Production control method and apparatus for high ductility mpp power pipe preparation

Abstract

This invention discloses a production control method and equipment for manufacturing high-toughness MPP power pipes, belonging to the field of power pipe manufacturing technology. It aims to solve the problems of insufficient pipe toughness caused by uneven fiber dispersion, weak interfacial bonding, and delayed cross-linking control in existing processes. The method includes: metering pre-activated composite glass fiber powder with polypropylene, high-density polyethylene, etc., according to a formula; measuring the distribution and reactivity of functional groups on the fiber powder surface to form an initial process parameter dataset, which is then input into a melt flow field and cross-linking reaction kinetics prediction model to generate a processing strategy; setting the batch addition sequence and melt mixing method of the active compatibilizer according to this strategy, and performing closed-loop control based on online monitoring of fiber dispersion state, interfacial tension, and cross-linking process data; and obtaining the finished product through zoned cooling and micro-controlled stretching after extrusion. This invention can achieve precise control of fiber dispersion and cross-linking reaction, significantly improving the toughness and impact resistance of the pipe.

Description

Technical Field

[0001] This invention relates to the field of power pipe manufacturing technology, specifically to a production control method and equipment for manufacturing high-toughness MPP power pipes. Background Technology

[0002] MPP power cable conduit is a type of power cable protection pipe made primarily of polypropylene (PP). It is widely used in power, telecommunications, and engineering construction, and features high temperature resistance, lightweight, good insulation, and corrosion resistance. To improve mechanical properties, existing technologies typically add high-density polyethylene (HDPE), glass fiber powder, or toughening agents to MPP pipes to enhance the material's rigidity and strength.

[0003] In engineering practice, when pipes are subjected to bending, impact, or low-temperature environments, micro-cracks or fractures are easily formed locally, making it difficult to effectively absorb external forces, leading to pipe breakage or failure and reducing the cable protection effect. Especially under laying conditions of high impact and high bending load, traditional MPP pipes cannot guarantee the continuous toughness and impact resistance of the pipe wall.

[0004] Furthermore, in traditional manufacturing processes, the interfacial bonding between glass fiber powder and the polymer matrix is ​​unstable. Fibers tend to agglomerate or distribute unevenly during the melt mixing process, leading to localized stress concentration zones within the pipe and resulting in inconsistent toughness. Existing processes largely rely on empirical parameter adjustments and lack predictive and control methods for melt flow, cross-linking reactions, and interfacial bonding. Therefore, it is difficult to systematically improve the microstructural uniformity and overall toughness of the pipe.

[0005] Therefore, existing MPP power pipes still have significant shortcomings in terms of high toughness, impact resistance, and low temperature adaptability, and it is necessary to improve the toughness of the material to enhance the reliability of the pipes under complex working conditions. Summary of the Invention

[0006] In view of the above-mentioned shortcomings mentioned in the background art, the purpose of this invention is to provide a production control method and equipment for the preparation of high-toughness MPP power pipes.

[0007] A first aspect of the present invention provides a production control method for manufacturing high-toughness MPP power pipes, the method comprising the following steps: Polypropylene, high-density polyethylene, composite glass fiber powder, active compatibilizer and auxiliary additives are measured separately according to the formula. Among them, the composite glass fiber powder has been pretreated with activating liquid, polyethyleneimine and modified flame retardant, and has active functional groups on its surface. The surface functional group distribution and reactivity of composite glass fiber powder were determined to form an initial process parameter dataset, which was then input into a melt flow field and crosslinking reaction kinetics prediction model to generate a processing strategy. Based on the processing strategy, the batch addition sequence of the active compatibilizer and the polymer melt mixing method are set, and online monitoring data of the extruder, including fiber dispersion state, interfacial tension and crosslinking process, are obtained. Closed-loop control is then performed based on the online monitoring data to optimize fiber dispersion and crosslinking reaction. After extrusion molding, the pipe is subjected to zoned cooling and micro-controlled stretching to obtain a high-toughness composite MPP power pipe.

[0008] A second aspect of the present invention provides a production control device for the fabrication of high-toughness MPP power pipes, the device comprising: The metering and preparation unit is used to meter polypropylene, high-density polyethylene, composite glass fiber powder, active compatibilizer and auxiliary additives according to the formula; wherein, the composite glass fiber powder has been pretreated with activating liquid, polyethyleneimine and modified flame retardant, and has active functional groups on its surface. The parameter generation and strategy calculation unit is used to determine the surface functional group distribution and reactivity of the composite glass fiber powder, form an initial process parameter dataset, and input it into the melt flow field and crosslinking reaction kinetics prediction model to generate a processing strategy. The setting and closed-loop control unit is used to set the batch addition sequence of the active compatibilizer and the polymer melt mixing method according to the processing strategy, obtain online monitoring data of the extruder including fiber dispersion state, interfacial tension and crosslinking process, and perform closed-loop control based on the online monitoring data to optimize fiber dispersion and crosslinking reaction. The post-processing unit is used to perform zoned cooling and micro-controlled stretching of the tube after extrusion molding to produce a high-toughness composite MPP power tube.

[0009] Compared with the prior art, the present invention has at least the following beneficial technical effects: This invention achieves precise measurement of the surface functional groups of composite glass fiber powder and the establishment of an initial process parameter dataset. Combined with a processing strategy generated by a melt flow field and crosslinking reaction kinetic prediction model, it realizes real-time control of the batch addition of active compatibilizer and the polymer melting and mixing method during the extruder melting stage. By monitoring the fiber dispersion state, interfacial tension, and crosslinking process online, it implements local disturbance, shear fine-tuning, and residence time control. Finally, it fixes the multi-component crosslinking network through partitioned cooling and micro-controlled stretching, thereby achieving uniform fiber dispersion, optimized interfacial bonding force, and balanced formation of the crosslinking network at the microscopic level. This results in a composite MPP power pipe that exhibits significantly improved toughness and mechanical stability under impact, bending, and low-temperature conditions. Attached Figure Description

[0010] Figure 1 This is a schematic flowchart of a production control method for manufacturing high-toughness MPP power pipes, as disclosed in an embodiment of the present invention. Figure 2This is a schematic diagram of the structure of the melt flow field and crosslinking reaction kinetics prediction model disclosed in an embodiment of the present invention; Figure 3 This is another structural schematic diagram of the melt flow field and crosslinking reaction kinetics prediction model disclosed in the embodiments of the present invention; Figure 4 This is a schematic diagram of a production control device for manufacturing high-toughness MPP power pipes, as disclosed in an embodiment of the present invention. Detailed Implementation

[0011] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below in conjunction with the embodiments of this invention. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0012] This invention provides a production control method for the preparation of high-toughness MPP power pipes. This method aims to solve the technical problems of local stress concentration and uneven toughness performance of pipes caused by weak interfacial bonding between composite glass fiber powder and polymer matrix, uneven fiber dispersion, and delayed control of crosslinking reaction in existing processes.

