Reverse prediction method for composite material additive manufacturing component with adjustable dielectric property
By preparing multiple composite additive manufacturing components with different physical parameters, and combining the Yamada dielectric theory model and the few-shot learning model of the meta-learning algorithm, the problem of rapid and efficient prediction of dielectric properties in composite additive manufacturing technology was solved, and precise control and low-cost prediction of dielectric properties were achieved.
Patent Information
- Application Number
- CN202511511669.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-22
- Publication Date
- 2026-03-20
AI Technical Summary
Existing additive manufacturing technologies for composite materials suffer from problems such as long R&D cycles, numerous iterations, and high costs in predicting dielectric properties, and lack fast and efficient reverse prediction methods.
Additive manufacturing technology was used to prepare multiple composite additive manufacturing components with different physical parameters. Data sets were obtained through testing and input into the prediction model for training. The dielectric properties were predicted using a few-shot learning model based on the Yamada dielectric theory model and the meta-learning algorithm. The dielectric constant and loss tangent were then precisely controlled by combining the third-order polynomial equations.
It enables rapid and accurate control and prediction of dielectric properties, reduces manufacturing costs, simplifies the process, shortens the R&D cycle, and improves prediction efficiency.
Smart Images
Figure CN121709066A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of additive manufacturing technology, and in particular to a reverse prediction method for additively manufactured composite material components with tunable dielectric properties. Background Technology
[0002] With the rapid development of 5G / 6G communications, flexible electronics (such as wearable sensors and flexible antennas), smart skins (such as adaptive stealth structures), and biomedical devices (such as biodegradable implantable sensors), various electronic devices are moving towards cutting-edge developments in high energy storage, miniaturization, lightweight, low cost, and high reliability. This places higher performance demands on the dielectric constant and loss tangent of dielectric materials, with urgent needs for tunable dielectric properties, lightweight structures, integrated structural functions, and wideband stability. Dielectric constant and loss tangent are two fundamental performance indicators describing dielectric properties. Different application fields have different requirements for the dielectric performance parameters of materials. Therefore, it is necessary to make the dielectric constant and loss tangent tunable and predictable to meet the dielectric performance requirements of materials in different fields, such as high-dielectric-constant materials or low-dielectric-constant materials, and dielectric materials with different orders of magnitude or sizes of loss tangents. The matrices of existing dielectric composite materials mainly include nonpolar polymers (e.g., polyethylene, polypropylene), polar polymers (e.g., polyvinylidene fluoride, epoxy resin), and bio-based materials (e.g., polyhydroxy fatty acids, polylactic acid). Filler systems primarily consist of ceramics (e.g., fillers, titanium dioxide), carbon-based materials (e.g., carbon nanotubes, graphene), and hybrid fillers. Single polymers typically have low dielectric constants, and dielectric properties are mainly improved through hybrid functional fillers. However, current technologies struggle to simultaneously achieve high dielectric constants, low loss tangents, good processability, lightweight, and low cost. Compared to traditional ceramic or semiconductor materials, polymer-based composite materials offer advantages such as lightweight construction, multi-processability, and designable dielectric properties.
[0003] Polymers, due to their diverse molecular structures and origins, exhibit vastly different material properties and application areas. Polypropylene, as a widely produced general-purpose plastic, has become an indispensable basic material in everything from everyday consumer goods to the automotive industry due to its low cost, excellent chemical resistance, and good overall performance. Nylon, on the other hand, is a high-performance engineering plastic. Its superior mechanical strength, wear resistance, and heat resistance make it ideal for manufacturing precision components and parts subjected to mechanical stress, and it is widely used in demanding industrial fields such as automobiles, electronics, and sporting goods. In short, from biodegradable bio-based plastics to cost-effective general-purpose plastics to high-performance engineering plastics, they meet the complex needs from environmental protection to daily life to high-end manufacturing. Inorganic fillers are mainly used to improve the dielectric constant of composite materials, while organic fillers focus more on maintaining or slightly improving the dielectric constant while significantly reducing dielectric loss and improving processability. Among inorganic fillers, barium titanate (BaTiO3) is a ferroelectric ceramic material with high dielectric constant and piezoelectric properties, which can improve the dielectric properties of composite materials.
