum pump assembly through a vacuum pump connecting pipe, and the gas source port is connected to a gas source control system through a gas source pipe. The superconducting quantum interference device detection array is fixed on the top of the vacuum chamber.

[0032] Furthermore, the synchronization controller includes a high-precision pulse generator, a power supply control system, and a 10ns precision time synchronizer. The high-precision pulse generator triggers the terahertz time-domain spectroscopy detection module, the power supply control system supplies rated current to the first superconducting coil and the second superconducting coil, and the 10ns precision time synchronizer aligns the signals of the terahertz time-domain spectroscopy detection module with the superconducting quantum interference instrument detection array. After spatial registration is completed by the SIFT algorithm, the signals are input into the three-dimensional reconstruction unit.

[0033] Furthermore, the terahertz time-domain spectroscopy detection module emits broadband pulses at a center frequency of 0.8THz and scans the coil surface along the embedded movable electric drive slider at a step size of 10μm to collect defect reflection signals. The time-domain signals are reconstructed into voxel data using the inverse Radon transform algorithm to achieve three-dimensional tomographic imaging.

[0034] Furthermore, the superconducting quantum interference device (SQUID) detection array is a DC-SQUID fabricated from a YBa2Cu3O7 superconducting thin film. Each unit contains a double Josephson junction and operates at a temperature of 77K. The data processing unit establishes ΔB=k·Δε based on the Biot-Savart law and Maxwell's equations. r / d 2 A depth-magnetic field quantitative model is established, where k is the electromagnetic coupling coefficient of the material, which is determined through calibration experiments. The SIFT algorithm achieves spatial alignment between terahertz data and superconducting quantum interference device data, with a registration error of less than 5 μm, and a ΔB-Vt three-dimensional database is established for maintenance decision-making.

[0035] This invention also provides a method for detecting defects in the insulation layer of armored superconducting coils used in nuclear fusion, implemented using the aforementioned terahertz detection system for the insulation layer of armored superconducting coils used in nuclear fusion, comprising the following steps:

[0036] The surface of the superconducting coil after vacuum pressure impregnation and curing is polished while controlling the ambient humidity to be below 5%RH.

[0037] The THz-TDS detection module is triggered to emit a broadband pulse to scan the surface of the coil. The power supply control system is simultaneously triggered to supply the rated current to the superconducting coil. The superconducting quantum interference instrument detection array is started to measure the magnetic field distribution. Timing alignment is achieved through a 10ns precision time synchronizer. Spatial registration is completed through the SIFT algorithm. The input to the three-dimensional reconstruction unit generates a preliminary defect distribution map.

[0038] Constructing a depth-magnetic field quantitative model ΔB=k·Δε r / d 2 The electromagnetic coupling coefficient k was determined through calibration experiments, and the change in dielectric constant Δε obtained from terahertz detection was used to determine the change in dielectric constant. r The theoretical magnetic field perturbation ΔBcal was calculated by substituting the defect depth d into the model, and the error rate was calculated by comparing it with the measured magnetic field perturbation ΔBmeas of the superconducting quantum interference device.

[0039] Based on the threshold set by the change in dielectric constant and magnetic field disturbance, the area of ​​insulation performance degradation is marked, and a full-parameter test report including defect location, size, change in dielectric constant, magnetic field disturbance and risk level is output.

[0040] Furthermore, the THz-TDS detection module scans along the embedded movable electric drive slider at a 10μm step size with a center frequency of 0.8THz to collect the amplitude attenuation and time delay of the defect reflection signal. The inverse Radon transform algorithm is used to reconstruct the time domain signal into voxel data, realizing three-dimensional tomographic imaging with a voxel accuracy of 20μm.

[0041] Furthermore, the calibration experiment used cyanate ester-epoxy resin composite insulation layer samples, with gradient air gap defects of 0.1-2 mm in diameter and 1-10 mm in depth implanted. The relationship between the defect depth d and the dielectric constant change Δε was measured using a terahertz time-domain spectroscopy detection module. r The corresponding magnetic field disturbance ΔB was measured by a superconducting quantum interference device detection array. The electromagnetic coupling coefficient k was fitted by the least squares method. The experiment was repeated 20 times and the mean was taken. The variance was ≤ ±5%.

