Particle size compounding method for enhancing conductivity of carbon fibers

Abstract

The invention provides a particle size compounding method for enhancing conductivity of carbon fibers. The particle size compounding method comprises the following steps: selecting short carbon fibers with different particle sizes, selecting resin compatible with the carbon fibers, and selecting a dispersing agent suitable for the carbon fibers and the resin; according to the carbon fibers with selected particle sizes, the proportion of the carbon fibers in the composite material is calculated so as to optimize the density, mixing and preparation of a conductive network, and the carbon fibers with different particle sizes are added into a high-speed mixer together with a dispersing agent according to the calculated proportion. Through the technical means of particle size compounding, intelligent control, multi-stage mixing, real-time monitoring and the like, the conductivity of the carbon fiber composite material is optimized, the stability and efficiency of the production process are improved, meanwhile, the breakage risk and the material cost of carbon fibers are reduced, and remarkable technical advantages and commercial value are achieved. The method not only solves the problems of poor conductivity, uneven dispersion and the like of the carbon fiber, but also has good innovativeness, practicability and wide application prospect.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon fiber, and specifically to a particle size compounding method for enhancing the electrical conductivity of carbon fiber. Background Art

[0002] Carbon fiber is widely used in fields such as aerospace, automotive manufacturing, and electronic products due to its excellent strength, light weight, and high temperature resistance. However, the electrical conductivity of carbon fiber is relatively poor, which limits its application in materials with higher electrical conductivity requirements. Therefore, how to improve the electrical conductivity of carbon fiber in composite materials has become an important research topic in the field of materials science.

[0003] In traditional methods, carbon fiber is mainly filled into the matrix material in a random distribution manner of short-cut fibers to form an electrical conduction network. Although this method can improve the electrical conductivity of the material, due to the single length of short-cut carbon fiber, the network structure is relatively simple, and the electrical conduction paths are limited, resulting in limited improvement of electrical conductivity. At the same time, if the length of the short-cut fiber is too long, it may form local accumulation in the matrix material, generating electrical conduction blind spots, further affecting the uniformity and overall effect of electrical conductivity.

[0004] Therefore, there is an urgent need for a new method in the application of carbon fiber in composite materials, which can significantly improve the electrical conductivity and effectively reduce electrical conduction blind spots without increasing the fiber volume and cost. To address this problem, the present application proposes a method of compounding short-cut carbon fibers with different particle sizes. Summary of the Invention

[0005] The purpose of the present invention is to provide a method of compounding short-cut carbon fibers with different particle sizes to optimize the filling performance of carbon fiber in composite materials and the construction of the electrical conduction network. By adjusting the proportion and distribution law of carbon fibers with different particle sizes, a more dense and complex electrical conduction network structure can be formed, effectively improving the electrical conductivity of the material and reducing the electrical conduction blind area.

[0006] To achieve the above purpose, the present invention proposes the following technical solution: A particle size compounding method for enhancing the electrical conductivity of carbon fiber, comprising the following steps:

[0007] Step 1, select short-cut carbon fibers with different particle sizes, select a resin compatible with the carbon fiber, and select a dispersant suitable for the carbon fiber and the resin;

[0008] Step 2, according to the selected carbon fibers with different particle sizes, calculate their proportion in the composite material to optimize the density of the electrical conduction network. Assume that n different lengths of carbon fibers are used, and their lengths are L 1 , L 2 , L 3 ,..., L n , and the proportions of each carbon fiber are P 1, P 2 , P 3 , ..., P n Then the mixing ratio formula is as follows:

[0009]

[0010] P i : The mass ratio or weight percentage of each type of carbon fiber;

[0011] n: The number of types of carbon fibers, representing how many different lengths of carbon fibers are involved in the compounding;

[0012] L i : The length of the i-th type of carbon fiber;

[0013] k i : The proportionality coefficient of the carbon fiber length to other carbon fiber lengths;

[0014] Step 3: Mixing and preparation. Add carbon fibers of different particle sizes and the calculated proportion of dispersant into a high-speed mixer. The mixing time is 10 - 20 minutes to ensure the uniform distribution of carbon fibers and dispersant and avoid agglomeration between fibers. Then add the uniformly mixed mixture of carbon fibers and dispersant into the thermoplastic resin, and use the main feeding port of a twin-screw granulator for mixing to ensure the uniform dispersion of carbon fibers and resin;

[0015] Set the temperature of the twin-screw granulator to 200 - 230 °C. During the extrusion process, ensure the temperature is stable to avoid excessive breakage or uneven dispersion of carbon fibers. Adjust the screw speed and pressure to ensure a stable production process. Cool and collect the extruded plastic pellets for the next processing or performance testing.

[0016] Furthermore, in the present invention, the resin in Step 1 is one of polypropylene, polyethylene, or polyamide. The mass proportion of the resin is 90% - 95%, and the proportion of carbon fibers is 5% - 10%.

[0017] Furthermore, in the present invention, the dispersant in Step 1 is one of polyethylene wax or polypropylene wax. The proportion of the dispersant is 1% - 3% of the total mass of carbon fibers.

