Additive ultrahigh-strength steel strengthening and toughening process based on dynamic phase change control

Abstract

The invention relates to the field of additive manufacturing, and discloses a dynamic phase change control additive ultrahigh-strength steel strengthening and toughening process which comprises the following steps: S1, constructing a phase change structure design target which comprises retained austenite distribution, a structure space structure and mechanical property requirements; s2, based on the phase change structure design target, establishing a free energy model taking a phase field variable as a core for representing a phase change driving behavior between austenite and martensite; s3, on the basis of the free energy model, constructing a multi-field partial differential control model coupling phase field evolution, element diffusion and heat conduction; and S4, in combination with the multi-field partial differential control model, determining an interface adsorption parameter, a diffusion coefficient and a thermophysical parameter based on experimental characterization data. According to the method, path parameters are optimized through a multi-field partial differential control model and tissue deviation feedback, and accurate control of the phase change tissue and dynamic closed-loop adjustment of the machining process are achieved.

Description

Technical Field

[0001] This invention relates to the field of additive manufacturing, specifically to a dynamic phase change controlled additive ultra-high strength steel strengthening process. Background Technology

[0002] In fields such as high-end equipment manufacturing, aerospace component production, and biomedical implant fabrication, highly customized metal components with microstructures are often required. These applications demand extremely high material properties, such as high strength and toughness, excellent fatigue life, or specific residual stress states. To achieve these properties, the microstructure evolution behavior during the additive manufacturing process must be precisely controlled.

[0003] In existing technologies, some researchers estimate the thermal history of materials using heat conduction models or simplified temperature field simulations. These methods can, to some extent, predict parameters such as peak temperatures and cooling rates, thereby controlling the occurrence of macroscopic defects. Additionally, some researchers employ scanning strategy optimization software to reduce residual stress or improve forming efficiency through path planning. Overall, existing technologies demonstrate effectiveness in improving the surface quality of parts and reducing deformation risks, and have good engineering application value.

[0004] However, existing technologies cannot achieve precise control at the level of phase transition microstructure formation mechanisms, and there is a lack of quantifiable correlation between microstructure prediction and processing parameters; simulations using models cannot clearly explain how the microstructure actually changes. This directly leads to a disconnect between microstructure and performance indicators, and microstructure control can only rely on a lot of experiments and chance. In addition, most existing technologies rely on fixed or empirical parameter inputs and cannot automatically feed back microstructure deviations to correct process plans, which decouples microstructure design from process formulation. Some path optimization methods only target geometric scanning strategies, completely ignoring the laws of microstructure evolution, and are basically unusable after process changes. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a dynamic phase transformation controlled additive ultra-high strength steel strengthening and toughening process, which solves the problem that existing technologies lack quantitative correlation between processing path parameters and phase transformation microstructure evolution, and cannot achieve dynamic optimization based on microstructure targets.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a dynamic phase change-controlled additive ultra-high strength steel strengthening and toughening process, comprising the following steps:

[0007] S1. Construct phase transformation microstructure design objectives, which include the distribution of retained austenite, microstructure spatial structure, and mechanical property requirements;

[0008] S2. Based on the phase transformation microstructure design objectives, establish a free energy model with phase field variables as the core to characterize the phase transformation driving behavior between austenite and martensite.

[0009] S3. Based on the free energy model, construct a multi-field partial differential control model that couples phase field evolution, element diffusion, and heat conduction to describe the evolution path of the tissue.

[0010] S4. Based on the experimental characterization data, determine the interfacial adsorption parameters, diffusion coefficient, and thermophysical parameters using the multi-field partial differential control model, and fit the model parameters.

[0011] S5. Based on the deviation relationship between the evolution output of the multi-field partial differential control model and the phase transition structure design target, optimize the processing path parameters, including laser power, scanning speed and interlayer cooling time.

[0012] Preferably, in step S1, the design objectives for constructing the phase change organization include:

[0013] The determination of the residual austenite distribution involves spatially planning the residual austenite distribution based on the strain localization characteristics of the target region, forming a three-dimensional directional distribution function.

[0014] The microstructure is set based on target parameters such as grain size, phase boundary orientation, and martensite lath width.

[0015] The establishment of mechanical performance requirements includes setting the target yield strength, elongation and fracture toughness ranges, and these settings, combined with the expected service load conditions, form constraint boundaries.

