A device and method for repairing defects in fiber reinforced composites

Abstract

The application provides a kind of fiber reinforced composite material defect repair equipment and batch repair method, the repair equipment includes atmospheric pressure plasma processing module, the detection analysis module of the repair site of the fiber reinforced composite material to be repaired, repair data monitoring and information flow module, post-repair detection module and repair operation process planning module.The batch repair method based on the repair equipment is also disclosed, including constructing repair sample data set, using machine learning method to construct and train composite material repair machine learning model;Based on machine learning model, generate personalized repair process and parameters based on repair data, batch repair, and constantly update iteration model.The application can repair multiple samples simultaneously by establishing batch repair equipment, combined with automated transmission and precise process control, to improve repair efficiency.

Description

Technical Field

[0001] The present invention belongs to the field of fiber reinforced composite materials, and in particular relates to a device for repairing defects in fiber reinforced composite materials and a batch repair method. Background Art

[0002] Fiber-reinforced composites (FRCs), with their exceptional performance, are widely used in aerospace, energy and power, electronics, biomedicine, and other fields, becoming a key foundational material for lightweight, functionalized construction. However, surface defects are inevitable during the manufacturing and use of FRCs. Therefore, surface repair of FRCs is a key technology for ensuring the stability of composite performance.

[0003] Currently, the main surface repair methods rely on manual grinding, sandblasting, and chemical treatment. Some existing technologies also use auxiliary technologies for repair. For example, Chinese invention patent CN116749561A discloses a method for repairing carbon fiber composite materials under microwave plasma surface treatment. The technical solution of this patent is to wash and dry the carbon fiber composite material to be repaired, place it on the workbench of a microwave plasma device, use air as the plasma gas in a normal pressure environment, make the workbench reciprocate, clean the area to be repaired, paste T800 prepreg on the area to be repaired, and microwave cure to complete the repair. Although this method restores the mechanical properties of the carbon fiber composite material to a certain extent, the use of the prepreg curing method still has certain unevenness in the repair material. At the same time, due to the diversity of composite material components, processes, structures, and defects, targeted repair cannot be carried out. In addition, the repair efficiency of this method is low. As repair technology develops towards low cost, greenness, and rapidity, existing technologies have been unable to meet actual needs. Summary of the Invention

[0004] The purpose of the present invention is to overcome the above-mentioned defects of the prior art and provide an intelligent repair device and a high-throughput repair method for fiber-reinforced composite materials.

[0005] One aspect of the present invention provides a device for repairing defects in fiber-reinforced composite materials, the device comprising a main module: an atmospheric pressure plasma treatment module, a detection and analysis module for the repaired part of the fiber-reinforced composite material to be repaired, a repair data monitoring and information flow module, a post-repair detection module, and a repair operation process planning module.

[0006] The atmospheric pressure plasma treatment module is used for cleaning and activation treatment and depositing polymer for repair;

[0007] The detection and analysis module for the part to be repaired can scan and analyze the surface of the fiber reinforced composite material to be repaired and obtain defect morphology data for initial defect morphology classification;

[0008] The post-repair inspection module is capable of scanning and analyzing the surface of the repaired fiber-reinforced composite material and obtaining inspection data for classifying the defect repair status;

[0009] The repair data monitoring and information flow module is used to monitor the repair process of the fiber reinforced composite material to be repaired, record and transmit the spatiotemporal data of the repair process, and confirm and adjust the parameters used by the atmospheric pressure plasma treatment module;

[0010] The repair operation process planning module is used to call information from other modules, plan and control the specific operation steps of the atmospheric pressure plasma treatment module during the repair process, and output repair parameters.

[0011] Furthermore, the atmospheric pressure plasma processing module in the fiber-reinforced composite material defect repairing equipment includes a discharge gas release unit, a silicone monomer supply unit, a flow control unit, a heating unit, and a power supply system; the flow control unit can control the volume ratio of the discharge gas released by the discharge gas release unit and the silicone provided by the silicone supply unit to be 1:0.01 to 1:0.1.

