A lightweight method for laser selective melting forming titanium alloy gradient lattice structure

Abstract

The application discloses a lightweight method for laser selective melting forming of a titanium alloy gradient lattice structure, relates to the technical field of titanium alloy gradient lattice structure manufacturing, and comprises the following steps: constructing a titanium alloy gradient lattice database; based on the data obtained from the database, an intelligent optimization model is established by using a generative adversarial network and reinforcement learning; the training target of the intelligent optimization model is to generate a feasible structure meeting the manufacturing requirements of SLM (selective laser melting); the lattice forming quality is optimized based on SLM process parameters, and the generated intelligent lattice structure is globally optimized by using a multi-objective optimization algorithm. The application combines the generative adversarial network and the reinforcement learning, accurately simulates and optimizes the lattice structure under different working conditions. Through the reinforcement learning of the cell size adjustment, the maximum lightweight can be realized under the premise of ensuring the structural strength.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of manufacturing of titanium alloy gradient lattice structures, and in particular to a lightweight method for laser selective melting forming of titanium alloy gradient lattice structures. BACKGROUND

[0002] With the continuous development of laser selective melting (SLM) technology, the application of titanium alloy and its gradient lattice structure has attracted widespread attention in many fields. Titanium alloy materials have excellent mechanical properties, corrosion resistance and relatively light density, and have become an important material for high-performance components. However, the traditional processing method of titanium alloy has problems such as high cost, long manufacturing cycle and uneven local performance, especially when manufacturing complex structures, it is difficult to balance the strength and weight of the structure.

[0003] Gradient lattice structure (GLS) as a new type of structure can optimize its mechanical properties and weight by changing the lattice size and density of different regions, and is especially suitable for additive manufacturing technologies such as 3D printing. However, designing and optimizing such structures still faces several technical challenges:

[0004] How to design a gradient lattice structure that can withstand external forces and achieve lightweight through appropriate parameterized modeling methods, and traditional lattice optimization methods usually only design based on a single objective (such as minimum mass or maximum strength), lacking consideration of multiple indicators such as thermal stress accumulation and complexity of laser scanning path. SUMMARY

[0005] The purpose of this section is to summarize some aspects of the embodiments of the present application and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the abstract and title of the specification to avoid obscuring the purpose of this section, the abstract and the title, and such simplifications or omissions cannot be used to limit the scope of the present application.

[0006] In view of the problems existing in the prior art, the present application is proposed.

[0007] To solve the above technical problems, the present application provides the following technical solutions:

[0008] The method comprises the following steps:

[0009] Constructing a titanium alloy gradient lattice database, the database comprising: parameters of different lattice topological structures, mechanical properties, and effects of various process parameters on the final forming quality;

[0010] Based on the data obtained from the database, an intelligent optimization model is established using a generative adversarial network + reinforcement learning method;

[0011] Among them, the generative adversarial network is responsible for simulating lightweight design schemes under various working conditions; the reinforcement learning is responsible for adjusting the cell parameters, and the lattice is optimized under the premise of guaranteeing the structural strength and minimizing the weight to the greatest extent, and the training target of the intelligent optimization model is to generate a feasible structure meeting the manufacturing requirements of SLM (selective laser melting) ;

[0012] The SLM process parameters are used to optimize the lattice forming quality, and a multi-objective optimization algorithm is used to globally optimize the generated intelligent lattice structure.

[0013] As a preferred scheme of the lightweight method for the laser selective melting forming titanium alloy gradient lattice structure, the generative adversarial network comprises a generator and a discriminator , wherein the generator generates a potential lattice structure according to input noise, that is, random data z , and the discriminator outputs a probability value to determine whether the generated lattice structure is real according to the generated lattice structure ;

[0014] The optimization objective function is minimized and maximized to promote the continuous improvement of the generator and the discriminator, and finally the generator generates a realistic lattice structure, wherein the expression of the optimization objective function is:

[0015]

[0016] Among them, represents the output of the discriminator, and represents the output of the generator.