[0016] Preferably, in step S2, establishing the free energy model with phase field variables as the core includes:

[0017] A phase field variable φ is introduced to describe the continuous evolution of the phase transformation state between austenite and martensite;

[0018] Construct the free energy density function f(φ,C,T), which includes the following components:

[0019]

[0020] In the formula, ε is the interfacial energy coefficient; W(φ) is the potential energy function; G γ (C,T) and G α (C,T) represent the free energy density terms for austenite and martensite, respectively; f(φ,C,T) is the free energy density function. φ is the square of the phase field gradient, representing the rate of change of the phase boundary in space; φ is the phase field variable.

[0021] Preferably, the continuous evolution process of the phase transition state includes:

[0022] The evolution function of phase field variables is used to characterize the change in the interface position between the martensite and austenite phases;

[0023] Record the gradient changes of phase field variables along grain boundaries, interfaces, and three-dimensional space at different evolution stages;

[0024] Based on the coupling relationship between phase field variables and tissue evolution time, phase boundary migration trajectories and transition velocity field distributions are generated.

[0025] Preferably, in step S3, the multi-field partial differential control model includes:

[0026] The phase transition evolution path is controlled by Allen–Cahn type phase-field equations, and the form of the evolution equations is as follows:

[0027]

[0028] In the formula, L φ δF represents the phase field mobility; δF and δφ are the variational derivatives of the free energy function with respect to φ; φ is the phase field variable; t is the time variable;

[0029] The migration behavior of solute elements is described by an unsteady diffusion equation, and the diffusion behavior is coupled with adsorption at the phase change interface.

[0030] The evolution of the temperature field over time is described by the heat conduction equation, in which a latent heat term is introduced for energy coupling between the thermal and structural fields.

[0031] Preferably, the introduction of latent heat terms for energy coupling between the thermal and structural fields includes:

[0032] The latent heat of phase transition is expressed as the product of the derivative of the phase field variable with respect to time and the latent heat coefficient;

[0033] The product term is added to the right side of the heat conduction equation as a source term distribution.

[0034] Based on the relationship between the rate of temperature change and the rate of local latent heat release, the time scale and local non-equilibrium response in the thermal field evolution process are corrected.

[0035] Preferably, in step S4, the model parameter fitting includes:

[0036] Based on the microstructure images, interface element enrichment behavior and measured thermophysical data obtained from the experiment, a parameter fitting function relationship was constructed.

[0037] Physical quantities such as interfacial adsorption parameters, diffusion coefficient, and thermal conductivity are retrieved using multiple sets of experimental data. The parameter retrieval adopts a fitting method based on nonlinear least squares.

[0038] The fitting results are solved in conjunction with the multi-field partial differential control model, and the model coefficients are adjusted until the error between the evolution output and the experimental data is less than the preset convergence tolerance threshold.

[0039] Preferably, the step of simultaneously solving the fitting results with the multi-field partial differential control model includes:

[0040] Embed the fitted expressions for diffusion coefficient, adsorption parameters, and thermophysical property functions into the multi-field control model;

[0041] Based on the given initial boundary conditions, multiphysics numerical solutions are performed to obtain the state function of phase change structure evolution over time.

[0042] The residual terms in the model are continuously adjusted through iterative methods until the deviations between the multi-field evolution output and the experimental reference data in terms of phase field variables, element concentrations, and temperature distributions are within acceptable limits.

[0043] Preferably, in step S5, optimizing the processing path parameters includes:

[0044] An objective function is constructed based on the spatial deviation between the phase change organization evolution trajectory output in the multi-field partial differential control model and the phase change organization design objective.

[0045] The phase boundary morphology shift, the degree of deviation of the retained austenite distribution, and the temperature gradient change included in the objective function are used as penalty terms.

[0046] Laser power, scanning speed, and interlayer cooling time were set as parameters to be optimized, and a genetic algorithm was used to solve for the optimal parameter combination.

[0047] Preferably, the spatial deviation between the design objectives constructs the objective function including:

[0048] The phase change organization design objectives and evolution results are respectively constructed as tensor data structures in three-dimensional space;

[0049] The difference between the phase boundary location and the austenite volume fraction at each grid point is calculated based on Euclidean distance.