[0004] However, current research and development on the dielectric properties of two-phase composite materials involves long cycles, numerous iterations, and high manufacturing costs. Furthermore, a rapid and efficient method for reverse-engineering the dielectric properties of materials is currently lacking. Therefore, existing additive manufacturing technologies for composite materials require further improvement and development in the design of high-dielectric-performance and low-loss components. Summary of the Invention
[0005] In view of the shortcomings of the prior art, the purpose of this invention is to provide a reverse prediction method for additively manufactured composite material components with adjustable dielectric properties, aiming to solve the problem of the lack of fast and efficient dielectric property prediction for additively manufactured composite material components.
[0006] The technical solution of the present invention is as follows: In a first aspect, the present invention provides a reverse prediction method for additively manufactured composite material components with tunable dielectric properties, comprising the following steps: Multiple composite material additive manufacturing components with different physical parameters were prepared using additive manufacturing technology; A dataset was obtained through testing, which included the physical parameters and dielectric properties of additively manufactured components made of different composite materials. The dataset is preprocessed and then input into the prediction model for training. The trained prediction model is then used to predict the dielectric properties of the composite additive manufacturing components corresponding to the physical parameters.
[0007] Optionally, the method for preparing each composite additive manufacturing component includes: The thermoplastic matrix material and filler are mixed evenly to form a premix, and the filler content in the premix is between 0-80 wt%. The premixed material is extruded into thermoplastic filaments, and then the thermoplastic filaments are water-cooled, rolled, air-dried, and wound into composite material wires. Composite material wires are printed into composite material additive manufacturing components using additive manufacturing technology.
[0008] Optionally, the thermoplastic matrix material is a thermoplastic polymer material, such as polypropylene (PP), nylon (PA), low-density polyethylene (LDPE), and different thermoplastic matrix materials are adapted to different thermal stability requirements; the filler is an inorganic filler or an organic filler, such as barium strontium titanate, titanium dioxide, polyetheretherketone powder.
[0009] It should be noted that dispersants and plasticizers may also be added to the premix.
[0010] Optionally, the premixed material is extruded into thermoplastic filaments using a single-screw extrusion process. The extrusion process parameters include: heating section temperature of 180-200℃, extrusion section temperature of 190-210℃, composite material wire diameter of 1.75±0.05mm, and extrusion speed of 300-550mm / min.
[0011] Optionally, the additive manufacturing technology uses fused deposition modeling (FDM), where key process parameters include a nozzle temperature of 200-220°C; and / or a heated bed temperature of 60-80°C; and / or a printing speed of 50-250 mm / min; and / or a printing extrusion rate of 100%-200%; and / or a nozzle diameter of 0.4 mm, 0.6 mm, or 0.8 mm.
[0012] Optionally, the physical parameters include one of filler content, filler density, or print layer height, or filler density and print layer height.
[0013] Optionally, the prediction model is the Yamada dielectric theory model, in which the filler shape factor n=0.56 of the composite material.
[0014] In this embodiment, the dataset is input into the Yamada dielectric theory model for training, resulting in a trained Yamada dielectric theory model. Using this trained model, the dielectric constant and loss tangent corresponding to the physical parameters (filler content) are predicted. This embodiment achieves accurate reverse prediction of filler content using the dielectric constant, with a prediction error of less than 3.25%. Therefore, the required filler content for the composite material can be predicted based on the desired dielectric constant.
[0015] Optionally, the prediction model is a third-order polynomial equation, wherein R in the third-order polynomial equation 2 =0.982.
[0016] In this embodiment, the dataset is fitted with a third-order polynomial equation to obtain a well-fitted third-order polynomial equation, and the well-fitted third-order polynomial equation is used to predict the dielectric properties (loss tangent) corresponding to the physical parameters (filler content).
[0017] Optionally, the prediction model is a few-shot learning model of a meta-learning algorithm. In this embodiment, the dataset is input into the few-shot learning model of the meta-learning algorithm for training to obtain a trained few-shot learning model of the meta-learning algorithm. The trained few-shot learning model of the meta-learning algorithm is then used to predict the dielectric constant and loss tangent of the composite additive manufacturing component corresponding to the infill density or printing layer height.
[0018] Optionally, the filling density is 10%-100%.