[0017] As a preferred scheme of the lightweight method for the laser selective melting forming titanium alloy gradient lattice structure, based on the lattice scheme generated by the generative adversarial network, the reinforcement learning is used to adjust the structure parameters, and the mass is minimized under the premise of guaranteeing the mechanical properties, and specifically:

[0018] S201: Definition: state is the current lattice structure parameter; state is the adjustment amount of the lattice structure parameter, and the reward function

[0019] ;

[0020] Among them, represents the constrained mechanical property; represents the limited weight condition, This represents the feasibility loss of the SLM lattice structure; , , These represent the corresponding weight coefficients;

[0021] S202: Optimize the policy using the policy gradient method Maximizing cumulative rewards can be represented as follows:

[0022]

[0023] in, This represents the discount factor, used to balance short-term and long-term returns.

[0024] As a preferred embodiment of the lightweight method for laser selective melting forming of titanium alloy gradient lattice structures described in this invention, wherein: the multi-objective optimization algorithm in step three is the NSGA-II algorithm; including:

[0025] S301: The multiple lightweight lattice structure design schemes obtained in step two are used as initial solutions;

[0026] S302: Calculate the performance of each solution across all objectives;

[0027] S303: The solutions are divided into multiple levels according to the relative superiority of the objective function, and the ParetoFront solution is selected;

[0028] S304: Calculate the crowding degree for each solution, that is, how dense the solution is in the Pareto Front, to avoid the algorithm getting trapped in local optima;

[0029] S305: Then, a new generation of solutions is generated through the mechanism of genetic algorithm. After multiple iterations of optimization, a set of Pareto optimal solutions is finally obtained.

[0030] As a preferred embodiment of the lightweight method for laser selective melting forming of titanium alloy gradient lattice structures described in this invention, wherein: the performance of each solution across all targets includes:

[0031] Calculate crystal lattice structure Manufacturing quality loss ; Calculate the crystal lattice structure Thermal stress accumulation due to laser scanning path ; Calculate the crystal lattice structure Complexity of laser scanning path Both should be as small as possible.

[0032] As a preferred scheme of the lightweight method for the laser selective melting forming titanium alloy gradient lattice structure, the calculation method of the crowding degree is the calculation of the crowding distance of the solution, and the calculation method is as follows:

[0033]

[0034] Wherein, M represents the number of initial solutions, 、 represents the value of the adjacent solution on the current target , represents the maximum value and the minimum value on the current target .

[0035] As a preferred scheme of the lightweight method for the laser selective melting forming titanium alloy gradient lattice structure, the final optimal lattice structure is tested, the test indexes include size accuracy test, mechanical property test, density detection and surface roughness measurement, the reliability of the optimization model is evaluated through the test, and if a certain index does not meet the standard, the feedback is fed back to the previous step for adjustment.

[0036] The system is applied to the lightweight method for the laser selective melting forming titanium alloy gradient lattice structure, and the system comprises:

[0037] A data collection module is configured to collect parameters of different lattice topological structures, mechanical properties and influence data of final forming quality under various process parameters, and construct a titanium alloy gradient lattice database;

[0038] An intelligent lattice generation and optimization module is configured to generate an optimized gradient lattice structure by an intelligent optimization model established by a generative adversarial network+reinforcement learning method;

[0039] An SLM process adaptation and optimization module is configured to calculate laser energy input, scanning path and scanning speed parameters in the SLM process, and ensure that the generated lattice structure has strong manufacturability;

[0040] A multi-objective optimization module is configured to adopt a multi-objective optimization algorithm to balance between multiple objectives and select the best lattice structure scheme;

[0041] And an intelligent verification and simulation module is configured to perform virtual testing on the finally optimized lattice structure by a multi-physics field simulation technology, so that the lattice structure meets the design requirements.