[0012] Furthermore, the heating unit can provide a heating temperature of 60° C.-150° C. for the fiber reinforced composite material.

[0013] Furthermore, the discharge gas release unit can release one of argon, nitrogen and air.

[0014] Furthermore, the organosilicon monomer supply unit can provide an organosilane monomer selected from one of alkoxysilane, chlorosilane, aminosilane, epoxysilane, and vinylsilane.

[0015] Furthermore, the power supply system is used to provide stable and qualified electric energy to the discharge gas release unit, so that the discharge gas release unit can ionize the gas to generate plasma.

[0016] Furthermore, the above-mentioned fiber reinforced composite material defect repairing device also includes a post-processing module, and the post-processing module is used to heat and polish the repaired fiber reinforced composite material.

[0017] Another aspect of the present invention provides a method for batch repairing defects in fiber-reinforced composite materials, comprising the following steps:

[0018] S1) Building equipment to repair defects in the aforementioned fiber-reinforced composite materials;

[0019] S2) performing a repair test using the above-mentioned fiber reinforced composite material defect repair equipment and collecting test data from the repair test; and constructing a repair sample data set based on the test data;

[0020] S3) Based on the repair sample dataset, a machine learning method is used to build and train a machine learning model for composite material repair;

[0021] S4) Based on the repair machine learning model obtained in step S3), as well as real-time data obtained by the repair part detection and analysis module, the repair data monitoring and information flow module, and the repair operation process planning module, a personalized repair process and parameters for repairing defects in the fiber-reinforced composite material are generated, and the atmospheric pressure plasma treatment module is called through the repair data monitoring and information flow module to batch repair the fiber-reinforced composite material;

[0022] S5) Collect data from the repair part detection and analysis module, repair data monitoring and information flow module, post-repair detection module, and repair operation process planning module during the repair process, and add the data to the repair database to update and iterate the repair machine learning model.

[0023] The repair test in step S2) includes the following steps:

[0024] The fiber-reinforced composite material to be repaired is placed in an atmospheric pressure plasma jet for treatment, wherein the atmospheric pressure plasma is generated by a discharge gas and an organosilicon monomer, and the volume ratio of the discharge gas to the organosilicon monomer is 1:0.01 to 1:0.1;

[0025] The treatment in the atmospheric-pressure plasma jet is followed by a post-treatment step, which is heating.

[0026] One of the organosilicon monomers selected from alkoxysilane, chlorosilane, aminosilane, epoxysilane and vinylsilane;

[0027] The organosilicon monomer is selected from one of alkoxysilane, chlorosilane, aminosilane, epoxysilane and vinylsilane;

[0028] The discharge gas is selected from one of argon, nitrogen and air.

[0029] Before the fiber-reinforced composite material to be repaired is placed under an atmospheric pressure plasma jet for treatment, a pretreatment step is also included, wherein the pretreatment is to clean and activate the fiber-reinforced composite material to be repaired;

[0030] The restoration also includes polishing.

[0031] Furthermore, the type of fiber in the fiber-reinforced composite material is one of glass fiber, basalt fiber, carbon fiber, aramid fiber, polyimide fiber, and polyarylate fiber.

[0032] Furthermore, the type of resin in the fiber-reinforced composite material is one of unsaturated polyester resin, epoxy resin, phenolic resin, and silicone resin.

[0033] Beneficial effects

[0034] This method uses atmospheric-pressure plasma generated by discharge gas and organosilicon monomers to deposit the polymer. By chemically bonding the organosilicon polymer to the resin matrix at a specific temperature, the two are composited, significantly shortening repair time and improving efficiency. Requiring no vacuum equipment or large amounts of reagents, the reaction occurs in an open environment, reducing energy consumption, pollution, and equipment maintenance costs. The method is compatible with a variety of composite materials, and effective repair can be achieved by adjusting parameters, broadening its application areas.