[0013] Reference Figure 1 As shown, the method of the present invention includes the following steps: S1, according to the formula, polypropylene, high-density polyethylene, composite glass fiber powder, active compatibilizer and auxiliary additives are measured separately. Among them, the composite glass fiber powder has been pretreated with activating liquid, polyethyleneimine and modified flame retardant, and has active functional groups on the surface. The formulation composition of high-toughness MPP power pipe is, for example: The composition includes 20-30 parts polypropylene, 15-20 parts high-density polyethylene, 7-10 parts composite glass fiber powder, 5-8 parts active compatibilizer, and 1-1.5 parts auxiliary additives. The molecular weight of the polypropylene is 600-700. The auxiliary additives consist of plasticizer, dispersant, lubricant, antioxidant, and antistatic agent in a ratio of 1g:1g:1g:1g:1g.

[0014] Polypropylene possesses high toughness, enabling it to absorb energy; high-density polyethylene exhibits high rigidity, which helps disperse impact forces. By controlling the composition ratio of these two components, a balance can be achieved in the material's performance, ensuring it has sufficient strength to withstand loads while also possessing sufficient rigidity to prevent deformation or bending, thereby mitigating the impact of shocks on the material.

[0015] The composite glass fiber powder is the core modified filler that imparts high toughness to the pipe, and its compatibility with the polypropylene and high-density polyethylene matrix directly affects the final mechanical properties of the pipe. In traditional processes, due to the structural differences between the additives and the matrix components, the additives have poor compatibility in the pipe, and excessive use will form a weak interface layer in the pipe, leading to a decrease in mechanical properties. To address this, the composite glass fiber powder used in this invention undergoes multi-stage surface functionalization treatment beforehand. The specific preparation process is as follows: First, glass fiber powder and activation solution are added to a three-necked flask and ultrasonically dispersed to obtain pretreated glass fiber powder. Then, the pretreated glass fiber powder and N,N-dimethylformamide are added to a nitrogen-protected three-necked flask, ultrasonically dispersed and stirred, and then polyethyleneimine solution is added dropwise. The temperature is raised to 60-70℃ and kept at that temperature for 3-5 hours to obtain modified glass fiber powder. Finally, the modified glass fiber powder, modified flame retardant, and tetrahydrofuran are added to a three-necked flask and stirred. The temperature is raised to reflux and kept at that temperature for 20-24 hours to obtain composite glass fiber powder.

[0016] After the above treatment, the fiber surface has active functional groups, which can effectively improve its chemical bonding ability with active compatibilizers and provide reaction sites for subsequent cross-linking reactions.

[0017] After completing the formulation measurement, the surface properties of the pretreated composite glass fiber powder were further quantitatively characterized to form an initial process parameter dataset.

[0018] S2, determine the surface functional group distribution and reactivity of composite glass fiber powder, form an initial process parameter dataset, and input it into the melt flow field and crosslinking reaction kinetics prediction model to generate a processing strategy; As an example, the surface functional group distribution and reactivity of the composite glass fiber powder were determined to form an initial process parameter dataset, including: S21, the types of active functional groups and grafting density on the surface of composite glass fiber powder were determined by infrared spectroscopy and chemical titration, respectively, and its surface energy was determined by dynamic contact angle method. To accurately determine the surface chemical state of each batch of composite glass fiber powder, this step automatically performs online quantitative characterization of the fiber powder's surface properties during the metering and conveying process. An online detection bypass is installed along the path of the composite glass fiber powder after metering, as it enters the extruder feed port through the conveying pipeline. This bypass automatically extracts trace amounts of fiber powder samples from the main stream at preset time intervals and sends them to the series-connected detection modules.

[0019] Specifically, the online infrared spectroscopy module acquires diffuse reflectance infrared spectra of the extracted fiber powder samples. This module incorporates a Fourier transform infrared spectrometer and is equipped with an automatic sample introduction and cleaning device, automatically updating the sample after each detection. It automatically identifies the NH bending vibration peak (approximately 1580-1650 cm⁻¹) introduced by polyethyleneimine in the spectrum. -1 ) and CN stretching vibration peak (approximately 1020-1250 cm⁻¹) -1 The system automatically determines the types of active functional groups present on the surface of the current batch of fibers by means of the characteristic absorption peaks of the modified flame retardant molecular skeleton and a pre-set spectral library comparison algorithm, without the need for manual analysis.

[0020] Subsequently, the fiber powder sample is automatically transferred to the online titration module. This module incorporates a miniature reaction chamber and an automated pipetting system, quantitatively injecting standard hydrochloric acid solution to perform acid-base back titration of the fiber powder. The titration endpoint is monitored in real time by a pH electrode, and the module automatically calculates the number of moles of active functional groups per gram of fiber powder, i.e., the grafting density. After each titration, the reaction chamber is automatically cleaned and vacuum-dried for the next test.

[0021] Furthermore, to evaluate the wettability and reactivity between the fiber surface and the polymer melt, an online dynamic contact angle module is included in the aforementioned detection bypass. This module automatically presses the fiber powder into a flat sheet under standard pressure, and then uses an automatic dropping system to titrate it with water and diiodomethane as probe liquids. A high-speed camera automatically captures images of the droplets spreading on the fiber sheet surface, and the static contact angle is extracted in real time using image processing algorithms. The total solid surface energy, its dispersive component, and polar component are automatically calculated according to the OWRK geometric mean equation, where the polar component is positively correlated with the density of active functional groups on the fiber surface. A higher surface energy value indicates a stronger thermodynamic wetting driving force between the fiber surface and the active compatibilizer and polymer matrix, resulting in higher reactivity.

