A method for weld joint detection for nuclear power plants

Abstract

This invention provides a weld inspection method for nuclear power plants, relating to the field of nuclear power engineering construction quality control technology. It includes collecting and processing weld construction data generated during nuclear power engineering construction to form an initial weld dataset. The collection and processing includes uniformly timestamping the weld construction data, spatial coordinate registration, and digital entry of construction information. This digital control method for weld sampling inspection in nuclear power engineering implements digital control of weld sampling inspection through steps such as collecting weld construction data, defining rules, strong logic control, and automatic batch processing during nuclear power engineering construction. Specifically, it includes measures such as timestamping, spatial coordinate registration, the appropriate application of non-destructive testing methods, and automatic feedback control, thereby ensuring the efficiency and standardization of the weld sampling inspection process.

Description

Technical Field

[0001] This invention relates to the field of nuclear power engineering construction quality control technology, specifically a weld inspection method for nuclear power plants. Background Technology

[0002] In nuclear power plant construction, welding, as a critical process, directly impacts the operational safety of equipment and pipelines. Currently, weld sampling inspection typically employs manual batching, with inspectors determining the sampling scope and proportion based on experience and site conditions. This method has been applied in several nuclear power plant reactor types, such as weld sampling based on the M310 reactor technical requirements. Some conventional pipeline sampling inspection software has emerged in the domestic market, offering basic data processing functions, but most fail to integrate with construction management systems, still requiring manual data maintenance across multiple systems. Foreign industrial software development is more advanced. GE's GEPredix platform integrates welding process parameters, weld information, and inspection data, using big data analysis and machine learning to perform sampling and quality prediction. Siemens' MindSphere platform connects welding and inspection equipment, enabling weld sampling and process optimization. Leveraging industrial internet and automation technologies, it possesses strong stability and data analysis capabilities, and has been applied in complex pipeline engineering. Overall, existing methods and tools can, to a certain extent, assist in weld sampling inspection management, providing quality monitoring and decision support for nuclear power plant construction.

[0003] However, current nuclear power construction projects are large-scale, with numerous and complexly distributed weld joints, making manual or general industrial software-based spot checks insufficient. The core issue is the lack of deep integration between traditional methods and nuclear power construction management systems, hindering the linkage of construction information, process control, and spot checks within a single data link. This leads to data gaps, preventing spot check results from being reflected in construction quality control in a timely manner, increasing the risk of information delays and errors. Given the extremely high safety and consistency requirements of nuclear power engineering, this deficiency makes it difficult for existing technologies to meet the demands of full-process digitalization and closed-loop management in nuclear power construction. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a weld inspection method for nuclear power plants. The technical problem this invention aims to solve is: how to address the issues of non-standard weld data management, complex sampling inspection processes, and excessive manual intervention in nuclear power engineering by adopting a digital management and control process.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a weld inspection method for nuclear power plants, comprising:

[0006] S1. Collect weld construction data during the construction process of nuclear power engineering to form an initial weld dataset; the collection includes performing unified timestamp marking, spatial coordinate registration, and digital entry of construction information on the weld construction data;

[0007] S2. The initial weld dataset is processed to define rules to form a batch sampling inspection rule library; the rule definition process sets the sampling ratio, full batch quantity, probing quantity, nuclear safety level, joint type, welding material and non-destructive testing method through a visual interface.

[0008] S3. The initial weld joint dataset and the batch sampling inspection rule base are correlated to form a set of weld joints to be inspected; the correlation process uses predefined logical verification rules to ensure data consistency.

[0009] S4. The set of weld joints to be inspected is automatically batched to generate batch data for inspection. The automatic batching process is based on nuclear safety level, joint type and welding material to select weld joints in proportion, and non-destructive testing methods are used to generate weld joint objects to be inspected.

[0010] S5. Perform non-destructive testing on the weld joint to be inspected and generate an inspection report; if the inspection is qualified, the data is fed back to the construction quality database; if the inspection is unqualified, a rework process is triggered, and the inspection is repeated after the rework is completed until it is qualified.