[0035] The present invention constructs batch repair equipment, combines automated transmission with precise process control, and can repair multiple samples simultaneously, thereby improving repair output.

[0036] Furthermore, the methods and equipment of this invention leverage extensive sample data to accurately generate personalized repair plans tailored to the composite material and defect profile, thereby increasing success rates. Real-time monitoring of the repair process allows for quality predictions compared to historical data, enabling timely process adjustments to ensure product quality and reduce scrap rates. Continuous optimization and data accumulation enable the discovery of new repair patterns and methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 The figure is a flow chart of the method for batch repairing defects of fiber-reinforced composite materials according to the present invention.

[0038] Figure 2 Microscope photograph of a fiber-reinforced composite material with surface defects.

[0039] Figure 3 Microscope photograph of fiber-reinforced composite material after high-throughput repair. DETAILED DESCRIPTION

[0040] Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that unless otherwise specifically stated, the relative arrangement of components and steps, numerical expressions, and values ​​set forth in these embodiments do not limit the scope of the present invention. The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses.

[0041] Techniques, methods, and apparatus known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and apparatus should be considered part of the specification. In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0042] The present invention aims to provide an intelligent repair device and high-throughput repair method for fiber-reinforced composite materials. Combining atmospheric pressure plasma technology with machine learning, the method offers a low-cost, automated, green, rapid, and personalized repair method for fiber-reinforced composite materials. This method is capable of performing targeted, batch repairs of different types of defects in different fiber-reinforced composite materials at a rapid, environmentally friendly, and pollution-free rate. This method addresses the existing issues of fiber-reinforced composite repair equipment, which is complex, requires vacuum pressurization, and cannot perform personalized, batch repairs.

[0043] Specifically, combined Figure 1 As shown, the method for repairing defects of fiber-reinforced composite materials in batches provided by the present invention comprises the following steps:

[0044] Step S1) building a repair device;

[0045] The repairing device is based on atmospheric pressure plasma technology, and utilizes the reaction of silane monomer material and discharge gas to produce organic silicon polymer to achieve the repair of defects in fiber-reinforced composite materials.

[0046] The repair equipment includes: an atmospheric pressure plasma processing module, a detection and analysis module for the repaired part of the fiber reinforced composite material to be repaired, a repair data monitoring and information flow module, a post-repair detection module, and a repair operation process planning module.

[0047] The atmospheric pressure plasma treatment module, which performs cleaning and activation treatments as well as polymer deposition for repair, includes a discharge gas release unit, an organosilicon monomer supply unit, a flow control unit, a heating module, and a power supply system. This device generates atmospheric pressure plasma by mixing discharge gas and organosilicon monomer in a specific ratio, with the gas ratio controlled between 1:0.01 and 1:0.1. The discharge gas can be selected from argon, nitrogen, or air, and the organosilane monomer can be selected from alkoxysilane, chlorosilane, aminosilane, epoxysilane, or vinylsilane.

[0048] The heating unit can heat the fiber reinforced composite material to a temperature of 60° C. to 150° C.

[0049] The various operating parameters of the atmospheric pressure plasma treatment module are controlled by the repair operation process planning module. At the same time, the various operating parameters related to the atmospheric pressure plasma treatment module are recorded and transmitted by the repair data monitoring and information flow module.

[0050] In a specific embodiment, the repair device further includes a post-processing module.

[0051] In a specific embodiment, the operating parameters include the regulation of the working voltage of the atmospheric pressure plasma treatment module, the discharge gas flow rate, the flow rate of the organic silicon monomer, the distance control between the atmospheric pressure plasma and the surface of the material to be repaired, and the heating temperature of the heating unit.