[0022] S22, the measurement results are correlated with the proportions of each component, melt viscosity and melt index, and the corresponding reference process condition range is obtained by retrieving the pre-established process parameter mapping table based on fiber surface characteristics. The results are then combined with the measurement results and the proportions of each component, melt viscosity and melt index to form an initial process parameter dataset.

[0023] After the fiber surface properties are measured, the control system automatically correlates and matches the active functional group types, grafting density and surface energy data obtained from the above measurements with the proportions of each component in the current batch formulation (specific parts of polypropylene, high-density polyethylene, composite glass fiber powder, active compatibilizer and auxiliary additives), as well as the melt index and apparent viscosity curves of each component that have been measured and stored in the database in advance.

[0024] After the association is completed, the measured fiber grafting density and surface energy are used as search criteria to automatically perform a matching query in a pre-calibrated process parameter mapping table stored in the control system. This mapping table is established through prior orthogonal experiments or historical production data, with fiber grafting density as the row and polymer rheological parameter combinations as the column. The table stores experimentally verified reference process condition ranges for the corresponding combinations. The current batch's measured fiber grafting density value automatically determines its corresponding row, and the rheological parameter combination column is automatically determined based on the formulation ratio and measured melt index. The corresponding reference process condition range is then retrieved, including the compatibilizer batch addition sequence range, the shear strength setpoint range for each section, and the material residence time range.

[0025] Finally, the types of active functional groups, grafting density, and surface energy measured online, along with the proportions of each component in the formulation, the melt index and apparent viscosity curves of each component, and the reference process condition range retrieved from the mapping table, are automatically combined and packaged according to a predefined data field structure to form an initial process parameter dataset.

[0026] The initial process parameter dataset is input into a pre-constructed melt flow field and crosslinking reaction kinetics prediction model. The flow field distribution of the melt inside the extruder screw and the crosslinking reaction process between active components are simulated numerically. This allows for the derivation of the optimal combination of operating parameters under the current formulation and process conditions—that is, the generation of a processing strategy—before the actual forming of the pipe. This processing strategy will guide the extruder in executing relevant process actions during the melt mixing stage.

[0027] As an example, such as Figure 2 As shown, the crosslinking reaction kinetic prediction model includes: Flow field The component mapper is used to calculate the axial distribution of fiber dispersion and active functional group concentration based on the initial process parameter dataset and real-time screw parameters. Flow field The component mapper consists of two cascaded subnetworks: The flow field analysis network employs a deep neural network structure based on physical information constraints, specifically comprising a three-layer fully connected network as its backbone. The input layer receives melt viscosity, melt flow index, and real-time screw parameters (rotation speed, barrel temperature at various stages) from the initial process parameter dataset. The output layer provides the velocity and shear rate fields at each discretized micro-element node along the screw axis. During the training phase, a simplified form of the Navier-Stokes equations is incorporated as a physical regularization term into the loss function, enabling the network to output flow field predictions that conform to fundamental fluid mechanics laws with lower computational overhead during forward inference.

[0028] The component transport network receives the velocity and shear rate fields output by the flow field analysis network. Simultaneously, it extracts fiber graft density, surface energy, formulation component ratios, and active functional group types from the initial process parameter dataset as input. Through a four-layer fully connected network containing gated linear units, it recursively calculates the concentration distribution of each component along the screw axis, element by element. The final output is the fiber dispersion and active functional group concentration along the screw axis. Specifically, fiber graft density and active functional group types determine the spatial distribution weights of initial reaction sites on the fiber surface; formulation component ratios determine the initial volume fraction of each polymer component in the melt; and surface energy corrects the wetting and diffusion tendency and initial dispersion rate of fiber particles in the melt, thereby influencing the migration and dispersion kinetics of fiber particles in different shear zones.

[0029] A crosslinking trend predictor is used to solve for the crosslinking degree baseline evolution curve along the screw axis based on the axial distribution of the fiber dispersion and the concentration of the active functional groups. In actual processing, the dispersion of composite glass fiber powder within the extruder is not ideally uniform. Due to differences in shear strength across different sections of the screw, the fibers exhibit alternating aggregated and dispersed zones during axial transport, resulting in a corresponding non-uniform concentration distribution of the active functional groups on the fiber surface. Traditional homogeneous reaction kinetic models assume equal concentrations and instantaneous mixing of all reactants throughout space. However, in the heterogeneous melt system of this invention, this assumption leads to a significant overestimation or underestimation of local reaction rates. Furthermore, while the organic coating layer formed on the fiber surface after multi-stage treatment endows the fiber with abundant active functional groups, it also means that these functional groups do not immediately participate in the crosslinking reaction upon contact with the active compatibilizer. Instead, they undergo an induced process of stretching, peeling, and exposing the coating layer due to local shear forces. This interfacial reaction hysteresis effect is particularly pronounced in the low-shear region. Without correction, the predicted crosslinking degree baseline evolution curve will exhibit a systematic leading deviation in the axial position.

[0030] To address the aforementioned technical problems, this invention further improves the crosslinking trend predictor by combining mass transfer theory from fluid mechanics, reaction kinetics from interfacial chemistry, and actual process data to construct the following network structure: As an example, such as Figure 3 As shown, the crosslinking trend predictor includes: diffusion The reaction operator layer is used to discretize the screw axis into multiple micro-elements with the axial distribution of fiber dispersion and active functional group concentration as input, and to correct the effective collision frequency under non-uniform concentration field using the local Peckley number in each micro-element. This layer receives output data from the flow-component mapper, specifically the axial distribution of fiber dispersion and active functional group concentration along the screw axis. Due to the axial variations in the extruder screw's geometry—including the feeding section conveying elements, the mixing section kneading blocks, and the homogenizing section reverse threads—the melt flow state and component mixing degree differ significantly at different locations. To accurately capture this spatial difference, this layer first disperses the screw axially into multiple continuous micro-element units at equal intervals. The length of each micro-element can be 0.5 to 1 times the screw diameter, ensuring that the flow state within each micro-element can be approximated as a quasi-steady state.

