Method for detecting internal defects of prebaked anode

Abstract

The invention provides a method for detecting internal defects of a prebaked anode. The method can solve the problems that an existing detection method is low in detection precision, insufficient in safety, incomplete and inaccurate in defect information acquisition and the like. The detection method comprises the following steps: arranging a plurality of electrodes on the surface of a prebaked anode; wherein excitation current is injected into two adjacent electrodes, and the voltage between the positions of any two adjacent electrodes on the surface of the prebaked anode is obtained and serves as a voltage measured value; according to the obtained voltage measured value, the electrical conductivity distribution in the prebaked anode is obtained through EIT inverse problem solving; and identifying the internal defects of the prebaked anode according to the obtained internal conductivity distribution of the prebaked anode.

Description

Technical Field

[0001] The present invention relates to a defect detection method, specifically to a method for detecting internal defects of prebaked anodes, and belongs to the technical field of defect detection. Background Art

[0002] In the field of detecting internal defects of prebaked anodes, traditional detection methods mainly include ultrasonic detection, ray detection, and some methods based on physical property testing.

[0003] Ultrasonic detection is one of the more commonly used non-destructive testing means. Its principle is that when ultrasonic waves propagate inside the prebaked anode, phenomena such as reflection, refraction, and scattering will occur when encountering different medium interfaces (such as the interface between a defect and normal material). By emitting ultrasonic waves on the surface of the prebaked anode and receiving the reflected ultrasonic wave signals, and based on characteristics such as the time delay and amplitude change of the signals, it is inferred whether there are internal defects and the location and approximate characteristics of the defects. However, ultrasonic detection has certain limitations. Since the prebaked anode material itself is not completely uniform, factors such as its internal particle structure and pore distribution will cause complex scattering and attenuation of ultrasonic waves during propagation, resulting in reduced detection sensitivity, especially for the detection accuracy of micro-defects (such as holes or cracks with a diameter less than a few millimeters) and deep defects. Moreover, the accuracy of ultrasonic detection results depends to a large extent on the experience and skills of the operator, and there may be differences in the detection results of the same prebaked anode by different operators.

[0004] Ray detection mainly includes X-ray detection and γ-ray detection. This method uses rays to penetrate the prebaked anode. Due to the different absorption degrees of rays by the defective part and the normal part, different gray-scale images are formed on the imaging plate or detector, thereby visually showing information such as the shape, size, and location of internal defects. Ray detection can provide a relatively clear internal structure image and has good detection ability for some defects with complex shapes. However, ray detection equipment is expensive, requires professional operators for operation and maintenance, and there is a radiation hazard, posing a potential risk to the health and safety of operators. In addition, the detection speed of ray detection is relatively slow and is not suitable for large-scale online detection. Generally, it can only be used for spot checks or for detecting specific important prebaked anodes.

[0005] Methods based on physical property testing, such as indirectly judging whether there are defects inside by measuring physical property parameters of pre-baked anodes, such as resistivity, density, hardness, etc. This method is relatively simple to operate and can quickly conduct a preliminary screening of a large number of pre-baked anodes. For example, when there are holes or loose areas inside the pre-baked anode, its density may decrease and the resistivity may change. However, this method can only provide a rough judgment, cannot accurately determine the specific location, shape, size and other detailed information of the defects, has limited ability for qualitative and quantitative analysis of the defects, and is prone to misjudgment.

[0006] In summary, there are some defects in traditional ultrasonic testing, radiographic testing, and physical property testing-based methods:

[0007] For ultrasonic testing, its detection accuracy is limited: For tiny defects and deep-layer defects, due to the severe scattering and attenuation of ultrasonic waves in pre-baked anode materials, the detection sensitivity is significantly reduced. The description of defect shapes and characteristics by ultrasonic testing is not precise enough. Since the ultrasonic reflection signals are interfered by various factors, it is difficult to accurately restore the true contour and complex shape of the defects, and there are large errors in judging the defect shapes. In addition, the result reliability is low: The inhomogeneity of pre-baked anode materials leads to complex and variable ultrasonic propagation characteristics, making the detection results largely dependent on the experience and skill level of the operators. The detection results of the same pre-baked anode by different operators may deviate greatly, with poor repeatability and low credibility of the data. Ultrasonic testing is difficult to distinguish the signal changes caused by normal structural changes inside the material (such as slight differences in local material density) from the signal anomalies truly caused by defects.

[0008] Radiographic testing: It is costly and inefficient: The purchase cost of radiographic testing equipment is extremely high, and the price of γ-ray testing equipment is even more expensive. At the same time, the maintenance and operation costs of the equipment are also very high, including regular replacement of radiation sources, equipment calibration, and maintenance costs of professional personnel, etc. The detection process is complex and slow, and the detection time for a single pre-baked anode is relatively long, making it difficult to meet the requirements of rapid detection on large-scale production lines, easily causing production bottlenecks and reducing production efficiency. In addition, there are also potential safety and environmental hazards: Radiographic testing has radiation hazards, posing a potential threat to the health of operators, and requires the equipped with special radiation protection facilities and strict operation specifications, increasing the safety management costs and difficulties of enterprises. The radiation leakage during the radiographic testing process may cause pollution to the surrounding environment, which does not conform to the development concept of modern industrial green environmental protection.