[0052] The repair site inspection and analysis module is used to scan and analyze the surface of the fiber-reinforced composite material to be repaired. It includes a detection module and an analysis module. The detection module is used to collect data on surface defects of the fiber-reinforced composite material to be repaired. The detection module is selected from a visual imaging element. The analysis module is used to analyze the data obtained by the detection module to determine the type, size, shape, and depth of the surface defects of the repaired fiber-reinforced composite material. The data obtained from the inspection and analysis are transmitted to the repair operation process planning module via the repair data monitoring and information flow module.

[0053] In a specific embodiment, the imaging element may be a high-resolution optical imaging unit, which can accurately image the surface of the material, obtain visual data such as macroscopic morphology information of the material surface, and identify obvious damaged areas, crack directions, etc.

[0054] In a specific embodiment, the imaging element uses the OPT-SC series camera produced by Guangdong Opto Technology Co., Ltd., which has a built-in image sensor for converting light signals into digital image signals. At the same time, when used with the SciVision visual development kit, image processing and analysis can be achieved. In order to improve the imaging effect, a supplementary light source can also be provided to provide uniform and sufficient lighting for the shooting scene. The intelligent optical component also plays a key role in the imaging process. It is mainly composed of a laser sensor and an optical lens. Among them, the laser sensor is used to assist in measuring the distance, shape and other information of the target object, providing additional data support for image analysis; the optical lens is responsible for focusing light to ensure that the image is clearly and accurately projected onto the camera's image sensor.

[0055] The post-repair inspection module is used to scan and analyze the surface of the repaired fiber-reinforced composite material to obtain inspection data. The post-repair inspection module includes a detection module and an analysis module. The detection module is used to collect data on the repaired fiber-reinforced composite material, and the detection module is selected from an imaging element. The analysis module is used to analyze the data obtained by the detection module to determine the repair status of the surface defects of the repaired fiber-reinforced composite material. The data obtained from the inspection and analysis are transmitted to the repair operation process planning module via the repair data monitoring and information flow module.

[0056] In a specific embodiment, the repair device includes a post-processing module, and the post-repair detection module is used to detect the repaired fiber-reinforced composite material that has been post-processed by the post-processing module.

[0057] In a specific embodiment, the repair device does not include a post-processing module, and the post-repair detection module is used to detect the repaired fiber-reinforced composite material treated by the atmospheric pressure plasma treatment module.

[0058] The repair data monitoring and information flow module monitors the repair process of the fiber reinforced composite material and confirms and adjusts the operating parameters used by the atmospheric pressure plasma treatment module;

[0059] In a specific embodiment, monitoring the repair process of the fiber-reinforced composite material to be repaired is to monitor the operating parameters adopted by the atmospheric pressure plasma treatment module during the repair process, and the operating parameters include the operating voltage of the atmospheric pressure plasma treatment module, the discharge gas flow rate, the regulation of the silicone monomer flow rate, and the control of the distance between the atmospheric pressure plasma and the surface of the material to be repaired.

[0060] For example, flow meters for monitoring the flow of discharge gas and silicone monomer, voltage sensors for measuring working voltage, and distance sensors for obtaining the distance between atmospheric pressure plasma and the surface of the material to be repaired can be used to continuously collect corresponding data information.

[0061] In a specific embodiment, the method for confirming and adjusting the operating parameters used by the atmospheric pressure plasma processing module is manual setting.

[0062] In a specific embodiment, the method for confirming and adjusting the operating parameters used by the atmospheric pressure plasma processing module is real-time control through a model obtained through machine learning.

[0063] Specifically, real-time data, such as defect type and size, is acquired from the repair data monitoring and information flow module and the repaired area detection and analysis module and transmitted to the machine learning model. Upon receiving this real-time data, the model uses a trained algorithm to analyze and calculate operating parameters such as the atmospheric pressure plasma treatment module's optimal operating voltage, discharge gas flow rate, silicone monomer flow rate, and the distance between the nozzle and the material to be repaired under the current operating conditions. The model then automatically issues commands to precisely control the relevant control components, ensuring the device is always operating optimally and achieving efficient repair of fiber-reinforced composite materials.