[0031] For each micro-element, a local Peckley number calculation unit is located within the layer. The Peckley number is a dimensionless number in fluid mechanics that characterizes the ratio of convective mass transfer rate to diffusion mass transfer rate. Under the constraint of only the axial distribution of fiber dispersion and active functional group concentration as input, this layer indirectly derives the local Peckley number by analyzing the axial variation gradient of fiber dispersion between adjacent micro-elements, specifically defined as: ;in, For the first The absolute value of the axial gradient of fiber dispersion at each micro-element reflects the driving intensity of convective transport at that point: when the fiber dispersion between adjacent micro-elements changes drastically, the gradient value is large, indicating that there is a strong convective drive at that point. For the first The absolute value of the axial gradient of the concentration of active functional groups at each micro-element; The characteristic scale of the infinitesimal element; is the molecular diffusion coefficient of the active functional group in the melt.

[0032] when When this occurs, it indicates that molecular diffusion dominates within the micro-element, and active functional groups can rapidly diffuse throughout the entire micro-element space via Brownian motion, leading to a homogenization of the local concentration. At this point, the theoretically effective collision frequency approaches the calculated value under the homogeneous model. This indicates that convective mass transfer is dominant, with active functional groups mainly migrating along the flow direction. Lateral diffusion is insufficient to homogenize the concentration field, resulting in a spatially depleted region of reactants within micro-elements with low fiber dispersion. Consequently, the actual effective collision frequency will be lower than the theoretical value.

[0033] To address the aforementioned physical mechanism, this layer uses a single hidden layer network containing a Sigmoid activation function to map the local Peckley number of each microelement to the corresponding effective collision frequency correction coefficient. : ;in, It is the Sigmoid activation function. and These are the trainable network parameters. The value of this correction coefficient ranges from (0,1), when... When it approaches 0, Approaching 1 indicates that the state is close to homogeneous and no reduction is needed; when As the value increases, the output of the sigmoid function smoothly approaches an asymptotic value less than 1. This layer uses this correction factor. Multiplying this by the theoretical collision frequency under the homogeneous model, the corrected effective collision frequency of the infinitesimal element is finally output as a product. : ;in, The intrinsic reaction rate constant is... and The first The concentration of functional groups and active compatibilizer in each micro-element. and denoted as the reaction order.

[0034] The shear-induced delay factor estimation layer is used to output an interfacial reaction delay factor that characterizes the time lag from the exposure of functional groups to the point where they can participate in crosslinking, based on the local shear rate and material residence time of each micro-element. In multi-stage modified composite glass fiber powder, the active functional groups on the surface are not exposed on the outermost layer of the fiber, but are encapsulated by a coating layer formed by the cross-linking of organic segments introduced during the activation treatment and modified flame retardants. Only when the fiber particles are in a sufficiently strong shear flow field can the shear stress applied to the coating layer by the polymer matrix cause deformation, stretching, and even local peeling, thus exposing the internal active functional groups to the melt and allowing them to contact the active compatibilizer. This exposure process is not instantaneous, but requires a certain accumulation of shear action time, thus constituting a time lag in the interfacial reaction.

[0035] The inputs to this layer are the local shear rate of each micro-element and the estimated residence time of the material in that micro-element. Specifically, the local shear rate... The fiber dispersion can be indirectly deduced from the rate of change of fiber dispersion along the axial direction. Specifically, when the fiber dispersion in a certain section increases rapidly, it indicates that the area is undergoing strong shear dispersion. Material residence time... This can be indirectly characterized by the rate of decay of the concentration of active functional groups along the axial direction. Specifically, a faster rate of decay indicates that the material has sufficient residence time in this section to complete the accumulation process of functional group exposure and reaction. This layer uses a time-series coding network based on gated cyclic units (GRUs), treating the screw axis as a pseudo-time series, and inputting the shear action and residence time characteristics obtained indirectly along the axial direction element by element. The hidden state update process of the GRU is as follows:

[0036] ;in, For the first The input vector of infinitesimal elements, This is the hidden state passed down from the previous infinitesimal. and These are the update gate and the reset gate, representing candidate hidden states. This represents the current hidden output state of the infinitesimal element. and For trainable parameters, This indicates element-wise multiplication.

[0037] Through the hidden state propagation mechanism of the gated recurrent unit, the network can accumulate the shearing history information of the upstream micro-element. Even if the instantaneous shearing intensity of the current micro-element is not high, if its upstream has accumulated sufficient shearing action, the corresponding exposure memory is still retained in the hidden state, thus outputting a smaller delay factor; conversely, when both the upstream and the current micro-element lack sufficient shearing accumulation, a larger delay factor is output.

[0038] Finally, a fully connected output layer maps the hidden states to the interface reaction delay factors of each infinitesimal element. :

[0039] in, Ensure output It is a non-negative value. When the cumulative shear force is sufficient and the functional groups are fully exposed, A value approaching 0 indicates that the cross-linking reaction can occur instantly; when the shear force is insufficient, A larger positive value indicates that the crosslinking reaction is actually delayed within this infinitesimal element.

[0040] The derivation integration layer is used to couple the corrected effective collision frequency and the interface reaction delay factor to the crosslinking kinetic ordinary differential equation system, and to recursively integrate along the screw axis to output the crosslinking degree baseline evolution curve.

[0041] This layer, as the final output layer of the crosslinking trend extrapolator, is responsible for integrating the physical correction factors of the first two layers into the reaction kinetics solution framework. This layer contains a set of elementary crosslinking kinetic ordinary differential equations pre-calibrated through model compound experiments, including at least three coupled equations: the functional group consumption rate equation, the crosslink bond formation rate equation, and the active compatibilizer concentration change equation.

[0042] Based on the standard homogeneous model, this layer makes the following key modifications to the dynamic equations of each infinitesimal element: 1) Modify the functional group consumption rate equation by directly introducing the modified effective collision frequency from the diffusion-reaction operator layer output into its reaction rate constant term. The revised functional group consumption rate equation is expressed as:

[0043] It should be understood that this modification enables the local reaction rate to be adaptively adjusted according to the actual non-uniformity of the concentration field at that location.