[0011] Preferably, the unified timestamp is based on the Network Time Protocol, the spatial coordinate registration performs coordinate reference transformation on the spatial location information of the weld construction data, and the digital entry of construction information includes entering the welder number, welding process card number, welding material batch number and welding completion date.

[0012] Preferably, the rule definition process includes automatically converting the sampling ratio and the full batch quantity, setting hierarchical logic for the probe quantity, parameterizing and storing the nuclear safety level and the connector type, and establishing a correspondence between the non-destructive testing method and the nuclear safety level. The automatic conversion includes that when the sampling ratio is 10%, the full batch quantity is 10, and when the sampling ratio is 5%, the full batch quantity is 20.

[0013] Preferably, the non-destructive testing method includes liquid penetrant testing, magnetic particle testing, X-ray testing, and ultrasonic testing.

[0014] Preferably, the association processing includes grouping the welds in the set of welds to be inspected. The grouping includes welds with the same welder number, welds with the same welding process card, welds with the same material, and welds with the same wall thickness. The strong logic control mechanism includes prohibiting the automatic batching processing when the welds have not completed visual inspection and when they are not bound to quality control documents.

[0015] Preferably, the automatic batch processing employs a weighted probability allocation formula, the calculation formula of which is:

[0016] .

[0017] in, Let be the dimensionless probability of selecting the i-th weld joint. This is the nuclear safety level factor, dimensionless, with a value range of 1-3. The material complexity factor is dimensionless and ranges from 1 to 5. This is a dimensionless weighting coefficient for nuclear safety levels, ranging from 0.4 to 0.6. This is a dimensionless weighting coefficient for material complexity, ranging from 0.4 to 0.6. This represents a weighted summation of all weld joints within a batch, and is dimensionless.

[0018] Preferably, the nuclear safety level factor is obtained by looking up a table in the nuclear power engineering standard document, and the material complexity factor is determined by a comprehensive score of the welding material strength grade and the welding process complexity.

[0019] Preferably, the qualified result is associated with the welder number, welding material batch number and inspection date during backfilling. When the inspection result is the unqualified result, the feedback processing generates an unqualified notification and pushes it to the construction team. The construction team performs the rework procedure according to the unqualified notification.

[0020] This invention provides a method for inspecting weld joints in nuclear power plants. It has the following beneficial effects:

[0021] This digital management and control method for weld sampling inspection in nuclear power engineering implements digital management and control of weld sampling inspection through steps such as data collection, rule definition, strong logic control, and automatic batch processing during the construction of nuclear power projects. Specifically, it includes measures such as timestamp marking, spatial coordinate registration, the appropriate application of non-destructive testing methods, and automatic feedback control, thereby ensuring the efficiency and standardization of the weld sampling inspection process.

[0022] The system employs automated batch processing, robust logic control, and weighted probability allocation formulas to ensure the accuracy and fairness of the sampling inspection process. Simultaneously, the pass / fail results generated through non-destructive testing methods are fed back to the construction quality database, forming a closed-loop management system and thereby enhancing the transparency and reliability of construction quality management. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating the overall process of implementing an invention.

[0024] Figure 2 It is a feedback processing flowchart for realizing an invention;

[0025] Figure 3 It is a flowchart of a non-destructive welding process for realizing an invention. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] Example 1

[0028] like Figure 1-3 As shown, this embodiment of the invention provides a weld inspection method for nuclear power plants, including: S1. Collecting and processing weld construction data generated during the construction of nuclear power engineering to form an initial weld dataset. The collection and processing includes uniform timestamp marking, spatial coordinate registration, and digital entry of construction information for the weld construction data. The uniform timestamp marking adopts the Network Time Protocol (NTP). Spatial coordinate registration performs coordinate reference transformation on the spatial location information of the weld construction data. Digital entry of construction information includes entering the welder's number, welding process card number, welding material batch number, and welding completion date.

