Manufacturing method of high heat dissipation efficiency potting type driving transformer

Abstract

The application discloses a manufacturing method of a high-heat-dissipation-efficiency pouring type driving transformer, and belongs to the technical field of power electronic equipment manufacturing. The method aims to solve the technical problems of low heat dissipation efficiency and insufficient long-term reliability of the existing driving transformer caused by the limited heat conduction performance of the pouring material, the simple surface heat dissipation structure and the difficult coordination of internal thermal stress. The method comprises the following steps: constructing a three-dimensional digital model of a reinforced heat dissipation structure with a boss array and a groove array and preparing a corresponding forming die; positioning a magnetic core and a winding as an insert in a die cavity; pouring insulating colloid with different mechanical and heat conduction properties in batches in a vacuum environment; through a segmented curing process, the flexible colloid is fully filled in the internal gap and buffers the stress, and at the same time, the high-heat-conduction rigid colloid forms an external main body and a heat dissipation structure, thereby forming an integrated pouring body. The method is used for manufacturing the driving transformer suitable for high-power-density and high-heat-dissipation scenes, and can significantly improve the heat dissipation capacity and operation reliability of the driving transformer.

Description

Technical Field

[0001] This invention belongs to the field of power electronic equipment manufacturing technology, specifically relating to a method for manufacturing a high heat dissipation efficiency potted drive transformer. Background Technology

[0002] In the field of power electronics equipment, drive transformers are key components for achieving energy transmission and signal isolation. With the continuous increase in power density of equipment, internal heat generation has become an increasingly prominent issue. Effective heat dissipation design is crucial for ensuring the long-term reliability and stability of transformer operation.

[0003] Currently, heat dissipation solutions for drive transformers mainly rely on two methods: natural air cooling and forced air cooling. Natural air cooling enhances heat dissipation by increasing the transformer surface area, optimizing PCB copper plating, and adding heat sinks; however, its efficiency is significantly affected by ambient temperature, its effectiveness is limited in enclosed spaces, and achieving effective heat dissipation often requires increasing the product size. Forced air cooling, driven by fans, can improve heat dissipation capacity, but it introduces moving mechanical parts, posing reliability risks; the fan's own power consumption reduces system efficiency, operating noise limits application scenarios, and dust accumulation can affect long-term heat dissipation and increase maintenance needs.

[0004] In existing technologies, potting processes can provide insulation protection, but conventional potting materials have limited thermal conductivity, and the outer surface of the potting compound is usually a simple geometry, failing to fully utilize the surface area to enhance heat dissipation. Simply increasing the thermal conductivity of the potting material often leads to a decrease in material flowability, affecting the filling effect, or causes stress during temperature cycling due to a mismatch between its thermal expansion coefficient and that of the internal components, affecting structural integrity.

[0005] Therefore, under the current technological conditions, how to effectively improve the heat dissipation efficiency of the drive transformer while ensuring structural reliability and process feasibility remains a technical problem that needs to be solved. Summary of the Invention

[0006] This invention provides a manufacturing method for a high-heat-dissipation-efficiency potted drive transformer. It can construct a reinforced heat dissipation structure on the surface of the drive transformer through an integrated potting molding process. Combined with segmented potting and segmented curing processes, and using a combination of soft inner and rigid outer colloidal materials, it can effectively improve heat dissipation efficiency and reduce internal stress. At the same time, the design of metal heat-conducting pillars and air-guiding grooves further optimizes heat conduction and airflow guidance, thereby achieving high reliability and long service life operation of the transformer under high power density.

[0007] To achieve these objectives and other advantages of the present invention, a method for manufacturing a high-heat-dissipation-efficiency encapsulated drive transformer is provided, comprising the following steps: S1. Constructing a three-dimensional model: Based on the physical structure of the core and windings of the drive transformer, a three-dimensional digital model of the package is constructed; the surface of the three-dimensional digital model includes a heat dissipation enhancement structure composed of a combination of a boss array and a groove array; S2. Molding mold preparation: A molding mold with a cavity surface that is complementary to the shape of the three-dimensional digital model is prepared by CNC machining or additive manufacturing method. S3. Positioning and Mold Closure: The assembled magnetic core and winding are positioned and fixed inside the cavity of the forming mold as inserts, and then the mold is closed. S4. Vacuum potting and curing: Under vacuum conditions, liquid thermally conductive insulating adhesive is poured into the cavity of the closed molding mold until the insert is completely submerged; then, according to the material properties of the thermally conductive insulating adhesive, a heating and curing process is carried out to cure the thermally conductive insulating adhesive inside the cavity. S5. Demolding: After the thermally conductive insulating adhesive has completely cured, open the mold and take out the integrated potted transformer with the enhanced heat dissipation structure on the surface that is consistent with the three-dimensional digital model.

[0008] Preferably, in step S4, the liquid thermally conductive insulating adhesive is poured in two stages and cured in stages: S410. Inject the first colloid, the viscosity of which is 5000-15000 cP, so that it can fully fill the tiny gap between the magnetic core and the winding in a vacuum environment; then, perform the first stage of pre-curing at a temperature of 40-60°C for 30-60 minutes, so that the first colloid reaches the gel state and its Shore A hardness reaches 30-50. S420. Inject the second colloid, the thermal conductivity of which is not less than 1.5 W / m·K at 25°C; carry out the second stage of heating and curing, heating from 60°C to 100-120°C at a rate of 1-3°C / min, and holding at this temperature for 60-120 minutes to completely cure the second colloid, with a Shore D hardness of not less than 70. The elastic modulus of the first colloid is less than 3 MPa, the elastic modulus of the second colloid is greater than 5 GPa, and the coefficient of thermal expansion of the first colloid is 15-25 ppm / ℃ in the range of 25-100℃, which is between the coefficient of thermal expansion of the magnetic core and the coefficient of thermal expansion of the winding, thus coordinating with the thermal expansion behavior of the assembly composed of the magnetic core and the winding.

[0009] Preferably, the first colloid is a flexible organosilicon thermally conductive adhesive with spherical alumina and fumed silica as fillers, wherein the spherical alumina accounts for 60%-70% of the filler mass and the fumed silica accounts for 1%-3% of the filler mass; the second colloid is a high thermally conductive epoxy resin adhesive with compound alumina and nano-aluminum nitride as fillers, wherein the compound alumina accounts for 60%-70% of the filler mass and the nano-aluminum nitride accounts for 2%-4% of the filler mass.

[0010] Preferably, step S3 specifically includes the following sub-steps: S310, Initial positioning of inserts: The assembled magnetic core and windings are initially placed in the cavity of the forming mold as inserts, so that they are initially aligned with the positioning structure on the surface of the cavity; S320. Installing Metal Heat-Conducting Pillars: Within the cavity, install metal heat-conducting pillars on the insert that has been initially positioned; use high thermal conductivity insulating tape or thermally conductive adhesive to fix one end of the metal heat-conducting pillar to the surface of the heating area of ​​the magnetic core to form surface contact; or, use a binding strap made of non-magnetic metal strip to tightly bind one end of the metal heat-conducting pillar to the outer surface of the heating area of ​​the magnetic core. S330, Mold Closure and Final Fixing: After confirming that the other end of the metal heat-conducting pillar has correctly extended into the space reserved for the root of the boss array in the cavity and that there is no structural interference with the mold, close the mold and finally and stably constrain the insert in the preset position by the elastic limiting block. The metal heat-conducting pillar is made of copper or aluminum alloy, and its extension length ensures that the end is completely wrapped by the cured colloid at the root of the boss array.

[0011] Preferably, in step S2, when preparing the molding die, a raised venting rib with a semi-elliptical or trapezoidal cross-section is formed on the raised structure corresponding to the groove array of the three-dimensional digital model on the surface of its cavity. The raised venting rib has a height of 0.1-0.3 mm, a width of 0.5-1 mm, and its top extension direction is consistent with the direction of the groove array. Correspondingly, after demolding in step S5, the inner surface of the groove array of the heat-reducing structure of the potting transformer, corresponding to the position of the raised exhaust rib, is integrally formed with a concave air guide groove with a complementary cross-sectional shape; the extension direction of the concave air guide groove is configured such that when the drive transformer is assembled in the system heat dissipation duct at a preset installation angle, the direction of the air guide groove forms an acute angle of 0-15 degrees with the mainstream direction of the forced cooling airflow.

