High-thermal-conductivity aluminum alloy design method and system

Abstract

The invention discloses a high-thermal-conductivity aluminum alloy design method and system, and belongs to the technical field of metal material science. The method comprises the steps that S1, aluminum alloy component elements are set, and a plurality of individuals with different component element contents are generated to form a population; s2, obtaining a fitness score F of each individual in the population; s3, performing genetic manipulation on individuals in the population to generate a next-generation population; and S4, repeatedly executing the steps S2-S4 until a preset number of iterations is reached or the fitness score is converged to a preset target score. According to the method, the basic principle of material science and the optimization strategy of data driving are combined, and the LPBF aluminum alloy material which has excellent heat conductivity and avoids hot cracks can be efficiently designed. Compared with a traditional single machine learning method, a more accurate and reliable alloy design scheme is provided, the thermal conductivity of the aluminum alloy and the adaptability of the aluminum alloy in the LPBF process are remarkably improved, meanwhile, the material development period is shortened, the cost is reduced, and the design accuracy and practicability are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of metal materials science, and particularly relates to a design method and system for high thermal conductivity aluminum alloy. Background Art

[0002] In the modern industrial field, especially in the automotive and aerospace industries, the demand for efficient thermal management systems is increasing. These systems require materials to rapidly and effectively dissipate heat while maintaining structural integrity, so as to improve the overall performance and reliability of the system. Aluminum alloy has become an ideal choice for thermal management applications due to its advantages such as high thermal conductivity, low density, and high specific strength.

[0003] Laser Powder Bed Fusion (LPBF) technology can achieve the integrated design of complex structures and improve the heat dissipation efficiency, providing technical support for the production of thermal management components. However, the rapid solidification characteristics in the LPBF process often lead to an increase in the thermal crack sensitivity of aluminum alloy, affecting the structural integrity and thermal conductivity of components. Although the generation of thermal cracks can be suppressed to a certain extent by optimizing the LPBF process parameters, the fundamental way to solve this problem is to develop new alloy systems. However, there is an inherent contradiction between thermal conductivity and strength in the performance design of aluminum alloy, because although alloying increases the strength of aluminum, it also increases electron and phonon scattering, significantly reducing the thermal conductivity.

[0004] Traditional alloy design relies on experience, which is time-consuming and costly. Computational alloy design, especially the Calculation of Phase Diagrams (CALPHAD) method based on thermodynamic databases, can effectively predict the grain refinement behavior and thermal cracking tendency of LPBF aluminum alloy, etc., but there are still limitations in dealing with complex non-linear problems and multi-objective optimization problems. The development of Machine Learning (ML) technology provides a new way for material property prediction and multi-objective optimization, but simply relying on data-driven ML methods may ignore the basic theories in materials science, resulting in prediction results deviating from actual requirements. Therefore, combining machine learning with the basic principles of materials science to construct a comprehensive material design framework is becoming a rapidly developing research direction. For example, through the deep combination of high-throughput thermodynamic calculations and ML technology, Materials Genome Engineering has successfully developed high-performance high-entropy alloys, copper alloys, and perovskite batteries, etc. However, so far, there is no technology that integrates high-throughput CALPHAD and ML technology for optimizing the thermal conductivity, printability, and strength of LPBF aluminum alloy.

[0005] For example, Chinese Patent Application No. 202210612466.1 in the prior art discloses a data-driven aluminum alloy composition design method. The data-driven aluminum alloy composition design method includes: characterizing and analyzing the alloy composition and performance parameters of the aluminum alloy formed in each micro-region, establishing a database, where the database includes the alloy composition of the samples in each micro-region, the corresponding preparation parameters, and the corresponding performance parameters; obtaining the samples in the database, using the alloy composition and preparation parameters as inputs and the corresponding performance parameters as outputs to train an artificial neural network model; using a genetic algorithm for intelligent optimization, using the alloy composition and preparation parameters of the aluminum alloy as population individuals, using the target performance as the optimization target, calling the trained artificial neural network model to obtain the performance parameters under different alloy compositions and preparation parameters and calculating the corresponding individual fitness, and after genetic evolution, outputting the alloy composition and preparation parameters whose individual fitness meets the requirements.

[0006] In the heating environment in the above prior art, controlled gradient heating is adopted (such as solution treatment at 350°C - 550°C and aging treatment at 50°C - 250°C), and different temperature distributions are achieved by independently adjusting the power of the heating rods and the cooling water channels. The purpose of using this method in this patent is to more efficiently obtain a large amount of effective alloy data (such as the volume fraction of MgZn2 precipitation phase, electrical conductivity, and other properties), and then use artificial neural networks and genetic algorithms to design the alloy composition. However, since there are no hot cracks in traditional cast / forged aluminum alloys, the above technology also cannot alleviate the hot crack problem existing in aluminum alloys in the LPBF process. Summary of the Invention

[0007] The purpose of the present invention is to provide a high thermal conductivity aluminum alloy design method and system, which partially solve or alleviate the problems of poor printability and the difficulty in balancing thermal conductivity and strength existing in aluminum alloys in the LPBF process in the prior art.