[0044] 2) For the cross-linking bond formation rate equation, a shear-induced delay factor is introduced into its time-progression term to estimate the interfacial reaction delay factor output by the layer. As an increment of the time constant, the corrected cross-linking bond formation rate equation is expressed as:

[0045] ;in, The concentration of cross-linked bonds already generated within the micro-element. The characteristic crosslinking time constant under no-delay conditions is determined by intrinsic reaction kinetic parameters and system viscosity. When When it approaches 0, The formation of cross-links almost instantaneously follows the consumption of functional groups, indicating that the functional groups at that location have been fully exposed by shearing; when When the value is large, The increase is significant, and the formation of cross-links lags behind the consumption of functional groups in time, reflecting the delay effect required for interface exposure.

[0046] The equation for the change in the concentration of the active compatibilizer maintains the stoichiometric balance: ;in, The stoichiometric ratio of the active compatibilizer to the functional groups on the fiber surface is predetermined by the number and type of active groups on the compatibilizer molecule that can participate in cross-linking.

[0047] After completing the above corrections, this layer begins spatial recursive integration along the screw axis. The integration solution takes the starting end of the screw feeding section as the first infinitesimal element ( The entrance boundary of ), at which point the functional group concentration Take the measured values ​​from the initial process parameter dataset, crosslinking bond concentration active compatibilizer concentration The amount of compatibilizer injected is determined by the amount corresponding to that micro-element. For the first... For each infinitesimal element, a fourth-order Runge-Kutta numerical integration method is used, with the estimated dwell time of this infinitesimal element as the basis. To determine the integration time, the solution to the three coupled equations is advanced in the time domain. After integration, the state at the exit of this infinitesimal element is used as the initial entry condition for the next adjacent infinitesimal element.

[0048] The recursive solution is as follows:

[0049] ;in, The total number of infinitesimal elements. Indicates and The fourth-order Runge-Kutta integral operator for correcting parameters.

[0050] After traversing the entire screw axial discrete sequence, the crosslinking degree value at the outlet of each micro-element is... Connecting the axial coordinates to form a smooth curve yields the baseline evolution curve of the crosslinking degree, where the crosslinking degree is defined as: In the formula, This represents the initial total concentration of functional groups.

[0051] Through the aforementioned network structure, the crosslinking trend extrapolator transforms two complex factors deviating from the homogeneous assumption—the spatial non-uniformity of fiber dispersion and the shear-induced interfacial reaction hysteresis—into quantifiable and calculable effective collision frequency correction coefficients. and interfacial reaction delay factor It is then coupled into the crosslinking dynamics solution framework, with the corrected effective collision frequency output by the diffusion-reaction operator layer. The interface response delay factor is estimated from the shear-induced delay factor of the layer output. The inductive integral layer integrates both into the dynamic equation and solves it recursively, ultimately outputting a baseline evolution curve of crosslinking degree that is closer to the actual physical process.

[0052] The processing strategy solver is used to minimize the deviation between the crosslinking degree baseline evolution curve and the preset target crosslinking degree curve as the optimization objective, and outputs the processing strategy, including the batch addition sequence of active compatibilizer, the shear strength setting value of each segment, the location and intensity of local disturbances, and the amount of material residence time adjustment.

[0053] The policy solver consists of a target bias encoder and a policy generation network connected in series: The target deviation encoder receives the point-by-point deviation values ​​along the axial direction between the crosslinking degree baseline evolution curve and the preset target crosslinking degree curve as input. It extracts the spatial distribution pattern of the deviation through a Transformer encoding block containing a multi-head self-attention mechanism to generate a deviation pattern feature vector.

[0054] The strategy generation network consists of a strategy decoder based on a multilayer perceptron. It receives the deviation pattern feature vector and decodes it step-by-step through multiple fully connected layers into output values ​​for each control variable, including the injection timing and rate parameters for the batch addition sequence of the active compatibilizer, the shear strength setpoints for each segment, the location and intensity parameters of local disturbances, and the material residence time adjustment. During the training phase, the network optimizes its parameters to minimize the hybrid loss function (which includes the deviation term between the actual crosslinking degree and the target curve, as well as energy consumption and constraint violation penalties), enabling it to directly output the processing strategy that conforms to process constraints and minimizes deviation during forward inference.

[0055] S3, according to the processing strategy, set the batch addition sequence of the active compatibilizer and the polymer melt mixing method, obtain online monitoring data of the extruder including fiber dispersion state, interfacial tension and crosslinking process, and perform closed-loop control based on the online monitoring data to optimize fiber dispersion and crosslinking reaction; During the actual extruder melt mixing stage, the aforementioned virtual control scheme is integrated with the actual physical production line, and a real-time feedback closed loop is formed through online monitoring.

[0056] As an example, the batch addition sequence of the active compatibilizer and the polymer melt mixing method are set according to the processing strategy, including: S31, multiple compatibilizer injection ports arranged along the screw axis are independently controlled in terms of opening time, injection rate and injection volume to achieve batch addition; On a twin-screw extruder, independent adjustable injection ports are arranged along the screw axis in multiple preset functional sections. The positions of these injection ports are configured according to the different functional zones of the screw: for example, the first injection port is set in the main conveying section after the feeding section, the second injection port is set at the beginning of the mixing section, and the third injection port is set at the inlet of the homogenization section. Each injection port is equipped with an independent metering pump and an actuator valve, and its opening, closing, and flow regulation are all independently and sequentially controlled by the central controller according to the batch addition sequence of the active compatibilizer output by the S2 processing strategy solver.

[0057] Specifically, the processing strategy specifies the batch injection ratio and timing schedule of the compatibilizer at each axial position. For example, the strategy can stipulate that 30% of the total compatibilizer dose is injected first in the main conveying section, another 50% is injected in the mixing section, and the remaining 20% ​​is added in the homogenization section. The actuator precisely controls each injection port accordingly: the valve is opened at a specified time, the metering pump continuously delivers the specified amount of compatibilizer according to the preset injection rate curve, and closes on time after the target injection amount is reached. It can be understood that the technical effect of this process route of injecting compatibilizer in groups, at different times, and in segments as needed is that it can actively adapt to the fiber dispersion process that evolves with the screw propulsion. When the fiber first enters the mixing section and the functional groups are not yet fully exposed, only a small amount of compatibilizer is injected to avoid ineffective consumption; after the high shear action of the mixing section peels off the fiber coating layer and the functional groups are fully exposed, the main batches of compatibilizer are injected in a concentrated manner, ensuring that fresh compatibilizer molecules are locally ready at the moment when the functional group activity is highest, thereby maximizing the crosslinking reaction efficiency.