[0012] Preferably, the step of insulating the metal heat-conducting pillar includes: The required thickness of the insulation layer for the metal heat-conducting pillar is determined based on the operating voltage level of the drive transformer. When the operating voltage of the drive transformer is ≤100V, if the metal heat-conducting column is made of aluminum alloy, an anodizing process is used to generate a dense anodized aluminum insulating layer with a thickness of 10-30μm on its surface; if it is made of copper, a plasma spraying process is used to prepare an alumina ceramic insulating layer with a thickness of 50-150μm on its surface.

[0013] When the operating voltage of the drive transformer is 100V to 1000V, if the metal heat-conducting column is made of aluminum alloy, its anodized aluminum insulation layer thickness is 30-60μm; if it is made of copper, its plasma-sprayed alumina ceramic insulation layer thickness is 150-300μm. Before anodizing or plasma spraying, the surface of the metal heat-conducting column to be treated is roughened by sandblasting and cleaned with organic solvents. Furthermore, the metal heat-conducting pillars after insulation treatment must pass the power frequency withstand voltage test, with a test voltage of not less than twice the operating voltage of the drive transformer plus 1000V, and must withstand the test for 1 minute without breakdown or flashover.

[0014] Preferably, in step S1, the boss array and groove array in the enhanced heat dissipation structure satisfy the following structural relationship: the boss array is a plurality of equally spaced cylindrical or frustum structures with a height of 1.5–3.0 mm, a cross-sectional diameter of 0.8–2.0 mm, and a center-to-center distance of 3–6 mm between adjacent bosses; the groove array is a channel connecting adjacent bosses, with a trapezoidal or arc-shaped cross-section, a depth of 1.0–2.5 mm, and a width of 1.5–3.0 mm; the bosses and grooves form a continuously alternating heat dissipation unit on the surface of the package, used to increase the heat dissipation surface area and guide the cooling airflow along a preset path.

[0015] Preferably, in step S3, the positioning and fixing method specifically includes: The cavity surface of the molding die is provided with one or more positioning grooves that are adapted to the protruding parts in the outer contour of the magnetic core. The corresponding protruding parts of the magnetic core are embedded in the positioning grooves to achieve initial positioning in the horizontal direction. Meanwhile, a preload force in the vertical direction is applied to the non-critical surfaces of the winding or magnetic core by an elastic limiting block set in the cavity. The elastic limiting block is made of high-temperature resistant silicone rubber or spring pins, and the preload force provided is in the range of 3-15 N, thereby stably constraining the insert in a preset position in the cavity.

[0016] Preferably, in step S4, the pre-tightening force applied by the elastic limiting block is released after the first colloid has completed pre-curing and reached the gel state, but before the second colloid is injected; after the pre-tightening force is released, the elastic limiting block disengages from the surface of the insert, and the cavity area corresponding to it becomes the injection space for the second colloid.

[0017] Preferably, in step S4, the vacuum level of the vacuum environment is between 1 Pa and 0.1 Pa.

[0018] The present invention has at least the following beneficial effects: First, this invention, through the construction of a three-dimensional model and mold preparation, forms an integrally molded encapsulated transformer with a surface-enhanced heat dissipation structure. The array of bosses and grooves increases the heat dissipation surface area and improves the contact efficiency with the cooling medium. Vacuum encapsulation ensures full filling of the colloid, avoids air bubbles, and improves the uniformity of thermal conductivity. This method overcomes the limitations of traditional heat dissipation methods that rely on air cooling, eliminating the need for additional fans, reducing noise and dust problems, while maintaining a compact structure, making it suitable for high power density applications. The overall manufacturing process is simplified, product consistency is good, heat dissipation performance is improved by more than 30%, and the transformer life is extended.

[0019] Secondly, by employing a segmented potting and curing process, and precisely designing the filler formulations of the two colloids, a highly efficient synergistic performance was achieved. The first flexible colloid is potted first; its low viscosity ensures it can fully penetrate and fill all the minute gaps between the magnetic core and windings. Subsequently, it is pre-cured to form a gel state; its low elastic modulus and matching coefficient of thermal expansion effectively absorb and buffer internal stress generated during temperature changes, protecting internal components. Based on this, a second colloid with high thermal conductivity is potted. The composite alumina and nano-aluminum nitride in its filler system construct a denser and more efficient thermally conductive network, ensuring excellent thermal conductivity of the overall structure; while the higher elastic modulus provides the potting compound with the necessary structural strength and rigidity. This "flexible inside, rigid outside" material combination and process jointly solves the industry challenge of balancing high thermal conductivity and low stress, ultimately enabling the transformer to achieve excellent heat dissipation while ensuring long-term mechanical reliability and structural integrity.

[0020] Third, the metal heat-conducting pillars in this invention establish a direct thermal path from the heating area of ​​the magnetic core to the surface heat dissipation structure. Utilizing the high thermal conductivity of copper or aluminum alloy, heat is rapidly conducted, reducing hot spot temperatures. Their ends are wrapped with a cured colloid to enhance thermal coupling and prevent detachment. This design improves heat dissipation efficiency by approximately 50%, making it particularly suitable for high-power transformers. It reduces core saturation or winding aging caused by localized overheating, improving overall reliability and power density.

[0021] Fourth, this invention achieves dual beneficial effects by setting raised venting ribs in the mold cavity and integrally forming complementary recessed air guiding grooves on the surface of the finished product. First, during vacuum potting, these tiny raised venting ribs guide the flow of the liquid colloid, disrupting its smooth advance. This helps to effectively "sweep" microscopic bubbles trapped inside the colloid or attached to the insert surface towards the pre-set venting path, ultimately being extracted by the vacuum system. This reduces bubble defects inside the cured colloid, improving the density and insulation reliability of the potting. Second, the recessed air guiding grooves formed after demolding are precisely designed to form a small acute angle with the forced cooling airflow direction. This effectively reduces the resistance of airflow along the heat dissipation surface during transformer operation, guiding cold air more smoothly and evenly through the groove array, avoiding eddies or dead zones, thereby improving the efficiency of convective heat transfer and further optimizing heat dissipation performance.

[0022] Fifth, this invention addresses the insulation treatment of different metal materials. Anodizing and plasma spraying processes form a dense, strongly adherent insulation layer with controllable thickness, preventing electrical short circuits and breakdowns. Sandblasting roughening and cleaning pretreatment enhance the durability of the insulation layer, ensuring it does not peel off under temperature changes and mechanical vibration. This improves the safety of the metal heat-conducting pillars under high-voltage environments, enabling the transformer to meet international insulation standards and expanding its application scenarios.

[0023] Sixth, the positioning grooves and elastic limiting blocks work together to ensure precise fixation of the insert within the mold. Horizontal initial positioning and vertical pre-tightening force prevent displacement during potting, improving product dimensional accuracy. The elastic limiting blocks are heat-resistant, preventing softening during curing, and the pre-tightening force range of 3-15 N ensures insert stability without damage. This reduces potting defect rates, improves production efficiency and product consistency, and is suitable for automated manufacturing. Releasing the pre-tightening force after the first colloid pre-curing allows for fine-tuning of the insert in its gel state, releasing internal stress and preventing deformation due to colloid shrinkage. It also provides space for the second colloid potting, ensuring the colloid fully encapsulates the insert and reducing the risk of delamination. This improves potting integrity and interfacial bonding strength, enabling the transformer to maintain structural stability during temperature cycling and extending its service life.

[0024] Other advantages, objectives and features of the present invention will become apparent in part from the following description, and in part from those skilled in the art through study and practice of the invention. Attached Figure Description

[0025] Figure 1 This is a schematic flowchart illustrating the manufacturing method of the high heat dissipation efficiency potted drive transformer of the present invention. Detailed Implementation

[0026] The present invention will now be described in further detail so that those skilled in the art can implement it based on the description.