[0009] Physical property testing method, incomplete defect information: It can only indirectly infer whether there are defects inside by measuring physical property parameters such as the resistivity, density, and hardness of pre-baked anodes, and it is impossible to directly determine detailed information such as the location, shape, and size of the defects. For example, when measuring the resistivity of pre-baked anodes, the four-wire voltammetry method is generally used. The four-wire voltammetry method applies a current and measures the overall voltage of the sample, and calculates the average resistivity of the material, which reflects the comprehensive conductivity characteristics of the current path. If there are local defects (such as cracks or holes) inside, although it may cause an increase in the overall resistivity, it is impossible to locate the specific location or shape of the defect through a single measurement. Even if the location of the defect can be roughly determined through multiple measurements, the shape of the defect cannot be determined. In addition, the qualitative and quantitative accuracy is not high: The accuracy of qualitative and quantitative analysis of defects is poor and is easily affected by fluctuations in the raw materials and production processes of pre-baked anodes. Summary of the Invention

[0010] In view of this, the present invention provides a method for detecting internal defects of pre-baked anodes, which can solve problems such as low detection accuracy, insufficient safety, and incomplete and inaccurate acquisition of defect information in existing detection methods.

[0011] The technical solution of the present invention is: A method for detecting internal defects of pre-baked anodes, the detection method includes the following steps:

[0012] Arrange a number of electrodes on the surface of the pre-baked anode;

[0013] Inject an excitation current between two adjacent electrodes, and obtain the voltage between the positions of any other two adjacent electrodes on the surface of the pre-baked anode as the measured voltage value;

[0014] According to the obtained measured voltage value, use the inverse problem of EIT to solve for the conductivity distribution inside the pre-baked anode;

[0015] Identify the internal defects of the pre-baked anode according to the obtained conductivity distribution inside the pre-baked anode.

[0016] As a preferred embodiment of the present invention: The arrangement method of the surface electrodes of the pre-baked anode is:

[0017] Arrange a number of electrode assemblies in an array on the surface of the pre-baked anode, the electrode assembly includes a voltage probe and an electrode rod as an electrode; one end of the electrode rod and the measuring end of the voltage probe are respectively in contact with the pre-baked anode.

[0018] As a preferred embodiment of the present invention: A current probe is further provided in the electrode assembly for measuring the current flowing through the surface electrode of the pre-baked anode.

[0019] As a preferred embodiment of the present invention: the pre-baked anode has a cuboid structure, and electrode fixing frames are arranged on all four side surfaces thereof; the area of the electrode fixing frame is the same as the side area of the corresponding pre-baked anode;

[0020] The electrode fixing frame is used to install electrode rods and voltage probes, so that one end of the electrode rod and the measuring end of the voltage probe are in stable contact with the pre-baked anode;

[0021] A number of electrode assemblies are arranged on each electrode fixing frame in an N×M array, where N and M are both integers greater than 1; and the electrode assemblies on two opposite side surfaces of the pre-baked anode are symmetrically arranged.

[0022] As a preferred embodiment of the present invention: the adjacent electrode method is used to measure the voltage between any two adjacent electrodes on the surface of the pre-baked anode:

[0023] First, an excitation current is injected into two adjacent electrodes to establish an internal electric field in the pre-baked anode, and the voltages between the positions of the other two adjacent electrodes on the surface of the pre-baked anode are measured in turn. After one round of measurement; then the excitation current is injected into the next pair of adjacent electrodes, and the above operation is repeated; until all adjacent electrodes are rotated.

[0024] As a preferred embodiment of the present invention: when identifying internal defects of the pre-baked anode based on the obtained conductivity distribution inside the pre-baked anode:

[0025] First, a conductivity threshold range of the pre-baked anode material is set, and regions inside the pre-baked anode where the conductivity is lower or higher than this conductivity threshold range are identified as defect regions.

[0026] As a preferred embodiment of the present invention: the positions, shapes and sizes of the identified internal defects of the pre-baked anode are presented in a visual manner.

[0027] As a preferred embodiment of the present invention: in the process of obtaining the conductivity distribution inside the pre-baked anode by solving the EIT inverse problem, the process of establishing the finite element model of the forward problem is as follows:

[0028] First, according to the actual geometric shape of the pre-baked anode and the arrangement of electrodes on its surface, a finite element model of the pre-baked anode with electrodes arranged thereon is constructed, and this finite element model is a finite element model of a pre-baked anode without internal defects;

[0029] Then, mesh generation and physical parameter setting are performed on the finite element model of the pre-baked anode:

[0030] The finite element model of the pre-baked anode is divided into a number of tetrahedral element meshes or hexahedral element meshes: among them, the part at the edge of the pre-baked anode is divided by hexahedral element meshes; the region inside the pre-baked anode is divided by tetrahedral element meshes;

[0031] Assign an initial conductivity to each unit grid according to the known conductivity characteristics of the pre-baked anode material;

[0032] At the same time, set boundary conditions and current excitation sources in the finite element model of the pre-baked anode according to the actual position of the electrodes and the application method of the excitation current;

[0033] Then, use the finite element method to simulate the forward problem of EIT for the constructed finite element model, and obtain the theoretical voltage value between the positions of any two adjacent electrodes on the surface of the pre-baked anode under the assumption that there are no defects inside the pre-baked anode.