[0058] S32 adjusts the rotation speed, kneading block combination, or throttle valve opening of the corresponding screw section according to the shear strength setting value of each section to change the polymer melt mixing mode.

[0059] For each section of the extruder whose shear strength can be independently controlled along the axial direction, the actuator makes mechanical adjustments based on the target set value of the shear strength of that section given in the processing strategy solver.

[0060] When the strategy requires a high shear dispersion effect in a certain mixing section, the screw speed of that section is increased through a servo drive mechanism. Simultaneously, the staggered angle combination of the kneading blocks at that point can be adjusted as needed; increasing the staggered angle enhances the stretching and shearing effect of the kneading blocks on the melt, promoting the breakup and dispersion of fiber agglomerates. Conversely, if the strategy requires a low-shear state in a certain area to prevent mechanical degradation of the formed cross-linked network due to excessive shear, the speed of that section is reduced, or the kneading blocks are replaced with conventional threaded conveying elements, switching to a gentler conveying and mixing mode.

[0061] Furthermore, by adjusting the opening of the throttling valves configured between each section or at the die head, the pressure difference before and after the throttling device can be changed, causing the molten material to be impeded to varying degrees when flowing through the throttling point, thereby altering the average residence time of the material in each melting section. When it is necessary to extend the residence time of a certain section to promote sufficient cross-linking reaction, the opening of the downstream throttling valve can be appropriately reduced; when it is necessary to shorten the residence time to prevent over-cross-linking, the opening can be increased.

[0062] Understandably, the combination and adjustment of the aforementioned differentiated screw speed, kneading block configuration, and throttle valve opening enable the same screw to create multiple functional zones along the axial direction, ranging from conveying, plasticizing, high-shear dispersion to low-shear uniform mixing, thereby achieving axial programmable control over fiber dispersion and matrix encapsulation force.

[0063] Regarding online monitoring and closed-loop control, the online monitoring system must be activated simultaneously with the execution of the aforementioned strategies to detect changes in the microscopic state of the melt in real time. Online rheological measurement units and impedance spectroscopy analysis units are integrated into the barrels of the mixing and homogenizing sections of the extruder, serving as the actuators for online monitoring.

[0064] The online rheological measurement units are arranged at intervals along the axial direction to acquire the melt storage modulus and its axial fluctuation amplitude, characterizing the fiber dispersion state. Its working principle is as follows: by applying small-amplitude oscillatory shear, the dynamic storage modulus G' of the melt is measured. In regions where fibers are uniformly dispersed, the melt microstructure is homogeneous, and the storage modulus exhibits a stable, gradual change trend along the axial direction. However, in regions where fibers agglomerate, local micro-flow field distortions occur, and the agglomerates, acting as stress concentration points, cause the storage modulus to exhibit significantly different values ​​and fluctuation characteristics compared to the uniformly dispersed regions. By monitoring whether the axial fluctuation amplitude of the storage modulus at each measuring point exceeds a preset threshold, it can be determined whether abnormal aggregation exists in that section.

[0065] The impedance spectroscopy unit utilizes the dielectric parameter changes induced by interfacial polarization to measure the composite dielectric constant and relaxation frequency of the melt. During the crosslinking reaction, chemical bonds are formed between the active compatibilizer molecules and the functional groups on the fiber surface, altering the charge distribution and polarization relaxation characteristics of the fiber-matrix interface. By monitoring changes in the dielectric relaxation spectrum online, the changes in the coupling degree between polymer segments and fiber particles caused by the crosslinking reaction between the active compatibilizer and fiber functional groups can be sensitively reflected, thereby indirectly characterizing the reduction in interfacial tension and the density of crosslinking points.

[0066] After acquiring the aforementioned real-time monitoring data, the control system performs closed-loop regulation based on this online monitoring data. When the axial fluctuation amplitude of the measured energy storage modulus in a certain axial section exceeds the preset threshold upper limit, or the dielectric parameter deviates from the target allowable range, it indicates that there may be significant fiber agglomeration or insufficient (or excessive) cross-linking in that area. The deviation signal is immediately fed back to the central control decision unit, and instead of invoking the original coarse-grained processing strategy, it directly responds to the deviation signal and automatically performs local closed-loop fine-tuning compensation control on the abnormal area.

[0067] Specific control actions include: applying intense but short-duration local disturbances to the melt in the abnormal micro-element zone through an auxiliary oscillation unit located near the barrel, using mechanical vibration energy to forcibly break up the newly formed soft fiber agglomerates and redistribute them in the melt; or, automatically making positive and negative fine adjustments to the screw speed in the unit zone according to the direction and amplitude of the deviation, or making small corrections to the current barrel temperature setpoint, thereby indirectly changing the local shear strength and melt viscosity to achieve the purpose of finely adjusting the micro-area dispersion conditions; when the total crosslinking degree is detected to be too low, the back pressure control unit at the end can also be automatically finely adjusted to appropriately extend the residence time of the material in the zone, so as to promote the reaction to further approach the target crosslinking degree.

[0068] S4, after extrusion molding, the pipe is subjected to zoned cooling and micro-controlled stretching to obtain a high-toughness composite MPP power pipe.

[0069] After the melt is formed into a tube blank through the extruder die, the temperature is still above the polymer melting point, and its internal microstructure (including fiber dispersion and cross-linking network distribution) is not yet fully fixed. If the cooling process is not properly controlled, new defects can easily be introduced due to factors such as thermal stress and crystallization shrinkage, causing the optimized microstructure to be destroyed during the shaping stage. Therefore, this step adopts a post-processing scheme that combines zoned cooling with micro-controlled stretching.

[0070] Zoned Cooling: The cooling water tank adopts a zoned, independently temperature-controlled design along the tube blank traction direction. Multiple temperature zones are sequentially set along the traction direction, with the cooling medium temperature in each zone decreasing gradually from the first zone near the die to the last zone. The temperature of the first zone is set slightly below the polypropylene crystallization temperature, allowing the outer layer of the tube blank to begin slow crystallization and shaping under relatively mild cooling conditions, avoiding internal stress caused by excessive differences in crystallinity between the inner and outer layers due to sudden cooling. As the tube blank is pulled forward, it successively enters subsequent cooling zones with progressively decreasing temperatures, allowing the tube wall to gradually cool and shape from the outside to the inside along the thickness direction according to a predetermined cooling curve. By adjusting the temperature and flow rate of the cooling medium in each zone, the heat transfer rate and crystallization process of the tube blank in each cooling stage can be precisely controlled, ensuring that the formed uniform fiber dispersion network and cross-linked structure are effectively locked in during the cooling process.