[0042] Furthermore, in the quantitative assessment of defects, the change in dielectric constant Δε is used as the criterion. r >0.05 is used as the criterion to mark the area of ​​insulation performance degradation. The magnetic field disturbance ΔB>0.5μT is set as the warning threshold to identify high-risk defects. The full parameter report in XML format is output, and a defect distribution heat map and magnetic field gradient line visualization image are generated. A ΔB-Vt three-dimensional database is established for maintenance decision-making.

[0043] Beneficial effects:

[0044] 1. Significant improvement in multi-physics field collaborative detection accuracy: Through THz time-domain spectroscopy (spatial resolution 20 μm) and SQUID quantum magnetic measurement (sensitivity 10 μm), the accuracy of detection has been greatly improved. -15The integration of ITER and CT enables dual verification of geometric location and electromagnetic validation of insulation defects. Experimental data show that compared with a single detection method: the defect false positive rate is reduced by 87% (ITER test data); the dielectric constant measurement accuracy is improved to ±0.01 (traditional method ±0.1); and the depth measurement error is compressed to ±0.02mm (industrial CT is ±0.1mm).

[0045] 2. The invention pioneers a quantitative model for depth-magnetic field (ΔB=k·Δε). r / d 2 A mapping relationship between defect depth d and magnetic field disturbance ΔB was established, and the repeatability error of the k value determined by calibration experiments was <3%, providing a theoretical tool for predicting the lifetime of superconducting devices.

[0046] 3. This invention employs a GPU-accelerated inverse Radon transform algorithm, achieving a reconstruction speed 15 times faster than traditional CPU computation. It supports real-time 3D imaging with a voxel precision of 20μm and can identify pixels as small as 0.1mm. 3 Micro-gaps (lower limit of conventional ultrasonic detection: 1 mm) 3 ).

[0047] 4. Nuclear fusion device safety assurance: Provides early warning of ITER-level coil insulation degradation up to 6 months in advance (verified by the French Atomic Energy Commission), avoiding losses from a single accidental shutdown (estimated savings). 2.3M / time).

[0048] 5. This invention establishes a ΔB-Vt three-dimensional database, improving the accuracy of maintenance decisions by 92%. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of the main structure of the terahertz detection system for the insulating layer of the armored superconducting coil used in nuclear fusion according to the present invention;

[0050] Figure 2 This is a schematic diagram showing the position markings of the vacuum port and gas source port in the detection system of the present invention;

[0051] Figure 3 This is a schematic diagram showing the position markings of the gas source pipe in the detection system of the present invention;

[0052] Figure 4 This is a schematic diagram of the synchronous control architecture and multimodal scanning process of the present invention.

[0053] The attached diagram is labeled as follows: 1-Base; 2-Electric rotary support platform; 3-Vacuum pump connecting pipe; 4-Vacuum pump assembly; 5-Signal and power feedback channel; 6-Gas source control system; 7-Rotating support column; 8-Electric rotating gear sleeve; 9-Square support bushing; 10-Superconducting coil support base; 11-First superconducting coil; 12-Second superconducting coil; 13-Second superconducting coil inlet; 14-First connecting cable; 15-First superconducting coil outlet; 16-First connector; 17-Power supply control system; 18-Second connector; 19-Second superconducting coil... 20-Coil outlet; 21-DC / AC insulation withstand voltage tester; 22-Second connecting cable; 23-Third connector; 24-First superconducting coil inlet; 25-Cylinder telescopic connecting shaft; 26-Third connecting cable; 27-Superconducting coil intermediate support frame; 28-First reinforcing rib; 29-Third reinforcing rib; 30-Embedded movable electric drive slider; 31-Terahertz detector; 32-Lifting rod; 33-Superconducting quantum interference device; 34-Vacuum chamber; 35-Second reinforcing rib; 36-Vacuum port; 37-Gas source port; 38-Gas source pipe. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0055] like Figures 1-3 As shown, the terahertz detection system for the insulating layer of the armored superconducting coil for nuclear fusion of the present invention includes a system base and support mechanism, a detection module and an auxiliary system.

[0056] The system base and support mechanism include:

[0057] The base 1 serves as the overall support foundation, with an electric rotary support platform 2 mounted on top. A rotating support column 7 is fixed on the electric rotary support platform 2, and an electric rotating gear sleeve 8 is fitted on the outside of the rotating support column 7. The electric rotating gear sleeve 8 is connected to the superconducting coil support base 10 through a square support bushing 9. The superconducting coil support base 10 and the superconducting coil intermediate support frame 27 (used to support the first superconducting coil 11 and the second superconducting coil 12) are reinforced by the first reinforcing rib 28, the second reinforcing rib 35, and the third reinforcing rib 29 to ensure structural stability.