[0019] Optionally, the layer height is 0.10-0.25mm, specifically 0.10, 0.15, 0.20, or 0.25mm.
[0020] It should be noted that the dataset is divided into a training set (60%), a validation set (20%), and a test set (20%).
[0021] Beneficial Effects: This invention employs an additive manufacturing process to prepare composite material additive manufacturing components. Using these components, combined with a predictive model, the corresponding dielectric properties (including dielectric constant and / or loss tangent) can be predicted through physical parameters. In other words, the dielectric properties of the composite material additive manufacturing components can be precisely controlled by adjusting the overall physical parameters of the composite material (including the filler content of the reactant material, the filling density of the reaction structure, and the printing layer height of the reaction process). The method of this invention is low-cost, simple in process, and provides accurate control and prediction of dielectric properties. Attached Figure Description
[0022] Figure 1 This is a flowchart of the method for predicting the dielectric properties of composite materials according to an embodiment of the present invention.
[0023] Figure 2 (a) is a scanning electron microscope image of the brittle fracture surface of pure PP printed wire, and (b) is a scanning electron microscope image of the brittle fracture surface of the PP / BaTiO3 (20wt%) composite material wire of the present invention.
[0024] Figure 3 The graph shows the relationship between the BaTiO3 filler content and the dielectric constant and loss tangent of the composite additive manufacturing component (PP / BaTiO3).
[0025] Figure 4A comparison of the theoretical model predictions and test values of the dielectric constants of additively manufactured components (PP / BaTiO3) with different BaTiO3 filler contents.
[0026] Figure 5 The diagram shows the influence of dielectric properties and additive filler density on additively manufactured components (PP / BaTiO3) of composite materials.
[0027] Figure 6 Inverse prediction region map of physical parameters for additive manufacturing of composite materials. Detailed Implementation
[0028] This invention provides a reverse prediction method for additive manufacturing components of composite materials with tunable dielectric properties. To make the objectives, technical solutions, and effects of this invention clearer and more explicit, the invention is further described in detail below. It should be understood that the specific embodiments described herein are only for explaining the invention and are not intended to limit the invention.
[0029] This embodiment provides a reverse prediction method for additively manufactured composite material components with tunable dielectric properties, such as... Figure 1 As shown, it includes the following steps: S1. Using additive manufacturing technology, multiple composite material additive manufacturing components with different physical parameters are prepared. S2. Obtain a dataset through testing, the dataset including the physical parameters and dielectric properties of different composite materials; S3. After preprocessing, the dataset is input into the prediction model for training, and the trained prediction model is used to predict the dielectric properties corresponding to the physical parameters.
[0030] This embodiment employs additive manufacturing technology (an integrated melt-laminated molding process, comprising: first preparing a premix, then preparing composite material wire, and finally preparing a composite material additive manufacturing component) to prepare a composite material additive manufacturing component. This composite material is a two-phase composite material. Combining dielectric constant theory and machine learning prediction models, the dielectric properties (including dielectric constant and loss tangent) of the composite material additive manufacturing component can be inversely predicted by adjusting physical parameters. This method can precisely control the dielectric properties of the composite material additive manufacturing component by adjusting the physical parameters of the composite material and the additive manufacturing process (including filler content, structural filler density, and process printing layer height). The method of this invention has low cost, simple process flow, and accurate dielectric property control and inverse prediction.
[0031] In some embodiments, the method for preparing each composite additive manufacturing component includes: The thermoplastic matrix material and filler are mixed evenly to form a premix, and the filler content in the premix is between 0-80 wt%. The premixed material is extruded into thermoplastic filaments, and then the thermoplastic filaments are water-cooled, rolled, air-dried, and wound into composite additive manufacturing component printing wires. Composite material wires are printed into composite material additive manufacturing components using additive manufacturing technology.
