Machine learning assisted design of a preparation process of high-performance copper alloy pipe

Abstract

The application discloses a kind of machine learning auxiliary design high-performance copper alloy pipe preparation process, belong to copper iron series alloy material preparation technical field.The application is optimized copper alloy composition by machine learning, and is combined with continuous casting, process such as deformation-heat treatment synergic control, solve high iron content Cu-Fe alloy solidification segregation, organization uneven and performance control problem, and the copper alloy pipe prepared has high strength, high thermal conductivity and corrosion resistance, and size precision, surface quality is excellent, it is suitable for CO2 Refrigerant and other new refrigeration equipment field.The application process controllability is strong, industrialization adaptability is high, significantly improve the comprehensive performance and production stability of copper alloy pipe.

Description

Technical Field

[0001] This invention belongs to the field of copper-based alloy material preparation technology, specifically involving a machine learning-aided design process for preparing high-performance copper alloy pipes. It is particularly suitable for the composition optimization and industrial preparation of high-iron-content Cu-Fe alloys, and can be applied to high-end manufacturing fields such as new refrigeration and heat exchange equipment. Background Technology

[0002] Copper and copper alloys, due to their excellent thermal conductivity, corrosion resistance, and processability, are core structural materials in the fields of refrigeration and heat exchange equipment. Among them, Cu-Fe alloys, through solid solution strengthening and precipitation strengthening of Fe, can significantly improve strength while maintaining high thermal conductivity, becoming a new type of low-cost, high-strength copper alloy to replace traditional copper and brass. They are particularly suitable for the high-pressure operating conditions of new environmentally friendly refrigeration technologies such as CO2 refrigerants. Typical high-iron content Cu-Fe alloys combine high strength, high thermal conductivity, and corrosion resistance, and have broad application prospects in heat exchange pipelines and high-pressure transmission pipelines of refrigeration equipment.

[0003] With the accelerated localization of high-end refrigeration equipment in my country, the performance requirements for copper alloy tubing are becoming increasingly stringent: high tensile strength and high yield strength are needed to meet the demands of high-pressure operating conditions, while also ensuring high thermal conductivity and corrosion resistance to improve heat exchange efficiency and service life. Strict requirements are also placed on the dimensional accuracy (outer diameter tolerance ±0.15mm, wall thickness tolerance ±10%) and surface quality (roughness Ra≤0.8μm). However, the preparation and industrial application of copper alloys still face many technical bottlenecks.

[0004] Firstly, the solidification process of high-iron-content Cu-Fe alloys is prone to liquid-liquid phase separation and Fe dendrite segregation, resulting in uneven ingot composition distribution, coarse microstructure, and localized enrichment of Fe elements forming coarse dendrites. This not only reduces the alloy's plasticity and machinability but also causes fluctuations in the subsequent pipe performance, making it difficult to meet the requirements of stable industrial production. Existing research mostly alleviates segregation by adjusting single process parameters such as casting temperature and cooling rate, but has not developed a systematic solution from the perspectives of crystallizer structure and process synergy control, thus the segregation problem remains unresolved.

[0005] Secondly, the design of copper alloy compositions relies on the traditional "trial and error" method, which involves screening the alloying element ratios through numerous orthogonal experiments. This approach is not only time-consuming and costly, but also makes it difficult to accurately grasp the interactive effects of elements such as Fe, Ni, and P on solidification behavior, precipitates, and multiple performance indicators. Different elements exhibit complex synergistic or antagonistic effects. For example, Ni can refine Fe precipitates, but excessive amounts can reduce thermal conductivity; P can improve corrosion resistance, but excessive amounts can embrittle grain boundaries. Traditional experimental methods cannot efficiently screen for the optimal balance point of multiple performance indicators, making it difficult to achieve a synergistic improvement in alloy performance, including strength, thermal conductivity, and corrosion resistance.

[0006] Third, the preparation of copper alloy pipes involves multiple coupled processes such as smelting, casting, deformation processing, and heat treatment. The control mechanism of each process parameter on the microstructure and final properties is still unclear. There is a lack of systematic research on the grain breakage and recrystallization behavior during deformation and the precipitation of strengthening phases and grain boundary migration during heat treatment. This easily leads to problems such as uneven pipe wall thickness, large roundness deviation, and many surface defects. At the same time, the uneven microstructure results in significant differences in the cross-sectional properties of the pipes, which cannot meet the application requirements of high-precision and high-performance pipes.

[0007] Fourth, existing manufacturing processes have not achieved deep integration of machine learning-aided design and industrial production, lacking a closed-loop control system from composition prediction and process optimization to performance testing. The application of technologies such as machine learning and data mining in alloy design is still in its early stages, and mature multi-objective optimization and interpretable analysis tools have not yet been developed. Performance testing data cannot be quickly fed back to the composition and process optimization stages, resulting in low R&D efficiency, slow product iteration, and difficulty in meeting the demand for customized high-performance pipes in high-end equipment.

[0008] To address the aforementioned issues, scholars both domestically and internationally have conducted extensive research. Some studies have improved casting microstructure and alleviated Fe segregation through external field techniques such as electromagnetic stirring and ultrasonic treatment; however, these external field equipment is costly and lacks industrial applicability. Other studies have refined the microstructure by adding trace alloying elements, but the synergistic mechanism between elements remains unclear, potentially introducing new performance defects. Still other studies have controlled precipitates through heat treatment, but these have not been combined with deformation processes to achieve synergistic optimization of microstructure and properties. Overall, existing technologies have not yet formed a complete, industrially applicable scheme for preparing high-performance copper alloy tubing, failing to simultaneously address core issues such as composition optimization, solidification segregation, microstructure control, and synergistic performance improvement, thus hindering the large-scale application of copper alloys in high-end refrigeration equipment.

[0009] Therefore, developing a high-performance copper alloy tube manufacturing process based on machine learning-aided design, achieving precise composition optimization through machine learning, and combining continuous casting and deformation-heat treatment synergistic control technologies to solve solidification segregation and microstructure inhomogeneity problems, and constructing a closed-loop control system from design to production, has become a key technical challenge that urgently needs to be solved in the field of copper-iron alloy materials. This is of great significance for promoting the localization of high-end refrigeration equipment in my country. Summary of the Invention

[0010] To address the problems of low composition design efficiency, severe solidification segregation, difficulty in microstructure control, insufficient synergistic performance improvement, and poor industrial adaptability in existing copper alloy preparation technologies, this invention provides a machine learning-aided design process for preparing high-performance copper alloy tubing, aiming to:

[0011] 1. Based on machine learning, multi-objective optimization of copper alloy composition is achieved, the interaction law of elements on performance is clarified, and the optimal balance ratio of strength, thermal conductivity and corrosion resistance is efficiently screened;

[0012] 2. Develop new continuous casting technology to solve the problems of Fe dendrite segregation and uneven microstructure during ingot solidification, and prepare copper alloy ingots with uniform composition and fine microstructure.

[0013] 3. Construct a deformation-heat treatment synergistic control technology to precisely control the dimensional accuracy and surface quality of the pipe, achieve uniform control of microstructure and properties, and improve the stability of the pipe cross-section performance;

[0014] 4. Establish a closed-loop system for machine learning-aided design and industrial production, feeding performance testing data back to model iteration and optimization to improve R&D efficiency and product consistency;

[0015] 5. To produce high-performance copper alloy tubing that meets the requirements of high-end refrigeration equipment, achieving a synergistic improvement in high strength, high thermal conductivity, corrosion resistance, high precision, and high surface quality.

[0016] To achieve the above objectives, this invention provides a machine learning-aided design process for manufacturing high-performance copper alloy tubing, comprising the following steps:

[0017] (1) Optimization design of copper alloy composition based on machine learning: Collect data on composition, solidification behavior, precipitates and properties of Cu-Fe alloys, construct a multi-objective prediction model, and optimize the proportion range of Cu, Fe, Ni, P and Nd elements through multi-objective Bayesian optimization and non-dominated genetic algorithm. The proportions are Fe 4.5%-6.5%, Ni 0.2%-1.0%, P 0.003%-0.006%, Nd 0.01%-0.15% by mass fraction, with the balance being Cu and unavoidable impurities. The total amount of impurities does not exceed 0.5%. SHAP analysis is used to clarify the marginal contribution of each element to strength, corrosion resistance and thermal conductivity.

[0018] (2) Raw material pretreatment: Select high-purity Cu, Fe, Ni, P and Nd according to the optimized composition ratio, remove the surface oxide layer and impurities, and dry them under vacuum for later use;

[0019] (3) Continuous casting to prepare copper alloy ingots: The pretreated raw materials are placed in a vacuum induction furnace and melted into alloy liquid at a set temperature of 1200-1350℃ and under a protective atmosphere. By optimizing the structure of the continuous casting crystallizer, the cooling rate is 50-200℃ / s and the billet pulling speed is 0.5-1.2m / min, the solidification structure and Fe element distribution are controlled to prepare copper alloy ingots with uniform composition and structure.

[0020] (4) Homogenization treatment of ingots: The ingots obtained by continuous casting are subjected to high-temperature homogenization annealing with a heating rate of 5-15℃ / min, heated to 850-950℃, held for 4-12h, cooled to below 300℃ in the furnace and then air-cooled to eliminate residual stress and compositional segregation in the ingots and refine the Fe dendrite structure.

[0021] (5) Deformation-heat treatment synergistic control: The homogenized ingot is subjected to alternating hot and cold processing, and after cold rolling, intermediate annealing is carried out at a temperature of 400-600℃ and a holding time of 1-3h. The uniformity of pipe wall thickness, inner and outer diameter tolerance and roundness are precisely controlled. Then, by setting the heating process, cooling rate and holding time, the grain nucleation and growth, the size distribution of strengthening phase and the grain boundary migration behavior are controlled.

[0022] (6) Finishing and surface treatment: The deformed and heat-treated pipes are straightened and cut. Mechanical polishing and chemical passivation are combined to control the roughness of the inner and outer surfaces and eliminate scratches and cracks. The chemical passivation uses an environmentally friendly composite chemical passivation solution. The environmentally friendly composite chemical passivation solution, in terms of mass fraction, includes the following raw materials: citric acid 6%-14%, benzotriazole 0.2%-0.4%, phytic acid 1.5%-3%, sodium molybdate 0.8%-1.5%, sodium silicate 0.5%-0.6%, sodium gluconate 2%-4%, monoethanolamine 0.3%-0.8%, ascorbic acid 0.1%-0.3%, fatty alcohol polyoxyethylene ether 0.05%-0.15%, deionized water: balance;

[0023] (7) Performance testing and iterative optimization: The mechanical properties, corrosion resistance, thermal conductivity and dimensional accuracy of the finished pipes are tested, and the test data are fed back to the machine learning model to iteratively optimize the composition and process parameters, so as to achieve precise control of the pipe performance.

[0024] Furthermore, the machine learning model in step (1) includes random forest, gradient boosting tree and neural network. The input features include alloy element content, solidification cooling rate and heat treatment temperature. The output targets are tensile strength, yield strength, elongation, thermal conductivity and corrosion rate.

[0025] Furthermore, in step (1), the ratio of Cu, Fe, Ni, P and Nd elements is optimized. In terms of mass fraction, Fe is 6.0%, Ni is 0.6%, P is 0.005%, Nd is 0.1%, the balance is Cu, and the total amount of impurities is 0.3%.

[0026] Furthermore, in step (3), the crystallizer vibration frequency is 20-80Hz.

[0027] Furthermore, the protective atmosphere described in step (3) is a mixture of argon and nitrogen, with the oxygen content controlled below 10 ppm.

[0028] Furthermore, the heat treatment process in step (5) is as follows: heating in sections to 500-750℃ and holding for 2-8 hours.

[0029] Furthermore, in step (5), water cooling, air cooling or furnace cooling is used, with a cooling rate of 10-100℃ / min.

[0030] Furthermore, in step (6), mechanical polishing is performed using diamond abrasive for step-by-step polishing, and the roughness Ra value is controlled below 0.8 μm.

[0031] Furthermore, in step (6), the passivation temperature for chemical passivation is 25-50℃, and the passivation time is 5-30min.

[0032] Furthermore, the chemical passivation uses an environmentally friendly composite chemical passivation solution, which, by mass fraction, comprises the following raw materials: citric acid 9%, benzotriazole 0.3%, phytic acid 2.3%, sodium molybdate 1.2%, sodium silicate 0.5%, sodium gluconate 3%, monoethanolamine 0.6%, ascorbic acid 0.2%, fatty alcohol polyoxyethylene ether 0.1%, and deionized water: balance.

[0033] Compared with the prior art, the technical advantages of the present invention are as follows:

[0034] Advantage 1: Achieve high efficiency and precision in ingredient design, overcoming the limitations of traditional trial-and-error methods.