[0071] Micro-controlled stretching: During the cooling and shaping process, the tube blank moves forward at a constant linear speed under the pull of the traction machine. Due to the thermal shrinkage of the polymer during cooling, the outer diameter and wall thickness of the tube blank change accordingly. To compensate for the dimensional deviations caused by cooling shrinkage, an online diameter gauge is installed in the traction section to measure the outer diameter and ellipticity of the tube in real time. The control system compares the real-time outer diameter data fed back by the diameter gauge with the preset target outer diameter and performs closed-loop fine-tuning of the stretching rate of the traction machine. When the measured outer diameter is too large, the traction rate is slightly increased to increase the stretching ratio, allowing the tube to achieve slight orientation during cooling, while simultaneously tightening the dimensions circumferentially; when the measured outer diameter is too small, the traction rate is appropriately reduced to decrease the stretching amount. It can be understood that this micro-controlled stretching not only serves as a dimensional calibration tool but also further enhances the strength and toughness of the tube along the traction direction through slight thermal stretching and orientation. Finally, the high-toughness composite MPP power pipe is obtained after being cut to a fixed length.

[0072] Please see Figure 4 This invention provides a production control device 300 for the preparation of high-toughness MPP power pipes, the device comprising: The metering and preparation unit 301 is used to meter polypropylene, high-density polyethylene, composite glass fiber powder, active compatibilizer and auxiliary additives according to the formula; wherein, the composite glass fiber powder has been pretreated with activating liquid, polyethyleneimine and modified flame retardant, and has active functional groups on its surface. The parameter generation and strategy calculation unit 302 is used to determine the surface functional group distribution and reactivity of the composite glass fiber powder, form an initial process parameter dataset, and input it into the melt flow field and crosslinking reaction kinetics prediction model to generate a processing strategy. The setting and closed-loop control unit 303 is used to set the batch addition sequence of the active compatibilizer and the polymer melt mixing method according to the processing strategy, obtain online monitoring data of the extruder including fiber dispersion state, interfacial tension and crosslinking process, and perform closed-loop control based on the online monitoring data to optimize fiber dispersion and crosslinking reaction. The post-processing unit 304 is used to perform zoned cooling and micro-controlled stretching of the tube after extrusion molding to produce a high-toughness composite MPP power tube.

[0073] As an example, the parameter generation and strategy calculation unit 302 is configured to implement: The types of active functional groups and grafting density on the surface of composite glass fiber powder were determined by infrared spectroscopy and chemical titration, respectively, and its surface energy was determined by dynamic contact angle method. The measurement results are correlated with the proportions of each component, melt viscosity, and melt index. A pre-established process parameter mapping table based on fiber surface properties is retrieved to obtain the corresponding reference process condition range. This data is then combined with the measurement results, the proportions of each component, melt viscosity, and melt index to form an initial process parameter dataset.

[0074] As an example, the crosslinking reaction kinetic prediction model includes: Flow field The component mapper is used to calculate the axial distribution of fiber dispersion and active functional group concentration based on the initial process parameter dataset and real-time screw parameters. A crosslinking trend predictor is used to solve for the crosslinking degree baseline evolution curve along the screw axis based on the axial distribution of the fiber dispersion and the concentration of the active functional groups. The processing strategy solver is used to minimize the deviation between the crosslinking degree baseline evolution curve and the preset target crosslinking degree curve as the optimization objective, and outputs the processing strategy, including the batch addition sequence of active compatibilizer, the shear strength setting value of each segment, the location and intensity of local disturbances, and the amount of material residence time adjustment.

[0075] As an example, the crosslinking trend extrapolator includes: diffusion The reaction operator layer is used to discretize the screw axis into multiple micro-elements with the axial distribution of fiber dispersion and active functional group concentration as input, and to correct the effective collision frequency under non-uniform concentration field using the local Peckley number in each micro-element. The shear-induced delay factor estimation layer is used to output an interfacial reaction delay factor that characterizes the time lag from the exposure of functional groups to the point where they can participate in crosslinking, based on the local shear rate and material residence time of each micro-element. The derivation integration layer is used to couple the corrected effective collision frequency and the interface reaction delay factor to the crosslinking kinetic ordinary differential equation system, and to recursively integrate along the screw axis to output the crosslinking degree baseline evolution curve.

[0076] As an example, the setting and closed-loop control unit 303 is configured to achieve: Multiple compatibilizer injection ports arranged along the screw axis are independently controlled in terms of opening time, injection rate and injection volume to achieve batch addition; Based on the shear strength setting value of each section, adjust the rotation speed of the corresponding screw section, the kneading block combination, or the throttle valve opening to change the polymer melting and mixing method.

[0077] 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 production control method for manufacturing high-toughness MPP power pipes, characterized in that, Includes the following steps: Polypropylene, high-density polyethylene, composite glass fiber powder, active compatibilizer and auxiliary additives are measured separately according to the formula; among them, the composite glass fiber powder has been pretreated with activation liquid, polyethyleneimine and modified flame retardant, and has active functional groups on the surface. The surface functional group distribution and reactivity of composite glass fiber powder were determined to form an initial process parameter dataset, which was then input into a melt flow field and crosslinking reaction kinetics prediction model to generate a processing strategy. Based on the processing strategy, the batch addition sequence of the active compatibilizer and the polymer melt mixing method are set, and online monitoring data of the extruder, including fiber dispersion state, interfacial tension and crosslinking process, are obtained. Closed-loop control is then performed based on the online monitoring data to optimize fiber dispersion and crosslinking reaction. After extrusion molding, the pipe is subjected to zoned cooling and micro-controlled stretching to obtain a high-toughness composite MPP power pipe.