[0034] As a preferred embodiment of the present invention: Before obtaining the conductivity distribution inside the pre-baked anode by solving the inverse problem of EIT, the following steps are first performed:

[0035] According to the actual geometric shape of the pre-baked anode and the arrangement of the electrodes on its surface, construct a finite element model of the pre-baked anode with electrodes arranged, and set defects with a preset shape and size at a preset position inside the finite element model;

[0036] Perform element mesh division on the finite element model, and assign conductivity to each unit grid to obtain the set conductivity distribution inside the finite element model; where the conductivity at the defect position is the air conductivity, and the conductivity at the remaining positions is the conductivity of the pre-baked anode material;

[0037] Solve the forward problem of EIT for the finite element model to obtain the theoretical voltage values of each electrode under the assumption that there are preset defects inside the pre-baked anode;

[0038] Substitute the theoretical voltage values of the electrodes into the calculation model for solving the inverse problem of EIT, and use the inverse problem of EIT to solve for the theoretical conductivity distribution inside the pre-baked anode;

[0039] Compare the theoretical conductivity distribution with the set conductivity distribution to determine whether they are consistent:

[0040] If they are consistent, perform the step of "obtaining the conductivity distribution inside the pre-baked anode by solving the inverse problem of EIT based on the measured voltage values of the electrodes on the surface of the obtained pre-baked anode";

[0041] If they are not consistent, modify the electrode layout in the finite element model, including increasing the number of electrodes and / or reducing the spacing between adjacent electrodes, and then repeat the above steps for the modified finite element model until the theoretical conductivity distribution is consistent with the set conductivity distribution; finally, modify the electrode layout on the surface of the pre-baked anode corresponding to the optimized electrode layout.

[0042] Beneficial effects:

[0043] (1) The present invention applies the EIT technology to the detection of internal defects in pre-baked anodes, which can comprehensively and accurately obtain detailed information such as the position, shape, and size of internal defects in pre-baked anodes, and improve the detection accuracy of internal micro-defects (such as holes, cracks, etc.) and deep defects in pre-baked anodes.

[0044] (2) The present invention outputs the calculated internal conductivity distribution of the pre-baked anode in a visual manner, which can intuitively display the internal defects of the pre-baked anode.

[0045] (3) By constructing a finite element model with preset defects inside and solving the forward problem for it, the layout of the surface electrodes of the pre-baked anode is verified, the electrode arrangement is optimized, and the detection accuracy of defects can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 is a three-dimensional schematic diagram of a pre-baked anode with electrode assemblies arranged on its surface;

[0047] Figure 2 is a top view of a pre-baked anode with electrode assemblies arranged on its surface;

[0048] Figure 3 is a flowchart of the method for detecting internal defects of the pre-baked anode of the present invention;

[0049] Wherein: 1 - pre-baked anode, 2 - electrode fixing frame, 3 - electrode rod. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] The following further describes the present invention in detail with reference to the drawings and embodiments.

[0051] Embodiment 1:

[0052] This embodiment provides a method for detecting internal defects of a pre-baked anode. Applying the EIT (Electrical Impedance Tomography, EIT) algorithm to the detection of internal defects of the pre-baked anode can comprehensively and accurately obtain detailed information such as the position, shape, and size of internal defects of the pre-baked anode, providing a reliable basis for the quality evaluation and production process improvement of the pre-baked anode.

[0053] EIT is fully called Electrical Impedance Tomography technology. At present, this technology is used in the medical field. Its principle is to apply a weak current to the surface electrodes of the human body and measure the voltage values on other electrodes, and reconstruct the conductivity or impedance distribution inside the human body according to the relationship between voltage and current. In this embodiment, this technology is applied to the detection of internal defects of the pre-baked anode; the pre-baked anode, that is, the anode carbon block itself has a very small conductivity. When there are defects inside it, the conductivity at the defect position will change suddenly (abnormally large). Based on this, the EIT technology can be applied to the detection of internal defects of the pre-baked anode.

[0054] As Figure 3 shown, the detection method includes the following steps:

[0055] First, arrange a number of electrodes on the surface of the pre-baked anode.