[0008] To solve the above-mentioned technical problems, the present invention specifically adopts the following technical solutions: A high thermal conductivity aluminum alloy design method includes: S1 Set the aluminum alloy composition elements and generate several individuals with different content of composition elements to form a population; S2 Obtain the fitness score F of each individual in the population, including: S21 When the theoretical thermal conductivity λ of the individual is lower than the thermal conductivity threshold, assign a fitness score F worse than the score threshold to this individual; or, When no heterogeneous nucleation phase is formed in the individual during simulated solidification in the initial stage of solidification, assign a fitness score F worse than the score threshold to this individual; When the individual's theoretical thermal conductivity λ is higher than or equal to the thermal conductivity threshold and a heterogeneous nucleation phase is formed, calculate the fitness score F of the individual based on the grain refinement and eutectic solidification metallurgical indicators; the grain refinement indicators include the initial slope and the initial solidification interval, and the eutectic solidification metallurgical indicators include the brittle solidification range and the crack sensitivity factor; S3 Perform genetic operations on the individuals in the population to generate the next generation population; S4 Repeat steps S2 to S4 until the preset number of iterations is reached or the fitness score converges to the preset target score.

[0009] As an improvement, the method for calculating the fitness score F of the individual based on the grain refinement and eutectic solidification metallurgical indicators includes using the formula: ; Calculate the fitness score; where F is the fitness score, IS is the initial slope, IS ref is the slope reference value, ΔT IFR is the initial solidification interval, ΔT IFR,ref is the solidification interval reference value, ΔT BTR is the brittle solidification range, ΔT BTR,ref is the brittle solidification range reference value, CSI is the crack sensitivity factor, CSI ref is the crack sensitivity factor reference value, ω IS 、ω dt,IFR 、ω dt,BTR and ω CSI are weight coefficients.

[0010] As an improvement, the calculation method of the initial slope is to use the formula: ; Calculate the initial slope; where IS is the initial slope, f s is the solid fraction, and T is the temperature.

[0011] As an improvement, the calculation method of the initial solidification interval includes using the formula: ; Calculate the initial solidification interval; where, ΔT IFR is the initial solidification interval, T fs1 and T fs2 are the temperatures corresponding to the alloy solidifying to a certain solid fraction at the initial stage of solidification (0 ≤ f s ≤ 0.4), respectively. Specifically, the calculation range of ΔT IFR can be flexibly set according to the performance target.

[0012] As an improvement, the calculation method of the brittle solidification range includes using the formula: ; Calculate the brittle solidification range; where, ΔT BTR is the brittle solidification range, T ZST is the zero-strength temperature, and T ZDT is the zero-ductility temperature.

[0013] As an improvement, the calculation method of the crack sensitivity factor includes using the formula: ; Calculate the crack sensitivity factor; where, CSI is the crack sensitivity factor, T is the temperature, and f s 1 / 2 is the square root of the solid fraction.

[0014] As an improvement, the slope reference value and the solidification interval reference value are derived from Scalmalloy alloy; the brittle solidification range reference value and the crack sensitivity factor reference value are derived from AlSi 10 Mg alloy.

[0015] As an improvement, the aluminum alloy composition elements include aluminum and at least one of rare earth elements, magnesium, silicon, iron, zinc, titanium, zirconium, manganese, copper, nickel, chromium, yttrium, molybdenum, and vanadium.

[0016] As an improvement, the steps of generating a number of individuals with different content ratios of composition elements include: Set the content range of each composition element in the aluminum alloy; Generate a number of individuals within the content range according to a preset content step size.

[0017] The present invention also provides a high thermal conductivity aluminum alloy design system, including: An initialization module, used to set the aluminum alloy composition elements and generate a population composed of a number of individuals with different content ratios of composition elements; A fitness score acquisition module, used to obtain the fitness score F of each individual in the population, including: S21 When the theoretical thermal conductivity λ of an individual is lower than the thermal conductivity threshold, assign a fitness score F worse than the score threshold to this individual; When in the simulated solidification, an individual does not form a heterogeneous nucleation phase for a long time during solidification, assign a fitness score F worse than the score threshold to this individual; When the individual's theoretical thermal conductivity λ is higher than or equal to the thermal conductivity threshold and a heterogeneous nucleation phase is formed, calculate the fitness score F of the individual based on the grain refinement and eutectic solidification metallurgical indexes; the grain refinement indexes include the initial slope and the initial solidification interval, and the eutectic solidification metallurgical indexes include the brittle solidification range and the crack sensitivity factor; A mutation and inheritance module, configured to perform genetic operations on individuals within a population to generate a next-generation population; An iteration module, configured to iterate until a preset number of iterations is reached or the fitness score converges to a preset target score.

[0018] Beneficial effects: Through the integration of high-throughput CALPHAD calculation and multi-objective genetic algorithm technology, the present invention develops a high-thermal-conductivity and crack-free aluminum alloy design method applicable to the LPBF process, and this method has the following technical effects: 1. Optimize the balance between thermal conductivity and strength: By precisely optimizing the alloy composition, the present invention can significantly improve the printability and strength of aluminum alloys while maintaining their high thermal conductivity, overcoming the problem of mutual restriction between thermal conductivity and strength in traditional aluminum alloy design.

[0019] 2. Reduce the generation of thermal cracks: Using Scheil-Gulliver solidification simulation, optimize the grain refinement performance and eutectic solidification effect of aluminum alloys, making the design results more in line with the principles of physical metallurgy, effectively reducing the generation of thermal cracks in the LPBF process, and improving the structural integrity of the alloy.

[0020] 3. Improve the alloy design efficiency: Through the combination of high-throughput CALPHAD multi-objective genetic algorithm, the present invention realizes rapid and efficient aluminum alloy composition design, avoiding the time and economic costs of traditional trial-and-error methods, and greatly improving the efficiency and accuracy of the design process.