[0035] Traditional copper alloy composition design relies on extensive orthogonal experiments and trial-and-error verification, resulting in long development cycles, high costs, and an inability to accurately grasp the interaction patterns between multiple elements, easily leading to performance issues where some aspects are prioritized at the expense of others. This invention utilizes machine learning to construct a multi-objective prediction model, rapidly predicting alloy properties under different compositions and processes through a data-driven approach. This eliminates the need for numerous repetitive experiments, significantly shortening the development cycle and reducing costs. Simultaneously, the SHAP interpretability analysis tool is introduced to quantify the marginal contributions of each element to strength, thermal conductivity, and corrosion resistance, clarifying the synergistic or antagonistic mechanisms between elements. For example, it reveals that Ni, while refining Fe precipitates and improving strength, reduces the rate of thermal conductivity decay by regulating its electronic structure; and that P, while inhibiting intergranular corrosion through grain boundary segregation, can cause grain boundary embrittlement if excessive. This allows for precise selection of optimal balanced ratios for multiple performance indicators, enabling targeted alloy composition design and avoiding the blind spots of traditional methods, providing scientific guidance for the development of high-performance copper alloys.

[0036] Advantage 2: It solves the solidification segregation problem at its root and improves the uniformity of the ingot structure.

[0037] Liquid-liquid phase separation and Fe dendrite segregation are prone to occur during the solidification process of high-iron Cu-Fe alloys, which are the core reasons for the uneven ingot microstructure and performance fluctuations. Existing technologies mostly alleviate segregation by simply adjusting the cooling rate or adding trace elements, without forming a systematic solution from the perspective of process synergy control. This invention optimizes the continuous casting crystallizer structure and combines the synergistic matching of cooling rate, vibration frequency and billet pulling speed. On the one hand, rapid cooling inhibits the nucleation and growth of Fe dendrites, refines the solidification microstructure, and shortens the element diffusion distance. On the other hand, crystallizer vibration breaks up primary dendrites, promotes uniform mixing of components, and inhibits liquid-liquid phase separation. At the same time, vacuum protective atmosphere melting and casting are used to avoid alloy liquid oxidation and impurity introduction, fundamentally solving the problems of Fe dendrite segregation and coarse microstructure. The resulting copper alloy ingot has a uniform composition and refined microstructure, laying the foundation for subsequent pipe processing and performance improvement. Compared with existing technologies, the ingot composition uniformity is significantly improved, there are no obvious segregation defects, and the processing formability is greatly improved.

[0038] Advantage 3: Constructing a deformation-heat treatment synergistic control system to achieve precise control of microstructure and properties.

[0039] In existing copper alloy tube manufacturing processes, deformation processing and heat treatment are mostly independent steps, failing to achieve synergistic optimization of microstructure control, which easily leads to problems such as poor dimensional accuracy and uneven cross-sectional properties. This invention constructs a synergistic control system of "alternating hot and cold processing - segmented heat treatment." Hot rolling breaks down the coarse as-cast microstructure, cold rolling refines grains and precisely controls dimensions, and intermediate annealing eliminates work hardening, ensuring tube formability. Segmented heat treatment specifically controls the precipitation behavior and grain boundary migration of the Fe strengthening phase. By matching heating temperature, holding time, and cooling rate, the size of the strengthening phase is controlled at the nanometer to submicrometer level, achieving a balance between strength and plasticity, while simultaneously inhibiting grain growth and ensuring uniform cross-sectional properties. Compared to existing technologies, this invention can precisely control the outer diameter tolerance, wall thickness tolerance, and roundness deviation of the tube, avoiding performance differences caused by microstructure inhomogeneity in traditional processes. It achieves simultaneous optimization of microstructure and macroscopic properties, meeting the demands of high-end equipment for high-precision, high-performance tubes.

[0040] Advantage 4: Achieve synergistic improvement in surface quality and corrosion resistance, extending the service life of pipes.

[0041] Existing surface treatments for copper alloy pipes often employ simple mechanical polishing or pickling, which easily leads to problems such as high surface roughness, numerous scratches, and a weak passivation film, failing to meet the corrosion resistance and surface quality requirements under refrigeration conditions. This invention employs a surface treatment process combining mechanical polishing and chemical passivation. Mechanical polishing, through progressively refined abrasives, effectively eliminates scratches and burrs on the pipe surface, controlling the roughness below 0.8μm and improving surface smoothness. Chemical passivation uses an environmentally friendly composite chemical passivation solution to form a uniform and dense passivation film on the pipe surface. This avoids the environmental pollution problems of traditional chromate passivation and effectively inhibits corrosion reactions in the refrigerant medium, improving corrosion resistance. Compared to existing technologies, the pipe surface produced by this invention has no obvious defects, the passivation film has strong adhesion and excellent corrosion resistance, significantly extending its service life under high pressure and corrosive conditions, while simultaneously meeting the surface quality requirements of heat exchange equipment and improving heat exchange efficiency.

[0042] Advantage 5: Establishing a closed-loop control system to improve the stability and consistency of industrial production.

[0043] Existing copper alloy manufacturing technologies lack deep integration of design, preparation, and testing, and lack data feedback and iterative optimization mechanisms, resulting in large fluctuations in product performance and poor industrial adaptability. This invention constructs a closed-loop control system of "machine learning design - continuous casting - deformation - heat treatment - performance testing - model iteration," feeding performance testing data back to the machine learning model in real time to update the database and re-optimize composition and process parameters, achieving continuous improvement in product performance. Compared to existing technologies, this invention offers strong process controllability, with quantifiable and replicable parameters for each step, effectively avoiding human error and improving the stability and consistency of industrial production. Furthermore, it allows for rapid customization of composition and processes according to different application scenarios, adapting to the personalized needs of high-end equipment and providing technical support for the large-scale production of copper alloy pipes.

[0044] Advantage 6: Balancing high strength and high thermal conductivity, breaking through the performance bottlenecks of traditional copper alloys.

[0045] Traditional high-strength copper alloys (such as beryllium copper and chromium-zirconium copper) are strengthened by adding expensive alloying elements, which is not only costly but also significantly reduces thermal conductivity. While pure copper has excellent thermal conductivity, its low strength cannot meet the requirements of high-pressure applications. This invention utilizes solid solution strengthening and nano-precipitation strengthening with Fe to significantly improve the alloy strength while maintaining the high thermal conductivity of the Cu matrix. Simultaneously, through the synergistic regulation of P, the size and distribution of the strengthening phase are optimized, avoiding excessive obstruction of the thermal conductivity pathway by the strengthening phase. Compared to existing technologies, the copper alloy tubing produced by this invention combines high strength (tensile strength > 650 MPa) and high thermal conductivity while maintaining good corrosion resistance, achieving a synergistic improvement of "high strength-high thermal conductivity-high corrosion resistance." This overcomes the bottleneck of traditional copper alloys where "increased strength leads to decreased thermal conductivity," providing an ideal structural-functional integrated material for new refrigeration equipment. Attached Figure Description

[0046] Figure 1 This is a process flow diagram of the preparation process of high-performance copper alloy tubing designed with machine learning assistance according to the present invention;

[0047] Figure 2 This is a diagram of the machine learning model architecture for optimizing copper alloy composition;

[0048] Figure 3 It is a closed-loop system diagram of performance testing data and model iterative optimization;

[0049] Figure 4 This is a process flow diagram for preparing the environmentally friendly composite chemical passivation solution of the present invention;

[0050] Figure 5 This is a metallographic microstructure of the copper alloy tubing prepared in Example 1;

[0051] Figure 6 This is a secondary electron (SE) morphology image of the copper alloy tubing obtained in Example 1, obtained by scanning electron microscopy (SEM).

[0052] Figure 7 This is a metallographic microstructure of the copper alloy tube prepared in Example 1 after etching for 20 seconds.

[0053] Figure 8 The image shows the secondary electron (SE) morphology of the copper alloy tubing prepared in Example 1 after immersion in 3.5 wt.% NaCl solution for 7 days.

[0054] Figure 9 The image shows the secondary electron (SE) morphology of the copper alloy tubing prepared in Example 1 after immersion in 3.5 wt.% NaCl solution for 15 days.

[0055] Figure 10The image shows the secondary electron (SE) morphology of the copper alloy tubing prepared in Example 1 after immersion in 3.5 wt.% NaCl solution for 15 days using a scanning electron microscope (SEM).

[0056] Figure 11 This is a scanning electron microscope image of the uniform and dense microstructure of the copper alloy tubing prepared in Example 1 after immersion in a 3.5 wt.% NaCl solution for 30 days.

[0057] Figure 12 This is a Nyquist comparison of the electrochemical impedance spectroscopy (EIS) of the copper alloy tubing prepared in Example 1 and Comparative Example 1 before immersion. Detailed Implementation

[0058] In embodiments of the present invention, such as Figure 1-3 As shown, a machine learning-aided design process for manufacturing high-performance copper alloy tubing includes the following steps:

[0059] Step 1: Machine Learning-Based Copper Alloy Composition Optimization Design

[0060] We collected publicly available literature, enterprise production data, and preliminary experimental data to construct a database covering Cu-Fe alloy composition (Cu, Fe, Ni, P, Nd element content), preparation process parameters (solidification cooling rate, heat treatment temperature, etc.), microstructure characteristics (precipitate size, grain size, etc.), and performance indicators (tensile strength, yield strength, thermal conductivity, corrosion rate, etc.). Based on this database, we constructed a multi-objective machine learning prediction model, including random forest, gradient boosting tree, and neural network models, using alloy element content and process parameters as input features, and strength, thermal conductivity, and corrosion resistance as output objectives.

[0061] A multi-objective Bayesian optimization algorithm and a non-dominated genetic algorithm were used to solve the model for multi-objective optimization, and the alloy element ratio range that meets the multiple performance constraints of "high strength, high thermal conductivity, and high corrosion resistance" was selected. The SHAP (SHapley Additive exPlanations) interpretability analysis tool was introduced to quantify the marginal contribution of each alloying element and its interactions to the performance indicators, clarifying key influencing factors and their mechanisms of action. For example, the contribution mechanism of Fe to strength, the synergistic effect of Ni on thermal conductivity and corrosion resistance, and the regulatory effect of P on grain boundary corrosion resistance were identified. Finally, the optimal chemical composition range for the copper alloy was determined as follows: Fe 4.5%-6.5%, Ni 0.2%-1.0%, P 0.003%-0.006%, Nd 0.01%-0.15%, with the balance being Cu and unavoidable impurities, with the total impurity content not exceeding 0.5%.

[0062] Step 2: Raw material pretreatment

[0063] According to the optimized composition ratio, electrolytic copper, high-purity iron, nickel powder, phosphorus copper master alloy, and copper-neodymium master alloy with a purity ≥99.95% were selected. A combination of mechanical grinding and pickling was used to remove the oxide layer, oil, and impurities from the raw material surface. The pickling solution was a 5%-10% (w / w) dilute sulfuric acid solution, at a temperature of 25-40℃ for 5-15 minutes. After pickling, the mixture was rinsed with deionized water until neutral and then placed in a vacuum drying oven at 80-120℃ for 2-4 hours to remove moisture and residual acid, preventing the introduction of impurities that could affect the alloy's properties.

[0064] Step 3: Continuous casting to prepare copper alloy ingots

[0065] The pretreated raw materials are loaded into a vacuum induction furnace and evacuated to a vacuum level of 1×10⁻⁶. -3 Below Pa, a protective gas mixture of argon and nitrogen (volume ratio 1:1) is introduced, and the temperature is raised to 1200-1350℃ for melting. After the raw materials are completely melted, the temperature is held for 10-30 minutes to ensure that the alloy liquid has a uniform composition.

[0066] An optimized continuous casting crystallizer (with built-in spiral cooling water channel and vibration device) is used to control the pouring temperature at 1200-1350℃, the cooling rate at 50-200℃ / s, the crystallizer vibration frequency at 20-80Hz, and the casting speed at 0.5-1.2m / min. Through the coordinated control of cooling rate and vibration parameters, liquid-liquid phase separation and Fe dendrite nucleation and growth are suppressed, the solidification structure is refined, and the Fe element distribution is uniform. Protective gas is continuously introduced during the casting process to prevent oxidation of the alloy liquid. The final product is a copper alloy ingot with a diameter of 80-200mm and a length of 2000-5000mm, with no obvious defects such as segregation, porosity, or inclusions.

[0067] Step 4: Ingot homogenization treatment

[0068] The ingots obtained from continuous casting are placed in a box-type resistance furnace and heated to 850-950℃ at a heating rate of 5-15℃ / min, and held at that temperature for 4-12 hours. High-temperature diffusion eliminates residual stress and compositional segregation within the ingot, dissolves coarse Fe dendrites, and refines the microstructure. After holding, the ingots are cooled to below 300℃ in the furnace, and then removed and air-cooled to room temperature to avoid rapid cooling causing new stress and microstructural defects.

[0069] Step 5: Coordinated regulation of deformation and heat treatment

[0070] The homogenized ingots undergo alternating hot and cold processing: First, hot rolling is performed at an initial rolling temperature of 850-900℃ and a final rolling temperature of 650-700℃, with a total deformation of 55%-65%, rolling the ingots into slabs; then, multiple passes of cold rolling are performed, with a pass deformation of 15%-25% and a total deformation of 65%-85%, achieving grain breakage and microstructure refinement through cold rolling, while precisely controlling the uniformity of pipe wall thickness, inner and outer diameter tolerances, and roundness; during the cold rolling process, intermediate annealing is inserted at an annealing temperature of 400-600℃ and held for 1-3 hours to eliminate work hardening and restore the alloy's plasticity.