[0050] The squared differences are summed to form the overall objective function expression, and this objective function is used as the fitness evaluation index of the processing path optimization algorithm.

[0051] This invention provides a dynamically phase-change-controlled additive ultra-high strength steel strengthening process. It offers the following advantages:

[0052] 1. This invention employs a parameter fitting mechanism that combines a multi-field partial differential control model with experimental data, achieving the technical effect of realistically reproducing the thermodynamic-kinetic behavior of phase transition processes. Compared to existing methods that rely solely on empirical formulas or single-parameter estimation, this invention solves the problems of model disconnect from actual tissue evolution and low prediction accuracy.

[0053] 2. This invention achieves dynamic adjustment of processing path parameters by constructing a target deviation function and coupling it with a multi-objective optimization process. Compared with the traditional static path setting method, it better matches the material's microstructure response characteristics and prevents excessive reliance on the tediousness and uncertainty of trial-and-error experiments.

[0054] 3. This invention adopts a functional modeling strategy oriented towards interface adsorption and diffusion characteristics, which can flexibly replace parameters to express the structure according to different material systems; thus breaking the limitations of previous models in terms of parameter universality and engineering portability, providing the possibility for rapid adaptation to different application scenarios, and making up for the lack of universality of traditional models.

[0055] 4. This invention uses a thermal-diffusion-phase-field fully coupled model to drive the path optimization process, forming a closed-loop feedback between the microstructure design and process input. Compared with the existing passive adjustment process of "process first, then evaluate", it completely improves the phenomenon of low optimization iteration efficiency and lag in microstructure control, and enhances the precision of material customization. Attached Figure Description

[0056] Figure 1 This is a schematic diagram of the process flow of the present invention. Detailed Implementation

[0057] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0058] Please see the appendix Figure 1 This invention provides a dynamic phase change controlled additive ultra-high strength steel strengthening and toughening process, including the following steps:

[0059] S1. Construct phase transformation microstructure design objectives, which include the distribution of retained austenite, microstructure spatial structure, and mechanical property requirements;

[0060] Constructing the phase transformation microstructure design target is the first step in the entire dynamic phase transformation controlled additive ultra-high strength steel strengthening process, and it has a significant impact on subsequent steps such as multi-field partial differential models and processing path optimization. By accurately setting the phase transformation microstructure design target, the distribution of retained austenite, the microstructure spatial structure, and mechanical properties can be effectively controlled, thereby ensuring that the final additively manufactured material has excellent mechanical properties and durability.

[0061] The core objectives of constructing phase change organization design goals include the following aspects:

[0062] Control of retained austenite distribution: The distribution of retained austenite is not only related to its volume fraction, but also to its directional distribution in three-dimensional space, especially the distribution of retained austenite in the stress region. By controlling its morphology, the mechanical properties of the material can be optimized.

[0063] Optimization of microstructure: The design of microstructure involves setting multiple parameters such as grain size, phase boundary orientation, and martensite lath width. By optimizing these structures, the properties of the microstructure can be effectively controlled, and the strength-toughness balance of ultra-high strength steel can be improved.

[0064] Setting mechanical performance requirements: Based on actual service requirements, the target values ​​of mechanical performance requirements, such as yield strength, elongation, and fracture toughness, need to be designed according to actual working load conditions to ensure that they meet the needs of actual use.

[0065] The design goal of retained austenite distribution is not merely to control the volume fraction of austenite, but to plan its spatial distribution in conjunction with the strain localization characteristics of the target region. In practice, finite element analysis was used to simulate the strain field in different regions, focusing on high-strain and low-strain regions, and the distribution pattern of retained austenite was determined based on strain localization characteristics. Typically, the spatial distribution of retained austenite is adjusted according to the strain gradient. The target region, under load, may require a higher proportion of retained austenite in strain-localized areas, while in low-strain regions, its proportion needs to be controlled to ensure optimal overall material mechanical properties. Through this method, we can achieve a directional distribution of retained austenite in three-dimensional space, forming a specific three-dimensional directional distribution function. The shape of this distribution function is designed based on actual load conditions and regional characteristics.