[0032] This embodiment lays the foundation for predicting the dielectric properties of additively manufactured composite components by preparing a two-phase composite material whose dielectric properties change accordingly with variations in physical parameters. Not all composite materials prepared by various methods can be used to predict the dielectric properties of additively manufactured composite components; only those prepared using additive manufacturing technology can be used to predict the dielectric constant and loss tangent. Experiments have shown that the BaTiO3 filler content in the composite wire, ranging from 0-80 wt% (including 0 wt%, 5 wt%, 10 wt%, 15 wt%, 20 wt%, 25 wt%, 30 wt%, 35 wt%, 40 wt%, 45 wt%, 50 wt%, 55 wt%, 60 wt%, 65 wt%, 70 wt%, 75 wt%, and 80 wt%), can meet the requirements for controlling and predicting the dielectric properties of additively manufactured composite components. This embodiment yields a two-phase composite material with a wide adjustable range of dielectric properties (dielectric constant and loss tangent). Specifically: the adjustable range of dielectric constant value meets the dielectric requirements of materials from low frequency to millimeter wave frequency band, with an adjustable range of 2.67 (filler content 0wt%) to 13.91 (filler content 80wt%); the adjustable range of loss tangent value meets the requirements from low loss to specific energy dissipation scenarios, with an adjustable range of 0.0097 (filler content 0wt%) to 0.0437 (filler content 80wt%).
[0033] Self-made PP and PP / BaTiO3 wire cross-sections are as follows Figure 2 As shown in (b), Figure 2 (a) is a scanning electron microscope image of the cross-section of pure PP printing filament, and (b) is a scanning electron microscope image of the cross-section of PP / BaTiO3 (20wt%) filament of the additive manufacturing component of the composite material of the present invention. It can be seen that the additive manufacturing component of the composite material of the present invention has a relatively regular morphology, uniform distribution of matrix and filler, and relatively uniform particle size distribution.
[0034] In some embodiments, the thermoplastic matrix material is a thermoplastic polymer material, such as polypropylene (PP), nylon (PA), or low-density polyethylene (LDPE). Different thermoplastic matrix materials are adapted to different thermal stability requirements. The filler is an inorganic filler or an organic filler, such as barium strontium titanate, titanium dioxide, or polyetheretherketone powder.
[0035] It should be noted that the lithium niobate and strontium titanate have higher dielectric constants, which can improve the dielectric response in high-frequency scenarios. The composite filler system can be a BaTiO3 / graphene (GO) core-shell structure filler, which enhances the dielectric constant through interfacial polarization (e.g., BaTiO3@GO increases the dielectric constant to 18.0), while using graphene to improve thermal conductivity. The inorganic filler: introduces ferrite (e.g., CoFe2O4, NiZnFe2O4) or metal nanoparticles (e.g., Fe, Ni) to construct a magnetoelectric coupling functional composite material.
[0036] It should be noted that an appropriate amount of dispersant or plasticizer may also be added to the premix.
[0037] The dispersant enhances the interfacial bonding between the nanofiller and the matrix, reducing dielectric loss caused by agglomeration. The plasticizer optimizes melt flowability, reducing the difficulty of composite additive printing processes and defects in composite additive manufactured components.
[0038] It should be noted that the polypropylene (PP) mentioned is the matrix material, with a density of 0.91 g / cm³. 3 The particle size is 100 mesh, and the dielectric constant is 2.36. Barium titanate (BaTiO3), as an inorganic filler, has a particle size of 110 nm and a density of 6.08 g / cm³. 3 The dielectric constant is 1500. Both the matrix and filler were dried in a 70°C oven for 12 hours before mixing to remove moisture and volatiles.
[0039] It should be noted that a planetary vacuum mixer was used for mechanical mixing of polypropylene (PP) and filler (BaTiO3) to ensure uniform mixing, with a speed of 500 r / min and a mixing time of 30 minutes.
[0040] In some embodiments, a single-screw extrusion process is used to extrude and roll the premix into a thermoplastic printing filament. The extrusion process parameters include: heating section temperature of 180-200℃, extrusion section temperature of 190-210℃, filament diameter of 1.75±0.05mm, and extrusion speed of 300-550mm / min.
[0041] In some implementations, fused deposition modeling (FDM) is used in additive manufacturing, with key process parameters including nozzle temperature: 200-220°C; heated bed temperature: 60-80°C; printing speed: 50-250 mm / min; printing extrusion rate: 100%-200%; and nozzle diameter: 0.4 mm, 0.6 mm, or 0.8 mm.