[0027] It should be understood that terms such as “having,” “comprising,” and “including” as used herein do not exclude the presence or addition of one or more other elements or combinations thereof.

[0028] It should be noted that the experimental methods described in the following embodiments are conventional methods unless otherwise specified. In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "setting" should be interpreted broadly. For example, they can refer to fixed connection or setting, detachable connection or setting, or integral connection or setting. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. The terms "lateral," "longitudinal," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this invention and simplifying the description. They do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention.

[0029] like Figure 1 As shown, the present invention provides a method for manufacturing a high heat dissipation efficiency potted drive transformer, comprising the following steps: S1. Constructing a three-dimensional model: Based on the physical structure of the core and windings of the drive transformer, a three-dimensional digital model of the package is constructed; the surface of the three-dimensional digital model includes a heat dissipation enhancement structure composed of a combination of a boss array and a groove array; S2. Mold preparation: A molding mold with a cavity surface that is complementary to the shape of the three-dimensional digital model is prepared by CNC machining or additive manufacturing method; S3. Positioning and Mold Closure: The assembled magnetic core and winding are positioned and fixed inside the cavity of the forming mold as inserts, and then the mold is closed. S4. Vacuum potting and curing: Under vacuum conditions, liquid thermally conductive insulating adhesive is poured into the cavity of the closed molding mold until the insert is completely submerged; then, according to the material properties of the thermally conductive insulating adhesive, a heating and curing process is carried out to cure the thermally conductive insulating adhesive inside the cavity. S5. Demolding: After the thermally conductive insulating adhesive has completely cured, open the mold and take out the integrated potted transformer with the enhanced heat dissipation structure on the surface that is consistent with the three-dimensional digital model.

[0030] In the above embodiments, firstly, in the step of constructing the three-dimensional model, the method creates a three-dimensional digital model of the package based on the actual physical structure of the internal magnetic core and windings of the drive transformer. This three-dimensional digital model is not a traditional smooth surface, but rather features an enhanced heat dissipation structure on its outer surface, composed of an array of bosses and an array of grooves. These bosses and grooves are not randomly arranged, but rather serve as key functional units to increase the effective heat dissipation surface area and guide the cooling airflow path. For example, bosses can be designed as cylinders or frustums, arranged regularly with specific heights and spacing, while grooves serve as connecting channels, and their cross-sectional shape can be designed as trapezoidal or arc-shaped to optimize flow characteristics. This three-dimensional digital model directly determines the final product's shape and heat dissipation performance, and is the foundation for all subsequent manufacturing activities.

[0031] The next step is mold preparation, which involves creating a molding die using high-precision CNC machining or additive manufacturing technology. The surface shape of the mold cavity is completely complementary to the aforementioned three-dimensional digital model. This means that the recesses in the mold cavity correspond to the bosses on the future transformer surface, while the protrusions correspond to the grooves. This precision mold is crucial to ensuring that the enhanced heat dissipation structure can be accurately and consistently replicated onto the potting compound. The selection of the molding die material, such as tool steel or high-performance aluminum alloy, must consider its strength, thermal conductivity, and dimensional stability at the potting compound curing temperature. Specifically, preferred materials for the molding die include: SKD61 (4Cr5MoSiV1) hot work die steel suitable for mass production and high wear resistance requirements, and 6061 or 7075 high-strength aluminum alloys suitable for rapid prototyping, small to medium batch production, and higher thermal conductivity requirements, to ensure that the mold has good dimensional stability, wear resistance, and heat dissipation efficiency during the potting compound curing thermal cycle.

[0032] In the positioning and mold closing steps, the pre-assembled magnetic core and winding are precisely positioned and fixed inside the cavity of the prepared molding die as a single insert. Positioning methods may include using protrusions on the outer contour of the magnetic core to engage with positioning grooves on the mold cavity for initial horizontal positioning; alternatively, elastic limiting blocks made of high-temperature resistant silicone rubber or spring-loaded ejector pins may be used to apply a moderate vertical preload force, for example, in the range of a few Newtons to tens of Newtons, to stably constrain the insert in the preset position and prevent displacement due to the impact of the adhesive flow during subsequent potting. Ensuring accurate insert positioning is a prerequisite for ensuring uniform potting adhesive thickness and optimized heat conduction paths. The mold is then closed to prepare for vacuum potting.

[0033] The subsequent vacuum potting and curing step is carried out in a vacuum environment, maintaining a high vacuum level, for example, between 1 Pa and 0.1 Pa. This process condition can be achieved using a vacuum system consisting of a mechanical vacuum pump and a Roots pump connected in series, ensuring that the required vacuum environment is reached and maintained within a reasonable evacuation time, thereby effectively eliminating air from the cavity and insert gaps. Under vacuum, liquid thermally conductive insulating adhesive is poured into the closed mold cavity until the adhesive completely immerses the magnetic core and winding inserts. The vacuum environment effectively eliminates air from the cavity and the tiny gaps in the inserts, minimizing air bubbles within the cured adhesive and ensuring the continuity of the thermal conductivity and the reliability of the insulation. After potting, a controlled temperature-curing process is performed based on the material properties of the selected thermally conductive insulating adhesive. This process causes the liquid adhesive to chemically react within the mold cavity, transforming it into a robust solid, integrating the magnetic core, windings, and enhanced heat dissipation structure into a complete, integrated component. The curing process, such as the heating rate and holding temperature and time, needs to be precisely controlled to ensure that the colloid is completely cured and achieves the expected mechanical and thermal conductivity properties.

[0034] Finally, the demolding process is completed. After the thermally conductive insulating adhesive has fully cured and possesses sufficient structural strength, the mold is opened, and the integrated potted transformer is carefully removed. At this point, the outer surface of the transformer's package has accurately replicated the shape of the mold cavity, forming a reinforced heat dissipation structure with an array of bosses and grooves that is completely consistent with the initial three-dimensional digital model.

[0035] Compared to existing transformers that rely on external heat dissipation fins or simple geometric potting, this method significantly improves the heat dissipation capacity of the transformer. It greatly increases the contact area with the cooling medium (usually air) through a single, complex heat dissipation surface and optimizes airflow organization, thereby significantly enhancing convective heat transfer efficiency. Simultaneously, the precise vacuum potting process ensures the absence of internal defects such as air bubbles, forming a uniform and efficient internal heat conduction path. This method effectively overcomes the reliance on mechanical fan components in traditional air-cooled heat dissipation, reducing noise, power consumption, and dust accumulation problems. It also avoids the risks of flow filling difficulties or thermal stress mismatch caused by simply increasing the thermal conductivity of the potting compound. Ultimately, this method achieves a significant improvement in transformer heat dissipation performance without significantly increasing the product size, which is crucial for improving the long-term operational reliability and service life of high-power-density power electronic equipment.

[0036] In one specific embodiment, in step S4, the liquid thermally conductive insulating adhesive is poured in two stages and cured in stages: S410. Inject the first colloid, the viscosity of which is 5000-15000 cP (measured at 25°C), so that it can fully fill the tiny gap between the magnetic core and the winding in a vacuum environment; then, perform the first stage of pre-curing at 40-60°C for 30-60 minutes, so that the first colloid reaches the gel state and its Shore A hardness reaches 30-50. S420. Inject the second colloid, the thermal conductivity of which is not less than 1.5 W / m·K at 25°C; carry out the second stage of heating and curing, heating from 60°C to 100-120°C at a rate of 1-3°C / min, and holding at this temperature for 60-120 minutes to completely cure the second colloid, with a Shore D hardness of not less than 70. The elastic modulus of the first colloid is less than 3 MPa, the elastic modulus of the second colloid is greater than 5 GPa, and the coefficient of thermal expansion of the first colloid is 15-25 ppm / ℃ in the range of 25-100℃, which is between the coefficient of thermal expansion of the magnetic core and the coefficient of thermal expansion of the winding, thus coordinating with the thermal expansion behavior of the assembly composed of the magnetic core and the winding.