[0064] The repair operation process planning module is used to call information from other modules to plan and control specific operation steps of the atmospheric pressure plasma treatment module during the repair process.

[0065] Specifically, the repair operation process planning module calls the data monitored and obtained in the repair data monitoring and information flow module to plan and control the specific operation steps of the atmospheric pressure plasma treatment module during the repair process, such as controlling the opening and closing time of the discharge gas release unit and the silicone monomer supply unit in the atmospheric pressure plasma treatment module, and the opening and closing time of the atmospheric pressure plasma treatment module.

[0066] In order to further optimize the repair effect, a post-processing module can be added. In a specific embodiment, the post-processing module is a polishing device, which includes a connecting rod, a polishing pad and a moving part. The moving part drives the connecting rod to move to the repair area, and the polishing pad fixed on the connecting rod polishes the surface of the fiber-reinforced composite material.

[0067] S2) performing a repair test using the aforementioned fiber-reinforced composite material defect repair equipment and collecting test data from the repair test; constructing a repair sample dataset based on the test data; the repair sample dataset supports access to multiple data types and multiple sources;

[0068] The multiple data types include structured data, unstructured data, multi-dimensional spatiotemporal data and real-time streaming data;

[0069] In some specific embodiments, the structured data includes parameter setting records recorded by the fiber-reinforced composite material defect repair equipment during the experiment, such as repair time, repair parameters, data collected by the detection and analysis module of the part to be repaired, and data obtained after analog-to-digital conversion and data sorting.

[0070] Unstructured data includes visual information such as images collected before, during, and after restoration, and can also include text information such as records and descriptions of the test process made by the test personnel.

[0071] The multi-dimensional spatiotemporal data include state change data of the fiber reinforced composite material to be repaired at different stages of repair.

[0072] Real-time streaming data includes data continuously generated by the equipment during operation. Sensors on repair equipment collect physical quantities such as temperature and flow in real time, transmitting them as a continuous data stream to continuously reflect the current status of the equipment.

[0073] Multi-source data access is achieved through various interfaces. This data can include the repair equipment's own settings and collected data, as well as test data transmitted through data connections between different components. The simulation software uses a specific data export function to output simulation data in a pre-set format. After various data sources are connected, they undergo pre-processing steps such as data cleaning and format conversion before being integrated into the repair sample dataset.

[0074] The repair test comprises the following steps:

[0075] Pre-treating the fiber-reinforced composite material to be repaired by cleaning and activation;

[0076] The fiber-reinforced composite material to be repaired is placed in an atmospheric pressure plasma jet for treatment, wherein the atmospheric pressure plasma is generated by a discharge gas and an organosilicon monomer, and the volume ratio of the discharge gas to the organosilicon monomer is 1:0.01 to 1:0.1;

[0077] The organosilicon monomer is selected from one of alkoxysilane, chlorosilane, aminosilane, epoxysilane and vinylsilane;

[0078] The discharge gas is selected from one of argon, nitrogen and air.

[0079] After treatment under the atmospheric pressure plasma jet, a post-treatment step is performed. This post-treatment involves heating at a temperature of 60-150°C, specifically the temperature at which the repaired fiber-reinforced composite material begins to change color. Heating is an essential step in the repair process, ensuring a tight bond between the repaired materials and a consistent appearance after the repair. Polishing can also be performed as needed.

[0080] S3) Based on the repair sample dataset, a machine learning method is used to build and train a machine learning model for composite material repair;

[0081] The specific process of constructing and training the machine learning model is to use the defect morphology data, repair process data, and post-repair inspection and detection data in the repair sample data set to construct and train the machine learning model.

[0082] It includes the following steps:

[0083] S301) Data preprocessing:

[0084] Different types of data are preprocessed separately, and image recognition technology is used to extract features from defect morphology data and visual data.

[0085] For the repair process data, normalization processing is used to handle different types of data. At the same time, differential processing can be performed on the time series repair process data to capture the dynamic characteristics of the repair process.