[0042] When measuring the dielectric constant and loss tangent of composite additive manufacturing components, an Agilent Technologies ENA series network analyzer (E5071C) with an OWED N150E split-column dielectric resonator (SPDR) was used for dielectric performance testing. The SPDR method is a high-precision method for measuring the dielectric constant of materials. The composite additive manufacturing components being tested are sheet-like, with a test frequency of 10 GHz. The maximum size of the composite additive manufacturing components is 80 × 80 mm, with a thickness of less than 0.90 mm. The upper and lower surfaces of the sample must be highly parallel and smooth to ensure good contact with the resonator's metal plate. The test was conducted in a constant temperature environment (e.g., 25 ± 1 °C), and the test standard was IEC 61189-2-721.
[0043] In some implementations, the physical parameters include filler content, filler density, and print layer height.
[0044] In one embodiment, the dielectric constant prediction model is the Yamada dielectric theory model, in which the filler shape factor n=0.56 of the composite material.
[0045] In this embodiment, the dataset is input into the Yamada dielectric theory model to determine the filler shape factor of the composite material. This model predicts the dielectric constant of the additively manufactured composite material component based on the physical parameters (filler content). In this embodiment, the target dielectric constant of the Yamada dielectric theory model is used to accurately predict the filler content of the additively manufactured composite material component with an error of 3.25%. Therefore, the filler content of the additively manufactured composite material component (PP / BaTiO3) can be predicted inversely based on the target dielectric constant. It should be noted that the Yamada dielectric theory model focuses on the theoretical model of polarization behavior in dielectric materials, describing the influence of dipole ordering on dielectric behavior. Using the dataset, a Yamada dielectric theory model suitable for this material is constructed to determine the filler shape factor n value, clarifying the mapping relationship between the BaTiO3 filler content and the dielectric constant.
[0046] like Figure 3 The figure shows the relationship between BaTiO3 filler content and the dielectric constant and loss tangent of the composite additive manufacturing component (PP / BaTiO3). Figure 3 In (a), the measured dielectric constant values are shown for the additive manufacturing components (PP / BaTiO3) of the present invention when the BaTiO3 filler content is 0-80 wt%. Figure 3 In (b), the test loss tangent value is the BaTiO3 filler content in the additive manufacturing component (PP / BaTiO3) of the present invention when the content is 0-80wt%.
[0047] Figure 4This figure compares the predicted and measured values of the dielectric constant of additively manufactured PP / BaTiO3 composite components under different BaTiO3 filler contents using three theoretical models. The measured values are compared with the predicted dielectric constant values from the Maxwell-Garnett, Yamada, and Effective Medium Theory models when the BaTiO3 filler content is 0-80 wt%. The figure shows that the Yamada dielectric theory model prediction has a good fit with the measured values, with a relative error of 3.25%.
[0048] In one implementation, the prediction model is a third-order polynomial equation, where R0 is a polynomial equation. 2 =0.982.
[0049] In this embodiment, the dataset is used to fit a third-order polynomial equation, and the fitted third-order polynomial equation is used to predict the dielectric properties (loss tangent) of the composite additive manufacturing component corresponding to the physical parameters (filler content ratio).
[0050] This invention combines dielectric theory models and meta-learning algorithms to significantly shorten the R&D cycle and improve efficiency compared to trial-and-error methods. For example, when the target dielectric constant of the dielectric substrate is 7.50, the prediction method directly outputs a filler content of 26.04 wt% for the (PP / BaTiO3) composite material. Experimental studies show that this invention uses the Yamada dielectric theory model to predict the dielectric constant of the additively manufactured composite material component, with a relative error of 1.93% between the predicted and tested values, and a predicted target dielectric constant of 7.36 ± 0.03. Furthermore, it uses a third-order polynomial fitting equation to predict the loss tangent of the additively manufactured composite material component, with a relative error of 6.15% between the predicted and tested values, and a predicted target loss tangent of (1.56 ± 0.01) × 10⁻⁶. -2 .
[0051] In one implementation, the prediction model is a few-shot learning model based on a meta-learning algorithm. In this embodiment, the dataset is input into the few-shot learning model of the meta-learning algorithm for training, and the trained few-shot learning model of the meta-learning algorithm is used to predict the dielectric properties of the composite additive manufacturing components corresponding to the linear infill density (10%-100%) and the printing layer height (0.10mm, 0.15mm, 0.20mm, 0.25mm).