[0071] Heat treatment control is performed after cold and hot processing: a segmented heating method is adopted. First, the temperature is raised to 300-400℃ at 10-20℃ / min and held for 1-2 hours to eliminate processing stress; then the temperature is raised to 500-750℃ and held for 2-8 hours to regulate the nucleation and growth of Fe strengthening phase, and to control the size, distribution and interfacial orientation of precipitated phase with the matrix; finally, water cooling, air cooling or furnace cooling is adopted at a cooling rate of 10-100℃ / min to suppress grain boundary migration, retain refined grains and uniform precipitated phase, and achieve uniform control of the cross-sectional properties of the pipe.

[0072] Step 6: Finishing and Surface Treatment

[0073] After deformation and heat treatment, the pipes are straightened using a multi-roller straightener to control the curvature to <2mm and ensure straightness to ≤1mm per meter; the pipes are then cut to the required length, and end defects are removed.

[0074] Surface treatment is carried out by combining mechanical polishing and chemical passivation: mechanical polishing uses diamond abrasive to polish in stages (grit size from 800 mesh to 2000 mesh) to control the roughness Ra of the inner and outer surfaces to ≤0.8μm and eliminate scratches and burrs; chemical passivation uses an environmentally friendly composite chemical passivation liquid with a passivation temperature of 25-50℃ and a passivation time of 5-30min to form a dense passivation film on the pipe surface, improve corrosion resistance, and avoid introducing new surface defects.

[0075] The environmentally friendly composite chemical passivation solution comprises, by mass fraction, the following raw materials: citric acid 6%-14%, benzotriazole 0.2%-0.4%, phytic acid 1.5%-3%, sodium molybdate 0.8%-1.5%, sodium silicate 0.5%-0.6%, sodium gluconate 2%-4%, monoethanolamine 0.3%-0.8%, ascorbic acid 0.1%-0.3%, fatty alcohol polyoxyethylene ether 0.05%-0.15%, and deionized water: balance.

[0076] like Figure 4 As shown, the preparation method of the environmentally friendly composite chemical passivation solution includes the following steps:

[0077] (1) Preheating of basic solvent and predissolution of inorganic components

[0078] Add the measured amount of deionized water to the reactor, turn on the stirring device and heat to 45°C, and stir at a constant temperature; first add the specified amounts of sodium molybdate and sodium silicate, and stir at 300 r / min for 12 min until the inorganic strong base salt is completely dissolved to form a transparent inorganic mother liquor, thus avoiding subsequent violent reaction with organic acids to precipitate.

[0079] (2) Organic complex corrosion inhibitors are compounded and dissolved in steps.

[0080] Keep the solution temperature (45℃) and stirring speed (300r / min) constant, and add citric acid, benzotriazole, phytic acid and sodium gluconate in sequence. Stir for 10 minutes after each addition to ensure that the single component is completely dissolved and has no residue. After all the materials are added, continue stirring for 25 minutes to allow the inorganic corrosion inhibitor and organic complexing agent to fully chelate and form a homogeneous composite inhibitory mother liquor, thereby improving the compatibility of the components.

[0081] (3) Addition of low-temperature additives and precise pH control

[0082] Turn off the heating device and cool the solution to 24°C. Add ascorbic acid and fatty alcohol polyoxyethylene ether and stir for 10 minutes until completely dissolved. Slowly add monoethanolamine dropwise while monitoring the pH with a precision pH meter in real time. Accurately adjust the pH of the solution to 7.0-7.5. After stopping the dropwise addition, continue stirring for 15 minutes to ensure that the pH of the solution is stable without fluctuations or local acid-base imbalance.

[0083] (4) Sealed aging and homogenization treatment

[0084] Maintain a low stirring speed of 200 rpm and place the solution in a sealed, light-proof environment for 40 minutes to completely eliminate air bubbles inside the solution, allowing all components to fully react synergistically, further improving the long-term stability of the passivation solution and preventing problems such as stratification, precipitation, and turbidity during use.

[0085] (5) Precision filtration purification and finished product storage

[0086] After aging, the passivation solution is pressure filtered using a 400-mesh precision filter cloth to remove trace amounts of insoluble impurities and microgel particles. The filtered passivation solution is then placed in a sealed plastic container and stored at room temperature away from light to obtain the finished environmentally friendly composite chemical passivation solution.

[0087] Step 7: Performance Testing and Iterative Optimization

[0088] Comprehensive performance testing is conducted on the finished pipes, and the test data is fed back to the machine learning model to update the database and re-optimize the composition and process parameters, thereby achieving closed-loop control of "design-preparation-testing-optimization" and continuously improving the performance stability and consistency of copper alloy pipes.

[0089] Technical principle of the invention:

[0090] This invention, centered on data-driven machine learning multi-objective optimization, breaks through the bottleneck of traditional "trial-and-error" R&D in copper alloy tubing. Through precise composition design, synergistic raw material blending, gradient control of process parameters, and closed-loop iteration throughout the entire process, it achieves a synergistic breakthrough in the high strength, high thermal conductivity, and high corrosion resistance of Cu-Fe copper alloy tubing. This completely solves the industry pain points of traditional copper alloy tubing, such as the trade-off between strength and thermal conductivity, insufficient corrosion resistance, and poor microstructure uniformity. The entire technical principle revolves around three core elements: "synergistic action of raw materials," "precise process parameters," and "intelligent machine learning," as detailed below:

[0091] I. Mechanism of Action and Synergistic Effects of Core Raw Materials

[0092] The alloy matrix elements, other elements, and passivation liquid components selected in this invention are not simply superimposed with single functions, but rather form synergistic gains through atomic-level interaction, interface regulation, and phase structure optimization, thus laying the foundation for high-performance pipes from the root.

[0093] (I) Copper alloy matrix and the role and synergy of elements

[0094] 1. Core role of a single raw material

[0095] Cu (matrix, balance): As the core matrix of copper alloys, it possesses excellent intrinsic thermal conductivity, plasticity, and formability, which is the basis for ensuring the thermal conductivity of pipes and their cold and hot processing. Controlling the total amount of impurities to ≤0.5% can prevent impurity phases from breaking the matrix and deteriorating thermal conductivity and mechanical properties.

[0096] Fe (4.5%-6.5%): The core strengthening element. During the melting and solidification process, nano-sized Fe intermetallic compound strengthening phases are precipitated. Through second-phase dispersion strengthening and dislocation pinning, the tensile strength and yield strength of the pipe are greatly improved. At the same time, Fe has strong lattice compatibility with Cu matrix, avoiding significant damage to the thermal conductivity of the matrix.

[0097] Ni (0.2%-1.0%): A dual-function synergistic element. On the one hand, it strengthens the Cu matrix through solid solution, helping to improve the strength of the alloy. On the other hand, it optimizes the electronic arrangement structure of the Cu matrix, reduces the thermal resistance at grain boundaries, and takes into account thermal conductivity. At the same time, Ni can passivate the alloy surface, form a dense oxide film, block the penetration of corrosive media, and improve corrosion resistance.

[0098] P (0.003%-0.006%): A grain boundary precisely controlled element. Trace amounts of P accumulate at grain boundaries, filling grain boundary defects and inhibiting grain boundary corrosion channels, thus solving the common problem of easy corrosion at grain boundaries in copper alloys. At the same time, P can purify the melt, remove oxygen impurities, avoid casting porosity and inclusion defects, and improve the density of the microstructure.

[0099] Nd (0.01%-0.15%): a microalloying refining element. Rare earth Nd has extremely strong nitrogen fixation and deoxidation capabilities, which can further purify the alloy melt. At the same time, Nd atoms are adsorbed at grain boundaries and phase interfaces, inhibiting grain growth and Fe dendrite coarsening, achieving dual refinement of solidification structure and precipitated phases, and balancing strength and plasticity.

[0100] 2. Synergistic effect mechanism among elements

[0101] The alloying elements do not act in isolation, but form a four-dimensional synergistic system of "Fe strengthening + Ni stabilization + P grain protection + Nd refinement": Nd refines the size of Fe precipitates and makes them dispersed, avoiding coarse Fe phases from cutting the matrix and reducing thermal conductivity; Ni forms a weak solid solution with Fe, alleviating the thermal loss caused by the Fe phase, and at the same time, it works with P to optimize the grain boundary structure, achieving "strength improvement without sacrificing thermal conductivity and simultaneous reinforcement of corrosion resistance"; P and Nd work together to purify the melt and eliminate harmful impurity phases, further ensuring that the functions of Ni and Fe elements are fully utilized, ultimately breaking the performance paradox of traditional copper alloys that "high strength must be low thermal conductivity and high thermal conductivity must be low strength".

[0102] (II) The role and synergy of components in environmentally friendly composite passivation liquid

[0103] 1. Core role of a single component

[0104] Citric acid: 6%-14%: The main complexing corrosion inhibitor gently etches the oxide layer on the surface of copper alloys, providing active sites for film formation, inhibiting the initial corrosion of the substrate, and retaining the core complexing advantages of the original system; combined with a weakly alkaline pH range, it avoids excessive etching that could damage the substrate.

[0105] Benzotriazole (BTA): 0.2%-0.4%: The main organic film-forming corrosion inhibitor, which forms stable coordination bonds with copper atoms to build a dense framework for the passivation film and improve the core anti-corrosion performance of the film.

[0106] Phytic acid: 1.5%-3%: a natural and environmentally friendly complexing agent, rich in polyhydroxy phosphate groups, which strengthens the bonding force between the film and the substrate, fills the micropores of the film, and improves the overall corrosion resistance.

[0107] Sodium molybdate: 0.8%-1.5%: An inorganic and environmentally friendly corrosion inhibitor that replaces toxic chromates. It forms a uniform oxide passivation layer on the copper surface and, in conjunction with organic components, significantly improves the corrosion resistance potential.

[0108] Sodium silicate: 0.5%-0.6%: Film-forming reinforcing agent, forming a dense inorganic silicon-oxygen network structure, enhancing the hardness and erosion resistance of the passivation film, and adapting to the medium flow conditions of pipes.

[0109] Sodium gluconate: 2%-4%: an auxiliary complexing agent that stabilizes free metal ions in the solution, prevents component aggregation and precipitation, and ensures the uniformity of the passivation film throughout.

[0110] Monoethanolamine: 0.3%-0.8%: pH adjuster and stabilizer, precisely controlling the solution to be nearly weakly alkaline, avoiding excessive acidity that corrodes the matrix and excessive alkalinity that causes the auxiliary agent to fail, thus extending the service life of the solution.

[0111] Ascorbic acid (vitamin C): 0.1%-0.3%: an antioxidant stabilizer that inhibits secondary oxidation on the surface of copper alloys, prevents deterioration of solution components, ensures consistent passivation effect by optimizing feeding conditions, and extends the shelf life of finished products.

[0112] Fatty alcohol polyoxyethylene ether (AEO-9): 0.05%-0.15%: Nonionic wetting agent, reduces the surface tension of the pipe, eliminates dead corners for film formation such as inner walls and edges, and achieves full coverage of the passivation film.

[0113] Deionized water: Balance: To prevent calcium and magnesium ions from interfering with film formation and to ensure solution purity and reaction stability.

[0114] 2. Synergistic effect of passivation fluid components

[0115] Inorganic corrosion inhibitors (sodium molybdate, sodium silicate) and organic complexing agents (citric acid, phytic acid) form an "inorganic-organic composite passivation film" that balances film density and adhesion. The corrosion inhibitors, wetting agents, and stabilizers work synergistically to achieve passivation film coverage without dead angles and without localized corrosion, while avoiding the pollution problems of traditional chromate passivation, thus achieving the dual goals of "improved corrosion resistance + environmental compliance".

[0116] II. Importance and Necessity of Selecting Preparation Process Parameters

[0117] The process parameters of this invention are not empirical values, but precise settings optimized by a machine learning model with multiple objectives and matched with the characteristics of alloy composition. Each parameter is designed to address structural defects and performance shortcomings, which is the core guarantee for realizing the functional application of alloy composition and high-performance forming of pipes.

[0118] (a) Raw material pretreatment parameters: the basic prerequisite for impurity control

[0119] Pickling with 5%-10% dilute sulfuric acid at 25-40℃ for 5-15 minutes, followed by vacuum drying at 80-120℃ for 2-4 hours, is essential to thoroughly remove the oxide layer, oil, and impurities from the raw material surface, preventing impurities from being introduced into the melt and forming inclusions and porosity defects. If the pickling concentration / temperature is insufficient, the impurities will not be thoroughly removed; if the pickling is excessive, it will corrode the substrate and deplete alloying elements. Precise parameter control is the first step in ensuring the purity of the alloy composition and the density of the microstructure.

[0120] (II) Vacuum Continuous Casting Parameters: The Core Key to Homogenization of Microstructure

[0121] Vacuum degree ≤1×10-3 The importance of the following parameters—Pa, argon-nitrogen mixed protective gas, 1200-1350℃ melting and casting, 50-200℃ / s cooling rate, 20-80Hz vibration frequency, and 0.5-1.2m / min casting speed—lies in three aspects: First, the vacuum and protective gas prevent oxidation of the alloy liquid, avoiding the burning of active elements such as Fe and Nd; second, the high cooling rate inhibits Fe dendrite coarsening and liquid-liquid phase separation, achieving uniform Fe distribution; and third, the synergy between the crystallizer vibration and the casting speed eliminates ingot segregation, porosity, and shrinkage defects, ensuring a uniform ingot structure without internal flaws, laying the foundation for subsequent deformation processing. If the cooling rate is too low, the microstructure will be coarse and the strength will decrease; an unbalanced vibration frequency will easily cause cracks, directly leading to the scrapping of the pipe.