[0058] The first superconducting coil 11 and the second superconducting coil 12 are respectively mounted on the superconducting coil support base 10 and the superconducting coil intermediate support frame 27.

[0059] The inlet of the first superconducting coil 11 (inlet 23) is connected to the DC / AC insulation withstand voltage tester 20 via the third connecting cable 26 and the third connector 22, and the outlet (outlet 15) is connected to the power supply control system 17 via the first connecting cable 14 and the first connector 16; the inlet of the second superconducting coil 12 (inlet 13) is directly connected to the power supply control system 17, and the outlet (outlet 19) is connected to the DC / AC insulation withstand voltage tester 20 via the second connecting cable 21 and the second connector 18; both the power supply control system 17 and the DC / AC insulation withstand voltage tester 20 are connected to the back-end data processing unit via the signal and power feedback channel 5.

[0060] The detection module and auxiliary system include:

[0061] A vacuum chamber 34 is located outside the superconducting coil. A vacuum port 36 (connected to a vacuum pump assembly 4 via a vacuum pump connecting pipe 3) and a gas source port 37 (connected to a gas source control system 6 via a gas source pipe 38) are opened on the side wall of the chamber. An embedded movable electric drive slider 30 is installed on the inner wall of the vacuum chamber. A lifting rod 32 is fixed on the embedded movable electric drive slider 30, and the top of the lifting rod 32 is connected to a terahertz detector 31 (i.e., a THz-TDS detection module). A superconducting quantum interference device 33 (i.e., a SQUID array) is installed on the top of the vacuum chamber 34, with a distance of 3-5 mm between it and the terahertz detector 31. A cylinder 24 is fixed to one side of the superconducting coil support base 10, and a piston rod (i.e., a cylinder telescopic connecting shaft 25) is connected to the side wall of the vacuum chamber 34 for adjusting the chamber position. The THz-TDS detection module includes a THz emission module.

[0062] This invention also provides a detection method for a terahertz detection system of the insulation layer of armored superconducting coils used in nuclear fusion. The method achieves defect detection through four stages: preprocessing, multimodal scanning, data fusion, and quantitative evaluation. Specifically, it includes the following steps:

[0063] Step 1. Pre-treatment of superconducting coils:

[0064] The surface of the VPI-cured superconducting coil is polished (ensuring a surface roughness Ra < 0.8 μm) to avoid surface impurities and protrusions from blocking THz pulses. At the same time, nitrogen is introduced into the vacuum chamber 34 through the gas source control system 6 to control the humidity inside the chamber to < 5%RH, eliminating the absorption interference of water molecules on the THz signal and providing a stable environment for subsequent detection.

[0065] Step 2. Multimodal scanning, including THz pre-scan and SQUID precision measurement:

[0066] like Figure 4 As shown, a synchronous control architecture is used to achieve collaborative detection of THz and SQUID. The process is as follows:

[0067] Step 2.1. Trigger the THz transmission module;

[0068] Step 2.2. The terahertz detector 31 scans the coil surface along the embedded movable electric drive slider 30 at 10μm step intervals to collect defect reflection signals;

[0069] Step 2.3. Synchronously trigger the power supply control system 17 to supply the rated current to the superconducting coil;

[0070] Step 2.4. Measure the magnetic field distribution generated by the coil current distortion using a SQUID array;

[0071] Step 2.5. Align the THz and SQUID signals using a 10ns precision time synchronizer, and complete spatial registration using the SIFT algorithm;

[0072] Step 2.6. Input 3D reconstruction elements to generate a preliminary defect distribution map.

[0073] Step 3. Data Fusion Analysis:

[0074] Step 3.1. Construct a quantitative depth-magnetic field model:

[0075] Based on the Biot-Savart law and Maxwell's equations, the defect depth d and the dielectric constant variation Δε are derived. r Relationship with magnetic field disturbance ΔB:

[0076] ΔB=k·Δε r / d²;

[0077] Step 3.2. Perform calibration testing:

[0078] A cyanate ester-epoxy resin composite insulation layer sample (same material as the ITER coil) was fixed to the superconducting coil support base 10, and a gradient air gap defect with a diameter of 0.1-2 mm and a depth of 1-10 mm was implanted. The defect depth d (error ±0.02 mm) and Δε were measured using a THz-TDS detection module. r The SQUID array measurement corresponds to ΔB; the electromagnetic coupling coefficient k is fitted using the least squares method (the mean value is taken after 20 repeated experiments, and the variance is ≤ ±5%).