[0066] Specifically, the distribution of retained austenite is described using a three-dimensional Gaussian function or a polynomial function. Assuming that within a specific region, the distribution function Φ(r) of retained austenite can be expressed by the following formula:

[0067]

[0068] In the formula, Φ(r) is the distribution function of retained austenite; A is the regularization factor; r0 is the distribution center point; σ is the standard deviation, controlling the distribution width of retained austenite; and |r-r0| is the Euclidean distance between the current position and the distribution center. The introduction of this distribution function can effectively control the spatial distribution of retained austenite in the material, allowing for adjustment of its content according to the needs of the actual strain region.

[0069] Regarding the spatial structure, specific target parameters include grain size, martensite lath width, and phase boundary orientation. Alternatively, the grain size can be set within a range of 3–5 μm, depending on the required size. This size range effectively improves yield strength for ultra-high-strength steels and allows for optimization of elongation and toughness through appropriate heat treatment. In practice, grain size can be adjusted by controlling the heating and cooling rates of the heat treatment process. For example, rapid cooling can reduce grain size, thereby increasing the material's strength.

[0070] In the design of microstructure, the width of the martensitic laths is also an important parameter, usually set to be less than 1 μm to ensure the strength and toughness of the material. The width of the martensitic laths directly affects the hardness and ductility of the material; therefore, a reasonable design can ensure strength while avoiding the material becoming too brittle.

[0071] Mechanical performance requirements are typically set based on actual operating loads. For example, the target yield strength should be no less than 1600 MPa, and the elongation no less than 10%. To achieve these requirements, the microstructure design must be optimized in conjunction with actual service loads. This is usually achieved through simulation calculations based on the target loads, and adjustments to the heat treatment process and the distribution of retained austenite are made based on the results.

[0072] In one possible implementation, the design of mechanical properties depends not only on the macroscopic structure but also on the combined influence of microscopic factors such as grain size and phase interfaces. For additive manufacturing processes, the expected service load conditions need to be based on simulation analysis, and optimized through multiple iterations by establishing stress-strain curves adapted to service conditions. This process is controlled in the model by setting the coupling relationship between local stress, temperature gradient, and phase transformation dynamics.

[0073] S2. Based on the phase transformation microstructure design objectives, establish a free energy model with phase field variables as the core to characterize the phase transformation driving behavior between austenite and martensite.

[0074] In the proposed additive ultra-high strength steel strengthening and toughening process controlled by dynamic phase transformation, mathematical modeling of the thermodynamic driving mechanism during the austenitic-martensite phase transformation is necessary to achieve precise control over the target microstructure and spatial structure. Generally, the formation and evolution of microstructure originate from the response of its free energy system to the path minimization. Especially in multiphase coexistence or interface migration processes, the free energy distribution directly determines the initiation conditions and evolution trend of phase transformation behavior. Therefore, establishing a reasonable, calculable, and physically consistent free energy description model based on the established phase transformation microstructure design goals is the core element of the process implementation in this invention.

[0075] By introducing continuous field variables, a thermodynamic free energy function with phase field variables as the core is established, thereby characterizing the transformation process between austenite and martensite without explicit interface tracking.

[0076] The free energy model uses phase field variables as the core quantities to describe the phase transition state.

[0077] In this embodiment, the constructed free energy density function f(φ,C,T) includes the interface energy term, the chemical potential energy term, and the free energy term within each phase, specifically expressed as follows:

[0078]

[0079] In the formula, ε is the interfacial energy coefficient; W(φ) is the potential energy function; G γ (C,T) and G α (C,T) represent the free energy density terms for austenite and martensite, respectively; f(φ,C,T) is the free energy density function. φ is the square of the phase field gradient, representing the rate of change of the phase boundary in space; φ is the phase field variable.

[0080] Under normal circumstances, G γ With G α The specific expression can be obtained from materials thermodynamic databases or based on the CALPHAD method, and its parameters depend on temperature, stress state, and alloy composition. In one possible implementation, the free energy density of each phase can be approximated using the following form:

[0081] G γ =H γ -TS γ G α =H α -TS α ;

[0082] In the formula, H γ H α The enthalpy of austenite and martensite, respectively; S γ ,S αThese represent the entropy of each phase; T is the absolute temperature; G... γ G is the molar free energy of the γ phase (austenite); α This represents the molar free energy of the α phase (martensite).