[0042] The application further discloses a computer device comprising a memory and a processor, the memory stores a computer program, and the processor realizes the steps of the lightweight method for the laser selective melting forming titanium alloy gradient lattice structure when the computer program is executed.

[0043] The application further discloses a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the steps of the method.

[0044] Advantages of the application:

[0045] The application adopts a combination of a generative adversarial network (GAN) and reinforcement learning (RL) to accurately simulate and optimize the lattice structure under different working conditions. The lattice size adjustment through reinforcement learning can ensure maximum lightweighting while ensuring structural strength. The optimization process takes into account the mechanical properties and lightweighting requirements of the structure in theory and practical application, which is a new design idea that cannot be provided by the prior art.

[0046] The application calculates the objective function values of individual manufacturing quality loss, thermal stress, scanning path optimization, reduces thermal stress generated by the laser scanning process, reduces the risk of deformation and cracks caused by thermal stress, ensures the stability of the structure, optimizes the manufacturing quality loss and the complexity of the laser scanning path, improves the manufacturing efficiency, reduces energy consumption, and improves the cost-effectiveness of production. BRIEF DESCRIPTION OF DRAWINGS

[0047] In order to more clearly illustrate the technical solutions of the embodiments of the application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the application, and for those skilled in the art, other drawings can be obtained without creative labor. Among them:

[0048] Fig. 1 A three-dimensional model construction diagram for the laser selective melting forming titanium alloy gradient lattice structure is provided for the application.

[0049] Fig. 2 A flowchart of the lightweighting method for the laser selective melting forming titanium alloy gradient lattice structure is provided for the application. DETAILED DESCRIPTION

[0050] In order to make the above-mentioned purposes, features and advantages of the application more apparent and easy to understand, the specific embodiments of the application will be described in detail below with reference to the drawings of the specification.

[0051] In the following description, many specific details are set forth in order to provide a thorough understanding of the application, but the application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the scope of the application. Therefore, the application is not limited to the specific embodiments disclosed below.

[0052] Second, the "one embodiment" or "an embodiment" described herein refers to a particular feature, structure, or characteristic included in at least one implementation of the disclosure. The appearances of "in one embodiment" or "in an embodiment" in various places in the specification are not necessarily all referring to the same embodiment, nor are they necessarily all referring to a particular feature, structure or characteristic that is necessarily included in every implementation of the disclosure.

[0053] Referring to Figs. 1-2 For one embodiment of the present application, a lightweight method for laser selective melting forming titanium alloy gradient lattice structure is provided, which comprises the following steps:

[0054] Step 1: Construct a titanium alloy gradient lattice database, which includes parameters of different lattice topologies, mechanical properties, and effects of various process parameters on the final forming quality;

[0055] Step 2: Based on the data obtained from the database, an intelligent optimization model is established using a generative adversarial network + reinforcement learning method.

[0056] Among them, the generative adversarial network is responsible for simulating lightweight design schemes under various working conditions. Specifically, the generative adversarial network includes a generator and a discriminator , wherein the generator generates a potential lattice structure according to the input noise, i.e., random data z , and the discriminator outputs a probability value to determine whether the generated lattice structure is real ;

[0057] By minimizing and maximizing the objective function , the generator and the discriminator are continuously improved, and the generator eventually generates realistic lattice structures. The expression of the optimization objective function is as follows:

[0058]

[0059] Among them, represents the output of the discriminator, and represents the output of the generator.