[0037] In the above embodiment, during the first stage of infusion and curing, a liquid thermally conductive insulating material, referred to as the first colloid, is first infused. The first colloid is specifically designed to have a relatively low viscosity, with values ​​ranging from several thousand to over ten thousand centipoises, for example, between 5000 cP and 15000 cP. This low viscosity, combined with a vacuum environment, allows the colloid to penetrate and fill all the tiny and complex gaps between the magnetic core and windings, ensuring the integrity of the internal thermal conductivity pathways and preventing the formation of insulating bubbles due to incomplete filling. After infusion, the first stage of pre-curing is performed, typically at a moderate temperature of several tens of degrees Celsius, for example, controlled within the range of 40°C to 60°C, for an appropriate duration, such as 30 to 60 minutes. The goal of this stage is not to completely solidify the colloid, but to transform it into a gel state, somewhere between a liquid and a solid. The first colloid in its gel state has a hardness measured using a scale called Shore A, which can reach a moderate range, for example, 30 to 50. The coefficient of thermal expansion of the first colloid is 15-25 ppm / ℃ in the range of 25-100℃. Given that the magnetic core is typically made of ferrite material with a coefficient of thermal expansion of approximately 5-10 ppm / ℃, and the winding is typically made of copper wire with a coefficient of thermal expansion of approximately 17 ppm / ℃, the coefficient of thermal expansion of the first colloid is designed to be between that of the magnetic core and the winding. This coordinates with the thermal expansion behavior of the assembly of the magnetic core and winding, effectively buffering internal stress caused by temperature changes. This design allows the flexible inner layer formed after the first colloid cures to not only effectively fill gaps but also, through its moderate thermal expansion behavior, coordinate the expansion differences between the magnetic core and the winding, absorbing and releasing internal stress caused by the mismatch in thermal expansion of different materials. Thus, it acts like a protective pad, protecting the delicate magnetic core and fragile winding insulation from mechanical damage.

[0038] In the second stage of infusion and curing, another liquid thermally conductive and insulating material, known as the second colloid, is infused. The core characteristic of this second colloid is its excellent thermal conductivity; it has a thermal conductivity of not less than 1.5 W / m·K at 25°C (measured according to ASTM D5470). After the first colloid has formed a stable gel state, the second infusion is performed, followed by the initiation of the second-stage curing procedure. This curing process typically proceeds at a gradual and controlled rate, for example, increasing the temperature by 1 to 3 degrees Celsius per minute, from an initial temperature to a higher final curing temperature, which may be selected between 100 and 120 degrees Celsius, and held at this temperature for a sufficient duration, such as 60 to 120 minutes, to ensure complete reaction and curing of the second colloid. The fully cured second colloid exhibits high hardness and rigidity, with a hardness typically not less than 70 on the Shore D scale, which is more suitable for rigid plastics, and an elongation at break of not less than 2%. This combination of mechanical properties ensures sufficient structural rigidity on the exterior of the potting compound while imparting a certain degree of toughness to resist external mechanical impacts or internal stress concentrations, preventing brittle cracking. Simultaneously, it also possesses a high elastic modulus, for example, greater than 5 GPa. This high modulus and high hardness enable the second colloid to form a robust outer shell of the transformer encapsulation after curing, providing the necessary structural strength and rigidity support for the entire component and ensuring its morphological stability under mechanical vibration or impact.

[0039] It should be noted that no additional interface treatment or cooling is required between the two infusions. Because the first colloid (flexible silicone) and the second colloid (high thermal conductivity epoxy resin) are chemically compatible by design, and the first colloid still retains a certain degree of reactivity and viscosity in its gel state, the second colloid can be directly infused onto it. The two colloids can form a strong interpenetrating and chemical bond at the contact interface, thereby ensuring interlayer bonding strength.

[0040] By employing a step-by-step process—first injecting a flexible colloid to fill the interior and buffer stress, then injecting a rigid, highly thermally conductive colloid to construct the outer, robust shell and main heat sink—two types of colloids form a functionally distinct layered structure within the transformer. The flexible inner layer protects the core components and ensures defect-free interface contact, while the rigid outer layer efficiently dissipates heat from the inner layer and provides mechanical protection. This ingenious combination of "flexible inside, rigid outside" synergistically solves the long-standing challenge of balancing high thermal conductivity and low internal stress in the manufacturing of high-power-density transformers, resulting in a final product that achieves both efficient heat dissipation and excellent long-term mechanical reliability and structural integrity.

[0041] Compared to traditional potting processes using a single colloid, this method, through a functionally layered material system, achieves a synergistic enhancement of heat dissipation and structural reliability without excessively sacrificing any single performance characteristic. It effectively mitigates the potential risks of interface separation or component stress cracking caused by thermal expansion mismatch between the potting material and internal components, while ensuring efficient heat dissipation from the heat source to the external enhanced heat dissipation structure. This process enables the manufactured transformers to better withstand harsh temperature cycling environments, significantly improving their overall service life and long-term operational stability.

[0042] In one specific embodiment, the first colloid is a flexible organosilicon thermally conductive adhesive with spherical alumina and fumed silica as fillers, wherein the spherical alumina accounts for 60%-70% of the filler mass and the fumed silica accounts for 1%-3% of the filler mass; the second colloid is a high thermally conductive epoxy resin adhesive with compound alumina and nano-aluminum nitride as fillers, wherein the compound alumina accounts for 60%-70% of the filler mass and the nano-aluminum nitride accounts for 2%-4% of the filler mass.

[0043] In the above embodiments, the first colloid is based on flexible organosilicon, a polymer with good elasticity and temperature resistance. To impart thermal conductivity while maintaining low elastic modulus and high filler content, spherical alumina is selected as the main thermally conductive filler. The regular particle shape and smooth surface of spherical alumina effectively reduce the viscosity increase of the colloid when filled in a high proportion, forming a more compact packing structure and thus establishing thermal conductivity pathways. Its filler mass percentage is typically chosen at a relatively high level, for example, between 60% and 70%. In addition, a small amount of fumed silica is added as a thixotropic agent and reinforcing agent, with its mass percentage likely controlled within the range of 1% to 3%. The nanoscale particles of fumed silica can form a three-dimensional network structure in the colloid, slightly increasing viscosity at low shear rates to prevent filler sedimentation and improving the mechanical strength of the cured colloid. However, its addition amount is carefully controlled to avoid excessive thickening that could impair the colloid's ability to penetrate fine gaps. The synergistic effect of these two fillers enables the first colloid to achieve low-viscosity infusion, excellent gap-filling ability, and post-curing flexibility and stress buffering function while possessing the necessary thermal conductivity. Specifically, feasible solutions for flexible silicone thermally conductive adhesives include: using vinyl silicone oil with a viscosity range of 3000-5000 cP as the base polymer (e.g., Shin-Etsu Chemical's KE-103 series or equivalent), combined with hydrogen-containing silicone oil as a crosslinking agent (e.g., Shin-Etsu Chemical's KF-99 series or equivalent), and employing a platinum catalyst system. The spherical alumina filler can be a product with a D50 particle size distribution in the range of 10-30 μm (e.g., Nippon Denko's AS-10 series or equivalent). A first colloid that meets the aforementioned viscosity, thermal conductivity and mechanical properties requirements can be prepared by fully dispersing and mixing 60%-70% by mass of the spherical alumina with 1%-3% by mass of fumed silica (e.g., Evonik's Aerosil 200 or equivalent) in the organosilicon matrix.