[0122] (III) Homogenization parameters: a necessary step to eliminate internal stress

[0123] The process involves heating at 5-15℃ / min, holding at 850-950℃ for 4-12 hours, and then cooling in the furnace to below 300℃ for air cooling. The purpose is to eliminate residual stress and compositional segregation in the ingot through high-temperature atomic diffusion, dissolve coarse Fe dendrites, and avoid stress concentration that could lead to cracking during subsequent hot and cold processing. Slow heating and cooling can prevent thermal stress from damaging the ingot, ensure uniform microstructure, and prepare for deformation to refine the microstructure.

[0124] (iv) Deformation-heat treatment synergistic parameters: the core means of performance regulation

[0125] Hot rolling (initial rolling 850-900℃, final rolling 650-700℃, deformation 55%-65%) and cold rolling (deformation per pass 15%-25%, total deformation 65%-85%) combined with segmented annealing (intermediate annealing at 400-600℃, final heat treatment at 500-750℃) is crucial because: alternating hot and cold deformation achieves grain breakage and microstructure refinement, improving strength and plasticity; intermediate annealing eliminates work hardening and ensures continuous forming; segmented heat treatment precisely controls the size and distribution of Fe precipitates, and the cooling rate (10-100℃ / min) suppresses grain boundary migration, preserving refined grains and dispersed strengthening phases, achieving a precise match between mechanical and thermal properties. Insufficient deformation results in poor microstructure refinement, while an imbalanced annealing temperature leads to coarsening of the strengthening phase and a sharp drop in performance.

[0126] (v) Finishing and passivation parameters: ensuring surface quality and corrosion resistance

[0127] Multi-roller straightening (bending degree <2mm), step-by-step polishing (Ra≤0.8μm), and passivation at 25-50℃ for 5-30min are necessary to eliminate processing deformation defects and ensure the dimensional accuracy of the pipes; mechanical polishing reduces surface roughness and eliminates stress concentration points, while environmentally friendly passivation forms a dense protective film, completely solving the problems of corrosion and scratches on the pipe surface and extending service life.

[0128] (vi) Closed-loop detection iteration parameters: a stable and long-term mechanism

[0129] The performance data of the finished product is fed back to the machine learning model, the database is updated, and the composition and process are iteratively optimized. The importance lies in overcoming uncertainties such as batch fluctuations and environmental interference, realizing a closed loop of "design-preparation-testing-optimization", continuously improving the consistency and stability of pipe performance, and adapting to industrial mass production.

[0130] III. Core Technical Principles of Machine Learning-Assisted Design

[0131] Machine learning is the "intelligent brain" of this invention, completely overturning the traditional copper alloy R&D model. Its core principle consists of four steps:

[0132] 1. Multi-dimensional database construction: Integrating data from all dimensions, including composition, process, organization, and performance, to provide accurate samples for model training and solve the problems of fragmented and irregular traditional R&D data.

[0133] 2. Multi-objective prediction model construction: Random forest, gradient boosting tree and neural network fusion model are adopted to achieve accurate mapping of "component-process-performance" and solve the problem of difficulty in quantifying the coupling relationship of multiple performance.

[0134] 3. Intelligent optimization and interpretability analysis: Bayesian optimization and non-dominated genetic algorithms are used to solve multiple performance-optimal solutions and screen suitable component ranges; SHAP tools are used to quantify the marginal contributions of elements and processes, clarify the performance regulation mechanism, avoid black-box optimization, and ensure the scientific nature of component and process design.

[0135] 4. Closed-loop iterative upgrade: Real-time feedback of detection data and continuous optimization of model parameters allow the technical solution to iterate with production data and continuously improve performance.

[0136] IV. Unexpected Breakthrough Technological Effects

[0137] This invention, through precise control via machine learning and synergistic processing of raw materials, breaks through the performance limits of traditional copper alloy pipes, achieving several unexpected technical effects:

[0138] Key Breakthrough 1: Achieving a Synergistic Balance Between High Strength and High Thermal Conductivity

[0139] Traditional Cu-Fe alloys often suffer from a sharp drop in thermal conductivity after strength improvement due to the coarsening of the Fe phase. This invention, through Nd refinement and high cooling rate control, results in a nanoscale dispersed distribution of the Fe precipitate phase. This significantly improves the tensile strength and yield strength of the pipe compared to traditional processes, while only slightly reducing the thermal conductivity. It perfectly balances structural strength and thermal conductivity, breaking through the long-standing performance bottleneck in the industry.

[0140] Key Breakthrough 2: Significantly Improved Corrosion Resistance, Complete Suppression of Grain Interval Corrosion

[0141] Synergistic grain boundary regulation with trace amounts of P, Ni, and Nd, combined with an environmentally friendly composite passivation film, significantly reduces the corrosion rate of the pipe compared to traditional copper alloys. There are no obvious corrosion traces at the grain boundaries, and the service life under humid, hot, acidic, and alkaline environments is significantly improved, solving the pain points of easy leakage and corrosion failure of copper alloy pipes during long-term service.

[0142] Key Breakthrough 3: Significantly shortened R&D cycle, green and environmentally compliant

[0143] Machine learning replaces traditional trial-and-error R&D, significantly shortening the cycle of component and process optimization; environmentally friendly passivation liquid is chromium-free and non-toxic, replacing traditional toxic passivation processes, while significantly improving raw material utilization, achieving a dual unity of high performance and green manufacturing.

[0144] To make the present invention more fully disclosed, more specific embodiments are described below.

[0145] Example 1

[0146] A machine learning-aided design process for manufacturing high-performance copper alloy tubing includes the following steps:

[0147] Step 1: Machine Learning-Based Copper Alloy Composition Optimization Design

[0148] (1) Construct a database containing 1000 sets of Cu-Fe alloy data. The input features are Fe, Ni, and P content and process parameters, and the output targets are tensile strength, yield strength, thermal conductivity, and corrosion rate. Construct a random forest prediction model, R 2 The value reached 0.92. Using multi-objective Bayesian optimization, the following composition was obtained: Fe 6.0%, Ni 0.6%, P 0.005%, Nd 0.15%, with the balance being Cu, and total impurities 0.3%.

[0149] Step 2: Raw material pretreatment

[0150] According to the optimized composition ratio, electrolytic copper (99.95% purity), high-purity iron, nickel powder, phosphorus copper master alloy, and copper-neodymium master alloy were selected. A combination of mechanical grinding and pickling was used to remove the oxide layer, oil, and impurities from the raw material surface. The pickling solution was an 8% (w / w) dilute sulfuric acid solution, at 32℃ for 10 minutes. After pickling, the mixture was rinsed with deionized water until neutral and placed in a vacuum drying oven at 100℃ for 3 hours to remove moisture and residual acid, preventing the introduction of impurities that could affect the alloy's properties.

[0151] Step 3: Continuous casting to prepare copper alloy ingots

[0152] The pretreated raw materials are loaded into a vacuum induction furnace and evacuated to a vacuum level of 1×10⁻⁶. -3 Pa, introduce a protective gas mixture of argon and nitrogen (volume ratio 1:1), heat to 1300℃ for melting, and hold at that temperature for 20 minutes after the raw materials are completely melted to ensure uniform composition of the alloy liquid.

[0153] An optimized continuous casting crystallizer (with built-in spiral cooling water channel and vibration device) was used to control the pouring temperature at 1280℃, the cooling rate at 100℃ / s, the crystallizer vibration frequency at 50Hz, and the casting speed at 1.0m / min. Through the coordinated control of cooling rate and vibration parameters, liquid-liquid phase separation and Fe dendrite nucleation growth were suppressed, refining the solidification structure and ensuring a uniform Fe element distribution. A protective gas was continuously introduced during the casting process to prevent oxidation of the alloy liquid. The final product was a copper alloy ingot with a diameter of 120mm and a length of 3000mm. The ingot was free of obvious defects such as segregation, porosity, and inclusions, and the oxygen content was controlled below 10ppm.

[0154] Step 4: Ingot homogenization treatment

[0155] The ingots obtained from continuous casting are placed in a box-type resistance furnace and heated to 900°C at a heating rate of 10°C / min. The furnace is held at this temperature for 8 hours to eliminate residual stress and compositional segregation within the ingot through high-temperature diffusion, dissolve coarse Fe dendrites, and refine the microstructure. After holding, the ingots are cooled to 250°C in the furnace and then air-cooled to room temperature to avoid rapid cooling causing new stress and microstructural defects.

[0156] Step 5: Coordinated regulation of deformation and heat treatment

[0157] The homogenized ingots are subjected to alternating hot and cold processing: First, hot rolling is performed with an initial rolling temperature of 850℃ and a final rolling temperature of 650℃, resulting in a total deformation of 60%, rolling the ingots into slabs; then, multiple cold rolling passes are performed with a deformation of 15% per pass and a total deformation of 75%, achieving grain breakage and microstructure refinement through cold rolling, while precisely controlling the uniformity of pipe wall thickness, inner and outer diameter tolerances, and roundness; during the cold rolling process, intermediate annealing is inserted at a temperature of 500℃ and held for 2 hours to eliminate work hardening and restore the alloy's plasticity.

[0158] Heat treatment control is carried out after cold and hot processing: a segmented heating method is adopted. First, the temperature is raised to 350℃ at 15℃ / min and held for 1.5h to eliminate processing stress; then the temperature is raised to 700℃ and held for 4h to control the nucleation and growth of Fe strengthening phase, control the size, distribution and interfacial orientation of precipitated phase with matrix; finally, air cooling is adopted at a cooling rate of 30℃ / min to suppress grain boundary migration, retain refined grains and uniform precipitated phase, and achieve uniform control of pipe cross-sectional properties.

[0159] Step 6: Finishing and Surface Treatment

[0160] After deformation and heat treatment, the pipes are straightened using a multi-roller straightener to control the curvature to <2mm and ensure straightness to ≤1mm per meter; the pipes are then cut to the required length, and end defects are removed.

[0161] Surface treatment is carried out by combining mechanical polishing and chemical passivation: mechanical polishing uses diamond abrasive to polish in stages (particle size from 800 mesh → 1200 mesh → 2000 mesh), with a roughness Ra=0.6μm, eliminating scratches and burrs; chemical passivation uses an environmentally friendly composite chemical passivation liquid, with a passivation temperature of 35℃ and a passivation time of 20min, forming a dense passivation film on the pipe surface, improving corrosion resistance, and avoiding the introduction of new surface defects.

[0162] The environmentally friendly composite chemical passivation solution comprises the following raw materials by mass fraction: citric acid 9%, benzotriazole 0.3%, phytic acid 2.3%, sodium molybdate 1.2%, sodium silicate 0.5%, sodium gluconate 3%, monoethanolamine 0.6%, ascorbic acid 0.2%, fatty alcohol polyoxyethylene ether 0.1%, and deionized water: balance.

[0163] The preparation method of the environmentally friendly composite chemical passivation solution includes the following steps:

[0164] (1) Preheating of basic solvent and predissolution of inorganic components

[0165] Add the measured amount of deionized water to the reactor, turn on the stirring device and heat to 45°C, and stir at a constant temperature; first add the specified amounts of sodium molybdate and sodium silicate, and stir at 300 r / min for 12 min until the inorganic strong base salt is completely dissolved to form a transparent inorganic mother liquor, thus avoiding subsequent violent reaction with organic acids to precipitate.

[0166] (2) Organic complex corrosion inhibitors are compounded and dissolved in steps.

[0167] Keep the solution temperature (45℃) and stirring speed (300r / min) constant, and add citric acid, benzotriazole, phytic acid and sodium gluconate in sequence. Stir for 10 minutes after each addition to ensure that the single component is completely dissolved and has no residue. After all the materials are added, continue stirring for 25 minutes to allow the inorganic corrosion inhibitor and organic complexing agent to fully chelate and form a homogeneous composite inhibitory mother liquor, thereby improving the compatibility of the components.

[0168] (3) Addition of low-temperature additives and precise pH control

[0169] Turn off the heating device and cool the solution to 24°C. Add ascorbic acid and fatty alcohol polyoxyethylene ether and stir for 10 minutes until completely dissolved. Slowly add monoethanolamine dropwise, monitoring the pH with a precision pH meter in real time as you add it. Adjust the pH of the solution to 7.3. After stopping the dropwise addition, continue stirring for 15 minutes to ensure that the pH of the solution is stable without fluctuations or local acid-base imbalance.

[0170] (4) Sealed aging and homogenization treatment

[0171] Maintain a low stirring speed of 200 rpm and place the solution in a sealed, light-proof environment for 40 minutes to completely eliminate air bubbles inside the solution, allowing all components to fully react synergistically, further improving the long-term stability of the passivation solution and preventing problems such as stratification, precipitation, and turbidity during use.

[0172] (5) Precision filtration purification and finished product storage

[0173] After aging, the passivation solution is pressure filtered using a 400-mesh precision filter cloth to remove trace amounts of insoluble impurities and microgel particles. The filtered passivation solution is then placed in a sealed plastic container and stored at room temperature away from light to obtain the finished environmentally friendly composite chemical passivation solution.