[0021] 4. Wide applicability and scalability: The method of the present invention is applicable to the development of multi-element (such as ternary, quaternary, quinary, etc.) aluminum alloy systems, and can meet the optimization design requirements of different alloy element combinations, with strong adaptability and scalability.

[0022] Compared with the prior art, by setting the grain refinement ability and eutectic solidification effect as constraints, the present invention takes into account the problem of thermal cracks while ensuring the strength of aluminum alloys. Description of the Drawings

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or in the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. In all the drawings, similar elements or parts are generally denoted by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale. Obviously, the following-described drawings are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0024] Figure 1 It is the flowchart of Embodiment 1 of the present invention; Figure 2 It is for the comparison of the Scheil-Gulliver solidification behavior of the Al-1.03Fe-0.39Zr alloy in Example 1 with that of the alloy in Comparative Example 1; where Figure 2 (a) is the cheil-Gulliver solidification behavior of the alloy in Example 1, Figure 2 (b) is the cheil-Gulliver solidification behavior of the alloy in Comparative Example 1; Figure 3 It is the Scheil-Gulliver solidification behavior of the Al-2.02Ni-0.13Sc-0.52Zr alloy in Example 2; Figure 4 It is the Scheil-Gulliver solidification behavior of the Al-0.45Mg-0.28Si-0.59Sc-0.56Zr alloy in Example 3; Figure 5 It is the structural diagram of Embodiment 2. Detailed implementation manners

[0025] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0026] In this article, suffixes such as "module", "component" or "unit" used to represent elements are only for the convenience of the description of the present invention and have no specific meaning in themselves. Therefore, "module", "component" or "unit" can be used interchangeably.

[0027] In this text, the orientation or positional relationship indicated by terms such as "upper", "lower", "inner", "outer", "front", "rear", "one end", "the other end", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be construed as a limitation on the present invention. In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0028] In this text, unless otherwise clearly defined and limited, terms such as "installed", "provided with", "connected", etc. shall be understood in a broad sense. For example, "connected" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection, a direct connection, or an indirect connection through an intermediate medium, and can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0029] In this text, "and / or" includes any and all combinations of one or more of the listed related items.

[0030] In this text, "a plurality of" means two or more, that is, it includes two, three, four, five, etc.

[0031] Example 1: To solve the problems of poor printability and difficulty in balancing thermal conductivity and strength of existing aluminum alloys in the LPBF process, the present invention provides an alloy design method that combines high-throughput CALPHAD calculations with a multi-objective genetic algorithm. This method combines the basic principles of materials science with data-driven optimization strategies and can efficiently design LPBF aluminum alloy materials that not only have excellent thermal conductivity but also avoid hot cracks. Compared with traditional single machine learning methods, the present invention provides a more accurate and reliable alloy design scheme, significantly improving the thermal conductivity of aluminum alloys and their adaptability in the LPBF process, while shortening the material development cycle, reducing costs, and improving the accuracy and practicality of the design.

[0032] As Figure 1 shown, the specific steps of this embodiment include: S1 Set the elemental composition of the aluminum alloy and generate a population consisting of several individuals with different elemental composition contents.

[0033] In the present invention, the setting of the composition elements of the aluminum alloy is carried out within the established alloy element system range. In addition to aluminum, the composition elements of the aluminum alloy may further include at least one of magnesium (Mg), silicon (Si), iron (Fe), zinc (Zn), titanium (Ti), zirconium (Zr), manganese (Mn), copper (Cu), nickel (Ni), chromium (Cr), yttrium (Y), molybdenum (Mo), vanadium (V), and rare earth elements (such as Sc, La, Ce, Nd, Sm).

[0034] For example, if the aluminum alloy system includes elements such as aluminum (Al), copper (Cu), magnesium (Mg), zinc (Zn), zirconium (Zr), etc., then during initialization, the computer program will randomly assign the mass or volume percentage of these elements to each individual according to the preset content range of each element. By setting the allocation step size, the range of the number of individuals N can be adjusted. That is, first, set the content range of each composition element in the aluminum alloy, and then generate a number of individuals within the content range according to the preset content step size.

[0035] Taking a simple ternary aluminum alloy system (Al-Fe-Zr) as an example for specific illustration. Set the content range of the Fe element to 0 - 2 wt%; Zr, as a trace alloy element, set its content range to 0 - 1 wt%. The alloy composition step size is 0.01wt%, that is, a population with 200 * 100 = 20000 individual numbers can be generated.

[0036] S2 Obtain the fitness score F of each individual in the population.

[0037] For each genetic individual, calculate the fitness score F according to the preset constraints and objectives. The fitness score F reflects the advantages and disadvantages of the alloy under multiple performance indicators (such as grain refinement ability, eutectic solidification effect). Specifically, it includes the following sub-steps: S21 Calculate the theoretical thermal conductivity λ of the individual. In the case where the theoretical thermal conductivity λ of the individual is lower than the thermal conductivity threshold, assign a fitness score F worse than the score threshold to this individual.

[0038] Specifically, use thermodynamic calculation software such as Thermo-calc, FactSage, MSE, or any other thermodynamic calculation software with similar functions to run a single-point equilibrium calculation, calculate the theoretical thermal conductivity λ of each individual, and compare it with the preset standard value. If λ is lower than the standard value, assign a poor value to the fitness score F.

[0039] In the present invention, the so-called "assign a fitness score F worse than the score threshold to this individual" means that in the case where the higher the fitness score F, the better, assign a value lower than the score threshold to the fitness score F of this individual. And in the case where the lower the fitness score F, the better, assign a value higher than the score threshold to the fitness score F of this individual.