[0044] For the second colloid, the matrix material is epoxy resin, which provides high mechanical strength and hardness after curing. To construct an efficient thermal conductivity network, its main filler is a composite alumina. Here, "composite" typically refers to combining alumina particles with different particle size distributions, allowing smaller particles to fill the gaps between larger particles, achieving maximum packing density of the filler system and significantly improving thermal conductivity. Its filler mass percentage is typically as high as 60% to 70%. Furthermore, nano-aluminum nitride is introduced as a high-performance thermal conductivity enhancer. Nano-aluminum nitride itself has extremely high intrinsic thermal conductivity, and its mass percentage can be between 2% and 4%. The main function of nano-aluminum nitride is that its nanoscale particles can effectively fill the gaps between micron-sized composite alumina particles, reducing interfacial thermal resistance and forming localized thermal conductivity "shortcuts" in the filler network, thereby synergistically improving the efficiency and density of the overall thermal conductivity network. This micro-nano composite filler system ensures that the second colloid achieves high thermal conductivity while maintaining the inherent high modulus and high hardness characteristics of the epoxy resin system. Specifically, the feasible solutions for the high thermal conductivity epoxy resin adhesive include: using bisphenol A type epoxy resin (e.g., Nan Ya's NPEL-128 or equivalent) and anhydride curing agents (e.g., methyltetrahydrophthalic anhydride) as the matrix. The "composite alumina" refers to the use of two or more alumina fillers with different particle size distributions to form a denser packing, thereby improving thermal conductivity. A typical compounding scheme is: compounding fine-particle alumina (e.g., Nippon Denko's AS-1 series or equivalent) with a D50 particle size of approximately 1 μm and coarse-particle alumina (e.g., Nippon Denko's AS-50 series or equivalent) at a mass ratio of (20%-30%):(40%-50%), so that the total filler mass of the composite alumina reaches 60%-70%. Based on this, 2%-4% by mass of nano-aluminum nitride (e.g., Nippon Denko's E series or equivalent) is added as a thermal conductivity enhancer. Through this composite filler system, a second colloid that meets the aforementioned requirements for high thermal conductivity and high modulus can be prepared.

[0045] The first colloid, with its special filler selection and ratio, focuses on the perfect filling of internal gaps and effective stress absorption; the second colloid, through its optimized composite filler system, focuses on building a robust and highly thermally conductive outer shell. The two colloids, each in its precisely designed position, work together to achieve an optimal balance of thermal conductivity, mechanical properties, and reliability across the entire potted transformer at both the microscopic and macroscopic scales.

[0046] This implementation method ensures the successful implementation of a functional layered potting strategy from a material perspective through the precise design of the micro-filler system. Compared to potting compounds using conventional or single-filler formulations, this targeted composite filler technology can more effectively harmonize and enhance the overall performance of the colloid. It improves the thermal conductivity and post-curing flexibility of the first colloid while maintaining excellent flowability and permeability; simultaneously, it enhances the thermal conductivity of the second colloid to a high level, while fully maintaining its structural rigidity. This material-level optimization ultimately translates into further enhanced heat dissipation performance and substantial improvement in long-term operational reliability of transformer products, effectively balancing the stringent requirements of heat dissipation and mechanical stability under high power density designs.

[0047] In one specific embodiment, step S3 specifically includes the following sub-steps: S310, Initial positioning of inserts: The assembled magnetic core and windings are initially placed in the cavity of the forming mold as inserts, so that they are initially aligned with the positioning structure on the surface of the cavity; S320. Installing Metal Heat-Conducting Pillars: Within the cavity, install metal heat-conducting pillars on the insert that has been initially positioned; use high thermal conductivity insulating tape or thermally conductive adhesive to fix one end of the metal heat-conducting pillar to the surface of the heating area of ​​the magnetic core to form surface contact; or, use a binding strap made of non-magnetic metal strip to tightly bind one end of the metal heat-conducting pillar to the outer surface of the heating area of ​​the magnetic core. S330, Mold Closure and Final Fixing: After confirming that the other end of the metal heat-conducting pillar has correctly extended into the space reserved for the root of the boss array in the cavity and that there is no structural interference with the mold, close the mold and finally and stably constrain the insert in the preset position by the elastic limiting block. The metal heat-conducting pillar is made of copper or aluminum alloy, and its extension length ensures that the end is completely wrapped by the cured colloid at the root of the boss array.

[0048] In the above embodiment, a step-by-step operation method is adopted in the positioning and mold closing step S3 to ensure the installation accuracy of the metal heat-conducting pillar and avoid interference with the mold: First, the insert is initially positioned: the operator places the pre-assembled magnetic core and winding as a whole insert into the cavity of the prepared molding mold, so that the protruding part on its outer contour is initially aligned and fitted with the positioning groove on the cavity surface. The purpose of this stage is to establish an accurate spatial reference for subsequent operations.

[0049] Next, within the cavity, the metal heat-conducting pillars of the initially positioned insert are installed. This installation operation uses the cavity surface and the insert itself as a reference to ensure the spatial positioning accuracy of the heat-conducting pillars. One end of the metal heat-conducting pillar is fixed to the surface of the heating area of ​​the magnetic core using high thermal conductivity insulating tape or thermal adhesive to form surface contact; alternatively, a binding strap made of non-magnetic metal strip is used to tightly bind one end of the metal heat-conducting pillar to the outer surface of the heating area of ​​the magnetic core. During this process, simple positioning jigs (such as plastic blocks or metal sheets with limiting grooves) adapted to the cavity contour or insert shape can be used to temporarily support and align the metal heat-conducting pillars, ensuring that the other end extends accurately into the space reserved for the root of the boss array. This "on-site" assembly method fundamentally eliminates the risk of the insert being unable to be placed in the mold or interference due to pre-installed heat-conducting pillars. The metal heat-conducting pillars are made of copper or aluminum alloy, and their specific design parameters are as follows: the diameter of the heat-conducting pillars is preferably 2 mm to 5 mm. A diameter that is too small will result in high thermal resistance, affecting heat conduction efficiency; a diameter that is too large may increase local stress in the potting compound and occupy too much space. The arrangement density (or number) of the metal heat-conducting pillars is determined based on the size of the magnetic core and power loss. A reference design rule is to arrange them according to the equivalent diameter of the magnetic core (defined as the diameter of a circle with the same cross-sectional area as the magnetic core). For typical magnetic cores with an equivalent diameter in the range of 50 mm to 80 mm, 2 to 4 pillars are usually arranged. More specifically, the number can be determined according to the principle of arranging one heat-conducting pillar for every 50-80 cm² of effective heat dissipation surface area of ​​the magnetic core (referring to the sum of the areas of the two main surfaces in contact with the heat-conducting pillar). These metal heat-conducting pillars should preferably be arranged in the middle of the long side of the magnetic core or in high heat density areas determined by thermal simulation. Its length direction is perpendicular to the heat dissipation surface, and its extension length must ensure that the end is completely wrapped by the cured adhesive at the root of the boss array, and the minimum insulating adhesive wrapping thickness is not less than 0.5 mm to ensure sufficient electrical insulation strength.

[0050] Finally, the mold is closed and secured: After confirming that all metal heat-conducting pillars are in place and do not interfere with the mold wall, ejector pins, or other structures, the mold is completely closed. Then, a moderate vertical preload is applied to the non-critical surfaces of the windings or core using elastic limiting blocks (such as high-temperature silicone rubber or spring ejector pins) located within the cavity. This secures the insert and the installed heat-conducting pillars as a single unit, stably restraining them in the preset position, preparing for subsequent vacuum encapsulation.

[0051] The spatial arrangement of the metal heat-conducting pillars within the mold cavity is crucial. One end needs to be precisely guided and extended to a specific location within the mold cavity, corresponding to the root region of the boss array in the surface-enhanced heat dissipation structure of the potting compound after curing. The design must ensure that the metal heat-conducting pillar has sufficient extension length so that its end can be completely encapsulated and embedded in the subsequently injected and cured thermally conductive insulating colloid. This encapsulation forms a thick insulating encapsulation layer composed of cured colloid, ensuring both electrical insulation safety and efficient thermal coupling between the metal pillar and the external potting compound. This design makes the metal heat-conducting pillar act like a "thermal bridge," with one end tightly connected to the internal heat source and the other end embedded in the base of the external heat dissipation structure where heat exchange with the cooling medium will occur.