[0052] It should be noted that the Meta-Learning (MAML) algorithm is used to set the initial parameters of the model, enabling it to quickly adjust to the optimal model parameters suitable for the current task, guided by a small number of gradient updates within the inner loop. The MAML algorithm optimizes the position of the initial parameters, placing them in a region sensitive to task changes. The model only requires a small number of samples for learning (FSL) to rapidly descend along the gradient direction of the loss function to the optimal solution for the new task. The trained few-sample meta-learning model is selected to predict the dielectric properties of the aforementioned composite additive manufacturing components. The few-sample meta-learning model is used for training and validation to minimize the error between the predicted and actual values; in this embodiment, the error between the measured and predicted dielectric properties is less than 5%.
[0053] It should be noted that the few-shot learning (FSL-MAML) model prediction method based on meta-learning algorithms predicts the dielectric properties of composite additive manufacturing components by adjusting the infill density. This model is relatively accurate in predicting the dielectric constant of linearly infilled composite additive manufacturing components, with a small relative error between the tested and predicted values. Experimental studies show that the dielectric constant and loss tangent of composite additive manufacturing components increase linearly with increasing infill density, and the dominant mechanism is the decrease in porosity and the enhancement of polarization. The printing layer height has a weaker impact on the dielectric properties of composite additive manufacturing components, exhibiting a nonlinear law, and the dominant mechanism is the synergistic effect of interface effects and porosity.
[0054] In some embodiments, the linear fill density is 10%-100%.
[0055] In some embodiments, the printed layer height is 0.10 mm, 0.15 mm, 0.20 mm, or 0.25 mm.
[0056] It should be noted that the dataset is divided into 60% training set, 20% validation set, and 20% test set.
[0057] Figure 5 The graph shows the influence of dielectric properties on additive manufacturing density of PP / BaTiO3 composite components, where (a) represents the dielectric constant of additive manufacturing components with linear filler densities of 10%-100%, and (b) represents the loss tangent of additive manufacturing components with linear filler densities of 10%-100%. Figure 5 It can be concluded that the dielectric constant and loss tangent of the additively manufactured composite material components increase linearly with the increase of linear filler density.
[0058] Figure 6The diagram shows the reverse prediction region for physical parameters in composite additive manufacturing. (a) illustrates the reverse design of the dielectric constant of the composite additive manufacturing component by coordinating the printing layer height and filler density. (b) illustrates the reverse design of the loss tangent of the composite additive manufacturing component by coordinating the printing layer height and filler density. Figure 6 It can be concluded that the filler density has the most significant effect on the dielectric constant and loss tangent, while the printing layer height has no significant effect on the dielectric constant and loss tangent. Therefore, the filler density is preferred for controlling the dielectric properties of additively manufactured composite components.
[0059] In summary, this invention employs an integrated melt-laminated deposition modeling (FLM) process (first preparing a premix, then preparing a printable composite filament, and finally preparing a composite additive manufacturing component) to fabricate composite additive manufacturing components. A few-shot learning (FSL-MAML) model using meta-learning algorithms is employed to predict the dielectric properties of the composite additive manufacturing components. This invention rapidly predicts the dielectric constant and loss tangent of the composite additive manufacturing components by adjusting physical parameters. Overall, additive physical parameters (mainly including material filler ratio, structural filler density, and process printing layer height) precisely control the dielectric properties of the composite additive manufacturing components. The method of this invention features low manufacturing cost, simple process flow, reverse performance control, and accurate performance prediction.
[0060] A few-shot learning (FSL-MAML) model prediction method based on meta-learning algorithm predicts the dielectric properties of additively manufactured composite components by adjusting the infill density. Experiments show that the relative error between the predicted dielectric constant and the measured value is 2.47%. Experimental results indicate that the measured dielectric constant and loss tangent of the additively manufactured composite component (PP / BaTiO3) exhibit a linear relationship with increasing infill density, primarily due to decreased porosity and enhanced polarization. The printing layer height has a weaker impact on the dielectric properties of the additively manufactured composite component, exhibiting a non-linear relationship, primarily due to the synergistic effect of interface effects and porosity.