[0086] In some specific solutions, sliding window technology can also be used to divide continuous time series data into windows of fixed length as input to the model so that the model can learn local time series patterns.

[0087] S302) Model training

[0088] The repair model is trained using a repair sample dataset. A model is constructed to establish a correspondence between the operating parameters of the plasma generator and the spatiotemporal data of the repair defect morphology. The input is the operating parameters, and the output is the spatiotemporal data of the defect repair morphology. This model then reveals the correspondence between the operating parameters and the defect morphology data, inspection data, and repair process data reflected in the repair sample dataset. Based on the characteristics of the data and the requirements of the repair task, an appropriate machine learning algorithm, such as a neural network, random forest, or support vector machine, is selected, and the corresponding model structure is designed.

[0089] In some specific embodiments, the operating parameters include the regulation of the atmospheric pressure plasma treatment module's operating voltage, discharge gas flow rate, and organosilicon monomer flow rate, as well as the distance between the atmospheric pressure plasma and the surface of the material to be repaired. The spatiotemporal data of the defect repair morphology includes the size, shape, and completeness of the defect repair at a specific moment.

[0090] S4) Based on the repair machine learning model obtained in step S3), the data obtained by the to-be-repaired part detection and analysis module, and the expected repair target, a personalized repair process and parameters for repairing defects in the fiber-reinforced composite material are generated, and the atmospheric pressure plasma treatment module is called through the repair data monitoring and information flow module to batch repair the fiber-reinforced composite material;

[0091] S401) analyzing the fiber reinforced composite material to be repaired using a detection and analysis module for the part to be repaired to obtain defect morphology data;

[0092] S402) obtaining repair process data based on the obtained defect morphology data and the expected repair data, and simultaneously invoking the atmospheric pressure plasma treatment module to batch repair the fiber reinforced composite material through the repair data monitoring and information flow module;

[0093] S5) Collect data from the repair part detection and analysis module, repair data monitoring and information flow module, post-repair detection module, and repair operation process planning module during the repair process, and add the data to the repair database to update and iterate the repair machine learning model.

[0094] It includes

[0095] S501) Real-time data collection of multiple modules: At different stages of the repair process, multiple sensors and data collection interfaces are used to complete the collection of multi-module data.

[0096] For example, the repair data monitoring and information flow module can monitor environmental parameters, repair equipment parameters, and dynamic material status data. It can also include data on human intervention events, such as manual parameter adjustments and repair pauses. Environmental parameters can include temperature and humidity during the repair process, while repair equipment parameters can include discharge power, discharge gas flow rate, silicone monomer flow rate, and nozzle spacing. Dynamic material status data can include imaging data, such as using a high-resolution camera to regularly capture the repair area (e.g., every 5 seconds) to record the defect filling process.

[0097] S502) Data preprocessing

[0098] The data obtained in S501) is preprocessed by data cleaning, format conversion, spatiotemporal alignment, and feature extraction.

[0099] Data cleaning involves identifying and removing outliers and noise from the data. For example, if the data collected by a flow sensor exceeds a preset threshold by 20%, it is identified as an outlier and corrected or deleted.

[0100] Format conversion is to convert heterogeneous data collected by different modules into a unified format to facilitate subsequent storage and analysis.

[0101] Spatiotemporal alignment involves aligning multi-source data by timestamps and associating them to the spatial coordinates of the repair area.

[0102] Feature extraction involves extracting parameter features related to model input and output from the collected data.

[0103] S503) Model iteration

[0104] The preprocessed data is stored in the repair sample dataset, triggering the update and iteration process of the repair machine learning model:

[0105] Data loading: Load the latest repair data from the repair sample dataset and merge it with the original repair sample dataset to form a new training dataset.

[0106] Model fine-tuning: Based on the new training dataset, transfer learning technology is used to fine-tune the existing repair machine learning model. During the fine-tuning process, most of the model parameters are fixed, and only a few parameters related to the repair task are optimized to improve the model's training efficiency, avoid overfitting, and achieve model iteration.