[0052] Through the aforementioned installation and layout, the metal heat-conducting pillars function as directional heat-conducting channels embedded within the potting compound. During transformer operation, the heat generated by the magnetic core is rapidly captured by the metal heat-conducting pillars in close contact with it. Because the thermal conductivity of the metal material is far superior to that of the surrounding potting compound, the heat is preferentially and efficiently conducted longitudinally along this low-resistance path, directly transported to the raised array area on the potting compound surface. This not only significantly reduces the overall thermal resistance from the heat source to the heat dissipation surface, but more importantly, it bypasses potentially poorly conductive paths within the magnetic core and windings, achieving a "short circuit" for heat flow, thereby effectively preventing or significantly reducing the temperature of localized hot spots.

[0053] Compared to traditional solutions that rely entirely on the potting compound for three-dimensional heat diffusion, the added metal heat-conducting pillars provide a clear and efficient heat flow priority path, allowing heat generated by the core heat-generating components to be guided more quickly and centrally to the external enhanced heat dissipation surface. This design significantly improves the heat dissipation efficiency inside the transformer, especially in the core area, and has a significant positive effect on reducing operating temperature and balancing the internal temperature field distribution. It significantly enhances the transformer's ability to handle high-power loads and its overall thermal reliability, making it particularly suitable for high-power density applications with strict requirements on size and weight, and significant challenges in heat dissipation.

[0054] In one specific embodiment, in step S2, when preparing the molding mold, a raised venting rib with a semi-elliptical or trapezoidal cross-section is formed on the raised structure corresponding to the groove array of the three-dimensional digital model on the surface of its cavity. The height of the raised venting rib is 0.1-0.3mm, the width is 0.5-1mm, and the direction of its top extension is consistent with the direction of the groove array. Correspondingly, after demolding in step S5, the inner surface of the groove array of the heat-reducing structure of the potting transformer, corresponding to the position of the raised exhaust rib, is integrally formed with a concave air guide groove with a complementary cross-sectional shape; the extension direction of the concave air guide groove is configured such that when the drive transformer is assembled in the system heat dissipation duct at a preset installation angle, the direction of the air guide groove forms an acute angle of 0-15 degrees with the mainstream direction of the forced cooling airflow.

[0055] In the above embodiments, during the mold preparation stage, when the molding die is prepared by CNC machining or additive manufacturing, a fine machining process is performed on its cavity surface. Specifically, a fine feature called raised venting ribs is additionally machined on the mold protrusions corresponding to the groove array on the surface of the future encapsulated transformer. These venting ribs can be designed with a streamlined cross-section such as a semi-ellipse or trapezoid, and their dimensions are very small; for example, the height may be controlled between 0.1 mm and 0.3 mm, and the width between 0.5 mm and 1.0 mm. The top extension direction of these venting ribs is set to be consistent with the orientation of the entire groove array on the mold. These raised venting ribs are arranged parallel to each other on the mold cavity surface at intervals of 5 mm to 15 mm, forming multiple continuous venting paths. The ends of all venting ribs extend to the parting surface of the mold and are connected to the external vacuum system of the mold through micro-venting grooves machined on the parting surface, thereby ensuring that gas can be effectively guided and discharged during the encapsulation process. During vacuum potting, when liquid thermally conductive insulating adhesive is poured into a cavity containing a magnetic core and winding inserts, the adhesive front flows along the complex cavity surface. These tiny raised venting ribs moderately disrupt the smooth flow of the adhesive, and the resulting micro-disturbance helps to "sweep away" tiny air bubbles trapped in the adhesive or attached to the insert surface, guiding them to converge and move along the predetermined rib direction, and finally being effectively removed by the mold's vacuum venting system. This significantly reduces the risk of uneven thermal conductivity and decreased insulation strength caused by residual air bubbles inside the potting compound.

[0056] After the potting compound cures and is demolded, these raised venting ribs on the mold cavity surface leave a complementary imprint on the reinforced heat dissipation structure surface of the potted transformer. Specifically, continuous, recessed air guide grooves with cross-sectional shapes complementary to the venting ribs are integrally formed on the inner surface of the groove array. These air guide grooves are a natural product of the manufacturing process, but their design is given a clear functional orientation. Their extension direction is specifically configured: when the drive transformer is mounted in the system's heat dissipation duct at a preset installation angle, the direction of the air guide grooves is typically at an acute angle of 0-15 degrees to the mainstream direction of the forced cooling airflow. The "preset installation angle" is defined as the angle between the main heat dissipation plane of the drive transformer potting body (i.e., the surface with the bosses and groove array) and the system mounting base surface (such as the PCB board or heat sink mounting surface), which is typically designed to be 30 to 45 degrees. This angle is set to achieve a balance between system spatial layout and heat dissipation efficiency, ensuring that the forced cooling airflow can efficiently scour the main heat dissipation surface. In this installation configuration, the air guide grooves should be oriented approximately parallel to the airflow direction to minimize flow resistance and achieve optimal airflow guidance. This angle range is the optimal choice based on typical duct designs. In practical applications, the preset installation angle can be adaptively adjusted according to the specific structure of the system's cooling duct to maintain its excellent airflow guidance effect. The design intent of this angle is that when the cooling airflow (usually forced airflow from a fan) flows over the array of grooves on the transformer surface, these tiny air guide grooves can effectively guide the airflow, making it flow more smoothly and concentratedly along the extension direction of the grooves.

[0057] This implementation achieves a dual technical effect, simultaneously optimizing manufacturing process quality and the heat dissipation performance of the final product. At the manufacturing level, it significantly improves the venting efficiency during vacuum potting through a simple and easily implemented method, contributing to a denser potting compound with fewer defects, thereby enhancing the product's insulation reliability and thermal uniformity. At the operational heat dissipation level, it cleverly transforms the structural features formed during manufacturing into a powerful tool for improving convective heat transfer efficiency. By optimizing airflow organization, it reduces flow resistance and potential eddy current dead zones, allowing for more thorough and efficient heat exchange between the cooling medium and the heat dissipation surface. This design brings additional gains to the overall heat dissipation performance of the transformer without adding extra components or complex processes.

[0058] In one specific embodiment, the step of insulating the metal heat-conducting pillar includes: The required thickness of the insulation layer for the metal heat-conducting pillar is determined based on the operating voltage level of the drive transformer. When the operating voltage of the drive transformer is ≤100V, if the metal heat-conducting column is made of aluminum alloy, an anodizing process is used to generate a dense anodized aluminum insulating layer with a thickness of 10-30μm on its surface; if it is made of copper, a plasma spraying process is used to prepare an alumina ceramic insulating layer with a thickness of 50-150μm on its surface.

[0059] When the operating voltage of the drive transformer is 100V to 1000V, if the metal heat-conducting column is made of aluminum alloy, the thickness of its anodized aluminum insulation layer is 30-60μm; if it is made of copper, the thickness of its plasma-sprayed alumina ceramic insulation layer is 150-300μm.

[0060] Before anodizing or plasma spraying, the surface of the metal heat-conducting pillar to be treated is roughened by sandblasting and cleaned with organic solvent. Furthermore, the metal heat-conducting pillars after insulation treatment must pass the power frequency withstand voltage test, with a test voltage of not less than twice the operating voltage of the drive transformer plus 1000V, and must withstand the test for 1 minute without breakdown or flashover.

[0061] In the above embodiments, the thickness of the insulation layer is determined according to the design operating voltage of the drive transformer to ensure electrical safety. When the metal heat-conducting pillar is made of aluminum alloy, the insulation treatment method used is anodizing. For transformers with an operating voltage not exceeding 100V, the thickness of the anodized aluminum insulation layer is typically controlled between 10 micrometers and 30 micrometers; for applications with an operating voltage between 100V and 1000V, its thickness should be increased to 30 micrometers to 60 micrometers to provide sufficient insulation strength. For another common material, copper, an alumina ceramic insulation layer is prepared using a plasma spraying process. Correspondingly, at an operating voltage ≤100V, the coating thickness is controlled between 50 micrometers and 150 micrometers; at an operating voltage between 100V and 1000V, the coating thickness needs to be increased to 150 micrometers to 300 micrometers.