[0029] S2. The initial weld dataset is processed to form a batch sampling inspection rule library. This rule definition process employs a visual maintenance method, which includes setting and invoking sampling ratios, full batch quantities, probe quantity, nuclear safety level, joint type, welding material, and non-destructive testing methods. The rule definition process includes automatic conversion between sampling ratios and full batch quantities, setting hierarchical logic for probe quantity, parameterizing and storing nuclear safety level and joint type, and establishing a correspondence between non-destructive testing methods and nuclear safety levels. Automatic conversion includes setting a full batch quantity of 10 when the sampling ratio is 10%, and 20 when the sampling ratio is 5%. Non-destructive testing methods include liquid penetrant testing, magnetic particle testing, radiographic testing, and ultrasonic testing.

[0030] S3. The initial weld dataset and the batch sampling rule base are correlated to form a set of welds to be sampled. The correlation processing adopts a strong logic control mechanism. The correlation processing includes grouping the welds in the set to be sampled. The groups include welds with the same welder number, welds with the same welding process card, welds with the same material, and welds with the same wall thickness. The strong logic control mechanism prohibits automatic batch processing when the welds have not completed visual inspection or are not bound to quality control documents.

[0031] S4. The set of weld joints to be sampled is automatically batched to form batch data. Automatic batching uses proportional sampling, which is combined with nuclear safety level, joint type, and welding material. The batch data is used to generate weld joint objects to be inspected using non-destructive testing methods. Automatic batching uses a weighted probability allocation formula, the calculation formula of which is:

[0032] .

[0033] in, Let be the dimensionless probability of selecting the i-th weld joint. This is the nuclear safety level factor, dimensionless, with a value range of 1-3. The material complexity factor is dimensionless and ranges from 1 to 5. This is a dimensionless weighting coefficient for nuclear safety levels, ranging from 0.4 to 0.6. This is a dimensionless weighting coefficient for material complexity, ranging from 0.4 to 0.6. This represents a weighted summation of all weld joints within a batch, and is dimensionless. The nuclear safety level factor is obtained from tables in nuclear power engineering standard documents, and the material complexity factor is determined by a comprehensive score of the welding material strength grade and welding process complexity.

[0034] S5. Non-destructive testing (NDT) is performed on the weld joints to be inspected to generate inspection results. NDT processing includes generating an inspection report, and the results include both pass and fail outcomes. Feedback processing of the inspection results forms a closed-loop control system. Feedback processing includes backfilling pass outcomes into the construction quality database and triggering rework procedures for fail outcomes. After the rework procedures are completed, NDT is re-performed until a pass outcome is obtained. Pass outcomes are associated with the welder's number, welding material batch number, and inspection date during backfilling. When an outcome is fail, feedback processing generates a fail notification and sends it to the construction team. The construction team then executes the rework procedure based on the fail notification.

[0035] By automatically calculating and optimizing sampling ratios and batch sizes using hierarchical logic, the accuracy and rationality of the testing plan are ensured. Establishing a correspondence between non-destructive testing methods and nuclear safety levels helps in selecting appropriate testing methods based on different safety levels, thereby improving safety. Through weighted probability allocation, automatic batch processing ensures that high-risk and complex welds are prioritized for sampling, thus optimizing the allocation of quality control resources. The closed-loop management mechanism, by providing feedback on pass and fail results, promptly corrects quality problems, ensuring that welds meet quality standards and enhancing the safety and reliability of nuclear power engineering.

[0036] Example 2

[0037] This embodiment is based on a digital management and control method for weld sampling inspection process in nuclear power engineering. Through rule definition processing, the initial weld dataset is transformed into a batch sampling inspection rule library, and it ensures that various parameters, rules and standards can be managed reasonably and automatically.

[0038] 1. Initialize the parameters of the weld joint dataset.

[0039] To effectively implement the rule definition, the weld joint dataset must first be initialized to ensure that it contains the following key information:

[0040] Table 1: Initialized weld joint data table.