[0040] During the population evolution process of the genetic algorithm, the fitness score is an important indicator for measuring the quality of individuals. Assigning a poor F value to individuals that do not meet the requirements helps to eliminate those individuals that perform poorly in terms of thermal conductivity, solidification characteristics, etc. in the selection operation, making it more likely for individuals with better performance potential to be retained and inherited to the next generation, thereby improving the overall quality of the population and accelerating the speed at which the algorithm converges to the optimal solution.

[0041] Similarly, to judge the heterogeneous nucleation condition, when an individual does not form a heterogeneous nucleation phase in the early stage of solidification in the simulation of solidification, assign a fitness score F to this individual that is worse than the score threshold.

[0042] Specifically, according to the Scheil-Gulliver solidification simulation results of the thermodynamic calculation software, judge whether each individual first forms a heterogeneous nucleation phase Al3X (where X is any element in the designed alloy that can form a heterogeneous nucleation phase, such as Zr, Sc, Ti, etc.) in the early stage of solidification. If this phase is not formed, assign a poor value to the fitness score F.

[0043] It can be foreseen that for the two constraint conditions: 1) the theoretical thermal conductivity λ of the individual is higher than or equal to the thermal conductivity threshold; 2) any one of the heterogeneous nucleation phases is not formed by the individual in the early stage of solidification in the simulation of solidification, a poor fitness score F will be directly assigned to the corresponding individual. Correspondingly, the individual assigned this fitness score will very likely not be screened in the subsequent process. And the order of calculating the theoretical thermal conductivity λ of the individual and judging the heterogeneous nucleation condition can be interchanged. For example, when it is judged that the theoretical thermal conductivity λ of the individual is higher than or equal to the thermal conductivity threshold, and then it is judged that the individual does not form a heterogeneous nucleation phase in the early stage of solidification in the simulation of solidification, a fitness score F that is worse than the score threshold will also be assigned to this individual. Or, when it is judged that the individual forms a heterogeneous nucleation phase in the early stage of solidification in the simulation of solidification, but it is judged that the theoretical thermal conductivity λ of the individual is lower than the thermal conductivity threshold, a fitness score F that is worse than the score threshold will also be assigned to this individual.

[0044] S22 When the theoretical thermal conductivity λ of the individual is higher than or equal to the thermal conductivity threshold and a heterogeneous nucleation phase is formed in the early stage of solidification, calculate the fitness score F of this individual based on the grain refinement and eutectic solidification metallurgical indexes.

[0045] When both the theoretical thermal conductivity λ of the individual and the judgment of the heterogeneous nucleation condition meet the requirements, calculate the performance index of the individual, that is, according to the Scheil-Gulliver solidification simulation results of the thermodynamic calculation software, obtain the temperature-solid fraction (T-f s ) curve and the square root of the temperature-solid fraction (T-f s 1 / 2The curve is used to calculate the grain refinement and eutectic solidification metallurgical indexes of each individual. Among them, the grain refinement indexes include the initial slope and the initial solidification range, and the eutectic solidification metallurgical indexes include the brittle solidification range and the crack sensitivity factor.

[0046] During the solidification process of the alloy, the initial slope IS, which is one of the grain refinement indexes, reflects the initial rate (r) of constitutional supercooling. A larger IS value corresponds to a higher r value, indicating a faster nucleation rate, which helps to generate more dispersed phases, thereby inhibiting the growth of columnar crystals and promoting the formation of equiaxed crystals. In this embodiment, the calculation method of the initial slope is to use the formula: ; Calculate the initial slope; where IS is the initial slope of the T - f s curve, f s is the solid fraction, and T is the temperature. The present invention aims to increase the IS value, and by optimizing the alloy composition, the grain refinement effect is improved and crack generation is inhibited.

[0047] The initial solidification range ΔT, which is the second grain refinement index IFR is defined as the temperature range of the alloy during the primary phase formation stage. The larger the solidification range, the higher the supercooling degree of the alloy, thus promoting grain refinement. In this embodiment, the calculation method of the initial solidification range includes using the formula: ; Calculate the initial solidification range; where ΔT IFR is the initial solidification range, T fs1 and T fs2 are the temperatures corresponding to the alloy solidifying to a certain solid fraction at the initial stage of solidification respectively, and its calculation range can be flexibly set according to the performance target.

[0048] Specifically, the definition of the initial solidification range ΔT IFR is adjustable within the range of 0 ≤ f s ≤ 0.4 at the initial stage of solidification, where the better range is 0 ≤ f s ≤ 0.2, and the recommended range is 0 ≤ f s ≤ 0.1 to optimize the grain refinement effect of the alloy.

[0049] The brittle solidification range ΔT, which is one of the eutectic solidification metallurgical indexes BTR is the temperature range that measures the transition of the material from toughness to brittleness, which is usually most significant at the end of solidification. The larger ΔT BTR , the higher the hot cracking tendency of the material. Therefore, it is crucial to reduce ΔT BTR by optimizing the alloy composition for improving printability. In this embodiment, the calculation method of the brittle solidification range includes using the formula: ; Calculate the brittle solidification range; where, ΔT BTR is the brittle solidification range, T ZST is the zero strength temperature, T ZDT is the zero ductility temperature.

[0050] The definition of the brittle solidification range ΔT BTR is adjustable within the range of 0.6 ≤ f s ≤ 1 at the end of solidification, where the preferred range is 0.7 ≤ f s ≤ 1, and the recommended range is 0.85 ≤ f s ≤ 0.95 to optimize the eutectic solidification behavior of the alloy.