[0018] Furthermore, in the present invention, in Step 3, the shear force and equipment speed of the high-speed mixer during the mixing process are monitored in real time through a torque sensor and a speed sensor. The data of the torque sensor and the speed sensor are used to train a machine learning model to predict and adjust the mixing speed in real time through the machine learning model to reduce the risk of carbon fiber breakage;

[0019] The calculation formula is as follows:

[0020] Where: is the fracture risk prediction value, and the prediction value is between 0 and 1;

[0021] T is the torque, with the unit of Nm;

[0022] S is the predicted rotational speed, with the unit of rpm;

[0023] P is the predicted particle size parameter, with the unit of mm;

[0024] V is the predicted temperature, with the unit of °C;

[0025] β 0 , β 1 , β 2 , β 3 , β 4 is the regression coefficient;

[0026] ∈ is the residual term, which is calculated through the model to dynamically adjust the mixing speed S so that the mixing speed is always within the optimal range for reducing the fracture risk.

[0027] Furthermore, in the present invention, multi-stage mixing is adopted in the third step. The multi-stage mixing process is divided into three stages: pre-mixing, preliminary mixing, and high-efficiency mixing. The rotational speed in the pre-mixing stage is set to 100 - 200 rpm) to evenly disperse the raw materials. The rotational speed in the preliminary mixing stage is 300 - 500 rpm. After uniform mixing, it enters the high-efficiency mixing stage, and the rotational speed in the high-efficiency mixing stage is 500 - 2000 rpm. A timer is used to monitor the mixing time of each stage to reduce damage to carbon fibers. A real-time sensor is used to monitor the uniformity of carbon fibers in each mixing stage, and image recognition or spectral analysis technology is used to evaluate the distribution of carbon fibers;

[0028] A reinforcement learning model is used to optimize the mixing speed and time of each stage. In each stage, the model adjusts the mixing strategy according to the current dispersion state of carbon fibers, thereby gradually optimizing the shear force and mixing effect in each stage. The calculation formula:

[0029] Where: R t is the reward for the current state, such as the dispersion effect and the carbon fiber fracture rate;

[0030] δ t is the immediate reward for the current state, such as the dispersion score of carbon fibers;

[0031] γ is the discount factor, and the value of the discount factor is 0.8 - 0.9;

[0032] Q t (s, a) is the state-action value function, which represents the expected return of taking a certain action in a certain state;

[0033] α is the learning rate, which is used to control the learning speed of the model;

[0034] s′ is the next state;

[0035] a is the currently taken action, that is, the adjusted mixing speed or time. Through continuous trials and feedback, reinforcement learning optimizes the mixing process at each stage, avoiding excessive shear force. Based on the training results of the reinforcement learning model, the system can dynamically adjust the mixing speed and duration at each stage.

[0036] Furthermore, in the present invention, in the third step, a thermal sensor is used to continuously monitor the temperature during the processing of the twin-screw granulator and transmit the data to the machine learning model. A neural network model is used to predict the relationship between the processing temperature and the carbon fiber properties, and the temperature is dynamically adjusted according to the current processing environment. The calculation formula is:

[0037] Where: is the predicted material property;

[0038] T is the processing temperature, in °C;

[0039] η is the resin viscosity, in Pa·s;

[0040] f is the neural network function, which learns the non-linear relationship through training. The neural network can predict the optimal processing temperature according to different temperature and resin viscosity conditions to reduce the risk of carbon fiber fracture. The system can adjust the processing temperature according to the output of the neural network and maintain the optimal temperature range through the temperature control system.

[0041] Furthermore, in the present invention, in the third step, a viscosity sensor is used to continuously measure the viscosity of the resin. A support vector machine is used to predict the influence of the resin viscosity on the carbon fiber fracture, and the viscosity is optimized by adjusting the addition amount of the viscosity increasing agent. The calculation formula is:

[0042] Where: is the predicted carbon fiber fracture risk;

[0043] v is the resin viscosity, in Pa·s;

[0044] C is the addition amount of the viscosity increasing agent, in wt%;

[0045] SVM is the support vector machine model. The SVM model predicts the possibility of carbon fiber fracture according to the resin viscosity and the proportion of the viscosity increasing agent, and adjusts the addition amount of the viscosity increasing agent according to the prediction result.

[0046] Further, in the present invention, in the third step, a dispersibility detector is used to monitor the effect of the dispersant in real time, and a random forest regression model is used to predict the relationship between the dosage of the dispersant and the dispersing effect. The calculation formula is:

[0047]

[0048] Where: is the predicted dispersibility score;

[0049] x 1 , x 2 , …, x n are input features, and the input features include the dosage of the dispersant, the particle size of the carbon fiber, and the resin type;

[0050] RF is the random forest regression model; the random forest model predicts the optimal dosage of the dispersant according to various input parameters.