[0041] Weld ID Welder number Welding process card number Welding material batch number Construction Date Nuclear safety level Connector type Welding material Wall thickness W001 W1001 C001 A123 2023-05-01 3 Butt welding Stainless steel 20mm W002 W1002 C002 A124 2023-05-03 2 Butt welding low alloy steel 25mm W003 W1001 C001 A125 2023-05-04 1 Corner weld Stainless steel 15mm

[0042] 2. Rule definition and parameter setting

[0043] Based on the weld joint dataset, rule definition processing is performed, mainly including the setting of parameters such as sampling ratio, full batch quantity, probing quantity, nuclear safety level, joint type, welding material, and non-destructive testing method.

[0044] Rules for calculating the sampling ratio and full batch quantity:

[0045] The sampling rate is set and automatically calculated to correlate with the full batch quantity. The sampling rate is adjusted based on the nuclear safety level and project requirements.

[0046] Calculation formula:

[0047]

[0048] When the sampling rate is 10%, the system automatically calculates that the full batch quantity is 10 weld joints.

[0049] When the sampling rate is 5%, the system automatically calculates that the full batch quantity is 20 weld joints.

[0050] The hierarchical logic of the number of exploration runs:

[0051] The number of probes required is adjusted based on the nuclear safety level. The higher the nuclear safety level, the more probes are required.

[0052] Rule settings:

[0053] The number of weld probing probes for nuclear safety level 1 is 0, for nuclear safety level 2 it is 2, and for nuclear safety level 3 it is 4.

[0054] The weld ID is W002, the nuclear safety level is 2, and the system automatically calculates and sets the probe quantity to 2.

[0055] The weld ID is W001, the nuclear safety level is 3, and the system automatically calculates and sets the probe quantity to 4.

[0056] The correspondence between nuclear safety levels and non-destructive testing methods:

[0057] The nuclear safety level directly determines the choice of non-destructive testing methods. Welds with different nuclear safety levels will use different testing technologies to ensure the effectiveness of the testing.

[0058] Non-destructive testing method setting:

[0059] Nuclear safety level 1 uses liquid penetration testing, nuclear safety level 2 uses magnetic particle testing or ultrasonic testing, and nuclear safety level 3 uses radiation testing.

[0060] Weld ID W003, nuclear safety level 1, uses liquid penetrant testing. Weld ID W002, nuclear safety level 2, uses magnetic particle testing or ultrasonic testing. Weld ID W001, nuclear safety level 3, uses radiographic testing.

[0061] 3. Generation of the batch sampling rule library

[0062] Based on the various attributes of the weld joint, the system will automatically generate a batch sampling inspection rule library. Weld joint batches will then be allocated according to the set sampling ratio, probing quantity, and other rules.

[0063] Weld ID W001, nuclear safety level 3, allocated to a sampling batch with a sampling rate of 10%, and 4 welds subjected to further investigation, using radiographic testing. Weld ID W002, nuclear safety level 2, allocated to a sampling batch with a sampling rate of 5%, and 2 welds subjected to further investigation, using magnetic particle testing or ultrasonic testing. Weld ID W003, nuclear safety level 1, allocated to a sampling batch with a sampling rate of 10%, and 0 welds subjected to further investigation, using liquid penetrant testing.

[0064] 4. Output sampling rule library

[0065] The system-generated sampling rule library records detailed information such as grouping information, sampling ratio, full batch quantity, probe quantity, and testing method for each weld joint, and forms the final testing plan.

[0066] Table 2: Output table of generated data.

[0067] Weld ID Nuclear safety level Sampling ratio Full batch quantity Expansion quantity Non-destructive testing methods W001 3 10% 10 4 X-ray inspection W002 2 5% 20 2 Ultrasonic testing W003 1 10% 10 0 Liquid Penetration Testing

[0068] 5. Implementation of visualization tools

[0069] To improve the efficiency and accuracy of rule setting, it is recommended to use a visual tool for rule input and maintenance. Visual tools provide an intuitive interface that allows for real-time viewing and modification of parameters. The visual interface features are as follows:

[0070] It allows users to select parameters such as sampling ratio and nuclear safety level, and automatically selects appropriate non-destructive testing methods by calculating the full batch quantity and the number of probes in real time.