[0056] On each outer side of the pre-baked anode 1, electrode assemblies are arranged. The electrode assembly includes: an electrode rod 3 (i.e., by arranging the electrode rod 3 on the surface of the pre-baked anode 1 as an electrode) and a voltage probe. In this example, the materials of the electrode rod 3 and the voltage probe are selected as copper (or silver) with good electrical conductivity and high chemical stability. One end of the electrode rod 3 and the measuring end of the voltage probe are in stable contact with the pre-baked anode 1. A number of electrode assemblies are distributed in an array on the surface of the pre-baked anode to form an electrode array. The use of the electrode array can fully cover the surface of the pre-baked anode 1 and avoid potential difference measurement errors or electrode interference caused by improper spacing; a sufficient number, evenly placed and stable electrode assemblies can contribute to the stability of subsequent operations and data collection.

[0057] As an example, the electrode assembly is installed on the electrode fixing frame 2.

[0058] As an example, the electrode rod 3 is fixed to the surface of the pre-baked anode 1 by screws and conductive paste (or special conductive glue) is applied to ensure stable low-resistance contact and reduce the risk of data fluctuations caused by poor contact.

[0059] As an example, the electrode rod 3 is fixed to the surface of the pre-baked anode 1 by a magnetic attraction fixing method. Magnetic materials are embedded on the surface of the electrode rod 3 and the corresponding positions of the pre-baked anode, and the electrode rod 3 is adsorbed on the surface of the pre-baked anode by magnetic force. This method is convenient for the installation and disassembly of the electrode rod 3, which is beneficial to the reuse of the electrode rod 3 and the maintenance of the detection system. However, it is necessary to ensure that the magnetic materials will not interfere with the detected electrical signals and can provide sufficient adsorption force to ensure the stability of the electrode rod 3 during the detection process.

[0060] As an example, the electrode rod 3 is fixed to the surface of the pre-baked anode 1 in an embedded manner. For example, a groove matching the diameter of the electrode rod 3 is reserved during the production of the pre-baked anode, the electrode rod 3 is embedded therein, and then a conductive and viscous material (such as special conductive resin) is filled to fix the electrode rod 3 and ensure good electrical contact. This method can make the electrode rod 3 fit more closely to the surface of the pre-baked anode, reduce the external force influence that may be caused by the protrusion of the electrode, and improve the stability of the detection.

[0061] As an example, for the cylindrical pre-baked anode 1, the electrode assemblies are distributed in a circular array; for the pre-baked anode 1 with a cuboid structure, the electrode assemblies are distributed in a matrix array, and the electrode assemblies on two opposite sides of the pre-baked anode 1 are symmetrically arranged; thereby improving the detection spatial resolution and defect capture ability and reducing the detection blind area.

[0062] As Figure 1 and Figure 2 shown, in this example, the pre-baked anode 1 is a cuboid structure, and electrode fixing frames 2 are arranged on all four side surfaces thereof. The area of the electrode fixing frame 2 is the same as the side surface area of the corresponding pre-baked anode 1. The electrode fixing frame 2 is used to install electrode rods 3 and voltage probes, so that one end of the electrode rod 3 and the measuring end of the voltage probe are stably in contact with the pre-baked anode 1. A plurality of electrode assemblies are distributed in an array on each electrode fixing frame 2; all the electrode assemblies are arranged in a three-dimensional layout on the surface of the pre-baked anode 1, that is, a plurality of electrode assemblies are arranged in an array of N×M (both N and M are integers greater than 1) on each electrode fixing frame 2, and the electrode assemblies on two opposite side surfaces of the pre-baked anode 1 are symmetrically arranged. In this example, 48 pairs of electrode rods are uniformly fixed on the electrode fixing plates 2 located on the left and right sides as Figure 2 shown (that is, 12 columns of electrode assemblies are uniformly arranged at equal intervals along the length direction on the electrode fixing plates 2 on the left and right sides, and each column includes 4 electrode assemblies uniformly distributed at equal intervals along the height direction), and 20 pairs of electrode rods are uniformly fixed on the electrode fixing plates 2 located on the front and back sides as Figure 2 shown (that is, 5 columns of electrode assemblies are uniformly arranged at equal intervals along the width direction on the electrode fixing plates 2 on the left and right sides, and each column includes 4 electrode assemblies uniformly distributed at equal intervals along the height direction).

[0063] Voltage data acquisition: Obtain the voltage between the positions of any two adjacent electrodes on the surface of the pre-baked anode as the measured voltage value.

[0064] When performing voltage data acquisition, start the excitation current source to generate a stable excitation current according to the set frequency and amplitude, and apply the excitation current to the selected electrode pair; then collect the voltage data generated on other electrode pairs due to the action of the excitation current.

[0065] As an example, the adjacent electrode method is used to measure the voltage of the electrodes on the surface of the pre-baked anode; that is, two adjacent electrodes form an electrode pair. This method injects an excitation current into two adjacent electrodes to establish an internal electric field in the pre-baked anode, and alternately measures the voltage between the positions of the other two adjacent electrodes on the surface of the pre-baked anode. After completing one round of measurement, inject the excitation current into the next pair of adjacent two electrodes, and repeat the above operation; until all adjacent electrodes are rotated. In this example, "adjacent electrodes" refer to two adjacent electrodes in the same horizontal plane.