[0079] Step 3.3. Integrate on-site data:

[0080] Extracting Δε from THz positioning points r And d, calculate the theoretical magnetic field perturbation:

[0081] ΔBcal=k·Δε r / d;

[0082] Compare the measured SQUID value ΔBmeas and calculate the error rate:

[0083] Error rate = |ΔBcal - ΔBmeas| / ΔBmeas × 100%;

[0084] When the error rate is ≤8%, the defect parameters are confirmed; if the error rate exceeds the limit, a re-inspection is triggered.

[0085] Step 4. Perform a quantitative assessment of defects and output the results of data fusion:

[0086] Location of dielectric anomaly region: based on Δε r A value >0.05 is used as a criterion to mark areas of deteriorated insulation performance;

[0087] Magnetic field disturbance analysis: Set ΔB>0.5μT as the warning threshold to identify high-risk defects;

[0088] Report generation: Outputs a full-parameter report in XML format, including defect location (x, y, z, accuracy ±5μm), dimension V (accuracy ±5%), and Δε. r ΔB and risk level (Level I < 10) -13 T·mm 3 Level II ≥10 -13 T·mm 3 );

[0089] Visualization: Generate a heat map of defect distribution (red-yellow-green corresponding to high-medium-low risk), and extract the maximum ΔB profile to display the field strength gradient line.

[0090] The core performance of this invention compared with traditional detection technologies is as follows:

[0091]

[0092] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A terahertz inspection system for the insulation layer of a superconducting magnet coil for nuclear fusion, characterized in that, Comprise: THz-TDS detection module, THz represents terahertz, TDS represents time domain spectrum; Superconducting quantum interference device detection array; Synchronous controller for triggering the THz-TDS detection module and the superconducting quantum interference device detection array and realizing time sequence alignment; Data processing unit for receiving signals of the THz-TDS detection module and the superconducting quantum interference device detection array, performing three-dimensional tomographic reconstruction and data fusion, and outputting defect detection report.

2. The nuclear fusion armored superconducting coil insulation layer terahertz detection system according to claim 1, characterized in that, Also include: The base is fixedly installed with an electric rotary support table on the top, a rotating support column is fixedly arranged on the electric rotary support table, an electric rotary gear sleeve is sleeved outside the rotating support column, the electric rotary gear sleeve is connected with a superconducting coil support base through a square support shaft sleeve, and the superconducting coil support base and a superconducting coil middle support frame are reinforced and connected through first, second and third reinforcing ribs; The first superconducting coil and the second superconducting coil are respectively installed on the superconducting coil support base and the superconducting coil middle support frame; An embedded movable electric drive sliding block is installed on the inner wall of the vacuum chamber, a lifting rod is fixedly arranged on the embedded movable electric drive sliding block, and the top end of the lifting rod is connected with the terahertz time domain spectrum detection module; A cylinder is fixed on one side of the superconducting coil support base, and a piston rod of the cylinder is connected with the side wall of the vacuum chamber.

3. The nuclear fusion armored superconducting coil insulation layer terahertz detection system according to claim 2, characterized in that, Vacuum pipe openings and gas source pipe openings are formed in the side wall of the vacuum chamber, the vacuum pipe openings are connected with a vacuum pump group through a vacuum pump connecting pipe, the gas source pipe openings are connected with a gas source control system through a gas source pipe, and the superconducting quantum interference device detection array is fixed on the top of the vacuum chamber.

4. The nuclear fusion armored superconducting coil insulation layer terahertz detection system according to any one of claims 1 to 3, characterized in that, The synchronous controller comprises a high-precision pulse generator, a power supply control system and a 10 ns precision time synchronizer, the high-precision pulse generator triggers the terahertz time domain spectrum detection module, the power supply control system supplies rated current to the first superconducting coil and the second superconducting coil, and the 10 ns precision time synchronizer aligns signals of the terahertz time domain spectrum detection module and the superconducting quantum interference device detection array, and inputs a three-dimensional reconstruction unit after spatial registration is completed through a SIFT algorithm.