[0107] Some specific implementations further include a model evaluation step: using the updated trained model to predict a portion of data in the repair database and comparing it with the actual repair results to evaluate the model's performance. The model's prediction accuracy is quantified by calculating metrics such as accuracy and mean squared error. If the model evaluation results meet the preset performance requirements, the updated trained model is deployed into the repair system, replacing the original model. If not, the model parameters are further adjusted or the model structure is optimized, and the training and evaluation process is repeated until the model performance meets the requirements, completing the model iteration.

[0108] Attachment Figure 2 The image data shown is obtained before the fiber reinforced composite material is repaired. Figure 3 What is shown is the image data collected after the fiber reinforced composite material is repaired. By comparison, it can be seen that the method of the present invention is capable of repairing the surface defects of the fiber reinforced composite material.

[0109] In addition to being used for glass fiber reinforced composites, the method of the present invention can also be used to repair surface defects of different fiber reinforced composites such as carbon fiber reinforced composites, polyarylate fiber reinforced composites, polyimide fiber reinforced composites, and natural fiber reinforced composites. Fiber reinforced composites are a type of high-performance material that is compounded by fibers and matrix materials through a specific process. Among them, the reinforcing fibers play the main load-bearing role, giving the material properties such as high strength and high modulus; the matrix material bonds the reinforcing fibers together, giving the composite material a certain shape, and protecting the fibers from erosion by the external environment, while transmitting stress when the material is subjected to force. The fibers include different fibers such as carbon fibers, glass fibers, natural fibers, and ceramic fibers. The matrix materials include thermoplastic resins and thermosetting resin matrices, etc.

[0110] In summary, this invention uses atmospheric pressure plasma technology to repair surface defects in fiber-reinforced composite materials. By chemically bonding the organosilicon polymer to the resin matrix, the organosilicon polymer and resin are composited, significantly shortening the repair time and improving efficiency. Furthermore, vacuum equipment and large amounts of chemical reagents are eliminated, and reactions are carried out in an open environment, reducing energy consumption, pollution, and equipment maintenance costs. In terms of application value, machine learning uses deep learning of big data sample data to accurately formulate personalized repair plans based on the characteristics of the composite material and the defect status, thereby improving the success rate. During the repair process, data is monitored in real time, and the quality of the prediction is compared with historical data, allowing timely process adjustments to ensure the quality of the repair.

Claims

1. A device for repairing defects in fiber-reinforced composite materials, characterized in that: The repair equipment includes a main module: an atmospheric pressure plasma treatment module, a detection and analysis module for the repaired part of the fiber reinforced composite material to be repaired, a repair data monitoring and information flow module, a post-repair detection module and a repair operation process planning module; The atmospheric pressure plasma treatment module is used for cleaning and activation treatment and depositing polymer for repair; The detection and analysis module for the part to be repaired can scan and analyze the surface of the fiber reinforced composite material to be repaired and obtain defect morphology data for initial defect morphology classification; The post-repair inspection module is capable of scanning and analyzing the surface of the repaired fiber-reinforced composite material and obtaining inspection data for classifying the defect repair status; The repair data monitoring and information flow module is used to monitor the repair process of the fiber reinforced composite material to be repaired, record and transmit the spatiotemporal data of the repair process, and confirm and adjust the parameters used by the atmospheric pressure plasma treatment module; The repair operation process planning module is used to call information from other modules to plan and control the specific operation steps of the atmospheric pressure plasma treatment module during the repair process for outputting repair parameters; The atmospheric pressure plasma treatment module in the repair equipment includes a discharge gas release unit, an organic silicon monomer supply unit, a flow control unit, a heating unit, and a power supply system; the flow control unit can control the volume ratio of the discharge gas released by the discharge gas release unit to the organic silicon provided by the organic silicon supply unit to be 1:0.01 to 1:0.1; The discharge gas release unit can release one of argon, nitrogen and air; The organosilicon monomer supply unit can provide an organosilane monomer selected from one of alkoxysilane, chlorosilane, aminosilane, epoxysilane, and vinylsilane.