[0083] Alternatively, the aforementioned thermodynamic potential function can also incorporate the alloy element concentration field and the local elastic strain energy density f. total This is extended to accommodate multi-component alloy systems and stress-induced phase transformation behavior. The total free energy density is then extended as follows:

[0084]

[0085] In the formula, f cgem (c) represents the free energy term related to the concentration function of the component field; f elast (φ,ε ij ε represents the elastic energy that varies with the phase transition state; ij f is the strain tensor; f(φ,C,T) is the free energy density function; f total This represents the local elastic strain energy density.

[0086] In some embodiments, to enhance the numerical solvability of the model, a smooth cubic or quintic polynomial interpolation form can be used for the potential energy function W(φ) to avoid numerical oscillations during discrete calculations.

[0087] Specifically, this free energy model will be embedded in the partial differential control model as the core driving function of the subsequent phase field evolution equations, and will be coupled in conjunction with the element diffusion and heat conduction equations. In the simulation of the microscale evolution of material microstructure, this model has the advantage of not relying on explicit phase boundary expressions, and is particularly suitable for predicting the microstructure morphology under complex thermal histories in additive manufacturing.

[0088] It should be noted that the free energy construction method in this embodiment is adapted to the martensitic phase transformation model represented by the ferrite-austenite system. For other possible multiphase metal systems, it can also be adapted by adjusting the structure of the free energy terms.

[0089] S3. Based on the free energy model, construct a multi-field partial differential control model that couples phase field evolution, element diffusion, and heat conduction to describe the evolution path of the tissue.

[0090] Based on the construction of a free energy model with phase field variables as the core and the clarification of the driving mechanism of austenite-martensite phase transformation, in order to further track the spatial evolution behavior of microstructure under the complex thermo-mechanical path of additive manufacturing, it is necessary to introduce a partial differential control model coupled with multiple physics fields, so as to truly restore the evolution path of the microstructure by solving multiple field quantities in a collaborative manner.

[0091] Generally, phase transition processes in microstructures do not evolve in isolation, but are strongly coupled with factors such as temperature field evolution, solute element diffusion, and latent heat release. This coupling effect is particularly pronounced under rapid thermal cycling and unsteady-state conditions. As an alternative, jointly modeling the phase field evolution equation with the diffusion and heat conduction equations is an effective means of achieving precise control over microstructures.

[0092] In this embodiment, the multi-field partial differential control model includes three core equations: phase field evolution equation, solute diffusion equation, and heat conduction equation.

[0093] Specifically, the phase transition evolution path is controlled by Allen–Cahn type phase-field equations, and the evolution equations are in the following form:

[0094]

[0095] In the formula, L φ δ is the phase field mobility; δF and δφ are the variational derivatives of the free energy function with respect to φ; φ is the phase field variable; t is the time variable.

[0096] To reflect the influence of alloying element distribution on the phase transformation path, an unsteady diffusion equation is introduced to model the solute migration behavior. The diffusion equation takes the following form:

[0097]

[0098] In the formula, D(φ,T) is the diffusion coefficient function, which depends on the current phase field state and temperature; S c (φ) represents the source terms coupled with the phase transition, reflecting the non-uniform source distribution of interfacial adsorption behavior; This represents the spatial gradient of solute concentration. Let be the partial derivative of the solute concentration with respect to time, and represent the diffusion evolution rate over time.

[0099] In one possible implementation, the diffusion coefficient D(φ,T) can be fitted experimentally as a piecewise continuous function, for example:

[0100] D(φ,T)=D γ (1-φ) 2 +D α φ 2 ;

[0101] In the formula, D γ D α φ and T are the effective diffusion coefficients in the austenitic and martensitic phases, respectively; D(φ,T) is the diffusion coefficient function, which depends on the current phase field state and temperature; φ is the phase field variable.

[0102] Meanwhile, to simulate the dynamic changes of the temperature field with time and space during laser scanning, a heat conduction equation is introduced:

[0103]

[0104] In the formula, ρ is the material density; c p ρ is the specific heat capacity; k is the thermal conductivity; Q is the specific heat capacity. latent This is a latent heat source term for phase change; Let be the partial derivative of temperature with respect to time, representing the rate of temperature change; is the gradient operator, representing the spatial derivative; This represents the spatial gradient of the temperature field.

[0105] To ensure the numerical stability and physical consistency of the coupled model, this embodiment employs an interleaved grid finite difference method during discretization to handle different evolution scales of the thermal and phase fields, and introduces a fractional-step method during time progression to maintain the synchronicity of the solutions for each field. Boundary values ​​are passed between various field quantities through interface functions, such as the temperature-diffusion coupling coefficient and the temperature-phase field driving force function.