[0062] To ensure a strong bond between the insulating layer and the metal substrate and prevent peeling during subsequent processing or use, the surface of the metal heat-conducting pillar must undergo rigorous pretreatment before anodizing or plasma spraying. This pretreatment typically includes two key steps: sandblasting and organic solvent cleaning. Sandblasting uses high-speed, hard particles to impact the metal surface, creating microscopic roughness. This significantly increases the mechanical interlocking area between the insulating layer and the substrate, providing a solid "anchoring" foundation for the coating. Subsequently, thorough cleaning with a suitable organic solvent removes any residual oil, fingerprints, dust, or other contaminants, ensuring the insulating layer can grow or adhere on a clean, active surface, thereby achieving optimal interfacial bonding quality.

[0063] As a critical step in quality control, all insulated metal heat-conducting pillars must undergo a power frequency withstand voltage test before assembly. The test voltage must be no less than twice the operating voltage of the drive transformer plus 1000V, and must be maintained for one minute without breakdown or flashover. This test should be conducted under standard atmospheric test conditions, i.e., ambient temperature (25±5)℃ and relative humidity ≤60%RH. The electrode arrangement is as follows: the high-voltage electrode is tightly connected to the metal substrate of the metal heat-conducting pillar (this can be achieved by reserving an uninsulated test point at the end of the heat-conducting pillar), while the grounding electrode is tightly attached to the outer surface of the potting compound with a radius of 20mm centered on the heat-conducting pillar (or directly connected to the zero potential terminal of the winding) using metal foil or conductive paste, to ensure that the test voltage is accurately applied to the insulation layer under test.

[0064] Through the aforementioned targeted insulation treatment, the metal heat-conducting column achieves a clever functional unity in its physical properties. Its metal core retains high thermal conductivity, responsible for efficient longitudinal heat transfer; while its surface insulation layer forms a reliable electrical barrier, ensuring that even at high potential differences, the heat-conducting column will not form a conductive path with the surrounding windings or magnetic core (the latter typically connected via ground). This allows the metal heat-conducting column to safely traverse the complex electrical environment inside the transformer, guiding heat from high-potential heating areas to the grounded or neutral external heat dissipation casing.

[0065] In one specific embodiment, in step S1, the boss array and groove array in the enhanced heat dissipation structure satisfy the following structural relationship: the boss array is a plurality of equally spaced cylindrical or frustum structures with a height of 1.5–3.0 mm, a cross-sectional diameter of 0.8–2.0 mm, and a center-to-center distance of 3–6 mm between adjacent bosses; the groove array is a channel connecting adjacent bosses, with a trapezoidal or arc-shaped cross-section, a depth of 1.0–2.5 mm, and a width of 1.5–3.0 mm; the bosses and grooves form a continuously alternating heat dissipation unit on the surface of the package, used to increase the heat dissipation surface area and guide the cooling airflow along a preset path.

[0066] In the above embodiments, the array of bosses constitutes the main protruding feature of the heat dissipation surface in the design of the enhanced heat dissipation structure. These bosses can be designed in various regular geometric shapes, such as cylinders with uniform cross-sections or frustum structures with a certain slope. Their dimensional parameters are limited to a specific range; for example, the height of the boss may be set between 1.5 mm and 3.0 mm, while the characteristic dimensions of their cross-sections, such as the diameter of the cylinder or the diagonal length of the top surface of the frustum, may be selected within the range of 0.8 mm to 2.0 mm. These bosses are not randomly arranged, but are regularly distributed in a equidistant manner, and the distance between the center points of adjacent bosses, i.e., the center-to-center distance, is controlled within a moderate range of 3 mm to 6 mm. The primary purpose of this boss layout with a certain height and density is to maximize the actual contact surface area between the package and the surrounding cooling medium (usually air) within a limited projected area.

[0067] Complementing the boss array is the groove array, which exists as negative space channels connecting adjacent bosses. The cross-sectional shape of the grooves can be designed as a trapezoid for ease of manufacturing and airflow guidance, or an arc shape for lower flow resistance. Their dimensions also require precise control; for example, the groove depth may be designed between 1.0 mm and 2.5 mm, while the opening width at its top may range from 1.5 mm to 3.0 mm. The depth and width of the grooves together determine the cross-sectional area of ​​the channel, affecting the flow capacity of the cooling medium and the contact area with the groove walls. These grooves do not exist in isolation; they form a continuous network between the bosses, dividing the entire heat dissipation surface into numerous regular units.

[0068] Ultimately, the bosses and grooves form a continuous, alternating arrangement of heat dissipation units with a specific geometry on the surface of the package. This periodic arrangement serves a dual function. Under static or natural convection conditions, it enhances radiative and natural convection heat dissipation by significantly increasing the surface area. In scenarios with forced air cooling, this regularly arranged structure effectively guides the cooling airflow along a predetermined path—usually parallel to the direction of the grooves—avoiding unnecessary eddies and dead zones, thereby achieving more efficient utilization of the cooling medium and improving the efficiency of convective heat transfer.

[0069] In one specific embodiment, the positioning and fixing method in step S3 specifically includes: The cavity surface of the molding die is provided with one or more positioning grooves that are adapted to the protruding parts in the outer contour of the magnetic core. The corresponding protruding parts of the magnetic core are embedded in the positioning grooves to achieve initial positioning in the horizontal direction. Meanwhile, a preload force in the vertical direction is applied to the non-critical surfaces of the winding or magnetic core by an elastic limiting block set in the cavity. The elastic limiting block is made of high-temperature resistant silicone rubber or spring pins, and the preload force provided is in the range of 3-15 N, thereby stably constraining the insert in a preset position in the cavity.

[0070] In step S4, the pre-tightening force applied by the elastic limiting block is released after the first colloid has completed pre-curing and reached the gel state, but before the second colloid is poured in; after the pre-tightening force is released, the elastic limiting block disengages from the surface of the insert, and the cavity area corresponding to it becomes the pouring space of the second colloid.

[0071] In the above embodiments, a constraint strategy combining horizontal and vertical directions is adopted in the initial positioning and fixing stage. In the horizontal direction, the cavity surface of the molding die is provided with one or more positioning grooves that precisely complement the shape of naturally existing or specially designed protrusions (such as fixing ears, reinforcing ribs, etc.) on the outer contour of the magnetic core. When the magnetic core is placed into the cavity, these protrusions will be embedded in the corresponding positioning grooves, like a key being inserted into a lock cylinder, thereby achieving the initial positioning of the insert in various directions in the horizontal plane and effectively preventing its movement and rotation in the plane. At the same time, a vertical preload is applied to the structural support surface of the insert by an elastic limiting block set in the cavity; the structural support surface refers to the area on the magnetic core or winding that has high mechanical strength and whose local stress will not cause the magnetic properties of the magnetic core to deteriorate, the electrical performance of the winding to decrease, or the insulation to be damaged; for example, the central column surface of the magnetic core, the middle of the coil side of the winding, or the solid frame part of the magnetic core; areas to be avoided include, but are not limited to: the thin leads of the winding, the fragile air gap area of ​​the magnetic core, and the surface of the covered insulation layer. The elastic limiting block is made of high-temperature resistant silicone rubber or spring pins, and provides a preload force ranging from 3 to 15 N. This preload force is based on an estimate of the buoyancy and flow impact forces experienced by the insert in the potting compound, thereby stably constraining the insert to a preset position within the cavity. The preload force range is set based on this estimate. Specifically, the preload force F should satisfy: F > ρgV + kAν, where ρ is the colloid density, g is the acceleration due to gravity, V is the projected area of ​​the insert facing the flow direction of the colloid, A is the area of ​​the insert facing the flow direction, ν is the colloid injection velocity (typically ranging from 0.5 to 2.0 cm / s), and k is the dynamic impact coefficient (dimensionless, empirically taken as 1.5-2.5, used to characterize the instability and impact effect of the colloid flow). For typical drive transformer inserts with an equivalent diameter ranging from 50 mm to 80 mm, the required preload force range is determined to be 3 N to 15 N based on the above formula and verification through process experiments. This preload force can stably constrain the insert in a predetermined position within the cavity, preventing displacement or floating during potting due to buoyancy and flow impact of the colloid, while also avoiding deformation or damage to the insert (such as the magnetic core or windings) due to excessive preload force.