[0066] As an example, a current probe is further provided in the electrode assembly for measuring the current flowing through the electrodes on the surface of the pre-baked anode 1 each time an excitation current is injected between two adjacent electrodes. The current probe and the voltage probe can be integrally provided as an electric measurement probe. By measuring the current flowing through the electrodes, the internal defects of the pre-baked anode 1 can be assisted in judgment. The principle is as follows: If there are no defects inside the pre-baked anode 1, the currents at the electrodes at two opposite positions on the surface of the pre-baked anode 1 are the same (for example, electrodes A and B are respectively provided at two opposite positions on the surface of the pre-baked anode 1, and the current flows in from electrode A. If there is no defect between electrode A and electrode B inside the pre-baked anode 1, theoretically, the current magnitude at electrode B is the same as that at electrode A); If there are defects inside the pre-baked anode 1, due to the extremely large resistance at the defect position, less current passes through the defect position and more current passes through the intact area near the defect, showing that the current flowing out from the electrode at the defect level is smaller, and the current flowing out from the electrode near this electrode is larger.

[0067] Based on the obtained measured voltage values, the conductivity distribution inside the pre-baked anode 1 is obtained by solving the EIT inverse problem.

[0068] Solving the EIT inverse problem refers to the process of calculating the conductivity distribution from the measured values of the surface voltage. Based on the above-mentioned measured voltage values on the surface of the pre-baked anode, the conductivity distribution inside the pre-baked anode 1 is obtained by solving the EIT inverse problem. That is, the EIT inverse problem can reconstruct the internal structure image of the pre-baked anode 1 according to the measured electrical data (voltage).

[0069] As an example, the EIT inverse problem is solved using an iterative algorithm of the traditional Gauss-Newton method.

[0070] Finally, based on the obtained conductivity distribution inside the pre-baked anode 1, the internal defects of the pre-baked anode 1 are identified.

[0071] The conductivity at the internal defect locations (such as holes, cracks, inclusions, etc.) of the pre-baked anode 1 will deviate from the normal value (i.e., the conductivity of the pre-baked anode 1 material itself) due to changes in the material structure. Using this as a judgment basis, the defect locations inside the pre-baked anode 1 can be accurately found.

[0072] As an example, first, a conductivity threshold range of the material of the pre-baked anode 1 is set, and the regions inside the pre-baked anode 1 where the conductivity is lower or higher than this conductivity threshold range are identified as defect regions. The conductivity threshold range is obtained by analyzing the conductivity distributions inside several pre-baked anode samples without defects.

[0073] As an example, the conductivity distribution inside the pre-baked anode 1 obtained by calculation is presented in a visual manner (such as the visualization software ParaView), and the defect positions, shapes, and sizes inside the pre-baked anode 1 can be intuitively seen; moreover, by increasing the distribution density of the electrode assembly, the accuracy of defect detection can be improved.

[0074] Using the above method can comprehensively and accurately obtain detailed information such as the position, shape, and size of the defects inside the pre-baked anode 1, providing a reliable basis for the quality assessment and production process improvement of the pre-baked anode 1. In this detection method, the entire detection process is based on the principle of electrical property detection, and the electrode is only in electrical contact with the pre-baked anode 1, and the structure and performance of the pre-baked anode 1 will not be damaged during normal detection operations, achieving non-invasive and non-destructive detection, ensuring the integrity of the pre-baked anode 1. The pre-baked anode 1 after detection can be directly put into use, eliminating product loss. And the detection process does not involve radiation sources such as rays at all, completely eliminating radiation hazards, removing potential threats to the health of operators, and eliminating the need for radiation protection facilities.

[0075] Therefore, using this method has the following advantages: it can enhance the reliability and repeatability of the detection results, reduce detection errors caused by differences in operator experience and material inhomogeneity; it has high detection efficiency, and the detection time for a single pre-baked anode 1 can be shortened to within 5 minutes, meeting the rapid detection requirements on large-scale production lines; it can achieve a detection process without radiation, safe and environmentally friendly, eliminate potential hazards to the health of operators, and avoid environmental pollution.

[0076] Example 2:

[0077] On the basis of the above Example 1, the following specific introduction is made to the solution process of the EIT inverse problem.

[0078] The core of the EIT inverse problem is to infer the internal conductivity distribution by measuring the voltage, and this process depends on the accurate modeling of the forward problem, that is, calculating the surface voltage when the conductivity distribution is known.

[0079] Based on this, it is necessary to construct a finite element model of the pre-baked anode with electrodes arranged (using the same electrode arrangement method as the above pre-baked anode physical object); that is, according to the actual geometric shape of the pre-baked anode and the arrangement of the electrodes on its surface, use finite element analysis software (such as COMSOL Multiphysics, Solidworks, etc.) to construct a finite element model of the pre-baked anode with electrodes arranged, and this finite element model is a finite element model of the pre-baked anode without internal defects.

[0080] Then, mesh generation and physical parameter settings are performed on the finite element model of the pre-baked anode:

[0081] Divide the finite element model of the pre-baked anode into a large number of small tetrahedral or hexahedral element meshes: among them, the part close to the edge of the pre-baked anode is divided by hexahedral elements to improve the calculation efficiency; for the area inside the pre-baked anode, tetrahedral elements are used to ensure the accuracy and adaptability of the model.