5. The nuclear fusion armored superconducting coil insulation layer terahertz detection system according to any one of claims 2-3, characterized in that, The terahertz time domain spectrum detection module emits a wideband pulse with 0.8 THz as a center frequency, scans the coil surface along the embedded movable electric drive sliding block at a step distance of 10 μm, collects defect reflection signals, reconstructs time domain signals into voxel data through an inverse Radon transform algorithm, and realizes three-dimensional tomographic imaging.

6. The nuclear fusion armored superconducting coil insulation layer terahertz detection system according to claim 1, wherein, The superconducting quantum interference detector array adopts YBa2Cu3O7 superconducting thin film to prepare DC-SQUID, each unit contains double Josephson junction and the working temperature is 77K, the data processing unit establishes the depth-magnetic field quantitative model of ΔB=k·Δε based on Biot-Savart law and Maxwell equation, wherein k is the electromagnetic coupling coefficient of the material, which is determined by calibration experiment, the SIFT algorithm realizes the spatial alignment of terahertz data and superconducting quantum interference detector data, the registration error is less than 5μm, and the ΔB-V-t three-dimensional database is established for maintenance decision r / d 2 ; d is the defect depth, Δε r is the dielectric constant change, and ΔB is the magnetic field disturbance.

7. A method for detecting defects in the insulation layer of a nuclear fusion armored superconducting coil, implemented using the terahertz detection system for the insulation layer of a nuclear fusion armored superconducting coil according to any one of claims 1-6, characterized in that, The following steps are included: Polish the surface of the superconducting coil after vacuum pressure impregnation and curing, and control the environmental humidity to be less than 5% RH; Trigger the THz-TDS detection module to emit a wideband pulse to scan the coil surface, synchronously trigger the power supply control system to supply rated current to the superconducting coil, start the superconducting quantum interference device detection array to measure the magnetic field distribution, realize time sequence alignment through the 10 ns precision time synchronizer, complete spatial registration through the SIFT algorithm, and input the three-dimensional reconstruction unit to generate a preliminary defect distribution map; Constructing a depth-magnetic field quantitative model ΔB=k·Δε r / d 2 The electromagnetic coupling coefficient k was determined through calibration experiments, and the change in dielectric constant Δε obtained from terahertz detection was used to determine the change in dielectric constant. r The theoretical magnetic field perturbation ΔBcal was calculated by substituting the defect depth d into the model, and the error rate was calculated by comparing it with the measured magnetic field perturbation ΔBmeas of the superconducting quantum interference device. According to the dielectric constant change and the magnetic field disturbance setting threshold, the insulation performance deterioration area is marked, and a full parameter detection report containing defect position, size, dielectric constant change, magnetic field disturbance and risk level is output.

8. A method of detecting defects in an insulation layer of an armored superconducting coil for nuclear fusion as defined in claim 7, characterized in that, The THz-TDS detection module scans along the embedded movable electric drive slider with a step of 10 μm, collects the amplitude attenuation and time delay of the defect reflection signal, and uses the inverse Radon transform algorithm to reconstruct the time domain signal into voxel data, so as to realize three-dimensional tomographic imaging with a voxel accuracy of 20 μm.

9. A method of detecting defects in an insulation layer of an armored superconducting coil for nuclear fusion as defined in claim 7, characterized in that, The calibration experiment adopts cyanate ester-epoxy resin composite insulation layer sample, implants gradient air gap defect with diameter of 0.1-2 mm and depth of 1-10 mm, measures defect depth d and dielectric constant change Δε through a terahertz time-domain spectroscopy detection module r , measures corresponding magnetic field disturbance ΔB through a superconducting quantum interference device detection array, adopts least square method to fit electromagnetic coupling coefficient k, repeats 20 groups of experiments to take average, and variance is ≤±5%.

10. A method of detecting defects in an insulation layer of an armored superconducting coil for nuclear fusion as defined in claim 7, characterized in that, In the quantitative evaluation of defects, the insulation performance degradation area is marked with the criterion of dielectric constant change Δε r > 0.05, the magnetic field disturbance ΔB> 0.5 μT is set as the early warning threshold to identify high-risk defects, the full-parameter report in XML format is output, the defect distribution thermodynamic map and the magnetic field gradient line visualization image are generated, and the ΔB-V-t three-dimensional database is established for maintenance decision.

Images