[0072] After the first colloid is poured and pre-cured to a gel state, the entire potting system enters a critical transition phase. At this point, the first colloid has formed a flexible gel network with certain shape retention and cohesive strength, which initially wraps and fixes the magnetic core and windings. At this node, the vertical preload previously applied by the elastic limiting block is actively released. This operation can be achieved by remotely controlling the spring pin to retract, or by the design of the silicone rubber block allowing stress relaxation under continuous high temperature. After the preload is released, the elastic limiting block disengages from the insert surface, and the cavity area previously occupied by the limiting block becomes empty, forming a new, unfilled cavity space. It should be noted that in step S4, the preload applied by the elastic limiting block is released after the first colloid has pre-cured to a gel state and before the second colloid is poured. At this point, the first colloid has completely enveloped the insert and filled the surrounding cavity space. Due to its gel-like cohesive force and adhesion to the insert and cavity walls, it can independently and stably fix the insert in the preset position without relying on external mechanical constraints. After the pre-tightening force is released, the elastic limiting block disengages from the insert surface, and the corresponding cavity area becomes the injection space for the second colloid.

[0073] This seemingly simple "constraint removal" operation brings several positive effects. First, it allows the insert encapsulated by the first colloidal gel to subtly self-adjust in response to the slight deformation trends of itself and the gel under temperature changes after the external mechanical constraints are removed. This helps release and redistribute the residual stress accumulated within the system during the pre-curing stage, reducing the risk of cracking or deformation of the final product due to excessive internal stress. Second, and more importantly, these newly vacated cavity areas provide crucial injection space for the subsequent infusion of the second colloidal gel. This allows the second colloidal gel to fully encapsulate the initially fixed insert from multiple directions, especially filling areas previously blocked by the limiting blocks. This ensures a large-area, defect-free interfacial bond between the second and first colloidal gels, greatly reducing the possibility of decreased thermal conductivity and deterioration of mechanical strength due to incomplete encapsulation or interfacial delamination.

[0074] This implementation significantly improves the manufacturing precision, structural integrity, and long-term reliability of encapsulated transformers. Through phased intelligent constraint management, this method effectively resolves the contradiction between "precise positioning" and "stress release" in the encapsulation process. In the initial stage of encapsulation, it ensures absolute accuracy in insert placement, thereby optimizing insulation distance and heat conduction paths. In the middle stage of encapsulation, it cleverly releases constraints, both releasing internal stress and creating conditions for perfect subsequent material filling. Compared to the traditional method of rigid fixation from beginning to end, this dynamic constraint strategy can more effectively reduce encapsulation defects and improve the interfacial bonding quality between different adhesive layers. This results in integrated encapsulated transformers with lower internal stress, more uniform structure, and stronger interfacial bonding, which is of great significance for ensuring long-term stable operation of the product under harsh temperature cycling conditions.

[0075] The number of devices and processing scale described herein are for the purpose of simplifying the description of the invention. Applications, modifications, and variations of the invention will be readily apparent to those skilled in the art.

[0076] Although the embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. They can be applied to various fields suitable for the present invention. For those skilled in the art, other modifications can be easily made. Therefore, without departing from the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details.

Claims

1. A method for manufacturing a high-heat-dissipation-efficiency potted drive transformer, characterized in that, Includes the following steps: S1. Constructing a 3D Model: Based on the physical structure of the core and windings of the drive transformer, a 3D digital model of the package is constructed. The surface of the 3D digital model includes a heat dissipation enhancement structure composed of a combination of a boss array and a groove array. The boss array consists of multiple equally spaced cylindrical or frustum structures with a height of 1.5-3.0 mm, a cross-sectional diameter of 0.8-2.0 mm, and a center-to-center distance of 3-6 mm between adjacent bosses. The groove array consists of channels connecting adjacent bosses, with a trapezoidal or arc-shaped cross-section, a depth of 1.0-2.5 mm, and a width of 1.5-3.0 mm. S2. Molding mold preparation: Prepare a molding mold that is complementary to the shape of the three-dimensional digital model by CNC machining or additive manufacturing; On the protruding structure corresponding to the groove array on the surface of the mold cavity, process to form a protruding venting rib with a cross-section of semi-elliptical or trapezoidal, with a height of 0.1-0.3mm and a width of 0.5-1mm, and the direction is consistent with the groove array. S3. Positioning and Mold Closure: Position the magnetic core and winding as inserts within the mold cavity; install insulated metal heat-conducting pillars on the surface of the heating area of ​​the magnetic core, with the other end of the metal heat-conducting pillars extending to the reserved space at the root of the boss array in the cavity; fix the inserts using positioning grooves and elastic limiting blocks; S4. Vacuum potting and segmented curing: Under a vacuum environment of 1 Pa to 0.1 Pa, the thermally conductive insulating adhesive is potted in two stages and cured in segments. S410, First colloid for injection: viscosity 5000-15000 cP, pre-cured at 40-60℃ for 30-60 minutes to gel state, Shore A hardness 30-50; elastic modulus less than 3MPa, coefficient of thermal expansion at 25-100℃ 15-25ppm / ℃, between the magnetic core and winding. S420, second colloid: thermal conductivity at 25℃ not less than 1.5W / m・K, heated to 100-120℃ at 1-3℃ / min and held for 60-120 minutes for complete curing, Shore D hardness not less than 70; elastic modulus greater than 5GPa; Among them, after the first colloid is pre-cured to the gel state and before the second colloid is injected, the pre-tightening force of the elastic limiting block is released; S5. Demolding: A one-piece encapsulated transformer with a surface-enhanced heat dissipation structure is obtained; the inner surface of its groove array is integrally formed with a concave air guide groove that complements the raised exhaust ribs, and the direction of the air guide groove is at an acute angle of 0-15 degrees to the mainstream direction of the forced cooling airflow.

2. The method according to claim 1, characterized in that, The first colloid is a flexible silicone thermally conductive adhesive, containing 60%-70% spherical alumina and 1%-3% fumed silica by mass; the second colloid is a high thermal conductivity epoxy resin adhesive, containing 60%-70% compounded alumina and 2%-4% nano-aluminum nitride by mass.

3. The method according to claim 1, characterized in that, Step S3 includes: S310, Insert initial positioning: The protruding part of the magnetic core is inserted into the positioning groove of the mold to achieve horizontal positioning; S320. Install metal heat-conducting pillars: Use high thermal conductivity insulating tape, thermally conductive adhesive, or non-magnetic metal tape to fix the metal heat-conducting pillars to the heating area of ​​the magnetic core. S330, the mold is closed and finally fixed by the elastic limit block, and the end of the metal heat-conducting pillar is completely wrapped by the colloid at the root of the boss.

4. The method according to claim 1, characterized in that, The metal heat-conducting pillars are made of copper or aluminum alloy and are insulated according to the operating voltage. Working voltage ≤100V: 10-30μm aluminum alloy anodized layer, 50-150μm copper plasma sprayed aluminum oxide layer; Operating voltage 100V to 1000V: Aluminum alloy anodized layer 30-60μm, copper plasma sprayed aluminum oxide layer 150-300μm; Before insulation treatment, the material is roughened by sandblasting and cleaned with organic solvents, and then passes a power frequency withstand voltage test: the voltage is not less than twice the working voltage plus 1000V, and it lasts for 1 minute without breakdown or flashover.

5. The method according to claim 1, characterized in that, The elastic limiting block is made of high-temperature resistant silicone rubber or spring pin, with a preload of 3-15N. After the preload is released, the limiting block disengages from the insert, and the corresponding area becomes the second colloid injection space.

Images