[0174] Step 7: Performance Testing and Iterative Optimization

[0175] The mechanical properties, corrosion resistance, thermal conductivity, and dimensional accuracy of the finished pipes were tested, and the results are shown in Table 9. The test data were fed back into a machine learning model to iteratively optimize the composition and process parameters, achieving precise control of the pipe performance. The metallographic microstructure of the copper alloy pipes prepared in Example 1 is shown in [Table 9]. Figure 5 ,Depend on Figure 5 It can be seen that the alloy matrix has a uniform structure and no metallurgical defects such as macrosegregation, coarse primary phase, microcracks, or pores were found. The black second phase particles are uniformly distributed in the copper matrix in a diffuse state, without local enrichment or agglomeration. This indicates that the composition and structure of the pipe are precisely controlled throughout the smelting, hot working, and heat treatment processes, and the metallurgical quality is stable and reliable, providing a basic guarantee for subsequent precision forming and service performance.

[0176] Figure 6 This is a scanning electron microscope (SEM) secondary electron (SE) morphology image of the copper alloy tubing prepared in Example 1, obtained by... Figure 6 It can be seen that the copper alloy pipe obtained in Example 1 has good surface integrity and no obvious macroscopic corrosion damage.

[0177] Figure 7 This is a metallographic microstructure image of the copper alloy tube prepared in Example 1 after etching for 20 seconds. Figure 7 It can be seen that the dendrite morphology is complete and the branches are well developed, indicating that the cooling rate during solidification is moderate and no ultrafine or equiaxed crystal structure is formed. The corrosion effect of the copper alloy tube is uniform after 20 seconds of etching, and the dendrite outline is clear, indicating that the 20-second etching time can effectively show the dendritic structure of the copper alloy and there is no over-corrosion or under-corrosion phenomenon.

[0178] Figure 8This is a scanning electron microscope (SEM) secondary electron (SE) morphology image of the copper alloy tubing prepared in Example 1 after immersion in a 3.5 wt.% NaCl solution for 7 days. Figure 8 It can be seen that the white particles on the surface are corrosion products generated by the copper alloy in the corrosive medium. The products exist in the form of particles and do not form a continuous and dense protective film; the products are unevenly distributed and do not completely cover the substrate.

[0179] Figure 9 These are secondary electron (SE) morphology images of the copper alloy tubing prepared in Example 1, obtained after immersion in a 3.5 wt.% NaCl solution for 15 days. Figure 9 It can be seen that the copper alloy pipe prepared in Example 1 formed a uniformly distributed fine particle / plate structure on its surface after corrosion. The particle distribution is dense and uniform, indicating that the structure has high coverage and can effectively block the penetration of corrosive media into the substrate, indicating that the structure has good protective properties.

[0180] Figure 10 This is a secondary electron (SE) morphology image of the copper alloy tubing prepared in Example 1 after immersion in a 3.5 wt.% NaCl solution for 15 days, obtained by scanning electron microscopy (SEM). Figure 10 It can be seen that after corrosion, a composite product layer is formed on the surface, mainly composed of cubic / plate-shaped crystals and supplemented by flocculent products. The dense covering layer formed by the mutual stacking of crystals can hinder the diffusion of corrosive media into the matrix, indicating that the product layer has good protective properties.

[0181] The copper alloy tubing prepared in Example 1 was immersed in a 3.5 wt.% NaCl solution for 30 days and then subjected to scanning electron microscopy. (See figure) Figure 11 ,from Figure 11 It can be seen that after 30 days of immersion in NaCl solution, the microstructure of the pipe remains uniform and dense, without obvious intergranular corrosion, pitting, or large-area corrosion peeling. This uniform and dense microstructure effectively inhibits the penetration and spread of corrosive media along grain boundaries or defects, significantly improving the long-term corrosion resistance of the copper alloy.

[0182] Example 2

[0183] A machine learning-aided design process for manufacturing high-performance copper alloy tubing includes the following steps:

[0184] Step 1: Machine Learning-Based Copper Alloy Composition Optimization Design

[0185] (1) A database containing 1000 sets of Cu-Fe alloy data was constructed. The input features were Fe, Ni, and P contents and process parameters. The output targets were tensile strength, yield strength, thermal conductivity, and corrosion rate. A random forest prediction model was constructed and multi-objective Bayesian optimization was used to screen the composition ratio as follows: Fe 4.8%, Ni 0.8%, P 0.003%, Nd 0.02%, balance Cu, and total impurities 0.4%.

[0186] Step 2: Raw material pretreatment

[0187] According to the optimized composition ratio, electrolytic copper (99.95% purity), high-purity iron, nickel powder, phosphorus copper master alloy, and copper-neodymium master alloy were selected. A combination of mechanical grinding and pickling was used to remove the oxide layer, oil, and impurities from the raw material surfaces. The pickling solution was a 5% (w / w) dilute sulfuric acid solution, at 26°C for 6 minutes. After pickling, the mixture was rinsed with deionized water until neutral and then placed in a vacuum drying oven at 80°C for 4 hours to remove moisture and residual acid, preventing the introduction of impurities that could affect the alloy's properties.

[0188] Step 3: Continuous casting to prepare copper alloy ingots

[0189] The pretreated raw materials are loaded into a vacuum induction furnace and evacuated to a vacuum level of 1×10⁻⁶. -3 Pa, introduce a protective gas mixture of argon and nitrogen (volume ratio 1:1), heat to 1230℃ for melting, and hold at that temperature for 28 minutes after the raw materials are completely melted to ensure uniform composition of the alloy liquid.

[0190] An optimized continuous casting crystallizer (with built-in spiral cooling water channel and vibration device) was used to control the pouring temperature at 1220℃, the cooling rate at 150℃ / s, the crystallizer vibration frequency at 20Hz, and the billet pulling speed at 0.6m / min. Through the coordinated control of cooling rate and vibration parameters, liquid-liquid phase separation and Fe dendrite nucleation growth were suppressed, the solidification structure was refined, and the Fe element distribution was uniform. Protective gas was continuously introduced during the casting process to prevent oxidation of the alloy liquid. Finally, a copper alloy ingot with a diameter of 100mm and a length of 2000mm was produced. The ingot had no obvious defects such as segregation, porosity, or inclusions.

[0191] Step 4: Ingot homogenization treatment

[0192] The ingots obtained from continuous casting are placed in a box-type resistance furnace and heated to 880°C at a heating rate of 6°C / min, and held at that temperature for 10 hours. High-temperature diffusion eliminates residual stress and compositional segregation within the ingot, dissolves coarse Fe dendrites, and refines the microstructure. After holding, the ingots are cooled to 280°C in the furnace and then air-cooled to room temperature to avoid rapid cooling causing new stress and microstructural defects.

[0193] Step 5: Coordinated regulation of deformation and heat treatment

[0194] The homogenized ingots are subjected to alternating hot and cold processing: first, hot rolling is performed at an initial rolling temperature of 860℃ and a final rolling temperature of 670℃, with a total deformation of 62%, rolling the ingots into slabs; then, multiple passes of cold rolling are performed, with a pass deformation of 18% and a total deformation of 80%, achieving grain breakage and microstructure refinement through cold rolling, while precisely controlling the uniformity of pipe wall thickness, inner and outer diameter tolerances, and roundness; during the cold rolling process, intermediate annealing is inserted at a temperature of 400℃ and held for 3 hours to eliminate work hardening and restore the alloy's plasticity.

[0195] After cold and hot processing, heat treatment is carried out to control the temperature: a segmented heating method is adopted. First, the temperature is raised to 305℃ at 12℃ / min and held for 2 hours to eliminate processing stress; then the temperature is raised to 650℃ and held for 3 hours to control the nucleation and growth of Fe strengthening phase, and to control the size, distribution and interfacial orientation of precipitated phase with the matrix; finally, water cooling is adopted at a cooling rate of 80℃ / min to suppress grain boundary migration, retain refined grains and uniform precipitated phase, and achieve uniform control of the cross-sectional properties of the pipe.

[0196] Step 6: Finishing and Surface Treatment

[0197] After deformation and heat treatment, the pipes are straightened using a multi-roller straightener to control the curvature to <2mm and ensure straightness to ≤1mm per meter; the pipes are then cut to the required length, and end defects are removed.

[0198] Surface treatment is carried out by combining mechanical polishing and chemical passivation: mechanical polishing uses diamond abrasive for progressive polishing (particle size from 800 mesh → 1200 mesh → 2000 mesh), with a roughness Ra = 0.7 μm, eliminating scratches and burrs; chemical passivation uses an environmentally friendly composite chemical passivation solution, with a passivation temperature of 28℃ and a passivation time of 20 min, forming a dense passivation film on the pipe surface, improving corrosion resistance, and avoiding the introduction of new surface defects.

[0199] The environmentally friendly composite chemical passivation solution comprises the following raw materials by mass fraction: 7% citric acid, 0.2% benzotriazole, 1.6% phytic acid, 0.8% sodium molybdate, 0.5% sodium silicate, 2% sodium gluconate, 0.3% monoethanolamine, 0.1% ascorbic acid, 0.06% fatty alcohol polyoxyethylene ether, and deionized water as the balance. The preparation method of the environmentally friendly composite chemical passivation solution is the same as in Example 1.

[0200] Step 7: Performance Testing and Iterative Optimization

[0201] The mechanical properties, corrosion resistance, thermal conductivity and dimensional accuracy of the finished pipes were tested. The results are shown in Table 9. The test data were fed back to the machine learning model to iteratively optimize the composition and process parameters, so as to achieve precise control of the pipe performance.

[0202] Example 3

[0203] A machine learning-aided design process for manufacturing high-performance copper alloy tubing includes the following steps:

[0204] Step 1: Machine Learning-Based Copper Alloy Composition Optimization Design

[0205] (1) A database containing 1000 sets of Cu-Fe alloy data was constructed. The input features were Fe, Ni, and P contents and process parameters. The output targets were tensile strength, yield strength, thermal conductivity, and corrosion rate. A random forest prediction model was constructed and multi-objective Bayesian optimization was used to screen the composition ratio: Fe 5.0%, Ni 0.3%, P 0.004%, Nd 0.04%, balance Cu, and total impurities 0.3%.

[0206] Step 2: Raw material pretreatment

[0207] According to the optimized composition ratio, electrolytic copper (99.95% purity), high-purity iron, nickel powder, phosphorus copper master alloy, and copper-neodymium master alloy were selected. A combination of mechanical grinding and pickling was used to remove the oxide layer, oil, and impurities from the raw material surface. The pickling solution was a 7% (w / w) dilute sulfuric acid solution, at 30℃ for 9 minutes. After pickling, the mixture was rinsed with deionized water until neutral and then placed in a vacuum drying oven at 90℃ for 3.5 hours to remove moisture and residual acid, preventing the introduction of impurities that could affect the alloy's properties.

[0208] Step 3: Continuous casting to prepare copper alloy ingots

[0209] The pretreated raw materials are loaded into a vacuum induction furnace and evacuated to a vacuum level of 1×10⁻⁶. -3 Pa, introduce a protective gas mixture of argon and nitrogen (volume ratio 1:1), heat to 1300℃ for melting, and hold at that temperature for 20 minutes after the raw materials are completely melted to ensure uniform composition of the alloy liquid.

[0210] An optimized continuous casting crystallizer (with built-in spiral cooling water channel and vibration device) was used to control the pouring temperature at 1320℃, the cooling rate at 120℃ / s, the crystallizer vibration frequency at 60Hz, and the billet pulling speed at 1.1m / min. Through the coordinated control of cooling rate and vibration parameters, liquid-liquid phase separation and Fe dendrite nucleation growth were suppressed, the solidification structure was refined, and the Fe element distribution was uniform. Protective gas was continuously introduced during the casting process to prevent oxidation of the alloy liquid. Finally, a copper alloy ingot with a diameter of 90mm and a length of 2600mm was produced. The ingot had no obvious defects such as segregation, porosity, or inclusions.

[0211] Step 4: Ingot homogenization treatment

[0212] The ingots obtained from continuous casting are placed in a box-type resistance furnace and heated to 920°C at a heating rate of 8°C / min, and held at that temperature for 6 hours. This process eliminates residual stress and compositional segregation within the ingot through high-temperature diffusion, dissolves coarse Fe dendrites, and refines the microstructure. After holding, the ingots are cooled to 290°C in the furnace and then air-cooled to room temperature to avoid rapid cooling causing new stress and microstructural defects.

[0213] Step 5: Coordinated regulation of deformation and heat treatment

[0214] The homogenized ingots are subjected to alternating hot and cold processing: first, hot rolling is performed at an initial rolling temperature of 860℃ and a final rolling temperature of 660℃, with a total deformation of 56%, rolling the ingots into slabs; then, multiple passes of cold rolling are performed, with a pass deformation of 19% and a total deformation of 75%, achieving grain breakage and microstructure refinement through cold rolling, while precisely controlling the uniformity of pipe wall thickness, inner and outer diameter tolerances, and roundness; during the cold rolling process, intermediate annealing is inserted at a temperature of 420℃ and held for 2.5 hours to eliminate work hardening and restore the alloy's plasticity.