[0106] In some embodiments, to enhance the applicability of the model, the introduction of thermo-elastic coupling terms is further considered, that is, the elastic strain energy density is added to the free energy expression, and an additional source term caused by thermal strain is introduced into the heat conduction equation, so as to adapt to the needs of tissue evolution path modeling under more complex working conditions.

[0107] S4. Based on the experimental characterization data, determine the interfacial adsorption parameters, diffusion coefficient, and thermophysical parameters using the multi-field partial differential control model, and fit the model parameters.

[0108] After the multi-field partial differential control model is coupled and constructed, to ensure that the simulation results match the actual tissue evolution path, it is necessary to numerically determine and quantitatively fit the interfacial adsorption, diffusion behavior, and heat conduction-related parameters involved in the model. Generally, these parameters have strong material dependence and environmental coupling characteristics, and accurate values ​​cannot be obtained through theoretical estimation. Therefore, inversion fitting based on experimental characterization data is a key means to establish a physically realistic microstructure evolution prediction model.

[0109] Specifically, to determine the adsorption term S at the phase change interface c (φ), In this embodiment, high-resolution transmission electron microscopy (HRTEM) combined with energy dispersive spectroscopy (EDS) line scanning technology is used to quantitatively calibrate the elemental enrichment phenomenon near the austenite / martensite interface. In one possible implementation, the following functional form can be established:

[0110] S c (φ)=S0exp(-β(φ-φ c )2 );

[0111] In the formula, S0 represents the maximum adsorption intensity; β is the adsorption distribution control parameter; φ c The peak adsorption intensity is located near the midpoint of the interface; S c (φ) represents the adsorption term at the phase change interface; φ represents the phase field variable.

[0112] The function exhibits a local peak at the martensite-austenite transformation interface, which is in high agreement with the experimentally measured interface concentration profile.

[0113] This embodiment derives D by heat-treating sample slices at temperature points and combining this with diffusion profile fitting. γ D α Numerical ranges at different temperatures, and construct temperature-dependent expressions:

[0114]

[0115] In the formula, D0 is the pre-exponential factor; Q is the diffusion activation energy; R is the gas constant; T is the absolute temperature; and D(T) represents the diffusion coefficient at temperature T.

[0116] As an alternative, further experiments measuring thermophysical parameters can be conducted to obtain the material constants required for the thermal conductivity term. Specifically, this includes:

[0117] Thermal diffusivity was measured using the laser flare method, and thermal conductivity was deduced by combining density and specific heat capacity.

[0118] Specific heat capacity was directly determined using differential scanning calorimetry.

[0119] The material density was determined by combining the volume / mass method with X-ray absorption correction;

[0120] The latent heat of phase transformation was also obtained by integrating over the martensite initiation temperature range using the DSC method.

[0121] This embodiment employs a multi-objective inversion optimization algorithm to jointly correct the aforementioned parameters. The objective function is an image structure similarity measure (e.g., based on the multi-scale SSIM index) between the simulated tissue evolution path and the experimental metallographic images, and is embedded in the phase-field simulation module to form a closed-loop iterative system.

[0122] S5. Based on the deviation relationship between the evolution output of the multi-field partial differential control model and the phase transition structure design target, optimize the processing path parameters, including laser power, scanning speed and interlayer cooling time.

[0123] Based on the constructed multi-field partial differential control model and the material parameter system determined through experimental fitting, further closed-loop optimization of the actual processing path parameters is needed to achieve the set phase transformation microstructure design target. In this invention, the microstructure design target typically includes microscopic indices such as retained austenite volume fraction, martensite lath spacing, and grain orientation consistency, and these structural features are highly dependent on the heat input history during processing. Therefore, iteratively correcting the processing path parameters based on the deviation information between the model simulation results and the target microstructure is a key step in improving the accuracy control capability of the microstructure.

[0124] In general, in laser selective melting additive manufacturing processes, laser power, scanning speed and interlayer cooling time are the main adjustable controllable quantities that determine the local thermal cycle characteristics. These three factors directly affect the peak temperature field of the molten pool, the thermal gradient and the cooling rate, thereby determining the phase transition initiation time and evolution path.