[0051] The crack sensitivity factor CSI, which is the second eutectic solidification metallurgical index, is a key index for evaluating the hot cracking tendency of materials, mainly reflecting the balance between the lateral grain growth rate and the liquid feeding capacity during solidification. When the liquid feeding rate is lower than the grain bridging rate, the material is more prone to cracking. In this embodiment, the calculation method of the crack sensitivity factor includes using the formula: ; Calculate the crack sensitivity factor; where, CSI is the crack sensitivity factor, T is the temperature, f s 1 / 2 is the square root of the solid fraction; d represents taking the differential, and correspondingly, dT is the temperature gradient, reflecting the temperature change rate at the solidification front; is the differential of the square root of the solid fraction, characterizing the rate of the solidification process.

[0052] After obtaining the grain refinement index and the eutectic solidification metallurgical index, calculate the fitness score F of this individual. In this implementation, a linear weighting method is used to convert the grain refinement index and the eutectic solidification metallurgical index into the fitness score F. The fitness score F is such that the smaller F is, the better the individual. The specific calculation method includes using the formula: ; Calculate the fitness score; where, F is the fitness score, IS is the initial slope, IS ref is the slope reference value, ΔT IFR is the initial solidification interval, ΔT IFR,ref is the solidification interval reference value, ΔT BTR is the brittle solidification range, ΔT BTR,ref is the brittle solidification range reference value, CSI is the crack sensitivity factor, CSI ref is the crack sensitivity factor reference value, ω IS 、ω dt,IFR 、ω dt,BTR and ωCSI is the weight coefficient. The weight coefficients ω IS , ω dt,IFR , ω dt,BTR and ω CSI of the fitness score F can be dynamically adjusted according to the target properties of the alloy to optimize the performance of the alloy in specific applications. For the characteristics of typical or atypical "grain refinement - eutectic solidification" systems, the magnitudes of the above four weight coefficients can be adjusted according to actual needs in practical applications. For example, when it is necessary to focus on optimizing " grain refinement ” this characteristic, the corresponding weight coefficients ω IS , ω dt,IFR are greater than the weight coefficients ω dt,BTR , ω CSI ; if it is necessary to focus on optimizing the "eutectic solidification" characteristic, the weight coefficients ω IS , ω dt,IFR are less than the weight coefficients ω dt,BTR , ω CSI。

[0053] The reference values of the grain refinement and eutectic solidification can be selected from existing alloy systems (such as Scalmalloy alloy and AlSi 10 Mg alloy), but other alloys with similar grain refinement and eutectic solidification characteristics can also be selected as references to ensure that the designed alloy can achieve optimized grain refinement effects and eutectic solidification behaviors, thereby improving its adaptability in the LPBF process.

[0054] S3 performs genetic operations on the individuals within the population to generate the next generation population.

[0055] Specifically, the selection, crossover, and mutation genetic operations are sequentially performed to generate the next generation population, enabling the population to continuously evolve. Among them, the selection operation adopts a combination of the elite strategy and the random selection strategy to avoid the search from prematurely falling into a local optimum. The crossover operation adopts the random crossover method, where a part is intercepted from each of the parental gene segments and spliced to generate new offspring individuals. The crossover operation can effectively maintain the diversity of the population and increase the ability of the genetic algorithm to search for the global optimum. To enhance the local optimization ability, gene mutation is appropriately performed.

[0056] S4 Repeatedly execute steps S2~S4 until the preset number of iterations is reached or the fitness score converges to the preset target score.

[0057] After each iteration, it is judged whether the stopping condition is satisfied. If the preset number of generations is reached or the fitness score of the population converges to a certain target score, the optimization process is stopped and the final optimal alloy composition is output; otherwise, steps S2~S4 are continued.

[0058] In this embodiment, the genetic algorithm is used to find the printable aluminum alloy composition and its optimal ratio range on the basis of high thermal conductivity, while taking into account the problem of hot crack sensitivity, so as to ensure appropriate strength on the basis of its high thermal conductivity, and prevent its strength from being too low due to the generation of cracks, and then obtain an aluminum alloy with high thermal conductivity and appropriate strength.

[0059] The present invention will be described below through three examples.

[0060] Example 1: In this example, a ternary Al-Fe-Zr alloy is selected as the design object. This alloy has the characteristics of a typical "grain refinement - eutectic solidification" system. In setting the alloy composition, the content range of the Fe element is set to 0 - 2 wt%; Zr is used as a trace alloy element, and its content range is set to 0 - 1 wt%. The alloy composition step size is 0.01 wt%.

[0061] S1 Randomly generate N = 2000 individuals to form the initial population of the genetic algorithm.

[0062] S2 For each genetic individual, calculate the fitness score F according to the preset constraints and objectives. In this embodiment, the weights ω IS , ω dt,IFR , ω dt,BTR and ω CSI are set to 0.2, 0.2, 0.3 and 0.3.

[0063] S3 Selection, crossover and mutation: Select the top 20% of the individuals with the best fitness scores from the current population as the parental individuals, and at the same time randomly select 0.05 of the remaining individuals as the parental individuals to avoid premature search getting trapped in local optima. The crossover operation adopts a random crossover method to generate new offspring individuals. The gene mutation rate is set to 10%.

[0064] (4) Iterative optimization, judge the stop condition.