[0051] Further, in the present invention, according to the 3-level ratio algorithm, the particle sizes 5mm: 2.5mm: 1.25mm = 1: 0.5: 1, and the ratio of the particle sizes converted into percentages is 40%, 20%, 40%;

[0052] According to the 5-level ratio algorithm, the particle sizes 5mm: 2.5mm: 1.25mm: 0.63mm: 0.31mm = 1: 0.5: 1: 2: 4, and the ratio of the particle sizes converted into percentages is 11.8%, 5.9%, 11.8%, 23.5%, 47.1%;

[0053] According to the 8-level ratio algorithm, the particle sizes 5mm: 2.5mm: 1.25mm: 0.63mm: 0.31mm: 0.16mm: 0.08mm: 0.04mm = 1: 0.5: 1: 2: 4: 8: 16: 32, and the ratio of the particle sizes converted into percentages is 1.6%, 0.8%, 1.6%, 3.1%, 6.2%, 12.4%, 24.8%, 49.6%.

[0054] Further, in the present invention, the torque sensor is connected to the rotating shaft of the mixer, and a digital or analog torque sensor is selected.

[0055] Beneficial effects: The technical solution of the present application has the following technical effects:

[0056] 1. The present invention can improve the electrical conductivity, optimize the particle size compounding of carbon fibers. By selecting short carbon fibers with different particle sizes and compounding them according to a specific ratio, a denser and more complex conductive network structure can be formed in the composite material. This compounding method enables carbon fibers with different particle sizes to provide more contact points and conductive paths within the same volume, thereby significantly improving the electrical conductivity of the composite material. In particular, carbon fibers with smaller particle sizes can fill the gaps between carbon fibers with larger particle sizes, forming a more efficient conductive path and solving the problem of conductive blind spots. The electrical conductivity is enhanced. Through reasonable particle size compounding, the distribution of carbon fibers in the resin is optimized, making the conductive channels more uniform and reducing the fluctuation of electrical conductivity. The electrical conductivity of the material is improved. Through the particle size compounding design of the present invention, the conductive effect can be improved without increasing the volume of carbon fibers, meeting a wider range of application requirements.

[0057] 2. The present invention reduces carbon fiber breakage and has precise mixing control. In the mixing process of the present invention, a machine learning model, reinforcement learning, and real-time sensors are introduced. Combining the real-time monitoring data of torque and rotational speed, the mixing speed is dynamically adjusted to reduce the mechanical stress on carbon fibers during the mixing process, thereby reducing the risk of their breakage. Specifically, the breakage risk is predicted through a machine learning model, and the mixing speed is dynamically adjusted according to the feedback, making the mixing process more stable and avoiding carbon fiber breakage caused by excessive shear force. The adjustment mechanism based on machine learning and reinforcement learning makes the mixing process more intelligent and automated, avoiding errors caused by manual operation and improving the controllability of the production process.

[0058] 3. The present invention improves the mixing uniformity and adopts multi-stage mixing optimization and dispersant control. The present invention adopts a multi-stage mixing method, dividing the mixing process into three stages: pre-mixing, preliminary mixing, and high-efficiency mixing. Each stage is adjusted according to the dispersibility of carbon fibers. At the same time, during the entire mixing process, image recognition or spectroscopic analysis technology is used to monitor the distribution of carbon fibers to ensure that carbon fibers are evenly dispersed in the resin. The addition amount of the dispersant is optimized through a random forest regression model, further improving the dispersion effect.

[0059] Through multi-stage mixing and precise control of the dispersant, the phenomenon of carbon fiber aggregation or accumulation is avoided, enabling carbon fibers to be more evenly distributed in the resin. The mixing effect is optimized. By dynamically optimizing the mixing speed and time of each stage through a reinforcement learning model, carbon fiber breakage is avoided, and the mixing uniformity is maximized.

[0060] 4. The present invention improves the stability of the processing process. In the processing of carbon fiber, by monitoring parameters such as temperature and resin viscosity in real time, and combining with a neural network model to predict the relationship between processing temperature and carbon fiber properties, the system can dynamically adjust the processing temperature and maintain it within the optimal range to avoid the fracture of carbon fiber during processing. At the same time, the resin viscosity is optimized by a support vector machine (SVM) model to control the addition amount of the viscosity-increasing agent, so that the resin has appropriate fluidity and adhesiveness.

[0061] Avoiding carbon fiber fracture, by precisely controlling the temperature and resin viscosity, reduces the risk of carbon fiber fracture and improves the stability of the processing process. Optimizing the resin viscosity, by adjusting the addition amount of the viscosity-increasing agent, optimizes the viscosity of the resin, further improves the bonding effect between carbon fiber and resin, and enhances the performance of the composite material.

[0062] 5. Improving production efficiency and quality consistency, intelligent production control The intelligent methods such as machine learning, reinforcement learning, and neural network adopted in the present invention enable the production process to be automatically and intelligently adjusted, avoiding errors and uncertainties brought by manual operations. Real-time sensors and feedback mechanisms ensure that each production link is carried out in the best state, greatly improving production efficiency and ensuring material consistency.

[0063] Improving production efficiency, by means of automated process control, reduces the need for manual intervention and makes the production process more efficient. Ensuring material consistency, through intelligent optimization, ensures that each batch of products can be produced under the same control conditions, reducing the fluctuation of material properties.