[0061] This invention employs the Yamada dielectric theory model to reverse-engineer the filler content of composite additive manufacturing components corresponding to the target dielectric constant, predicting the dielectric constant of the composite additive manufacturing components. Based on a third-order polynomial fitting equation, it predicts the loss tangent value of the composite additive manufacturing components corresponding to the filler content. Combining the dielectric theory model and the fitting equation significantly shortens the design and development cycle of the dielectric properties of composite additive manufacturing components, significantly improving efficiency compared to the trial-and-error method. For example, when the target dielectric constant is 7.50, this prediction method directly and quickly obtains a composite filler content of 26.04 wt%. Experimental studies show that the Yamada dielectric theory model can accurately predict the dielectric constant of two-phase composite additive manufacturing components, with a relative error of only 1.93% between the predicted and tested values, and a predicted dielectric constant of 7.36 ± 0.03. The third-order polynomial fitting equation is used to predict the loss tangent value of the composite additive manufacturing components, with a relative error of 6.15% between the predicted and tested values, and a predicted loss tangent value of (1.56 ± 0.01) × 10⁻⁶. -2 .
[0062] It should be clearly stated that the scope of application of this invention is not limited to the examples above. For those skilled in the art or R&D personnel, the process flow can be optimized or improved based on the above description, and all such optimized or improved process flows should fall within the protection scope of the appended claims of this invention.
Claims
1. A reverse prediction method for additively manufactured composite material components with tunable dielectric properties, characterized in that, Includes the following steps: Multiple composite material additive manufacturing components with different physical parameters were prepared using additive manufacturing technology; A dataset was obtained through testing, which included the physical parameters and dielectric properties of additively manufactured components made of different composite materials. The dataset is preprocessed and then input into the prediction model for training. The trained prediction model is then used to predict the dielectric properties of the composite additive manufacturing components corresponding to the physical parameters.
2. The reverse prediction method for additively manufactured composite material components with tunable dielectric properties according to claim 1, characterized in that, Each composite additive manufacturing component fabrication method includes: The thermoplastic matrix material and filler are mixed evenly to form a premix, and the filler content in the premix is between 0-80 wt%. The premixed material is extruded into thermoplastic filaments, and then the thermoplastic filaments are water-cooled, rolled, air-dried, and wound into composite material wires. Composite material wires are printed into composite material additive manufacturing components using additive manufacturing technology.
3. The reverse prediction method for additively manufactured composite material components with tunable dielectric properties according to claim 2, characterized in that, The thermoplastic matrix material is a thermoplastic polymer material; the filler is an inorganic filler or an organic filler.
4. The reverse prediction method for additively manufactured composite material components with tunable dielectric properties according to claim 2, characterized in that, The premixed material is extruded into thermoplastic filaments using a single-screw extrusion process. The extrusion process parameters include: heating section temperature of 180-200℃, extrusion section temperature of 190-210℃, composite material filament diameter of 1.75±0.05mm, and extrusion speed of 300-550mm / min.
5. The reverse prediction method for additively manufactured composite material components with tunable dielectric properties according to claim 2, characterized in that, In additive manufacturing, fused deposition modeling is selected, and key process parameters include a nozzle temperature of 200-220℃. The temperature of the heated bed is 60-80℃; Printing speed is 50-250 mm / min; The printing extrusion rate is 100%-200%; The nozzle diameter is 0.4mm, 0.6mm or 0.8mm.
6. The reverse prediction method for additively manufactured composite material components with tunable dielectric properties according to claim 1, characterized in that, The physical parameters include one of filler content, filler density, or print layer height, or filler density and print layer height.
7. The reverse prediction method for additively manufactured composite material components with tunable dielectric properties according to claim 1, characterized in that, The prediction model is the Yamada dielectric theory model, in which the filler shape factor n=0.56 of the composite material.
8. The reverse prediction method for additively manufactured composite material components with tunable dielectric properties according to claim 1, characterized in that, The prediction model is a third-order polynomial equation, where R... 2 =0.
982.
9. The reverse prediction method for additively manufactured composite material components with tunable dielectric properties according to claim 1, characterized in that, The prediction model is a few-shot learning model based on the meta-learning algorithm.
10. The reverse prediction method for additively manufactured composite material components with tunable dielectric properties according to claim 6, characterized in that, The filling density is 10%-100%; And / or, the printed layer height is 0.10mm, 0.15mm, 0.20mm or 0.25mm.