[0065] After 8 rounds of cyclic iteration, the finally obtained alloy composition is Al-1.03Fe-0.39Zr, and its Scheil-Gulliver solidification behavior is as Figure 2 shown in (a): During the solidification process, as the temperature decreases, the aluminum alloy liquid gradually forms Al3Zr (as the heterogeneous nucleation phase), FCC and Al 13 Fe4. Specifically, as Figure 2During the cooling stage from the medium gray line to the red line in (a), as the temperature decreases, some aluminum alloy liquid metal gradually forms Al3Zr (which is a heterogeneous nucleation phase); as the temperature further decreases, some aluminum alloy liquid metal forms Al3Zr (heterogeneous nucleation phase) and FCC aluminum (i.e., aluminum crystal with face-centered cubic (fcc) structure) respectively; as shown in Figure 2 the blue line in (a), when the temperature further decreases, FCC aluminum continues to form; as shown in Figure 2 the yellow line in (a), when the temperature further decreases, on the basis of continuing to form FCC aluminum, some aluminum alloy liquid metal will gradually form Al 13 Fe4. For the Al-1.02Fe-1.05Zr alloy of Comparative Example 1 ( Figure 2 b) from doi.org / 10.1016 / j.actamat.2023.119199, after direct aging treatment, this alloy exhibits excellent yield strength (310 MPa) and thermal conductivity (180 W / m·K), which is a typical example of optimizing strength-thermal conductivity in current LPBF aluminum alloys. As shown in Figure 2 As shown in (b), the solidification curves of the Al-1.02Fe-1.05Zr alloy are all in the shape of "L": as shown in Figure 2 from the medium gray line to the red line in (b), as the temperature decreases, some aluminum alloy liquid metal gradually forms Al3Zr (heterogeneous nucleation phase); as shown in Figure 2 the blue line in (b), as the temperature further decreases, some aluminum alloy liquid metal will gradually form FCC aluminum; as shown in Figure 2 the yellow line in (b), when the temperature further decreases, in addition to forming FCC aluminum, Al 13 Fe4 will also gradually form. The results show that the solidification curves of the alloy in Example 1 and Comparative Example 1 are both in the shape of "L", indicating that their grain refinement and eutectic solidification effects are similar.

[0066] Table 1 Comparison of indexes and thermal conductivity between the designed alloy Al-1.03Fe-0.39Zr and the literature alloy Al-1.02Fe-1.05Zr Table 1 lists the various indexes of the alloy and the calculation results of thermal conductivity. The results show that the fitness score F of the alloy in Example 1 is lower than that of the alloy in Comparative Example 1, indicating that the designed alloy is superior in the comprehensive mechanism of "grain refinement - eutectic solidification". At the same time, the theoretical thermal conductivity of the alloy in Example 1 is 232.7 W·m -1 ·K -1 , which is higher than 231.7 W·m -1 ·K -1 of the alloy in Comparative Example 1. In addition, the alloy elements content of the alloy in Example 1 is lower, which has a greater cost advantage. The above results verify the advantages of the method proposed by the present invention from multiple dimensions.

[0067] Example 2: In this example, a quaternary Al-Ni-Sc-Zr alloy is selected as the design object. This alloy also has the characteristics of a typical "grain refinement - eutectic solidification" system. The difference from Example 1 is that its elemental types are more abundant. In setting the alloy composition, the content range of Ni element is set to 0 - 7 wt%; the contents of Sc and Zr are set at 0 - 1 wt% to control costs. The alloy composition step is 0.01 wt%.

[0068] This example includes the following steps: (1) Initialization: Randomly generate N = 2000 individuals to form the initial population of the genetic algorithm.

[0069] (2) Calculate the individual fitness score F: For each genetic individual, calculate the fitness score F according to the preset constraints and objectives. In this embodiment, the weights ω IS , ω dt,IFR , ω dt,BTR and ω CSI are respectively set to 0.2, 0.2, 0.3 and 0.3.

[0070] (3) Selection, crossover and mutation. Select the top 20% of the individuals with the best fitness scores from the current population as parental individuals, and at the same time randomly select from the remaining individuals as parental individuals with a probability of 0.05 to avoid premature search getting trapped in local optima. The crossover operation adopts a random crossover method to generate new offspring individuals. The gene mutation rate is set to 10%.

[0071] (4) Iterative optimization, judge the stopping condition.

[0072] After 12 rounds of cyclic iteration, the finally obtained alloy composition is Al-2.02Ni-0.13Sc-0.52Zr, and its Scheil-Gulliver solidification behavior is as Figure 3 shown. This curve also presents an "L" shape feature: as Figure 3 from the medium gray line to the red line, as the temperature decreases, part of the aluminum alloy liquid gradually forms Al3Zr (heterogeneous nucleation phase); as the temperature further decreases, on the basis of continuously forming Al3Zr, FCC aluminum will gradually form; as Figure 3 the blue line, when the temperature further decreases, FCC aluminum continues to form in this cooling stage; as Figure 3 the yellow line, when the temperature further decreases, on the basis of continuously forming FCC aluminum, part of the aluminum alloy liquid will gradually form Al3Ni; as Figure 3 the purple line, when the temperature further decreases, on the basis of continuously forming the heterogeneous nucleation phases of FCC aluminum and Al3Ni, Al3Sc gradually forms.

[0073] Table 2 Comparison of indicators and properties between the designed alloy Al-2.02Ni-0.13Sc-0.52Zr and the reference alloy Table 2 lists the indicators of the alloy and the calculated results of thermal conductivity. The results show that the grain refinement index (initial slope IS 、 of the initial solidification interval ΔT IFR ） of the alloy in Example 2 is equivalent to that of Comparative Example 2. Compared with Comparative Example 3, the eutectic solidification index (brittle solidification range ΔT BTR , crack sensitivity factor CSI) of the alloy in Example 2 is significantly optimized, and it has stronger resistance to hot cracks. In addition, the thermal conductivity of the alloy in Example 2 λ is greater than that of Comparative Example 4, a common high thermal conductivity aluminum alloy. The above results verify the applicability of the method proposed by the present invention in a more complex quaternary system.