[0215] After cold and hot processing, heat treatment is controlled: a segmented heating method is adopted. First, the temperature is raised to 300℃ at 12℃ / min and held for 2 hours to eliminate processing stress; then the temperature is raised to 550℃ and held for 6 hours to control the nucleation and growth of Fe strengthening phase, and to control the size, distribution and interfacial orientation of precipitated phase with the matrix; finally, furnace cooling is adopted at a cooling rate of 15℃ / min to suppress grain boundary migration, retain refined grains and uniform precipitated phase, and achieve uniform control of the cross-sectional properties of the pipe.

[0216] Step 6: Finishing and Surface Treatment

[0217] After deformation and heat treatment, the pipes are straightened using a multi-roller straightener to control the curvature to <2mm and ensure straightness to ≤1mm per meter; the pipes are then cut to the required length, and end defects are removed.

[0218] Surface treatment is carried out by combining mechanical polishing and chemical passivation: mechanical polishing uses diamond abrasive for progressive polishing (particle size from 800 mesh → 1200 mesh → 2000 mesh) to control the surface roughness Ra=0.5μm of the inner and outer surfaces, eliminating scratches and burrs; chemical passivation uses an environmentally friendly composite chemical passivation liquid, with a passivation temperature of 35℃ and a passivation time of 12min, to form a dense passivation film on the pipe surface, improving corrosion resistance and avoiding the introduction of new surface defects.

[0219] The environmentally friendly composite chemical passivation solution comprises the following raw materials by mass fraction: citric acid 8%, benzotriazole 0.2%, phytic acid 1.8%, sodium molybdate 0.9%, sodium silicate 0.5%, sodium gluconate 2.2%, monoethanolamine 0.4%, ascorbic acid 0.1%, fatty alcohol polyoxyethylene ether 0.07%, and deionized water: balance. The preparation method of the environmentally friendly composite chemical passivation solution is the same as in Example 1.

[0220] Step 7: Performance Testing and Iterative Optimization

[0221] The mechanical properties, corrosion resistance, thermal conductivity and dimensional accuracy of the finished pipes were tested. The results are shown in Table 9. The test data were fed back to the machine learning model to iteratively optimize the composition and process parameters, so as to achieve precise control of the pipe performance.

[0222] Example 4

[0223] A machine learning-aided design process for manufacturing high-performance copper alloy tubing includes the following steps:

[0224] Step 1: Machine Learning-Based Copper Alloy Composition Optimization Design

[0225] (1) A database containing 1000 sets of Cu-Fe alloy data was constructed. The input features were Fe, Ni, and P contents and process parameters. The output targets were tensile strength, yield strength, thermal conductivity, and corrosion rate. A random forest prediction model was constructed and multi-objective Bayesian optimization was used to screen the composition ratio as follows: Fe 5.2%, Ni 0.7%, P 0.005%, Nd 0.08%, balance Cu, and total impurities 0.5%.

[0226] Step 2: Raw material pretreatment

[0227] According to the optimized composition ratio, electrolytic copper (99.95% purity), high-purity iron, nickel powder, phosphorus copper master alloy, and copper-neodymium master alloy were selected. A combination of mechanical grinding and pickling was used to remove the oxide layer, oil, and impurities from the raw material surface. The pickling solution was an 8% (w / w) dilute sulfuric acid solution, at 30℃ for 9 minutes. After pickling, the mixture was rinsed with deionized water until neutral and placed in a vacuum drying oven at 100℃ for 3 hours to remove moisture and residual acid, preventing the introduction of impurities that could affect the alloy's properties.

[0228] Step 3: Continuous casting to prepare copper alloy ingots

[0229] The pretreated raw materials are loaded into a vacuum induction furnace and evacuated to a vacuum level of 1×10⁻⁶. -3 Pa, introduce a protective gas mixture of argon and nitrogen (volume ratio 1:1), heat to 1260℃ for melting, and hold at that temperature for 18 minutes after the raw materials are completely melted to ensure uniform composition of the alloy liquid.

[0230] An optimized continuous casting crystallizer (with built-in spiral cooling water channel and vibration device) was used to control the pouring temperature at 1200℃, the cooling rate at 80℃ / s, the crystallizer vibration frequency at 26Hz, and the casting speed at 1.1m / min. Through the coordinated control of cooling rate and vibration parameters, liquid-liquid phase separation and Fe dendrite nucleation growth were suppressed, the solidification structure was refined, and the Fe element distribution was uniform. Protective gas was continuously introduced during the casting process to prevent oxidation of the alloy liquid. Finally, a copper alloy ingot with a diameter of 130mm and a length of 3000mm was produced. The ingot had no obvious defects such as segregation, porosity, or inclusions.

[0231] Step 4: Ingot homogenization treatment

[0232] The ingots obtained from continuous casting are placed in a box-type resistance furnace and heated to 850°C at a heating rate of 5°C / min. The furnace is held at this temperature for 12 hours to eliminate residual stress and compositional segregation within the ingot through high-temperature diffusion, dissolve coarse Fe dendrites, and refine the microstructure. After holding, the ingots are cooled to 280°C in the furnace and then air-cooled to room temperature to avoid rapid cooling causing new stress and microstructural defects.

[0233] Step 5: Coordinated regulation of deformation and heat treatment

[0234] The homogenized ingots are subjected to alternating hot and cold processing: first, hot rolling is performed at an initial rolling temperature of 880℃ and a final rolling temperature of 650℃, with a total deformation of 58%, rolling the ingots into slabs; then, multiple cold rolling passes are performed, with a deformation of 20% per pass and a total deformation of 78%, achieving grain breakage and microstructure refinement through cold rolling, while precisely controlling the uniformity of pipe wall thickness, inner and outer diameter tolerances, and roundness; during the cold rolling process, intermediate annealing is inserted at a temperature of 450℃ and held for 2 hours to eliminate work hardening and restore the alloy's plasticity.

[0235] Heat treatment control is carried out after cold and hot processing: a segmented heating method is adopted. First, the temperature is raised to 350℃ at 13℃ / min and held for 1.5h to eliminate processing stress; then the temperature is raised to 700℃ and held for 2h to control the nucleation and growth of Fe strengthening phase, control the size, distribution and interfacial orientation of precipitated phase with matrix; finally, air cooling is adopted at a cooling rate of 40℃ / min to suppress grain boundary migration, retain refined grains and uniform precipitated phase, and achieve uniform control of pipe cross-sectional properties.

[0236] Step 6: Finishing and Surface Treatment

[0237] After deformation and heat treatment, the pipes are straightened using a multi-roller straightener to control the curvature to <2mm and ensure straightness to ≤1mm per meter; the pipes are then cut to the required length, and end defects are removed.

[0238] Surface treatment is carried out by combining mechanical polishing and chemical passivation: mechanical polishing uses diamond abrasive for progressive polishing (particle size from 800 mesh → 1200 mesh → 2000 mesh), with a roughness Ra = 0.8 μm, eliminating scratches and burrs; chemical passivation uses an environmentally friendly composite chemical passivation liquid, with a passivation temperature of 45℃ and a passivation time of 10 min, forming a dense passivation film on the pipe surface, improving corrosion resistance, and avoiding the introduction of new surface defects.

[0239] The environmentally friendly composite chemical passivation solution comprises the following raw materials by mass fraction: citric acid 9%, benzotriazole 0.3%, phytic acid 1.7%, sodium molybdate 1%, sodium silicate 0.5%, sodium gluconate 2.6%, monoethanolamine 0.4%, ascorbic acid 0.2%, fatty alcohol polyoxyethylene ether 0.09%, and deionized water: balance. The preparation method of the environmentally friendly composite chemical passivation solution is the same as in Example 1.

[0240] Step 7: Performance Testing and Iterative Optimization

[0241] The mechanical properties, corrosion resistance, thermal conductivity and dimensional accuracy of the finished pipes were tested. The results are shown in Table 9. The test data were fed back to the machine learning model to iteratively optimize the composition and process parameters, so as to achieve precise control of the pipe performance.

[0242] Example 5

[0243] A machine learning-aided design process for manufacturing high-performance copper alloy tubing includes the following steps:

[0244] Step 1: Machine Learning-Based Copper Alloy Composition Optimization Design

[0245] (1) A database containing 1000 sets of Cu-Fe alloy data was constructed. The input features were Fe, Ni, and P contents and process parameters. The output targets were tensile strength, yield strength, thermal conductivity, and corrosion rate. A random forest prediction model was constructed and multi-objective Bayesian optimization was used to screen the composition ratio as follows: Fe 6.0%, Ni 0.4%, P 0.004%, Nd 0.1%, balance Cu, and total impurities 0.3%.

[0246] Step 2: Raw material pretreatment

[0247] According to the optimized composition ratio, 99.95% electrolytic copper, high-purity iron, nickel powder, phosphorus copper master alloy, and copper-neodymium master alloy were selected. A combination of mechanical grinding and pickling was used to remove the oxide layer, oil, and impurities from the raw material surface. The pickling solution was a 7.2% (w / w) dilute sulfuric acid solution, at 32℃ for 7 minutes. After pickling, the mixture was rinsed with deionized water until neutral and then placed in a vacuum drying oven at 90℃ for 3.5 hours to remove moisture and residual acid, preventing the introduction of impurities that could affect the alloy's properties.

[0248] Step 3: Continuous casting to prepare copper alloy ingots

[0249] The pretreated raw materials are loaded into a vacuum induction furnace and evacuated to a vacuum level of 1×10⁻⁶. -3 Pa, introduce a protective gas mixture of argon and nitrogen (volume ratio 1:1), heat to 1240℃ for melting, and hold at that temperature for 16 minutes after the raw materials are completely melted to ensure uniform composition of the alloy liquid.

[0250] An optimized continuous casting crystallizer (with built-in spiral cooling water channel and vibration device) was used to control the pouring temperature at 1250℃, the cooling rate at 40℃ / s, the crystallizer vibration frequency at 30Hz, and the billet pulling speed at 1.2m / min. Through the coordinated control of cooling rate and vibration parameters, liquid-liquid phase separation and Fe dendrite nucleation growth were suppressed, the solidification structure was refined, and the Fe element distribution was uniform. Protective gas was continuously introduced during the casting process to prevent oxidation of the alloy liquid. Finally, a copper alloy ingot with a diameter of 120mm and a length of 4000mm was produced. The ingot had no obvious defects such as segregation, porosity, or inclusions.

[0251] Step 4: Ingot homogenization treatment

[0252] The ingots obtained from continuous casting are placed in a box-type resistance furnace and heated to 900°C at a heating rate of 9°C / min, and held at that temperature for 7 hours. This process eliminates residual stress and compositional segregation within the ingot through high-temperature diffusion, dissolves coarse Fe dendrites, and refines the microstructure. After holding, the ingots are cooled to 280°C in the furnace and then air-cooled to room temperature to avoid rapid cooling causing new stress and microstructural defects.

[0253] Step 5: Coordinated regulation of deformation and heat treatment

[0254] The homogenized ingots are subjected to alternating hot and cold processing: First, hot rolling is performed with an initial rolling temperature of 900℃ and a final rolling temperature of 700℃, resulting in a total deformation of 65%, rolling the ingots into slabs; then, multiple cold rolling passes are performed with a deformation of 20% per pass and a total deformation of 85%, achieving grain breakage and microstructure refinement through cold rolling, while precisely controlling the uniformity of pipe wall thickness, inner and outer diameter tolerances, and roundness; during the cold rolling process, intermediate annealing is inserted at a temperature of 400℃ and held for 3 hours to eliminate work hardening and restore the alloy's plasticity.

[0255] Heat treatment control is carried out after cold and hot processing: a segmented heating method is adopted. First, the temperature is raised to 300℃ at 11℃ / min and held for 2 hours to eliminate processing stress; then the temperature is raised to 580℃ and held for 3 hours to control the nucleation and growth of Fe strengthening phase, control the size, distribution and interfacial orientation of precipitated phase with the matrix; finally, water cooling, air cooling or furnace cooling is adopted at a cooling rate of 30℃ / min to suppress grain boundary migration, retain refined grains and uniform precipitated phase, and achieve uniform control of the cross-sectional properties of the pipe.

[0256] Step 6: Finishing and Surface Treatment

[0257] After deformation and heat treatment, the pipes are straightened using a multi-roller straightener to control the curvature to <2mm and ensure straightness to ≤1mm per meter; the pipes are then cut to the required length, and end defects are removed.

[0258] Surface treatment is carried out by combining mechanical polishing and chemical passivation: mechanical polishing uses diamond abrasive for progressive polishing (particle size from 800 mesh → 1200 mesh → 2000 mesh), with a roughness Ra=0.8μm, eliminating scratches and burrs; chemical passivation uses an environmentally friendly composite chemical passivation liquid, with a passivation temperature of 40℃ and a passivation time of 23min, forming a dense passivation film on the pipe surface, improving corrosion resistance, and avoiding the introduction of new surface defects.

[0259] The environmentally friendly composite chemical passivation solution comprises the following raw materials by mass fraction: 11.2% citric acid, 0.3% benzotriazole, 2.1% phytic acid, 1.1% sodium molybdate, 0.6% sodium silicate, 2.8% sodium gluconate, 0.4% monoethanolamine, 0.2% ascorbic acid, 0.11% fatty alcohol polyoxyethylene ether, and deionized water as the balance. The preparation method of the environmentally friendly composite chemical passivation solution is the same as in Example 1.