[0082] Assign conductivity to each element according to the known conductivity characteristics of the pre-baked anode material (obtained through experimental measurement of material samples).

[0083] At the same time, according to the actual position of the electrodes and the application method of the excitation current, accurately set the boundary conditions and excitation sources in the model. Set the outer surface of the pre-baked anode as an insulating boundary or set the electrical contact boundary conditions with the surrounding environment according to the actual situation. At the same time, input the position and parameters of the excitation current source (the same as the position and parameters of the actual excitation current source) as excitation conditions into the finite element model. For example, at the model boundary corresponding to the electrode connected to the excitation current source, set the current inflow boundary condition, and the current magnitude is the value of the actually applied excitation current; at the boundaries of the remaining electrodes used to measure voltage, set the voltage measurement boundary condition to simulate the real measurement environment.

[0084] Then use the finite element method to perform the forward problem simulation of EIT on the constructed finite element model (the forward problem simulation of EIT is the process of using the finite element method to solve the node electric potential), and obtain the theoretical voltage value between the positions of any two adjacent electrodes on the surface of the pre-baked anode under the assumption that there are no defects inside the pre-baked anode. The solution of the forward problem of EIT refers to the process of calculating the surface voltage from the conductivity distribution.

[0085] As an example, when solving the forward problem of EIT, use the built-in solver of the finite element software to perform iterative calculations. The number of iterations is determined according to the complexity of the model and the requirement of convergence accuracy until the convergence condition is met. The convergence condition can be set as the norm of the change in the voltage value obtained from two adjacent iterative calculations being less than the set threshold to ensure that the calculated theoretical voltage value is accurate enough to provide a reliable reference for subsequent steps.

[0086] Since a finite element model is constructed during the EIT calculation process, the inverse problem of EIT returns the inversely derived conductivity to each unit node, and then a numerical value is calculated by weighted averaging the conductivity on the tetrahedral unit nodes through visualization means. This numerical value represents the conductivity of this grid unit; and then the size and shape of this defect are reflected by the grid unit. Based on this, when performing the result visualization output: for the position of the defect, it is displayed by converting its coordinate information in the finite element model of the prebaked anode into actual three-dimensional coordinates (taking the center of the prebaked anode as the coordinate origin, establishing a Cartesian coordinate system, and clarifying the coordinate values in each direction); for the size of the defect, the volume value is obtained by cumulative calculation according to the number of units involved in the defect area and the volume of a single unit (the size of each unit has been determined during model construction, and the volume can be calculated), or the area value on the two-dimensional cross-section is calculated by geometric methods such as projection to represent it; for the shape of the defect, the edge detection algorithm in image processing technology (such as the Canny edge detection algorithm) is used to extract the contour of the defect area, and then the curve fitting algorithm (such as the least squares method for fitting curves) is used to fit the contour to clearly present the shape characteristics of the defect. Finally, a three-dimensional defect image or a two-dimensional cross-sectional view is generated. For example, in the three-dimensional image, the normal conductivity area and the defect area are distinguished by different colors, and at the same time, the key parameter information of the defect (such as volume, coordinates, etc.) is marked. In the two-dimensional cross-sectional view, the defect contour and related parameters at a specific cross-section are shown, which is convenient for the operator to intuitively understand the defect situation inside the prebaked anode and provides a strong basis for quality assessment and subsequent processing decisions.

[0087] Embodiment 3:

[0088] Based on the above Embodiment 1 or Embodiment 2, the layout of the surface electrodes of the prebaked anode 1 (the number of electrode arrangements, the spacing between adjacent electrodes, etc.) directly affects the accuracy of defect detection. In this example, before using the inverse problem of EIT to solve for the conductivity distribution inside the prebaked anode 1, the following steps are first performed to verify whether the current electrode layout can meet the measurement requirements (let this step be the electrode layout verification step). Specifically:

[0089] Construct a finite element model of the prebaked anode with electrodes arranged (using the same electrode arrangement method as when obtaining the actual measured voltage value above); and set a defect with a preset shape and size at a preset position inside the finite element model of the prebaked anode, that is, the position, shape, and size of this defect are all known values.

[0090] Perform mesh division and physical parameter setting on the physical model of the prebaked anode:

[0091] Divide the physical model of the pre-baked anode into a large number of small tetrahedral or hexahedral element meshes: Among them, the part close to the edge of the pre-baked anode is divided by hexahedral element meshes to improve the calculation efficiency; for the area inside the pre-baked anode, tetrahedral element meshes are used to ensure the accuracy and adaptability of the model.

[0092] Assign conductivity to each element mesh (i.e., the physical parameters here include conductivity): Among them, assign the inherent conductivity of the pre-baked anode material (obtained through experimental measurement of material samples) to the other element meshes except for the preset defects; the preset defects are equivalent to air, and assign the air conductivity to the element meshes at the preset defects. That is, the internal conductivity distribution of the physical model of this pre-baked anode is known, and let it be the set conductivity distribution.