[0060] Reinforcement learning is responsible for adjusting the lattice parameters. Under the premise of ensuring the strength of the structure and minimizing the weight, the lattice is optimized. The training goal of the intelligent optimization model is to generate a feasible structure that meets the requirements of SLM (selective laser melting) manufacturing ; Specifically, based on the lattice scheme generated by the generative adversarial network, reinforcement learning is used to adjust the structure parameters, and the mass is minimized under the premise of ensuring the mechanical properties, specifically:

[0061] S201: Define: State is the current lattice structure parameter; State is the adjustment amount of the lattice structure parameter, reward function ;

[0062] Where, represents the constrained mechanical properties (such as loss of mechanical strength); represents the limited weight condition (such as the value of exceeding the target weight), represents the feasibility loss of the lattice structure SLM (such as manufacturing error); , , respectively represent the corresponding weight coefficients;

[0063] S202: Optimize the strategy by policy gradient method , maximize the cumulative reward, the process is represented as:

[0064]

[0065] Where, represents the discount factor, which is used to balance short-term and long-term returns.

[0066] Step three: optimize the lattice forming quality based on SLM process parameters, in this embodiment, the SLM forming gradient lattice process parameters are set as: laser power 210W, scanning speed 1200mm / s, no contour scanning in the scanning strategy, spot diameter 0.1mm, scanning spacing 0.12mm, powder layer thickness 0.03mm, steel scraper one-way powder laying, forming bin in the environment of high-purity argon as protective gas, under the process condition, the forming density can reach 99.96%, the surface roughness is about 8um, and a multi-objective optimization algorithm is used to globally optimize the generated intelligent lattice structure.

[0067] Specifically, the multi-objective optimization algorithm is NSGA-II algorithm; including:

[0068] S301: The multiple lightweight lattice structure design schemes obtained by step two are used as initial solutions.

[0069] S302: Calculate the performance of each solution on all targets, respectively:

[0070] Calculate the manufacturing quality loss of the lattice structure ;

[0071]

[0072] Where, represents the total number of lattice units; represents the first​ defect volume of a unit cell; total volume of a unit cell; total volume of a unit cell;

[0073] calculating the lattice structure accumulation of thermal stress caused by the laser scanning path

[0074]

[0075] wherein, Ω represents the volume of the lattice structure, represents the temperature gradient;

[0076] calculating the lattice structure complexity of the laser scanning path , are all the smaller the better.

[0077]

[0078] wherein, represents the number of scanning segments, represents the scanning length of the i-th segment, represents the energy density of the i-th segment. Assuming that step two generates three lightweight design schemes, then the three design schemes correspond to three candidate solutions as initial solutions.

[0079] S303: divide the solutions into multiple levels according to the relative superiority of the objective function; in terms of the above table data, calculate the objective function value of each solution, and perform non-dominated sorting to divide the solutions into multiple Pareto layers:

[0080] Layer 1: no solution can completely dominate them (optimal solution set);

[0081] Layer 2: dominated by the solutions in layer 1, but better than layer 3.

[0082] Layer 2: dominated by the solutions in layer 1, but better than layer 3.

[0083] S304: calculate the crowding degree of each solution, that is, the density of the solution in the Pareto Front, to avoid the algorithm falling into local optimum. The calculation method of the crowding degree is to calculate the crowding distance of the solution, and the calculation method is:

[0084]

[0085] wherein, M represents the number of initial solutions, , represent the values of adjacent solutions on the current objective , represent the maximum and minimum values on the current objective .​​​

[0086] S305: Then generate a new generation of solutions through the mechanism of genetic algorithm, and the genetic operation generally includes: selection: select good solutions from the Pareto layer, and the better solutions are more likely to be selected, crossover: randomly select two solutions, combine part of the parameters to generate new solutions, mutation: apply a small random disturbance to part of the parameters to increase the diversity of the population.

[0087] After multiple iterations of optimization: repeat non-dominated sorting → calculate congestion → selection, crossover and mutation until the algorithm converges, NSGA-II converges to a Pareto front (Pareto Front), that is, a set of optimal design schemes, from which the user can choose the most suitable scheme.

[0088] Test the final optimal lattice structure, test indicators include: size accuracy test, mechanical property test, density detection and surface roughness measurement, through the test, evaluate the reliability of the optimization model, if a certain index is not up to standard, feedback to the previous step for adjustment.