[0260] Step 7: Performance Testing and Iterative Optimization

[0261] The mechanical properties, corrosion resistance, thermal conductivity and dimensional accuracy of the finished pipes were tested. The results are shown in Table 9. The test data were fed back to the machine learning model to iteratively optimize the composition and process parameters, so as to achieve precise control of the pipe performance.

[0262] Example 6

[0263] A machine learning-aided design process for manufacturing high-performance copper alloy tubing includes the following steps:

[0264] Step 1: Machine Learning-Based Copper Alloy Composition Optimization Design

[0265] (1) A database containing 1000 sets of Cu-Fe alloy data was constructed. The input features were Fe, Ni, and P contents and process parameters. The output targets were tensile strength, yield strength, thermal conductivity, and corrosion rate. A random forest prediction model was constructed and multi-objective Bayesian optimization was used to screen the composition ratio: Fe 5.2%, Ni 0.5%, P 0.006%, Nd 0.12%, balance Cu, and total impurities 0.4%.

[0266] Step 2: Raw material pretreatment

[0267] According to the optimized composition ratio, electrolytic copper (99.95% purity), high-purity iron, nickel powder, phosphorus copper master alloy, and copper-neodymium master alloy were selected. A combination of mechanical grinding and pickling was used to remove the oxide layer, oil, and impurities from the raw material surface. The pickling solution was a 9% (w / w) dilute sulfuric acid solution, at 38℃ for 6 minutes. After pickling, the mixture was rinsed with deionized water until neutral and placed in a vacuum drying oven at 120℃ for 2 hours to remove moisture and residual acid, preventing the introduction of impurities that could affect the alloy's properties.

[0268] Step 3: Continuous casting to prepare copper alloy ingots

[0269] The pretreated raw materials are loaded into a vacuum induction furnace and evacuated to a vacuum level of 1×10⁻⁶. -3 Pa, introduce a protective gas mixture of argon and nitrogen (volume ratio 1:1), heat to 1350℃ for melting, and hold at that temperature for 10 minutes after the raw materials are completely melted to ensure uniform composition of the alloy liquid.

[0270] An optimized continuous casting crystallizer (with built-in spiral cooling water channel and vibration device) was used to control the pouring temperature at 1260℃, the cooling rate at 180℃ / s, the crystallizer vibration frequency at 50Hz, and the casting speed at 0.9m / min. Through the coordinated control of cooling rate and vibration parameters, liquid-liquid phase separation and Fe dendrite nucleation growth were suppressed, the solidification structure was refined, and the Fe element distribution was uniform. Protective gas was continuously introduced during the casting process to prevent oxidation of the alloy liquid. Finally, a copper alloy ingot with a diameter of 200mm and a length of 5000mm was produced. The ingot had no obvious defects such as segregation, porosity, or inclusions.

[0271] Step 4: Ingot homogenization treatment

[0272] The ingots obtained from continuous casting are placed in a box-type resistance furnace and heated to 950°C at a heating rate of 14°C / min. The furnace is held at this temperature for 4 hours to eliminate residual stress and compositional segregation within the ingot through high-temperature diffusion, dissolve coarse Fe dendrites, and refine the microstructure. After holding, the ingots are cooled to 290°C in the furnace and then air-cooled to room temperature to avoid rapid cooling causing new stress and microstructural defects.

[0273] Step 5: Coordinated regulation of deformation and heat treatment

[0274] The homogenized ingots are subjected to alternating hot and cold processing: First, hot rolling is performed with an initial rolling temperature of 870℃ and a final rolling temperature of 700℃, resulting in a total deformation of 63%, rolling the ingots into slabs; then, multiple cold rolling passes are performed with a deformation of 17% per pass and a total deformation of 80%, achieving grain breakage and microstructure refinement through cold rolling, while precisely controlling the uniformity of pipe wall thickness, inner and outer diameter tolerances, and roundness; during the cold rolling process, intermediate annealing is inserted at a temperature of 600℃ and held for 1 hour to eliminate work hardening and restore the alloy's plasticity.

[0275] Heat treatment control is carried out after cold and hot processing: a segmented heating method is adopted. First, the temperature is raised to 360℃ at 20℃ / min and held for 1.5h to eliminate processing stress; then the temperature is raised to 620℃ and held for 3.5h to control the nucleation and growth of Fe strengthening phase, control the size, distribution and interfacial orientation of precipitated phase with the matrix; finally, air cooling is adopted at a cooling rate of 80℃ / min to suppress grain boundary migration, retain refined grains and uniform precipitated phase, and achieve uniform control of the cross-sectional properties of the pipe.

[0276] Step 6: Finishing and Surface Treatment

[0277] After deformation and heat treatment, the pipes are straightened using a multi-roller straightener to control the curvature to <2mm and ensure straightness to ≤1mm per meter; the pipes are then cut to the required length, and end defects are removed.

[0278] Surface treatment is carried out by combining mechanical polishing and chemical passivation: mechanical polishing uses diamond abrasive to polish in stages (particle size from 800 mesh → 1200 mesh → 2000 mesh), with a roughness Ra=0.7μm, eliminating scratches and burrs; chemical passivation uses an environmentally friendly composite chemical passivation liquid, with a passivation temperature of 50℃ and a passivation time of 5min, forming a dense passivation film on the pipe surface, improving corrosion resistance, and avoiding the introduction of new surface defects.

[0279] The environmentally friendly composite chemical passivation solution comprises the following raw materials by mass fraction: 13.5% citric acid, 0.4% benzotriazole, 2.8% phytic acid, 1.4% sodium molybdate, 0.6% sodium silicate, 3.8% sodium gluconate, 0.8% monoethanolamine, 0.3% ascorbic acid, 0.13% fatty alcohol polyoxyethylene ether, and deionized water as the balance. The preparation method of the environmentally friendly composite chemical passivation solution is the same as in Example 1.

[0280] Step 7: Performance Testing and Iterative Optimization

[0281] The mechanical properties, corrosion resistance, thermal conductivity and dimensional accuracy of the finished pipes were tested. The results are shown in Table 9. The test data were fed back to the machine learning model to iteratively optimize the composition and process parameters, so as to achieve precise control of the pipe performance.

[0282] Comparative Example 1

[0283] The difference from Example 1 is that machine learning was not used to optimize the composition; instead, a traditional trial-and-error method was used for the composition ratio: Fe 6.0%, Ni 0.6%, P 0.005%, Nd 0.1%, with the balance being Cu, and total impurities of 0.3%. The remaining steps are the same as in Example 1. Figure 12It can be seen that the electrochemical impedance of the copper alloy tube prepared in Example 1 is much higher than that of the copper alloy tube prepared in Comparative Example 1. Under the same test conditions, the corrosion resistance of the copper alloy tube in Example 1 is significantly better than that of the copper alloy tube in Comparative Example 1.

[0284] Comparative Example 2

[0285] The difference from Example 1 is that traditional metal mold casting is used instead of continuous casting, the pouring temperature is 1280℃, and solidification is carried out in the air. The remaining steps are the same as in Example 1.

[0286] Comparative Example 3

[0287] The difference from Example 1 is that no homogenization treatment was performed; the ingot was directly deformed. The remaining steps are the same as in Example 1.

[0288] Comparative Example 4

[0289] The difference from Example 1 is that the total deformation was 75%, and intermediate annealing was not performed. The remaining steps are the same as in Example 1.

[0290] Comparative Example 5

[0291] The difference from Example 1 is that the heat treatment temperature is 800℃, the holding time is 10 hours, and the product is air-cooled. The remaining steps are the same as in Example 1.

[0292] Comparative Example 6

[0293] The difference from Example 1 is that the surface treatment only involves mechanical polishing, without chemical passivation. The remaining steps are the same as in Example 1.

[0294] Single-factor experiment for screening key process parameters:

[0295] This invention conducts the following multiple single-factor experiments, based on Example 1, fixing other parameters and changing only the target parameter, detecting changes in performance indicators, and clarifying the optimal parameter range.

[0296] Experiment 1: The effect of Fe element content on performance

[0297] Based on Example 1, the Fe content was changed to 4.0%, 4.5%, 5.0%, 5.5%, 6.0%, 6.5%, and 7.0%, and the tensile strength and thermal conductivity were tested. The results are shown in Table 1.

[0298]

[0299] As shown in Table 1:

[0300] When the Fe content is less than 4.5%, solid solution strengthening and precipitation strengthening are insufficient, and the tensile strength is less than 650.1 MPa, which cannot meet the requirements of high-pressure working conditions; at the same time, the Fe content is too low, which cannot effectively refine the microstructure and slightly reduces the corrosion resistance.

[0301] When the Fe content is greater than 6.5%, liquid-liquid phase separation and Fe dendrite segregation are easily triggered, resulting in coarse ingot structure, poor machinability, and although the tensile strength is slightly improved, the thermal conductivity decreases to below 300 W / (m·K). At the same time, the corrosion resistance decreases and the corrosion rate increases.

[0302] The optimal Fe content range is 4.5%-6.5%, within which a better balance is achieved between strength, thermal conductivity and plasticity.

[0303] Experiment 2: Effect of Rare Earth Nd Addition on Performance

[0304] Based on Example 1, only the amount of Nd added was changed to 0.00%, 0.01%, 0.05%, 0.10%, and 0.15%, and the hardness and conductivity were tested. The results are shown in Table 2.

[0305]

[0306] The results in Table 2 show that:

[0307] Hardness performance: The addition of Nd can significantly improve the hardness of the alloy. The hardness reaches its peak (160.6 HV) at 0.01% Nd, which is 15.0% higher than the group without Nd. The hardness drops and tends to stabilize slightly at 0.05%-0.10% Nd. The hardness rises again to 158.2 HV at 0.15% Nd, which is still significantly higher than the group without Nd.

[0308] Conductivity: In the initial stage of Nd addition (0.01%), the conductivity decreased slightly due to the scattering of solid solution atoms; as the Nd content increased to 0.15%, Nd atoms precipitated in the form of a second phase, and the conductivity rebounded significantly to near the level of the unadded group, thus effectively preserving the conductivity.

[0309] Overall performance: 0.15% Nd is the optimal addition amount, which can significantly improve the hardness of the alloy with almost no loss of conductivity, thus balancing the mechanical and electrical properties of copper-based alloys and providing key process parameters for the preparation of high-performance copper-based conductive materials.

[0310] Experiment 3: Effect of continuous casting cooling rate on ingot segregation

[0311] Based on Example 1, only the cooling rate was changed to 30℃ / s, 50℃ / s, 100℃ / s, 150℃ / s, 200℃ / s, and 250℃ / s, and the Fe element segregation degree (maximum Fe content / average Fe content) was detected. The results are shown in Table 3.

[0312]

[0313] As shown in Table 3:

[0314] When the cooling rate is <50℃ / s, the solidification rate is slow, Fe atoms diffuse sufficiently, dendrite segregation is severe, segregation degree >1.5, the ingot structure is uneven, and subsequent processing is prone to cracking.

[0315] When the cooling rate is greater than 200℃ / s, the cooling is too fast, the internal stress of the ingot is concentrated, and cracks and porosity defects are easily generated. At the same time, the microstructure is too fine, and it is difficult to control the precipitated phase in subsequent heat treatment, resulting in a decrease in performance stability.

[0316] The optimal cooling rate range is 50%-200℃ / s. Within this range, the segregation is controlled below 1.4, the ingot has no obvious defects, and the microstructure is well uniform.

[0317] Experiment 4: Effect of homogenization treatment temperature on tissue

[0318] Based on Example 1, only the homogenization temperature was changed to 800℃, 850℃, 900℃, 950℃, and 1000℃. The grain size and segregation residue were detected, and the results are shown in Table 4.

[0319]

[0320] As shown in Table 4:

[0321] When the temperature is <850℃, diffusion is insufficient, segregation residue is >5%, coarse Fe dendrites are not completely dissolved, grain size is still >50μm, and subsequent processing plasticity is poor.

[0322] When the temperature is greater than 950℃, the grains grow excessively, the grain size is greater than 80μm, the grain boundaries are weakened, and subsequent deformation processing is prone to grain boundary cracking. At the same time, alloying elements volatilize, and the composition deviation increases.

[0323] The optimal homogenization temperature range is 850-950℃, within which the segregation residue is <2%, the grain size is refined to 32-40μm, and the processing plasticity is good.

[0324] Experiment 5: The Influence of Total Deformation on Mechanical Properties

[0325] Based on Example 1, the tensile strength and elongation were tested by changing the total deformation to 45%, 55%, 65%, 75%, 85%, and 90%, and the results are shown in Table 5.

[0326]

[0327] As shown in Table 5:

[0328] When the total deformation is less than 65%, the grains are not sufficiently broken, the microstructure is poorly refined, and the tensile strength is less than 650 MPa, which fails to meet the performance requirements.

[0329] When the total deformation exceeds 85%, work hardening is severe. Although the tensile strength increases to over 680 MPa, the elongation drops to below 3%, making the pipe prone to cracking and reducing the yield.

[0330] The optimal total deformation range is 65%-85%, within which the tensile strength is >650MPa and the elongation is >20%, balancing strength and plasticity.