[0093] Then use the finite element method to solve the forward EIT problem of the constructed finite element model with preset defects inside, and obtain the theoretical voltage values of each electrode under the assumption of preset defects inside the pre-baked anode. Solving the forward EIT problem refers to the process of calculating the surface electrode voltage from the conductivity distribution. As an example, when solving the forward EIT problem, use the built-in solver of the finite element software for iterative calculation.

[0094] Then substitute the theoretical voltage values of the electrodes obtained by solving the forward EIT problem into the calculation model for solving the inverse EIT problem, use the inverse EIT problem to solve to obtain the conductivity distribution inside the pre-baked anode 1, let this conductivity distribution be the theoretical conductivity distribution, and compare this theoretical conductivity distribution with the set conductivity distribution to judge whether the theoretical conductivity distribution is consistent with the set conductivity distribution:

[0095] If they are consistent (that is, based on the calculated theoretical conductivity distribution, the preset defects inside the finite element model of the pre-baked anode can be judged, and the differences in the position, shape, and size of the preset defects from the preset values are within the set error range), it indicates that the current electrode layout can meet the measurement requirements, and the step of "using the inverse EIT problem to solve to obtain the conductivity distribution inside the pre-baked anode 1 based on the measured voltage values of the surface electrodes of the pre-baked anode obtained" can be carried out;

[0096] If it is inconsistent with the set conductivity distribution (that is, based on the calculated theoretical conductivity distribution, the preset defects inside the physical model of the pre-baked anode can be judged, and the differences in the position, shape, and size of the preset defects from the preset values exceed the set error range), it indicates that the measurement error of the current electrode layout is large, and the electrode layout needs to be optimized (correct the electrode layout in the finite element model), such as increasing the number of electrodes and / or reducing the adjacent electrode spacing, etc.; then re-perform the above electrode layout verification steps on the optimized finite element model until the theoretical conductivity distribution is consistent with the set conductivity distribution. Then modify the surface electrode distribution of the pre-baked anode 1 correspondingly according to the optimized electrode layout.

[0097] Example 4:

[0098] Based on the above Example 1 or Example 2, the Gauss-Newton method iteration + regularization method is used to solve the EIT inverse problem.

[0099] The specific iteration process is as follows:

[0100] At the k-th iteration, based on the current conductivity distribution σ k， Calculate the Jacobian matrix J of the inverse problem k .

[0101] Note: The Jacobian matrix J is a matrix that reflects the sensitivity of the boundary measurement voltage to small changes in conductivity.

[0102] After that, construct the error vector: Calculate the error between the actual measured voltage vector V measure d and the theoretical voltage vector corresponding to the current conductivity distribution : where is the theoretical voltage vector obtained by solving the k-th iteration of the EIT forward problem.

[0103] Define the objective function: Construct an objective function that includes a data fitting term and a regularization term:

[0104]

[0105] where: The data fitting term ||e k || 2 is the sum of the squares of the error vector, that is, ||e k || 2 =(e k ) T e k , which characterizes the matching degree between the model prediction and the measured data; The regularization term introduces a regularization parameter λ and a regularization matrix L (such as a Tikhonov matrix or a prior information matrix), which is used to constrain the ill-posedness of the solution and improve the stability. It is a kind of optimization method. σ prior is the conductivity estimate.

[0106] Then, update the conductivity distribution by minimizing the objective function Φ k :

[0107] σ k+1 =σ k +Δσ k

[0108]

[0109] This equation is derived from the Gauss-Newton method, combining the linearized approximation of the Jacobian matrix and the regularization constraint.

[0110] Iterations are continuously performed until a preset convergence condition is met, such as the norm of the change in the conductivity distribution obtained in two consecutive iterations (the norm is also measured by calculating the square root of the sum of the squares of the elements of the conductivity difference vector) is less than a certain threshold, or the number of iterations reaches a preset maximum value.

[0111] Although the present invention has been described in detail above with general descriptions and specific embodiments, based on the present invention, some modifications or improvements can be made, which are obvious to those skilled in the art. Therefore, these modifications or improvements made without departing from the spirit of the present invention all fall within the scope of the present invention claimed.

Claims

1. A method for detecting internal defects of pre-baked anodes, characterized in that, The detection method includes the following steps: Arrange a number of electrodes on the surface of the pre-baked anode; Inject an excitation current into two adjacent electrodes among them, and obtain the voltage between the positions of any other two adjacent electrodes on the surface of the pre-baked anode as the measured voltage value; According to the obtained measured voltage value, use the inverse problem solution of EIT to obtain the conductivity distribution inside the pre-baked anode; Identify the internal defects of the pre-baked anode according to the obtained conductivity distribution inside the pre-baked anode.

2. The internal defect detection method of the pre-baked anode according to claim 1, wherein, The arrangement method of the electrodes on the surface of the pre-baked anode is as follows: Arrange a number of electrode assemblies in an array on the surface of the pre-baked anode. The electrode assembly includes a voltage probe and an electrode rod serving as an electrode; one end of the electrode rod and the measurement end of the voltage probe are respectively in contact with the pre-baked anode.