[0074] Example 3: Traditional 6063 aluminum alloy is famous for its high thermal conductivity, but it has the problem of being non-processable by LPBF. In this example, the Sc and Zr modified 6063 aluminum alloy (quinary Al-Mg-Si-Sc-Zr alloy) is designed as an example, and this alloy has the characteristics of an atypical "grain refinement - eutectic solidification" system. Compared with Example 1 and Example 2, it has more elements and a more complex design. In setting the alloy composition, according to the national standard requirements of traditional 6063 aluminum, the content range of Mg element is set to 0.45 - 0.9 wt%, and the content range of Si element is set to 0.2 - 0.6 wt%; the content ranges of Sc and Zr are 0 - 1 wt% to control the cost. The alloy composition step size is 0.01 wt%.

[0075] This example includes the following steps: (1) Initialization: Randomly generate N = 2000 individuals to form the initial population of the genetic algorithm.

[0076] (2) Calculate the individual fitness score F: For each genetic individual, calculate the fitness score F according to the preset constraints and objectives. Weights ω IS , ω dt,IFR , ω dt,BTR and ω CSI are respectively set to 0.3, 0.3, 0.2 and 0.2, focusing on optimizing the ability of grain refinement to inhibit cracks.

[0077] (3) Selection, crossover and mutation. Select the top 20% of the individuals with the best fitness scores in the current population as the parental individuals, and at the same time select individuals with relatively poor fitness scores with a probability of 0.05 to avoid premature convergence of the search to a local optimum. The crossover operation adopts a random crossover method to generate new offspring individuals. The gene mutation rate is set to 10%.

[0078] (4) Iterative optimization to determine the stopping condition.

[0079] After 8 rounds of cyclic iteration, the finally obtained alloy composition is Al-0.45Mg-0.28Si-0.59Sc-0.56Zr, and its Scheil-Gulliver solidification behavior is as Figure 4 shown: As Figure 4 from the medium gray line to the red line, as the temperature decreases, part of the aluminum alloy liquid metal gradually forms Al3Zr (heterogeneous nucleation phase); as the temperature further decreases, on the basis of forming Al3Zr, Al3Sc will gradually form; as Figure 4 the sky blue line in the figure, when the temperature further decreases, Al3Sc will continue to form; as Figure 4 the yellow line in the figure, when the temperature further decreases, on the basis of forming Al3Sc, FCC aluminum will gradually form; as Figure 4 the purple line in the figure, when the temperature further decreases, on the basis of continuing to form FCC aluminum, AlSc2Si2 will gradually form; as Figure 4 the blue-green line in the figure, when the temperature further decreases, on the basis of continuing to form FCC aluminum and AlSc2Si2, Mg2Si will gradually form; as Figure 4 the brown line in the figure, when the temperature gradually decreases to or approaches 550 °C, on the basis of continuing to form FCC aluminum, AlSc2Si2, and Mg2Si, Si will gradually form. The results show that no typical eutectic solidification reaction occurs in this system, resulting in a relatively steep downward trend in the solidification end curve. Compared with the traditional 6063 aluminum alloy (Comparative Example 4), it is expected that the in-situ precipitated Al3(Sc,Zr) nanophase in this alloy can inhibit the generation of cracks.

[0080] Table 3 Comparison of indicators and properties between the new Al-0.45Mg-0.28Si-0.59Sc-0.56Zr alloy and 6063 alloy Table 3 lists the various indicators of the alloy and the calculation results of the thermal conductivity. It can be seen from Table 3 that compared with Comparative Example 4, Example 3 has been significantly improved in terms of grain refinement, and at the same time, the eutectic solidification index has decreased, indicating that the solidification curve has been significantly optimized. In addition, the thermal conductivity of Example 3 is still higher than that of the common high-thermal-conductivity aluminum alloy Comparative Example 4. The above results fully illustrate that the method proposed by the present invention has strong adaptability and scalability.

[0081] Example Two: As Figure 5 shown, the present invention also provides a high-thermal-conductivity aluminum alloy design system, including: An initialization module for setting the elemental composition of aluminum alloy and generating a population consisting of a number of individuals with different elemental composition contents; A fitness score acquisition module for obtaining the fitness score F of each individual in the population, including: S21 When the theoretical thermal conductivity λ of an individual is lower than the thermal conductivity threshold, assign a fitness score F worse than the score threshold to this individual; When an individual does not form a heterogeneous nucleation phase during the long-term solidification in the simulation of solidification, assign a fitness score F worse than the score threshold to this individual; S22 When the theoretical thermal conductivity λ of an individual is higher than or equal to the thermal conductivity threshold and a heterogeneous nucleation phase is formed, calculate the fitness score F of this individual based on the grain refinement and eutectic solidification metallurgical indexes; the grain refinement indexes include the initial slope and the initial solidification interval, and the eutectic solidification metallurgical indexes include the brittle solidification range and the crack sensitivity factor; A mutation and inheritance module for performing genetic operations on the individuals in the population to generate the next generation population; An iteration module for iterating until a preset number of iterations is reached or the fitness score converges to a preset target score.

[0082] It should be noted that in this article, the term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including a..." does not exclude the existence of additional identical elements in the process, method, article or device including that element.