[0064] 6. This method improves the electrical conductivity and overall quality of carbon fiber composites without significantly increasing material and production costs by optimizing the compounding ratio of carbon fibers with different particle sizes, improving the mixing process, and introducing intelligent control technology. This makes the method not only have technical advantages but also have high economic benefits in commercial applications.

[0065] Reducing production costs, by optimizing the compounding ratio, adjusting the mixing process, and intelligent control system, can improve product performance without significantly increasing costs. Enhancing market competitiveness, the present invention provides a low-cost and high-performance solution, meeting the growing market demand for carbon fiber composites with high electrical conductivity.

[0066] 7. Flexible application scope and wide adaptability The particle size compounding method proposed in the present invention is not only applicable to carbon fiber but also can be extended to other reinforcing materials such as glass fiber and ceramic fiber. By precisely controlling the particle size and proportion compounding, different composite materials can be designed according to different application requirements to meet the different requirements for electrical conductivity in fields such as aerospace, automotive manufacturing, and electronic products.

[0067] Wide range of applications: This method is applicable to different types of fiber materials and can be widely used in the production of various composite materials, with good versatility. Customized design allows adjusting the compounding ratio and mixing process of carbon fibers according to specific application requirements, providing customized solutions.

[0068] In summary, through technical means such as particle size compounding, intelligent control, multi-stage mixing, and real-time monitoring, the present invention optimizes the electrical conductivity of carbon fiber composites, improves the stability and efficiency of the production process, reduces the fracture risk of carbon fibers and material costs, and has significant technical advantages and commercial value. This method not only solves the problems of poor electrical conductivity and uneven dispersion of carbon fibers but also has good innovation, practicability, and broad application prospects.

[0069] It should be understood that all combinations of the foregoing concepts and additional concepts described in greater detail below can be regarded as part of the inventive subject matter of the present disclosure as long as such concepts do not conflict with each other.

[0070] The foregoing and other aspects, embodiments, and features of the teachings of the present invention can be more fully understood from the following description in conjunction with the accompanying drawings. Other additional aspects of the present invention, such as features and / or beneficial effects of exemplary embodiments, will be apparent from the following description or will be learned through practice of specific embodiments according to the teachings of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] The drawings are not intended to be drawn to scale. In the drawings, each identical or nearly identical component shown in various figures can be represented by the same reference numeral. For clarity, not every component is labeled in each figure. Now, embodiments of various aspects of the present invention will be described by way of example and with reference to the drawings, wherein:

[0072] Figure 1 This is a simulation distribution diagram of the present invention's method for overlapping a triangular conductive network according to the distribution probability.

[0073] Figure 2 This is a simulation distribution diagram of the present invention's method for overlapping a triangular conductive network according to the distribution probability.

[0074] Figure 3 This is a simulation distribution diagram of the present invention's method for overlapping a triangular conductive network according to the distribution probability.

[0075] Figure 4 This is a simulation distribution diagram of the present invention's method for overlapping a triangular conductive network according to the distribution probability

[0076] Figure 5 This is a comparison chart of carbon fiber fracture rates. DETAILED DESCRIPTION OF THE INVENTION

[0077] To better understand the technical content of the present invention, specific embodiments are given below in conjunction with the accompanying drawings. In the present disclosure, aspects of the present invention are described with reference to the drawings, and many illustrative embodiments are shown in the drawings. The embodiments of the present disclosure do not necessarily define all aspects of the present invention. It should be understood that the various concepts and embodiments introduced above, as well as those described in more detail below, can be implemented in any of many ways, because the concepts and embodiments disclosed in the present invention are not limited to any implementation manner. Additionally, some aspects of the present invention can be used alone, or in any suitable combination with other aspects disclosed in the present invention.

[0078] This embodiment provides a particle size compounding method for enhancing the electrical conductivity of carbon fiber, including the following steps:

[0079] Step 1: Select short carbon fibers with different particle sizes, select a resin compatible with the carbon fiber, and select a dispersant suitable for the carbon fiber and the resin; Carbon fibers with different particle sizes have different electrical conductivity characteristics. Shorter carbon fibers can improve the filling property and improve the network structure, while long fibers contribute to improving the continuity of the conductive path. By reasonably selecting carbon fibers with different particle sizes, the density of the conductive network can be optimized, and the electrical conductivity can be enhanced.

[0080] As the matrix material, the resin needs to have good fluidity, formability, and compatibility with the carbon fiber. The resin in Step 1 is one of polypropylene, polyethylene, or polyamide. The mass ratio of the resin is 90%-95%, and the carbon fiber accounts for 5%-10%. Resins such as polypropylene, polyethylene, and polyamide have good comprehensive properties in such applications.

[0081] The dispersant in Step 1 is one of polyethylene wax or polypropylene wax. The proportion of the dispersant is 1%-3% of the total mass of the carbon fiber. The dispersant is used to improve the dispersibility of the carbon fiber in the resin, avoid agglomeration, ensure good bonding between the carbon fiber and the resin, and improve the overall performance of the composite material. Polyethylene wax or polypropylene wax has good dispersion effects and can effectively improve the dispersion degree of the fibers. By reasonably selecting these materials, while ensuring the electrical conductivity of the carbon fiber and the processing performance of the resin, good bonding between the fiber and the resin can be ensured, and the overall performance of the composite material can be improved.