2. The repair device according to claim 1, characterized in that The repair equipment further comprises a pretreatment module, which is used to clean and activate the fiber-reinforced composite material to be repaired; The repair equipment further comprises a post-processing module, which is used to perform polishing on the repaired fiber-reinforced composite material.

3. A batch repair method for defects in fiber-reinforced composite materials, characterized in that: It includes the following steps: S1) constructing the repair device according to any one of claims 1 to 2; S2) performing a repair test using the repair equipment and collecting test data from the repair test; Construct a repair sample dataset using experimental data; S3) Based on the repair sample dataset, a machine learning method is used to build and train a machine learning model for composite material repair; S4) Based on the repair machine learning model obtained in step S3), as well as real-time data obtained by the repair part detection and analysis module, the repair data monitoring and information flow module, and the repair operation process planning module, a personalized repair process and parameters for repairing defects in the fiber-reinforced composite material are generated, and the atmospheric pressure plasma treatment module is called through the repair data monitoring and information flow module to batch repair the fiber-reinforced composite material; S5) Collect data from the repair part detection and analysis module, repair data monitoring and information flow module, post-repair detection module, and repair operation process planning module during the repair process, and add the data to the repair database to update and iterate the repair machine learning model.

4. The batch repair method according to claim 3, characterized in that: The repair test in step S2) includes the following steps: The fiber-reinforced composite material to be repaired is placed in an atmospheric pressure plasma jet for treatment, wherein the atmospheric pressure plasma is generated by a discharge gas and an organosilicon monomer, and the volume ratio of the discharge gas to the organosilicon monomer is 1:0.01 to 1:0.1; The treatment in the atmospheric pressure plasma jet is followed by a post-treatment step, which is heating.

5. The batch repair method according to claim 4, characterized in that: The organosilicon monomer is selected from one of alkoxysilane, chlorosilane, aminosilane, epoxysilane and vinylsilane; The discharge gas is selected from one of argon, nitrogen and air.

6. The batch repair method according to claim 4, characterized in that: Before the fiber-reinforced composite material to be repaired is placed under an atmospheric pressure plasma jet for treatment, a pretreatment step is also included, wherein the pretreatment is to clean and activate the fiber-reinforced composite material to be repaired; The restoration also includes polishing.

7. The batch repair method according to claim 3, characterized in that: Step S3) comprises the following steps: S301) data preprocessing; S302) Model training The repair sample dataset is used as the training dataset to train the repair model, and a correspondence model between the operating parameters of the plasma generator and the spatiotemporal data of the repair of defect morphology is constructed; wherein the input is the operating parameters and the output is the spatiotemporal data of the defect repair morphology.

8. The batch repair method according to claim 3, characterized in that: Step S5) comprises the following steps: S501) Real-time data collection of multiple modules, using a variety of sensors and data acquisition interfaces to complete the collection of data from multiple modules at different stages of the repair process; S502) data preprocessing; S503) Model iteration The pre-processed data is stored in the repair sample dataset, triggering the update and iteration process of the repair machine learning model; Data loading: Load the latest repair data from the repair sample dataset and merge it with the original repair sample dataset to form an iterative training dataset; Model fine-tuning: Based on the iterative training dataset, transfer learning technology is used to fine-tune the existing repair machine learning model. During the fine-tuning process, most of the model parameters are fixed, and only a few parameters related to the repair task are optimized to improve the model's training efficiency, avoid overfitting, and achieve model iteration.

9. The batch repair method according to claim 3, characterized in that: The repaired sample data set in step S2) supports multiple data types and multi-source data access; the multiple data types include structured data, unstructured data, multi-dimensional spatiotemporal data and real-time streaming data.

Images