[0331] Experiment 6: Effect of heat treatment temperature on precipitated phases

[0332] Based on Example 1, only the heat treatment temperature was changed to 450℃, 500℃, 600℃, 700℃, 750℃, and 800℃. The size of the precipitated phase and the tensile strength were tested, and the results are shown in Table 6.

[0333]

[0334] As shown in Table 6:

[0335] When the temperature is <500℃, the nucleation of Fe precipitates is insufficient, the size of the precipitate is <20nm, the strengthening effect is limited, and the tensile strength is <650MPa.

[0336] When the temperature is greater than 750℃, the precipitated phase coarsens to more than 100nm, the pinning ability of dislocations decreases, and the grains grow, resulting in poor performance stability.

[0337] The preferred heat treatment temperature range is 500-750℃, within which the precipitated phase size is 30-82nm and the tensile strength is 150-162HV.

[0338] Experiment 7: Effect of Cooling Rate on Grain Boundary Migration

[0339] Based on Example 1, only the cooling rate was changed to 5℃ / min, 10℃ / min, 30℃ / min, 60℃ / min, 100℃ / min, and 150℃ / min. The grain size and performance uniformity (difference in cross-sectional hardness) were tested, and the results are shown in Table 7.

[0340]

[0341] As shown in Table 7:

[0342] When the cooling rate is <10℃ / min, grain boundary migration is sufficient, grains grow to more than 60μm, the cross-sectional hardness difference is >10HV, the performance is uneven, and the elongation decreases.

[0343] When the cooling rate is greater than 100℃ / min, internal stress concentrates, the pipe is prone to deformation, and the precipitated phase becomes excessively refined, resulting in decreased corrosion resistance.

[0344] The optimal cooling rate range is 10-100℃ / min, within which the grain size is 32-40μm, the cross-sectional hardness difference is 4-5HV, and the performance uniformity is good.

[0345] Experiment 8: Effect of passivation time on corrosion resistance

[0346] Based on Example 1, only the passivation time was changed to 5 min, 10 min, 15 min, 20 min, 30 min, and 40 min, and the corrosion rate and surface roughness were detected. The results are shown in Table 8.

[0347]

[0348] As shown in Table 8:

[0349] When the passivation time is less than 10 minutes, the passivation film is incomplete, the corrosion rate is greater than 0.05 mm / a, and the corrosion resistance is insufficient.

[0350] When the passivation time is greater than 30 min, the passivation film becomes too thick, and the surface roughness increases to more than 1.0 μm, which affects the heat exchange efficiency. At the same time, the film layer is easy to peel off, and the corrosion resistance decreases.

[0351] The optimal passivation time range is 10-30 min, within which the corrosion rate is ≤0.030 mm / a and the surface roughness Ra is ≤0.75 μm, thus balancing corrosion resistance and surface quality.

[0352] Performance testing and results analysis:

[0353] 1. Detection Method

[0354] Mechanical properties: tested according to GB / T 228.1-2021, with an original gauge length of 50 mm and a tensile rate of 1 mm / min;

[0355] Vickers hardness: Tested according to GB / T 4340.1-2024, test force 5 kgf (HV5), holding time 15 s;

[0356] Corrosion resistance: Periodic immersion corrosion test was conducted according to GB / T 19746-2018, with 3.5% (mass fraction) NaCl solution (simulating refrigeration medium) as the medium, and the test period was 30 days to test the corrosion rate;

[0357] Thermal conductivity: The high-temperature thermal conductivity of metals is measured by preparing a metal rod-shaped sample, passing a direct current through the sample to make it self-heat, and establishing a stable longitudinal heat flow and temperature gradient under high-temperature conditions; the temperature of the working section of the sample, the current, the voltage and the geometric dimensions of the sample are accurately measured, and the high-temperature thermal conductivity of the material is calculated by substituting the measured parameters into the steady-state heat conduction theory formula.

[0358] Dimensional accuracy: Vernier calipers (accuracy 0.01mm) are used to measure the outer diameter / wall thickness, and a roundness tester is used to measure roundness;

[0359] Surface quality: Ra value measured by a roughness tester.

[0360] 2. Test Results

[0361] The copper alloy tubing prepared in Examples 1-6 and Comparative Examples 1-6 was tested, and the test results are shown in Table 9.

[0362]

[0363] Based on Table 9, data comparison and theoretical analysis are performed:

[0364] 1. Performance comparison between the examples and comparative examples:

[0365] Examples 1-6 all exhibit tensile strength > 650 MPa, yield strength > 520 MPa, elongation > 20%, and hardness > 150 HV, meeting the core mechanical performance requirements. Comparative Examples 1-5, due to unreasonable composition design and process defects, all fail to meet mechanical performance standards. Comparative Example 1 shows excessive Fe content leading to segregation; Comparative Example 2 shows outdated casting process causing uneven microstructure; Comparative Example 3 shows incomplete homogenization resulting in residual segregation; Comparative Example 4 shows insufficient deformation leading to inadequate strengthening; and Comparative Example 5 shows excessively high heat treatment temperature leading to coarsening of the precipitates. All these examples demonstrate the necessity of the process described in this invention.

[0366] Regarding thermal conductivity, the thermal conductivity of the embodiments is all >315W / (m·K), which is much higher than that of the comparative examples of 275-300W / (m·K). This indicates that the present invention effectively suppresses the blockage of the relatively strong thermal conductivity pathway through composition optimization and process control, and achieves a balance between high strength and high thermal conductivity.

[0367] Regarding corrosion resistance, the corrosion rate of the examples was ≤0.03 mm / a, while that of Comparative Example 6, which lacked passivation, was >0.05 mm / a, demonstrating the crucial role of chemical passivation. Comparative Examples 1-5 exhibited inferior corrosion resistance compared to the examples due to uneven microstructure, indicating that microstructure uniformity is fundamental to improving corrosion resistance. In terms of dimensional accuracy and surface quality, the examples had an outer diameter tolerance of ±0.10-0.13 mm, a wall thickness tolerance of ±7-9%, a roundness deviation of <1%, and Ra <0.8 μm, all meeting the technical specifications. The comparative examples, due to process defects, exhibited large dimensional deviations and poor surface quality, failing to meet the requirements of high-end equipment.

[0368] 2. Theoretical Analysis:

[0369] This invention optimizes the composition through machine learning, achieving synergistic matching of Fe, Ni, P, and Nd elements. The strengthening effect of Fe, the stabilizing effect of Ni, the grain boundary regulation effect of P, and the grain refinement effect of Nd are fully utilized, avoiding defects caused by excessive amounts of a single element and ensuring the fundamental performance.

[0370] The synergistic effect of continuous casting and homogenization treatment eliminated Fe segregation and coarse dendrites, produced uniform ingots, provided high-quality billets for subsequent deformation-heat treatment, and avoided performance fluctuations caused by uneven microstructure.

[0371] The deformation-heat treatment synergistic regulation achieves grain refinement and precipitate control. By breaking the grains through large deformation and precisely controlling the size of the precipitates through heat treatment, the balance between strength and plasticity is ensured, while inhibiting grain boundary migration and achieving uniform cross-sectional properties.

[0372] Surface treatment processes improve surface quality and corrosion resistance. The formation of a passivation film effectively inhibits corrosion reactions while ensuring heat exchange efficiency and extending the service life of the pipe.

[0373] In summary, the preparation process of this invention, through multi-stage synergistic optimization, produces alloy pipes that significantly outperform existing technologies in terms of mechanical properties, thermal conductivity, corrosion resistance, dimensional accuracy, and surface quality, fully meeting the application requirements of high-end refrigeration equipment. This demonstrates that the technology of this invention represents a significant advancement.

[0374] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A manufacturing process for high-performance copper alloy tubing designed with machine learning assistance, characterized in that, Includes the following steps: (1) Optimization design of copper alloy composition based on machine learning: Collect data on composition, solidification behavior, precipitates and properties of Cu-Fe alloys, construct a multi-objective prediction model, and optimize the proportion range of Cu, Fe, Ni, P and Nd elements through multi-objective Bayesian optimization and non-dominated genetic algorithm. The proportions are Fe 4.5%-6.5%, Ni 0.2%-1.0%, P 0.003%-0.006%, Nd 0.01%-0.15% by mass fraction, with the balance being Cu and unavoidable impurities. The total amount of impurities does not exceed 0.5%. SHAP analysis is used to clarify the marginal contribution of each element to strength, corrosion resistance and thermal conductivity. (2) Raw material pretreatment: Select high-purity Cu, Fe, Ni, P and Nd according to the optimized composition ratio, remove the surface oxide layer and impurities, and dry them under vacuum for later use; (3) Continuous casting to prepare copper alloy ingots: The pretreated raw materials are placed in a vacuum induction furnace and melted into alloy liquid at a set temperature of 1200-1350℃ and under a protective atmosphere. By optimizing the structure of the continuous casting crystallizer, the cooling rate is 50-200℃ / s and the billet pulling speed is 0.5-1.2m / min, the solidification structure and Fe element distribution are controlled to prepare copper alloy ingots with uniform composition and structure. (4) Homogenization treatment of ingots: The ingots obtained by continuous casting are subjected to high-temperature homogenization annealing with a heating rate of 5-15℃ / min, heated to 850-950℃, held for 4-12h, cooled to below 300℃ in the furnace and then air-cooled to eliminate residual stress and compositional segregation in the ingots and refine the Fe dendrite structure. (5) Deformation-heat treatment synergistic control: The homogenized ingot is subjected to alternating hot and cold processing, and after cold rolling, intermediate annealing is carried out at a temperature of 400-600℃ and a holding time of 1-3h. The uniformity of pipe wall thickness, inner and outer diameter tolerance and roundness are precisely controlled. Then, by setting the heating process, cooling rate and holding time, the grain nucleation and growth, the size distribution of strengthening phase and the grain boundary migration behavior are controlled. (6) Finishing and surface treatment: The deformed and heat-treated pipes are straightened and cut. Mechanical polishing and chemical passivation are combined to control the roughness of the inner and outer surfaces and eliminate scratches and cracks. The chemical passivation uses an environmentally friendly composite chemical passivation solution. The environmentally friendly composite chemical passivation solution, in terms of mass fraction, includes the following raw materials: citric acid 6%-14%, benzotriazole 0.2%-0.4%, phytic acid 1.5%-3%, sodium molybdate 0.8%-1.5%, sodium silicate 0.5%-0.6%, sodium gluconate 2%-4%, monoethanolamine 0.3%-0.8%, ascorbic acid 0.1%-0.3%, fatty alcohol polyoxyethylene ether 0.05%-0.15%, deionized water: balance; (7) Performance testing and iterative optimization: The mechanical properties, corrosion resistance, thermal conductivity and dimensional accuracy of the finished pipes are tested, and the test data are fed back to the machine learning model to iteratively optimize the composition and process parameters, so as to achieve precise control of the pipe performance.

2. The manufacturing process for high-performance copper alloy tubing assisted by machine learning according to claim 1, characterized in that, In step (1), the machine learning model includes random forest, gradient boosting tree and neural network. The input features include alloy element content, solidification cooling rate and heat treatment temperature. The output targets are tensile strength, yield strength, elongation, thermal conductivity and corrosion rate.

3. The manufacturing process for high-performance copper alloy tubing assisted by machine learning according to claim 1, characterized in that, In step (1), the ratio of Cu, Fe, Ni, P and Nd elements is optimized. The ratio is expressed as mass fraction: Fe 6.0%, Ni 0.6%, P 0.005%, Nd 0.1%, with Cu as the balance and 0.3% total impurities.

4. The manufacturing process for high-performance copper alloy tubing assisted by machine learning according to claim 1, characterized in that, In step (3), the crystallizer vibration frequency is 20-80Hz.

5. The manufacturing process for machine learning-aided design of high-performance copper alloy tubing according to claim 1, characterized in that, The protective atmosphere described in step (3) is a mixture of argon and nitrogen, with the oxygen content controlled below 10 ppm.

6. The manufacturing process for machine learning-aided design of high-performance copper alloy tubing according to claim 1, characterized in that, The heat treatment process in step (5) is as follows: heating in sections to 500-750℃ and holding for 2-8 hours.

7. The manufacturing process for machine learning-aided design of high-performance copper alloy tubing according to claim 1, characterized in that, In step (5), water cooling, air cooling or furnace cooling is used, with a cooling rate of 10-100℃ / min.

8. The manufacturing process for machine learning-aided design of high-performance copper alloy tubing according to claim 1, characterized in that, In step (6), mechanical polishing is performed using diamond abrasive in stages, with the roughness Ra value controlled below 0.8 μm.

9. The manufacturing process for machine learning-aided design of high-performance copper alloy tubing according to claim 1, characterized in that, In step (6), the passivation temperature for chemical passivation is 25-50℃ and the passivation time is 5-30min.

10. The manufacturing process for machine learning-aided design of high-performance copper alloy tubing according to claim 8, characterized in that, The chemical passivation uses an environmentally friendly composite chemical passivation solution, which, by mass fraction, comprises the following raw materials: citric acid 9%, benzotriazole 0.3%, phytic acid 2.3%, sodium molybdate 1.2%, sodium silicate 0.5%, sodium gluconate 3%, monoethanolamine 0.6%, ascorbic acid 0.2%, fatty alcohol polyoxyethylene ether 0.1%, and deionized water: balance.

Images