[0071] Through the above steps, precise grouping and random inspection of weld joints were achieved. The automatic selection of non-destructive testing methods ensured the compliance and safety of each weld joint inspection. The use of visual tools for rule setting and maintenance simplified the operational process and improved management efficiency and transparency. This effectively reduced manual intervention and enhanced the level of quality control and safety management in nuclear power engineering.

[0072] Example 3

[0073] This embodiment is based on a digital management and control method for weld sampling inspection in nuclear power engineering. It uses a weighted probability allocation formula to automatically select welds to be inspected, ensuring that the weld sampling inspection process meets quality control standards. The specific implementation method is as follows:

[0074] 1. Determine the set of welds to be inspected.

[0075] The nuclear safety level factor is obtained by referring to tables in nuclear power engineering standard documents, while the material complexity factor is determined by a comprehensive score of welding material strength grade and welding process complexity. The nuclear safety level factor ranges from 1 to 3, and the material complexity factor ranges from 1 to 5.

[0076] Based on nuclear power engineering data, data from five weld joints to be inspected were collected. The specific data are as follows:

[0077] Table 3: Weld joint data table.

[0078] Weld joint number Nuclear safety level Material complexity 1 2 4 2 3 5 3 1 3 4 2 4 5 3 2

[0079] 2. Determine the parameters of the weighted probability allocation formula.

[0080] According to the quality inspection standards for nuclear power engineering, the weighting coefficient for nuclear safety level ranges from 0.4 to 0.6, and the weighting coefficient for material complexity ranges from 0.4 to 0.6.

[0081] Setting weights for nuclear safety levels Weight of material complexity .

[0082] Automatic batch processing uses a weighted probability allocation formula, which is calculated as follows:

[0083] .

[0084] in, Let be the dimensionless probability of selecting the i-th weld joint. This is the nuclear safety level factor, dimensionless, with a value range of 1-3. The material complexity factor is dimensionless and ranges from 1 to 5. This is a dimensionless weighting coefficient for nuclear safety levels, ranging from 0.4 to 0.6. This is a dimensionless weighting coefficient for material complexity, ranging from 0.4 to 0.6. This represents a weighted summation of all weld joints within a batch, and is dimensionless. The nuclear safety level factor is obtained from tables in nuclear power engineering standard documents, and the material complexity factor is determined by a comprehensive score of the welding material strength grade and welding process complexity.

[0085] 3. Calculate the sampling probability.

[0086] Calculate the weighted sum of all weld joints:

[0087] .

[0088] .

[0089] Calculate the probability of each weld joint:

[0090] Weld 1:

[0091]

[0092] Weld joint 2:

[0093]

[0094] Weld joint 3:

[0095]

[0096] Weld joint 4:

[0097]

[0098] Weld joint 5:

[0099]

[0100] 4. Generation of sampling batches

[0101] Determine the sampling ratio:

[0102] When the sampling rate is 10%, one weld joint is randomly selected from every five weld joints for testing.

[0103] Generate sampling batches:

[0104] Based on the calculated probabilities, the weld joint with the highest probability is selected for random inspection. Since weld joint 2 has the highest inspection probability of 0.276, it will be selected for inspection.

[0105] 5. Practical Operation

[0106] Batch processing: Once weld joint 2 is selected, the system will automatically assign weld joint 2 to the batch to be inspected and select an appropriate non-destructive testing method for inspection.

[0107] Non-destructive testing: Ultrasonic testing was selected for weld joint 2. The weld joint will undergo non-destructive testing, and a test result report will be generated.

[0108] 6. Feedback of test results

[0109] Generate test report:

[0110] Based on the test results, an inspection report for weld 2 is generated. If weld 2 is qualified, the report will record information such as welder number, welding material batch number, and inspection date. If it is unqualified, the system will automatically trigger a rework process and push feedback information to the construction team.

[0111] Data backfilling:

[0112] The qualified results will be backfilled into the construction quality database and linked with the relevant construction data.