3. The internal defect detection method of the pre-baked anode according to claim 2, characterized in that, A current probe is further arranged in the electrode assembly for measuring the current flowing through the electrodes on the surface of the pre-baked anode.

4. The internal defect detection method of the pre-baked anode according to claim 2 or 3, characterized in that, The pre-baked anode is of a cuboid structure, and electrode fixing frames are arranged on all four sides thereof; the area of the electrode fixing frame is the same as the area of the corresponding side of the pre-baked anode; The electrode fixing frame is used to install the electrode rod and the voltage probe, so that one end of the electrode rod and the measurement end of the voltage probe are stably in contact with the pre-baked anode; A number of electrode assemblies are arranged in an N×M array on each of the electrode fixing frames, where N and M are both integers greater than 1; and the electrode assemblies on the two opposite sides of the pre-baked anode are symmetrically arranged.

5. The method for detecting internal defects of pre-baked anodes according to claim 1, characterized in that The adjacent electrode method is used to measure the voltage between the positions of any two adjacent electrodes on the surface of the pre-baked anode: First, inject an excitation current into two adjacent electrodes, establish an internal electric field of the pre-baked anode, and alternately measure the voltage between the positions of the other two adjacent electrodes on the surface of the pre-baked anode. After completing one round of measurement; then inject the excitation current into the next pair of adjacent electrodes, and repeat the above operation; until all adjacent electrodes are rotated.

6. The internal defect detection method of the pre-baked anode according to claim 1, characterized in that, When identifying the internal defects of the pre-baked anode according to the obtained conductivity distribution inside the pre-baked anode: First, set the conductivity threshold range of the pre-baked anode material, and identify the regions where the conductivity inside the pre-baked anode is lower or higher than this conductivity threshold range as defect regions.

7. The internal defect detection method of the pre-baked anode according to claim 1, characterized in that, Present the positions, shapes and sizes of the identified internal defects of the pre-baked anode in a visual way.

8. The internal defect detection method of the pre-baked anode according to claim 1, wherein During the process of using the inverse problem solution of EIT to obtain the conductivity distribution inside the pre-baked anode, the process of establishing the finite element model of the forward problem is as follows: First, according to the actual geometric shape of the pre-baked anode and the arrangement of the electrodes on its surface, construct a finite element model of the pre-baked anode with electrodes arranged thereon. This finite element model is a finite element model of the pre-baked anode without internal defects; Then, perform mesh division and physical parameter setting on the finite element model of the pre-baked anode: Divide the finite element model of the pre-baked anode into several tetrahedral element meshes or hexahedral element meshes: among them, the part at the edge of the pre-baked anode is divided by hexahedral element meshes; the region inside the pre-baked anode is divided by tetrahedral element meshes; Assign an initial conductivity to each unit mesh according to the known conductivity characteristics of the pre-baked anode material; At the same time, set the boundary conditions and current excitation sources in the finite element model of the pre-baked anode according to the actual positions of the electrodes and the application method of the excitation current; Then, the finite element method is used to simulate the forward problem of EIT for the constructed finite element model, and the theoretical voltage values between the positions of any two adjacent electrodes on the surface of the pre-baked anode are obtained under the assumption that there are no defects inside the pre-baked anode.

9. The internal defect detection method of the pre-baked anode according to claim 1, wherein, Before using the inverse problem of EIT to solve for the conductivity distribution inside the pre-baked anode, the following steps are first performed: According to the actual geometric shape of the pre-baked anode and the arrangement of the electrodes on its surface, a finite element model of the pre-baked anode with electrodes arranged is constructed, and defects with a set shape and set size are preset at a set position inside the finite element model; The finite element model is divided into unit meshes, and each unit mesh is assigned a conductivity to obtain the set conductivity distribution inside the finite element model; Among them, the conductivity at the defect position is the air conductivity, and the conductivity at the remaining positions is the conductivity of the pre-baked anode material; The forward problem of EIT is solved for the finite element model to obtain the theoretical voltage values of each electrode under the assumption of preset defects inside the pre-baked anode; The theoretical voltage values of the electrodes are substituted into the calculation model for solving the inverse problem of EIT, and the theoretical conductivity distribution inside the pre-baked anode is obtained by using the inverse problem of EIT; The theoretical conductivity distribution is compared with the set conductivity distribution to determine whether they are consistent: If they are consistent, perform the step of "using the inverse problem of EIT to solve for the conductivity distribution inside the pre-baked anode based on the measured voltage values of the electrodes on the surface of the obtained pre-baked anode"; If they are not consistent, modify the electrode layout in the finite element model, including increasing the number of electrodes and / or reducing the spacing between adjacent electrodes, and then repeat the above steps for the modified finite element model until the theoretical conductivity distribution is consistent with the set conductivity distribution; finally, modify the electrode layout on the surface of the pre-baked anode corresponding to the optimized electrode layout.

Images