[0089] In summary, the present application adopts the combination of generative adversarial network (GAN) and reinforcement learning (RL) to accurately simulate and optimize the lattice structure under different working conditions. Through the adjustment of the unit cell size by reinforcement learning, the maximum lightweight can be realized under the premise of ensuring the structural strength. This optimization process takes into account the mechanical properties and lightweight requirements of the structure in theory and practical application, which is a new design idea that existing technologies cannot provide. The present application calculates the objective function values of individual manufacturing quality loss, thermal stress, scanning path optimization, etc., reduces the thermal stress generated by the laser scanning process, reduces the risk of deformation and cracking caused by thermal stress, and ensures the stability of the structure. It also optimizes the manufacturing quality loss and the complexity of the laser scanning path, improves the manufacturing efficiency, reduces energy consumption, and improves the cost-effectiveness of production.

[0090] The present embodiment also provides a system applied to a lightweight method for laser selective melting forming titanium alloy gradient lattice structure, comprising:

[0091] The data collection module collects parameters of different lattice topologies, mechanical properties, and influence data of various process parameters on the final forming quality, and constructs a titanium alloy gradient lattice database; the intelligent lattice generation and optimization module generates an optimized gradient lattice structure through an intelligent optimization model established by a generative adversarial network + reinforcement learning method; the SLM process adaptation and optimization module is used to calculate laser energy input, scanning path, and scanning speed parameters in the SLM process, to ensure that the generated lattice structure has strong manufacturability; the multi-objective optimization module adopts a multi-objective optimization algorithm to weigh among multiple objectives and select the best lattice structure scheme; and the intelligent verification and simulation module performs virtual testing on the finally optimized lattice structure through multi-physics field simulation technology, to ensure that the lattice structure meets the design requirements.

[0092] The embodiment also provides a computer device suitable for the case of the lightweight method for laser selective melting forming of a titanium alloy gradient lattice structure, including a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions to realize the lightweight method for laser selective melting forming of a titanium alloy gradient lattice structure as described in the above embodiment.

[0093] The computer device can be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner. The wireless manner can be achieved through WIFI, a carrier network, NFC (near field communication) or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device. In addition, the input device can also be an external keyboard, touchpad or mouse, etc.

[0094] The embodiment also provides a storage medium, which stores a computer program, and the computer program is executed by a processor to implement the lightweight method for forming a gradient lattice structure of a titanium alloy by laser selective melting as proposed in the above embodiment. The storage medium can be implemented by any type of volatile or non-volatile storage devices or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic storage, a flash memory, a magnetic disk, or an optical disk.

[0095] It should be noted that the above embodiment is only used to illustrate the technical solutions of the present application but not limit the present application. Although the present application is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical solutions of the present application, and all of them should be covered in the scope of the claims of the present application.