[0083] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-described embodiment methods can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions for causing a computer terminal (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0084] The embodiments of the present invention have been described above in conjunction with the accompanying drawings. However, the present invention is not limited to the above specific embodiments. The above specific embodiments are merely illustrative rather than restrictive. Under the inspiration of the present invention, those of ordinary skill in the art can also make many forms without departing from the spirit of the present invention and the scope protected by the claims. All of these fall within the protection scope of the present invention.

Claims

1. A method for designing a high thermal conductivity aluminum alloy, characterized in that: include: S1 sets the aluminum alloy component elements and generates a number of individuals with different component element contents to form a population; S2 obtains the fitness score F of each individual in the population, including: S21, when the individual theoretical thermal conductivity λ is lower than the thermal conductivity threshold, assigning the individual a fitness score F that is lower than the score threshold; or, In the case that an individual does not form a heterogeneous nucleation phase at the initial stage of solidification during the simulated solidification, a fitness score F that is worse than the score threshold is assigned to the individual; S22, when the individual theoretical thermal conductivity λ is higher than or equal to the thermal conductivity threshold and a heterogeneous nucleation phase is formed, the fitness score F of the individual is calculated based on grain refinement and eutectic solidification metallurgical indicators; the grain refinement indicators include the initial slope and the initial solidification range, and the eutectic solidification metallurgical indicators include the brittle solidification range and the crack sensitivity factor; S3 performs genetic operations on individuals in the population to generate the next generation of population; S4 repeats steps S2 to S4 until a preset number of iterations is reached or the fitness score converges to a preset target score.

2. A method for designing a high thermal conductivity aluminum alloy according to claim 1, characterized in that: The method for calculating the fitness score F of the individual based on the grain refinement and eutectic solidification metallurgical indicators includes using the formula: ； Calculate the fitness score; where F is the fitness score, IS is the initial slope, and IS ref is the slope reference value, ΔT IFR is the initial solidification interval, ΔT IFR,ref is the solidification interval reference value, ΔT BTR is the brittle solidification range, ΔT BTR,ref is the reference value of the brittle solidification range, CSI is the crack sensitivity factor, CSI ref is the reference value of crack sensitivity factor, ω IS ,ω dt,IFR ,ω dt,BTR and CSI IS, ΔT IFR、 ΔT BTR、 The weight coefficient of CSI.

3. A method for designing a high thermal conductivity aluminum alloy according to claim 1, characterized in that The initial slope is calculated using the formula: ； Calculate the initial slope; where IS is the initial slope, f s is the solid fraction and T is the temperature.

4. A method for designing a high thermal conductivity aluminum alloy according to claim 2, characterized in that The calculation method of the initial solidification interval includes using the formula: ； Calculate the initial solidification interval; where ΔT IFR is the initial solidification interval, T fs1 and T fs2 They are the temperatures corresponding to the alloy solidifying to a certain solid phase fraction in the initial stage of solidification.

5. A method for designing a high thermal conductivity aluminum alloy according to claim 2, characterized in that The calculation method of the brittle solidification range includes using the formula: ； Calculate the brittle solidification range; where ΔT BTR is the brittle solidification range, T ZST is the zero intensity temperature, T ZDT is the zero ductility temperature.

6. A method for designing a high thermal conductivity aluminum alloy according to claim 2, characterized in that The calculation method of the crack sensitivity factor includes using the formula: ； Calculate the crack sensitivity factor; where CSI is the crack sensitivity factor, T is the temperature, and f s 1 / 2 is the square root of the solid fraction.

7. A method for designing a high thermal conductivity aluminum alloy according to claim 2, characterized in that: The slope reference value and solidification interval reference value are derived from Scalmalloy alloy; the brittle solidification range reference value and crack sensitivity factor reference value are derived from AlSi 10 Mg alloy.

8. A method for designing a high thermal conductivity aluminum alloy according to claim 1, characterized in that: The aluminum alloy component elements include aluminum and at least one of rare earth elements, magnesium, silicon, iron, zinc, titanium, zirconium, manganese, copper, nickel, chromium, yttrium, molybdenum, and vanadium.

9. A method for designing a high thermal conductivity aluminum alloy according to claim 1, characterized in that The steps of generating several individuals with different proportions of component elements include: Set the content range of each component element in the aluminum alloy; Generate several individuals within the content range according to the preset content step size.

10. A high thermal conductivity aluminum alloy design system, characterized in that: include: The initialization module is used to set the aluminum alloy component elements and generate a number of individuals with different component element contents to form a population; The fitness score acquisition module is used to obtain the fitness score F of each individual in the population, including: S21, when the individual theoretical thermal conductivity λ is lower than the thermal conductivity threshold, assigning the individual a fitness score F that is lower than the score threshold; In the case that an individual does not form a heterogeneous nucleation phase at the initial stage of solidification during the simulated solidification, a fitness score F that is worse than the score threshold is assigned to the individual; S22, when the individual theoretical thermal conductivity λ is higher than or equal to the thermal conductivity threshold and a heterogeneous nucleation phase is formed, the fitness score F of the individual is calculated based on grain refinement and eutectic solidification metallurgical indicators; the grain refinement indicators include the initial slope and the initial solidification range, and the eutectic solidification metallurgical indicators include the brittle solidification range and the crack sensitivity factor; The mutation genetic module is used to perform genetic operations on individuals in the population to generate the next generation of population; The iteration module is used to iterate until a preset number of iterations is reached or the fitness score converges to a preset target score.

Images