[0082] Step 2: According to the selected carbon fiber with a certain particle size, calculate its ratio in the composite material to optimize the density of the conductive network. Assume that n different lengths of carbon fibers are used, and their lengths are L 1 , L 2 , L 3 ,..., L n , and the proportions of each carbon fiber are P 1 , P2 ,P 3 ,...,P n ,then the mixing ratio formula is as follows:

[0083]

[0084] P i : the mass ratio or weight percentage of each type of carbon fiber;

[0085] n: the number of types of carbon fibers, representing how many different lengths of carbon fibers are involved in the compounding;

[0086] L i : the length of the i-th type of carbon fiber;

[0087] k i : the proportionality coefficient of the carbon fiber length to other carbon fiber lengths;

[0088] Step 2 can improve the electrical conductivity. A reasonable particle size distribution can enable the carbon fibers to form a denser and more uniform conductive network in the resin, thereby reducing the resistance of the material and improving the electrical conductivity. It can reduce blind spots. An optimized mixing ratio can reduce the gaps between carbon fibers, avoid blind spots, and improve the electrical properties of the material. It can improve the mechanical strength of the material. By reasonably distributing the length and ratio of carbon fibers, the mechanical properties of the composite material can be enhanced, and the instability of mechanical properties caused by uneven fiber length can be avoided.

[0089] For example:

[0090] According to the 3-level mixing ratio algorithm, the particle sizes 5mm:2.5mm:1.25mm = 1:0.5:1, and the ratio of particle sizes converted to percentages is 40%, 20%, 40%;

[0091] According to the 5-level mixing ratio algorithm, the particle sizes 5mm:2.5mm:1.25mm:0.63mm:0.31mm = 1:0.5:1:2:4, and the ratio of particle sizes converted to percentages is 11.8%, 5.9%, 11.8%, 23.5%, 47.1%;

[0092] According to the 8-level mixing ratio algorithm, the particle sizes 5mm:2.5mm:1.25mm:0.63mm:0.31mm:0.16mm:0.08mm:0.04mm = 1:0.5:1:2:4:8:16:32, and the ratio of particle sizes converted to percentages is 1.6%, 0.8%, 1.6%, 3.1%, 6.2%, 12.4%, 24.8%, 49.6%.

[0093] Such as Figure 1 - 4, Calculated according to this rule, the compounded conductive network is divided into finer segments. The more developed the conductive network is, the conductivity can be increased exponentially, and the corresponding blind spots are also greatly reduced. This solves the shortcoming of blind spots that occur when blindly pursuing large-size transactions in the application of carbon fiber.

[0094] Step 3: Mixing and Preparation. Add carbon fibers of different particle sizes and a dispersant into a high-speed mixer according to the calculated ratio. The mixing time is 10 - 20 minutes to ensure the uniform distribution of carbon fibers and the dispersant, avoiding agglomeration between fibers. Then add the uniformly mixed mixture of carbon fibers and the dispersant into the thermoplastic resin, and use the main feeding port of a twin-screw granulator for mixing to ensure the uniform dispersion of carbon fibers and the resin.

[0095] Set the temperature of the twin-screw granulator to 200 - 230 °C. During the extrusion process, ensure the temperature is stable to avoid excessive breakage or uneven dispersion of carbon fibers. Adjust the screw speed and pressure to ensure a stable production process. Cool and collect the extruded plastic pellets for the next processing or performance testing.

[0096] In the above Step 3, the shear force and the equipment speed of the high-speed mixer during the mixing process are monitored in real time through a torque sensor and a speed sensor. The torque sensor is connected to the rotating shaft of the mixer. Select a digital or analog torque sensor. The data of the torque sensor and the speed sensor are used to train a machine learning model to predict and adjust the mixing speed, and the mixing speed is adjusted in real time through the machine learning model to reduce the fracture risk of carbon fibers.

[0097] The calculation formula is as follows:

[0098] Where: is the fracture risk prediction value, and the prediction value is between 0 and 1;

[0099] T is the torque, unit Nm;

[0100] S is the predicted speed value, unit rpm;

[0101] P is the predicted particle size parameter value, unit mm;

[0102] V is the predicted temperature value, unit °C;

[0103] β 0 , β 1 , β 2 , β 3 , β 4 are regression coefficients;

[0104] ∈ is the residual term. Through model calculation, the mixing speed S is dynamically adjusted so that the mixing speed is always within the optimal range for reducing the fracture risk.