Claims

1. A lightweight method for laser selective melting forming of gradient lattice structures in titanium alloys, characterized in that, Includes the following steps: Step 1: Construct a gradient lattice database for titanium alloys. The database includes parameters, mechanical properties, and the influence of various process parameters on the final forming quality of different lattice topologies. Step 2: Based on the data obtained from the database, an intelligent optimization model is built using generative adversarial networks and reinforcement learning. In this system, the generative adversarial network (GAN) is responsible for simulating lightweight design schemes under various working conditions; reinforcement learning is responsible for adjusting cell parameters to optimize the lattice while ensuring structural strength and minimizing weight. The training objective of the intelligent optimization model is to generate feasible structures that meet the requirements of SLM manufacturing. ; Step 3: Optimize the lattice forming quality based on SLM process parameters, and use a multi-objective optimization algorithm to perform global optimization on the generated smart lattice structure; In step two, based on the lattice scheme generated by the generative adversarial network, reinforcement learning is used to adjust the structural parameters to minimize the mass while ensuring mechanical performance. Specifically: S201: Definition: State The current lattice structure parameters; state It is the adjustment amount of the lattice structure parameters, the reward function. ; in, Indicates the constrained mechanical properties; Indicates a weight restriction condition. This represents the feasibility loss of the SLM lattice structure; , , These represent the corresponding weight coefficients; S202: Optimize the policy using the policy gradient method Maximizing cumulative rewards can be represented as follows: in, This represents the discount factor, used to balance short-term and long-term returns; The multi-objective optimization algorithm in step three is the NSGA-II algorithm; it includes: S301: The multiple lightweight lattice structure design schemes obtained in step two are used as initial solutions; S302: Calculate the performance of each solution across all objectives; S303: The solutions are divided into multiple levels according to the relative superiority of the objective function, and the Pareto Front solution is selected. S304: Calculate the crowding degree for each solution, that is, how dense the solution is in the Pareto Front, to avoid the algorithm getting trapped in local optima; S305: Then, a new generation of solutions is generated through the mechanism of genetic algorithm. After multiple iterations of optimization, a set of Pareto optimal solutions is finally obtained.

2. The lightweight method for laser selective melting forming of gradient lattice structures in titanium alloys according to claim 1, characterized in that: In step two, the generative adversarial network includes a generator. With discriminator generator It generates a potential lattice structure based on the input noise, i.e., random data z. The discriminator Based on the generated crystal structure Output a probability value to determine whether the lattice structure is real; Optimize the objective function by minimization and maximization. This promotes continuous improvement of both the generator and the discriminator, ultimately leading to the generator producing realistic lattice structures, where the objective function is optimized. The expression is: in, This represents the output of the discriminator. This represents the output of the generator.

3. The lightweight method for laser selective melting forming of gradient lattice structures in titanium alloys according to claim 2, characterized in that: The performance of each solution across all objectives includes: Calculate crystal lattice structure Manufacturing quality loss ; Calculate the crystal lattice structure Thermal stress accumulation due to laser scanning path ; Calculate the crystal lattice structure Complexity of laser scanning path Both should be as small as possible.

4. The lightweight method for laser selective melting forming of gradient lattice structures in titanium alloys according to claim 3, characterized in that: The congestion degree is calculated by calculating the congestion distance of the solution, and the calculation method is as follows: Where M represents the number of initial solutions. , Indicates the current goal The values ​​of the adjacent solutions above, Indicates the current goal The maximum and minimum values ​​on.

5. The lightweight method for laser selective melting forming of gradient lattice structures in titanium alloys according to claim 4, characterized in that: The final optimal crystal structure is tested, and the test indicators include: dimensional accuracy test, mechanical property test, density detection and surface roughness measurement. Through testing, the reliability of the optimization model is evaluated. If any indicator fails to meet the standard, it is fed back to the previous step for adjustment.

6. The system for a lightweight method of laser selective melting forming of gradient lattice structures in titanium alloys according to claim 5, characterized in that: The system includes: The data collection module collects data on parameters, mechanical properties, and the impact of various process parameters on the final forming quality of different lattice topologies, and constructs a titanium alloy gradient lattice database. The intelligent lattice generation and optimization module generates optimized gradient lattice structures through an intelligent optimization model established by generative adversarial networks and reinforcement learning. The SLM process adaptation and optimization module is used to calculate the laser energy input, scanning path, and scanning speed parameters during the SLM process to ensure that the generated lattice structure has strong manufacturability. The multi-objective optimization module employs a multi-objective optimization algorithm to weigh multiple objectives and select the optimal lattice structure scheme. It also includes an intelligent verification and simulation module, which uses multiphysics simulation technology to virtually test the final optimized lattice structure to ensure that it meets design requirements.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the lightweight method for laser selective melting forming of titanium alloy gradient lattice structures as described in any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the lightweight method for laser selective melting forming of titanium alloy gradient lattice structures as described in any one of claims 1 to 5.

Images