[0125] The tissue variable φ was simulated by comparing phase-field models. sim (r,t f ) and target organizational variable φ target The spatial deviation between (r) is constructed using the following deviation function:

[0126] ε φ =∫ v [φ sim (r,t f )-φ target (r)] 2 dV;

[0127] In the formula, ε φ t is the organizational evolution deviation function; f Indicates the termination time of the model simulation; r is the three-dimensional spatial coordinate; φ sim The final simulated phase field distribution; φ target An ideal structural field is set according to the organization's design goals.

[0128] Alternatively, the deviation function can be superimposed with weighting terms such as elemental concentration deviation, temperature hysteresis, or morphological roughness to form a multi-objective optimization function.

[0129] This embodiment employs a hybrid strategy of Newton's gradient method and heuristic genetic algorithm for parameter optimization. The optimization process is implemented through the following feedback flow:

[0130] First, fix a set of processing path parameters;

[0131] The data is input into a thermal-diffusion-phase-field coupling model for tissue evolution simulation.

[0132] Calculate the tissue deviation function ε φ ;

[0133] Update the parameter set based on gradient information or fitness score;

[0134] Repeat the iteration until convergence to the set threshold.

[0135] In some embodiments, to improve optimization efficiency, the scanning speed–laser power–cooling time parameter space is divided into three-dimensional mapping blocks, and an initial global search is performed through Latin hypercube sampling (LHS) to construct a surrogate model between parameters and tissue response, thereby further improving search efficiency.

[0136] This embodiment specifically considers the effect of the Gaussian heat flux density distribution in the laser heat source model on the width and depth of the molten pool, which is described by the following expression:

[0137]

[0138] In the formula, η is the laser absorption coefficient; r0 is the laser beam radius; P is the laser power; q(r) is the heat flux per unit area; and r is the three-dimensional spatial coordinate.

[0139] During the simulation, the location of the heat source center is updated in real time along the scanning path, forming time-segmented thermal boundary conditions, which are coupled into the heat conduction equation.

[0140] In some embodiments, the impact of process fluctuations or external disturbances on path stability is also considered. A Bayesian optimization framework is used to model the uncertainty of the parameter results in each round and use it as the basis for decision-making to update the processing strategy, thereby improving the consistency between the model and the actual product.

[0141] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A dynamic phase transformation controlled additive ultra-high strength steel strengthening and toughening process, characterized in that, Includes the following steps: S1. Construct phase transformation microstructure design objectives, which include the distribution of retained austenite, microstructure spatial structure, and mechanical property requirements; S2. Based on the phase transformation microstructure design objectives, establish a free energy model with phase field variables as the core to characterize the phase transformation driving behavior between austenite and martensite. S3. Based on the free energy model, construct a multi-field partial differential control model that couples phase field evolution, element diffusion, and heat conduction to describe the evolution path of the tissue. S4. Based on the experimental characterization data, determine the interfacial adsorption parameters, diffusion coefficient, and thermophysical parameters using the multi-field partial differential control model, and fit the model parameters. S5. Based on the deviation relationship between the evolution output of the multi-field partial differential control model and the phase transition structure design target, optimize the processing path parameters, including laser power, scanning speed and interlayer cooling time.

2. The additive ultra-high strength steel strengthening and toughening process with dynamic phase transformation control according to claim 1, characterized in that, In step S1, the design objectives for constructing the phase change organization include: The determination of the residual austenite distribution is based on the spatial distribution planning of the residual austenite distribution according to the strain localization characteristics of the target region, forming a three-dimensional directional distribution function; The microstructure is set based on target parameters such as grain size, phase boundary orientation, and martensite lath width. The establishment of mechanical performance requirements includes setting the target yield strength, elongation and fracture toughness ranges, and these settings, combined with the expected service load conditions, form constraint boundaries.

3. The additive ultra-high strength steel strengthening and toughening process with dynamic phase transformation control according to claim 1, characterized in that, In step S2, establishing the free energy model with phase field variables as the core includes: A phase field variable φ is introduced to describe the continuous evolution of the phase transformation state between austenite and martensite; Construct the free energy density function f(φ,C,T), which includes the following components: In the formula, ε is the interfacial energy coefficient; W(φ) is the potential energy function; G γ (C,T) and G α (C,T) represent the free energy density terms for austenite and martensite, respectively; f(φ,C,T) is the free energy density function. φ is the square of the phase field gradient, representing the rate of change of the phase boundary in space; φ is the phase field variable.