[0105] In step three, multi-stage mixing is adopted. The multi-stage mixing process is divided into three stages: pre-mixing, preliminary mixing, and high-efficiency mixing. The rotation speed in the pre-mixing stage is set to 100 - 200 rpm) to evenly disperse the raw materials. The rotation speed in the preliminary mixing stage is 300 - 500 rpm. After uniform mixing, it enters the high-efficiency mixing stage, and the rotation speed in the high-efficiency mixing stage is 500 - 2000 rpm. A timer is used to monitor the mixing time of each stage to reduce damage to the carbon fiber. A real-time sensor is used to monitor the uniformity of the carbon fiber in each mixing stage, and image recognition or spectral analysis technology is used to evaluate the distribution of the carbon fiber;

[0106] A reinforcement learning model is used to optimize the mixing speed and time of each stage. In each stage, the model adjusts the mixing strategy according to the current carbon fiber dispersion state, thereby gradually optimizing the shear force and mixing effect in each stage. The calculation formula:

[0107] Where: R t is the reward for the current state, such as the dispersion effect and the carbon fiber fracture rate;

[0108] δ t is the immediate reward for the current state, such as the carbon fiber dispersion score;

[0109] γ is the discount factor, and the value of the discount factor is 0.8 - 0.9;

[0110] Q t (s,a) is the state-action value function, representing the expected return of taking a certain action in a certain state;

[0111] α is the learning rate, used to control the learning speed of the model;

[0112] s′ is the next state;

[0113] a is the action taken currently, that is, the adjusted mixing speed or time. Through continuous trials and feedback, reinforcement learning optimizes the mixing process in each stage, avoiding excessive shear force. Based on the training results of the reinforcement learning model, the system can dynamically adjust the mixing speed and duration of each stage.

[0114] In step three, a thermal sensor is used to continuously monitor the temperature during the processing of the twin-screw granulator and transmit the data to the machine learning model. A neural network model is used to predict the relationship between the processing temperature and the carbon fiber performance, and the temperature is dynamically adjusted according to the current processing environment. The calculation formula:

[0115] Where: is the predicted material performance;

[0116] T is the processing temperature, in °C;

[0117] η is the resin viscosity, with the unit of Pa·s;

[0118] f is the neural network function. By training to learn the non-linear relationship, the neural network can predict the optimal processing temperature according to different temperature and resin viscosity conditions to reduce the risk of carbon fiber fracture. The system can adjust the processing temperature according to the output of the neural network and maintain the optimal temperature range through the temperature control system.

[0119] In the third step, a viscosity sensor is used to measure the viscosity of the resin in real time, a support vector machine is used to predict the influence of resin viscosity on carbon fiber fracture, and the viscosity is optimized by adjusting the addition amount of the viscosity increasing agent. The calculation formula is:

[0120] Where: is the predicted risk of carbon fiber fracture;

[0121] v is the resin viscosity, with the unit of Pa·s;

[0122] C is the addition amount of the viscosity increasing agent, with the unit of wt%;

[0123] SVM is the support vector machine model. The SVM model predicts the possibility of carbon fiber fracture according to the ratio of resin viscosity and viscosity increasing agent, and adjusts the addition amount of the viscosity increasing agent according to the prediction result.

[0124] In the third step, a dispersibility detector is used to monitor the effect of the dispersant in real time, and a random forest regression model is used to predict the relationship between the addition amount of the dispersant and the dispersing effect. The calculation formula is:

[0125]

[0126] Where: is the predicted dispersibility score;

[0127] x 1 , x 2 , …, x n are input features, and the input features include the addition amount of the dispersant, the carbon fiber particle size, and the resin type;

[0128] RF is the random forest regression model; the random forest model predicts the optimal addition amount of the dispersant according to the input various parameters.

[0129] This embodiment also gives the following experiment

[0130] Experiment 1: Influence of compounding of carbon fibers with different particle sizes on electrical conductivity

[0131] Experiment purpose: To verify whether the electrical conductivity of the composite material is optimized after compounding short carbon fibers with different particle sizes.

[0132] Experimental materials: Carbon fiber: with different particle sizes of 5mm, 2.5mm, 1.25mm, 0.63mm, 0.31mm, etc.

[0133] Resin: Polypropylene, polyethylene or polyamide.

[0134] Dispersant: Polyethylene wax or polypropylene wax.

[0135] Experimental method: Compound carbon fibers with different particle sizes according to a predetermined ratio, and use 3-level, 5-level and 8-level mixing algorithms to prepare different composite material samples. Conduct conductivity tests on each sample, and measure the conductivity of the material using the four-probe method. Record the conductive properties under different ratios and compare them with the samples without compounded carbon fibers.

[0136] Experimental data and analysis:

[0137] Sample number Ratio (5mm∶2.5mm∶1.25mm) Conductivity (S / m) Sample A 1∶0∶0 0.002 Sample B 1∶1∶0 0.004 Sample C 1∶0.5∶1 0.007 Sample D 1∶0.5∶1∶0.63 0.010 Sample E 1∶0.5∶1∶2∶4 0.015

[0138] It can be seen from the above data that the samples after compounding carbon fibers with different particle sizes have better conductive properties than the samples with a single particle size. Especially under the 5-level and 8-level ratios, the conductivity is significantly improved.

[0139] Experiment 2: Influence of mixing process on carbon fiber fracture

[0140] Experimental purpose: Verify whether the fracture situation of carbon fibers is improved after adopting a multi-stage mixing process and machine learning to optimize the mixing speed.