[0113] In summary, by applying a weighted probability allocation formula and comprehensively considering factors such as nuclear safety level and material complexity, the sampling priority of weld joints can be scientifically and rationally determined. This ensures the fairness and efficiency of the sampling process, reduces the influence of human factors, further improves the accuracy and reliability of weld joint quality control, and provides strong support for the safety and stability of nuclear power projects.

[0114] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for inspecting weld joints in nuclear power plants, characterized in that, include: S1. Collect weld construction data during the construction process of nuclear power engineering to form an initial weld dataset; the collection includes performing unified timestamp marking, spatial coordinate registration, and digital entry of construction information on the weld construction data; S2. The initial weld dataset is processed to define rules to form a batch sampling inspection rule library; the rule definition process sets the sampling ratio, full batch quantity, probing quantity, nuclear safety level, joint type, welding material and non-destructive testing method through a visual interface. S3. The initial weld joint dataset and the batch sampling inspection rule base are correlated to form a set of weld joints to be inspected; the correlation process uses predefined logical verification rules to ensure data consistency. S4. The set of weld joints to be inspected is automatically batched to generate batch data for inspection. The automatic batching process is based on nuclear safety level, joint type and welding material to select weld joints in proportion, and non-destructive testing methods are used to generate weld joint objects to be inspected. S5. Perform non-destructive testing on the weld joint to be inspected and generate an inspection report; if the inspection is qualified, the data is fed back to the construction quality database; if the inspection is unqualified, a rework process is triggered, and the inspection is repeated after the rework is completed until it is qualified.

2. The weld inspection method for nuclear power plants according to claim 1, characterized in that: The unified timestamp marker adopts the network time protocol, the spatial coordinate registration performs coordinate reference transformation on the spatial location information of the weld construction data, and the digital entry of construction information includes entering the welder number, welding process card number, welding material batch number and welding completion date.

3. The weld inspection method for nuclear power plants according to claim 1, characterized in that: The rule definition process includes automatically converting the sampling ratio and the full batch quantity, setting hierarchical logic for the probe quantity, parameterizing and storing the nuclear safety level and the connector type, and establishing a correspondence between the non-destructive testing method and the nuclear safety level. The automatic conversion includes that when the sampling ratio is 10%, the full batch quantity is 10, and when the sampling ratio is 5%, the full batch quantity is 20.

4. The weld inspection method for nuclear power plants according to claim 1, characterized in that: The non-destructive testing methods include liquid penetrant testing, magnetic particle testing, X-ray testing, and ultrasonic testing.

5. The weld inspection method for nuclear power plants according to claim 1, characterized in that: The association processing includes grouping the welds in the set of welds to be inspected. The grouping includes welds with the same welder number, welds with the same welding process card, welds with the same material, and welds with the same wall thickness. The strong logic control mechanism includes prohibiting the automatic batch processing when the welds have not completed visual inspection and when they are not bound to quality control documents.

6. The weld inspection method for nuclear power plants according to claim 1, characterized in that: The automatic batch processing uses a weighted probability allocation formula, and the calculation formula for the weighted probability allocation formula is as follows: ， in, Let be the probability of selecting the i-th weld joint. This is the nuclear safety level factor, with a value ranging from 1 to 3. This is the material complexity factor, with a value ranging from 1 to 5. This is the nuclear safety level weighting coefficient, with a value ranging from 0.4 to 0.

6. This is the material complexity weighting coefficient, with a value ranging from 0.4 to 0.

6. This represents a weighted sum of all weld joints within a batch.

7. A weld inspection method for nuclear power plants according to claim 6, characterized in that: The nuclear safety level factor is obtained by looking up tables in nuclear power engineering standard documents, and the material complexity factor is determined by a comprehensive score of welding material strength grade and welding process complexity.

8. The weld inspection method for nuclear power plants according to claim 1, characterized in that: When the qualified result is backfilled, it is associated with the welder number, welding material batch number and inspection date. When the inspection result is the unqualified result, the feedback processing generates an unqualified notice and pushes it to the construction team. The construction team performs the rework procedure according to the unqualified notice.

Images