4. The additive ultra-high strength steel strengthening and toughening process with dynamic phase transformation control according to claim 3, characterized in that, The continuous evolution process of the phase transition state includes: The evolution function of phase field variables is used to characterize the change in the interface position between the martensite and austenite phases; Record the gradient changes of phase field variables along grain boundaries, interfaces, and three-dimensional space at different evolution stages; Based on the coupling relationship between phase field variables and tissue evolution time, phase boundary migration trajectories and transition velocity field distributions are generated.

5. The additive ultra-high strength steel strengthening and toughening process with dynamic phase transformation control according to claim 1, characterized in that, In step S3, the multi-field partial differential control model includes: The phase transition evolution path is controlled by Allen–Cahn type phase-field equations, and the form of the evolution equations is as follows: In the formula, L φ δF represents the phase field mobility; δF and δφ are the variational derivatives of the free energy function with respect to φ; φ is the phase field variable; t is the time variable; The migration behavior of solute elements is described by an unsteady diffusion equation, and the diffusion behavior is coupled with adsorption at the phase change interface. The evolution of the temperature field over time is described by the heat conduction equation, in which a latent heat term is introduced for energy coupling between the thermal and structural fields.

6. The additive ultra-high strength steel strengthening and toughening process with dynamic phase change control according to claim 5, characterized in that, The introduction of latent heat terms for energy coupling between thermal and structural fields includes: The latent heat of phase transition is expressed as the product of the derivative of the phase field variable with respect to time and the latent heat coefficient; The product term is added to the right side of the heat conduction equation as a source term distribution. Based on the relationship between the rate of temperature change and the rate of local latent heat release, the time scale and local non-equilibrium response in the thermal field evolution process are corrected.

7. The additive ultra-high strength steel strengthening and toughening process with dynamic phase transformation control according to claim 1, characterized in that, In step S4, the process of fitting the model parameters includes: Based on the microstructure images, interface element enrichment behavior and measured thermophysical data obtained from the experiment, a parameter fitting function relationship was constructed. Physical quantities such as interfacial adsorption parameters, diffusion coefficient, and thermal conductivity are retrieved using multiple sets of experimental data. The parameter retrieval adopts a fitting method based on nonlinear least squares. The fitting results are solved in conjunction with the multi-field partial differential control model, and the model coefficients are adjusted until the error between the evolution output and the experimental data is less than the preset convergence tolerance threshold.

8. The additive ultra-high strength steel strengthening and toughening process with dynamic phase transformation control according to claim 7, characterized in that, The step of simultaneously solving the fitting results with the multi-field partial differential control model includes: Embed the fitted expressions for diffusion coefficient, adsorption parameters, and thermophysical property functions into the multi-field control model; Based on the given initial boundary conditions, multiphysics numerical solutions are performed to obtain the state function of phase change structure evolution over time; The residual terms in the model are continuously adjusted through iterative methods until the deviations between the multi-field evolution output and the experimental reference data in terms of phase field variables, element concentrations, and temperature distributions are within acceptable limits.

9. The additive ultra-high strength steel strengthening and toughening process with dynamic phase transformation control according to claim 1, characterized in that, In step S5, optimizing the processing path parameters includes: An objective function is constructed based on the spatial deviation between the phase change organization evolution trajectory output in the multi-field partial differential control model and the phase change organization design objective. The phase boundary morphology shift, the degree of deviation of the retained austenite distribution, and the temperature gradient change included in the objective function are used as penalty terms. Laser power, scanning speed, and interlayer cooling time were set as parameters to be optimized, and a genetic algorithm was used to solve for the optimal parameter combination.

10. The additive ultra-high strength steel strengthening and toughening process with dynamic phase transformation control according to claim 9, characterized in that, The spatial deviation between the design objectives constructs the objective function, which includes: The phase change organization design objectives and evolution results are respectively constructed as tensor data structures in three-dimensional space; The difference between the phase boundary location and the austenite volume fraction at each grid point is calculated based on Euclidean distance. The squared differences are summed to form the overall objective function expression, and this objective function is used as the fitness evaluation index of the processing path optimization algorithm.

Images