[0141] Experimental method: Use different mixing processes to prepare carbon fiber composites. Set conventional mixing (without machine learning optimization) and mixing optimized based on a reinforcement learning model respectively. Observe the fracture situation of carbon fibers after mixing through a scanning electron microscope (SEM). Compare and analyze the fracture rates of carbon fibers under the two processes.

[0142] Experimental data and analysis:

[0143] Sample number Mixing process Carbon fiber fracture rate (%) Sample F Conventional mixing 15.3 Sample G Reinforcement learning optimized mixing 6.5

[0144] Through the microscope images Figure 5 And the data in the above table, it can be seen that for the samples with mixing optimized by machine learning, the fracture rate of carbon fibers is significantly lower than that of the samples with conventional mixing, proving that the optimization of the mixing process effectively reduces the fracture of carbon fibers.

[0145] Experiment 3: Influence of temperature and viscosity on carbon fiber dispersion

[0146] Experimental purpose: Verify whether the optimization of temperature and resin viscosity can improve the dispersion of carbon fibers in the composite material.

[0147] Experimental method: The temperature and resin viscosity during the processing are monitored in real time through sensors. The mixing and processing of carbon fibers are carried out under different processing temperatures (such as 200 °C, 210 °C, 220 °C) and resin viscosity conditions. Image recognition technology or spectroscopic analysis technology is used to monitor the distribution of carbon fibers in the resin. The dispersion of carbon fibers is observed by SEM and a dispersion score is given.

[0148] Experimental data and analysis:

[0149] Sample number Temperature (°C) Viscosity (Pa·s) Dispersion score Sample H 200 800 80 Sample I 210 750 85 Sample J 220 700 90

[0150] It can be seen from the above data that with the increase in temperature and the appropriate optimization of resin viscosity, the dispersion of carbon fibers is significantly improved and the dispersion score increases.

[0151] These experimental results prove that the particle size compounding method proposed by the present invention can effectively optimize the performance of carbon fiber composites, significantly improve production efficiency and quality consistency, and has strong technical advantages and commercial value.

[0152] Although the present invention has been disclosed above with preferred embodiments, it is not intended to limit the present invention. Those with ordinary knowledge in the technical field to which the present invention pertains can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention shall be subject to what is defined by the claims.

Claims

1. A particle size compounding method for enhancing the electrical conductivity of carbon fiber, characterized in that: The steps include: Step 1: Select chopped carbon fibers of different particle sizes, select a resin compatible with the carbon fibers, and select a dispersant suitable for the carbon fibers and the resin; Step 2: Calculate the proportion of carbon fiber in the composite material according to the selected particle size to optimize the density of the conductive network. Assume that n carbon fibers of different lengths are used, and their lengths are L1, L2, L3, ..., L n The proportions of each carbon fiber are P1, P2, P3, ..., P n , the ratio formula is as follows: P i : mass ratio or weight percentage of each carbon fiber; n: The number of carbon fiber types, representing how many different lengths of carbon fibers are involved in the compounding; L i : the length of the i-th carbon fiber; k i : The ratio coefficient of carbon fiber length to other length carbon fibers; Step 3: Mixing and preparation: Add carbon fibers of different particle sizes into a high-speed mixer together with a dispersant according to a calculated ratio. The mixing time is 10-20 minutes to ensure that the carbon fibers and the dispersant are evenly distributed and to avoid agglomeration between the fibers. Add the evenly mixed mixture of carbon fibers and the dispersant into the thermoplastic resin and mix them using the main feed port of a twin-screw granulator to ensure even dispersion of the carbon fibers and the resin. Set the temperature of the twin-screw granulator to 200-230℃. During the extrusion process, ensure the temperature is stable to avoid excessive breakage or uneven dispersion of the carbon fiber. Adjust the screw speed and pressure to ensure a stable production process. Cool and collect the extruded plastic particles for the next processing or performance test.

2. A particle size compounding method for enhancing the electrical conductivity of carbon fiber according to claim 1, characterized in that: The resin in step 1 is one of polypropylene, polyethylene or polyamide, the mass proportion of the resin is 90%-95%, and the mass proportion of carbon fiber is 5%-10%.

3. A particle size compounding method for enhancing the electrical conductivity of carbon fiber according to claim 1, characterized in that: The dispersant in step 1 is one of polyethylene wax and polypropylene wax, and the proportion of the dispersant is 1%-3% of the total mass of the carbon fiber.

4. A particle size compounding method for enhancing the electrical conductivity of carbon fiber according to claim 1, characterized in that: In the step three, the shear force and the speed of the high-speed mixer during the mixing process are monitored in real time by a torque sensor and a speed sensor. The data of the torque sensor and the speed sensor are used to train a machine learning model to predict and adjust the mixing speed. The machine learning model is used to adjust the mixing speed in real time to reduce the risk of carbon fiber breakage; The calculation formula is as follows: in: is the fracture risk prediction value, the prediction value is between 0 and 1; T is torque, in Nm; S is the predicted speed value, in rpm; P is the predicted value of particle size parameter, unit: mm; V is the predicted temperature value, unit is °C; β0, β1, β2, β3, β4 are regression coefficients; ∈ is the residual term. Through model calculation, the mixing speed S is dynamically adjusted so that the mixing speed is always within the optimal range to reduce the risk of fracture.