2. The production control method for preparing high-toughness MPP electric power pipes according to claim 1, characterized in that: The surface functional group distribution and reactivity of the composite glass fiber powder were determined to form an initial process parameter dataset, including: The types of active functional groups and grafting density on the surface of composite glass fiber powder were determined by infrared spectroscopy and chemical titration, respectively, and its surface energy was determined by dynamic contact angle method. The measurement results are correlated with the proportions of each component, melt viscosity, and melt index. A pre-established process parameter mapping table based on fiber surface properties is retrieved to obtain the corresponding reference process condition range. This data is then combined with the measurement results, the proportions of each component, melt viscosity, and melt index to form an initial process parameter dataset.

3. The production control method for preparing high-ductility MPP electric power pipes according to claim 1, characterized by: The crosslinking reaction kinetic prediction model includes: Flow field a component mapper for calculating axial distribution of fiber dispersion and active functional group concentration based on initial process parameter data set and real-time screw parameters; A crosslinking trend predictor is used to solve for the crosslinking degree baseline evolution curve along the screw axis based on the axial distribution of the fiber dispersion and the concentration of the active functional groups. The processing strategy solver is used to minimize the deviation between the crosslinking degree baseline evolution curve and the preset target crosslinking degree curve as the optimization objective, and outputs the processing strategy, including the batch addition sequence of active compatibilizer, the shear strength setting value of each segment, the location and intensity of local disturbances, and the amount of material residence time adjustment.

4. The production control method for manufacturing high-toughness MPP power pipes according to claim 3, characterized in that: The crosslinking trend predictor includes: diffusion The reaction operator layer is used to discretize the screw axis into multiple micro-elements with the axial distribution of fiber dispersion and active functional group concentration as input, and to correct the effective collision frequency under non-uniform concentration field using the local Peckley number in each micro-element. The shear-induced delay factor estimation layer is used to output an interfacial reaction delay factor that characterizes the time lag from the exposure of functional groups to the point where they can participate in crosslinking, based on the local shear rate and material residence time of each micro-element. The derivation integration layer is used to couple the corrected effective collision frequency and the interface reaction delay factor to the crosslinking kinetic ordinary differential equation system, and to recursively integrate along the screw axis to output the crosslinking degree baseline evolution curve.

5. The production control method for preparing high-ductility MPP electric power pipes according to claim 1, characterized by: The batch addition sequence of the active compatibilizer and the polymer melt mixing method are set according to the processing strategy, including: Multiple compatibilizer injection ports arranged along the screw axis are independently controlled in terms of opening time, injection rate and injection volume to achieve batch addition; Based on the shear strength settings of each section, adjust the rotation speed of the corresponding screw section, the kneading block combination, or the throttle valve opening to change the polymer melting and mixing method.

6. A production control device for the preparation of high ductility MPP electrical power pipes, characterized by: The device includes: The metering and preparation unit is used to meter polypropylene, high-density polyethylene, composite glass fiber powder, active compatibilizer and auxiliary additives according to the formula; wherein, the composite glass fiber powder has been pretreated with activating liquid, polyethyleneimine and modified flame retardant, and has active functional groups on its surface. The parameter generation and strategy calculation unit is used to determine the surface functional group distribution and reactivity of the composite glass fiber powder, form an initial process parameter dataset, and input it into the melt flow field and crosslinking reaction kinetics prediction model to generate a processing strategy. The setting and closed-loop control unit is used to set the batch addition sequence of the active compatibilizer and the polymer melt mixing method according to the processing strategy, obtain online monitoring data of the extruder including fiber dispersion state, interfacial tension and crosslinking process, and perform closed-loop control based on the online monitoring data to optimize fiber dispersion and crosslinking reaction. The post-processing unit is used to perform zoned cooling and micro-controlled stretching of the tube after extrusion molding to produce a high-toughness composite MPP power tube.

7. The production control apparatus for producing a high-ductility MPP electric power tube according to claim 6, characterized by, The parameter generation and strategy calculation unit is configured to implement: The types of active functional groups and grafting density on the surface of composite glass fiber powder were determined by infrared spectroscopy and chemical titration, respectively, and its surface energy was determined by dynamic contact angle method. The measurement results are correlated with the proportions of each component, melt viscosity, and melt index. A pre-established process parameter mapping table based on fiber surface properties is retrieved to obtain the corresponding reference process condition range. This data is then combined with the measurement results, the proportions of each component, melt viscosity, and melt index to form an initial process parameter dataset.

8. The production control apparatus for producing a high-ductility MPP electric power tube according to claim 6, characterized by, The crosslinking reaction kinetic prediction model includes: Flow field a component mapper for calculating axial distribution of fiber dispersion and active functional group concentration based on initial process parameter data set and real-time screw parameters; A crosslinking trend predictor is used to solve for the crosslinking degree baseline evolution curve along the screw axis based on the axial distribution of the fiber dispersion and the concentration of the active functional groups. The processing strategy solver is used to minimize the deviation between the crosslinking degree baseline evolution curve and the preset target crosslinking degree curve as the optimization objective, and outputs the processing strategy, including the batch addition sequence of active compatibilizer, the shear strength setting value of each segment, the location and intensity of local disturbances, and the amount of material residence time adjustment.

9. A production control device for manufacturing high-toughness MPP power pipes according to claim 8, characterized in that: The crosslinking trend predictor includes: diffusion The reaction operator layer is used to discretize the screw axis into multiple micro-elements with the axial distribution of fiber dispersion and active functional group concentration as input, and to correct the effective collision frequency under non-uniform concentration field using the local Peckley number in each micro-element. The shear-induced delay factor estimation layer is used to output an interfacial reaction delay factor that characterizes the time lag from the exposure of functional groups to the point where they can participate in crosslinking, based on the local shear rate and material residence time of each micro-element. The derivation integration layer is used to couple the corrected effective collision frequency and the interface reaction delay factor to the crosslinking kinetic ordinary differential equation system, and to recursively integrate along the screw axis to output the crosslinking degree baseline evolution curve.

10. The production control device for preparing a high-ductility MPP electric power tube according to claim 6, characterized by: The setting and closed-loop control unit is configured to achieve: Multiple compatibilizer injection ports arranged along the screw axis are independently controlled in terms of opening time, injection rate and injection volume to achieve batch addition; Based on the shear strength setting value of each section, adjust the rotation speed of the corresponding screw section, the kneading block combination, or the throttle valve opening to change the polymer melting and mixing method.

Images