5. A particle size compounding method for enhancing the electrical conductivity of carbon fiber according to claim 4, characterized in that: In the step 3, multi-stage mixing is adopted, and the multi-stage mixing process is divided into three stages: pre-mixing, preliminary mixing and efficient mixing. The speed of the pre-mixing stage is set to 100-200rpm), which is used to evenly disperse the raw materials. The speed of the preliminary mixing stage is 300-500rpm. After uniform mixing, the efficient mixing stage is transferred to the efficient mixing stage. The speed of the efficient mixing stage is 500-2000rpm. A timer is used to monitor the mixing time of each stage to reduce damage to the carbon fiber. A real-time sensor is used to monitor the uniformity of the carbon fiber in each mixing stage, and image recognition or spectral analysis technology is used to evaluate the distribution of the carbon fiber; A reinforcement learning model is used to optimize the mixing speed and time of each stage. At each stage, the model adjusts the mixing strategy according to the current carbon fiber dispersion state, thereby gradually optimizing the shear force and mixing effect of each stage. The calculation formula is: Where: R t Rewards for the current state, such as dispersion effect and carbon fiber breakage rate; δ t Immediate rewards for the current state, such as scoring the dispersion of carbon fiber; γ is the discount factor, and the discount factor is 0.8-0.9; Q t (s,a) is the state-action value function, which represents the expected return of taking a certain action in a certain state; α is the learning rate, which is used to control the speed of model learning; s′ is the next state; a is the current action taken, that is, the adjusted mixing speed or time. Reinforcement learning optimizes the mixing process at each stage through continuous experiments and feedback to avoid excessive shear force. Based on the training results of the reinforcement learning model, the system can dynamically adjust the mixing speed and duration of each stage.

6. A particle size compounding method for enhancing the electrical conductivity of carbon fiber according to claim 5, characterized in that: In step three, a thermal sensor is used to monitor the temperature of the twin-screw pelletizer in real time during processing, and the data is transmitted to the machine learning model. The neural network model is used to predict the relationship between the processing temperature and the carbon fiber performance, and the temperature is dynamically adjusted according to the current processing environment. The calculation formula is: in: To predict material properties; T is the processing temperature, unit is °C; η is the resin viscosity, unit: Pa·s; f is a neural network function. By training and learning nonlinear relationships, the neural network can predict the optimal processing temperature based on different temperature and resin viscosity conditions to reduce the risk of carbon fiber breakage. The system can adjust the processing temperature based on the output of the neural network and maintain the optimal temperature range through the temperature control system.

7. A particle size compounding method for enhancing the electrical conductivity of carbon fiber according to claim 6, characterized in that: In step 3, a viscosity sensor is used to measure the viscosity of the resin in real time, and a support vector machine is used to predict the effect of resin viscosity on carbon fiber fracture, and the viscosity is optimized by adjusting the amount of tackifier added. The calculation formula is: in: To predict the risk of carbon fiber fracture; v is the resin viscosity, in Pa·s; C is the amount of tackifier added, in wt%; SVM is a support vector machine model. The SVM model predicts the possibility of carbon fiber breakage based on the ratio of resin viscosity and tackifier, and adjusts the amount of tackifier added according to the prediction results.

8. A particle size compounding method for enhancing the electrical conductivity of carbon fiber according to claim 7, characterized in that: In step 3, a dispersibility detection instrument is used to monitor the effect of the dispersant in real time, and a random forest regression model is used to predict the relationship between the amount of dispersant added and the dispersing effect. The calculation formula is: in: Score the dispersion of the forecasts; x1,x2,…,x n is the input feature, which includes the amount of dispersant added, carbon fiber particle size, and resin type; RF is a random forest regression model; the random forest model predicts the optimal dispersant addition amount based on various input parameters.

9. The particle size compounding method for enhancing the conductivity of carbon fiber according to claim 1, characterized in that: According to the three-level ratio algorithm, the particle size is 5mm:2.5mm:1.25mm=1:0.5:1, and the particle size ratio is converted into percentages of 40%, 20%, and 40%; According to the 5-level ratio algorithm, the particle size is 5mm: 2.5mm: 1.25mm: 0.63mm: 0.31mm = 1: 0.5: 1: 2: 4, and the particle size ratio is converted into percentages of 11.8%, 5.9%, 11.8%, 23.5%, and 47.1%; According to the 8-level ratio algorithm, the particle size is 5mm:2.5mm:1.25mm:0.63mm:0.31mm:0.16mm:0.08mm:0.04mm=1:0.5:1:2:4:8:16:32, and the particle size ratio converted into percentage is 1.6%, 0.8%, 1.6%, 3.1%, 6.2%, 12.4%, 24.8%, and 49.6%.

10. The particle size compounding method for enhancing the electrical conductivity of carbon fiber according to claim 4, characterized in that: The torque sensor is connected to the rotating shaft of the mixer, and a digital or analog torque sensor is selected.

Images