A grain morphology screening method for laser additive manufacturing of titanium alloys

Abstract

The application discloses a grain morphology screening method for laser additive manufacturing of titanium alloy, and comprises the following steps: 1, determining the additive elements and the adding range of the laser additive manufacturing of titanium alloy; 2, constructing a solidification behavior data set of a laser micro zone metallurgical sample of the titanium alloy to be laser additive manufactured; 3, preparing the laser micro zone metallurgical sample under different laser action process schemes; 4, collecting the grain morphology of the laser micro zone metallurgical sample corresponding to different thermal behavior conditions; 5, collecting the solidification temperature interval under the non-equilibrium solidification condition; 6, data denoising; 7, machine learning; 8, screening the grain morphology of the titanium alloy to be laser additive manufactured. The application takes the composition and content of the laser micro zone metallurgical sample, the solidification temperature interval, the temperature gradient and the solidification rate as the input layer nodes of the machine learning model, takes the grain morphology as the output layer nodes of the machine learning model, predicts the grain morphology, and screens the composition of the titanium alloy to be laser additive manufactured to meet the equiaxed structure, and has a wide application range.

Description

Technical Field

[0001] This invention belongs to the field of grain morphology screening technology for titanium alloys, and specifically relates to a method for screening the grain morphology of laser additive manufacturing titanium alloys. Background Technology

[0002] Additively manufactured titanium alloys are increasingly used in aerospace, deep-sea submersibles, and other fields, showing great potential for development. However, to date, only a limited number of titanium alloy grades have been successfully applied in engineering, such as TC4, TC21, TC18, and TA15. Furthermore, the extremely high cooling rate and temperature gradient inherent in metal additive manufacturing (AM) typically cause titanium alloys to produce coarse columnar β grains along the deposition direction. This directly leads to anisotropy and strength-ductility mismatch in their properties. Therefore, obtaining fine equiaxed β grains is a necessary condition for improving alloy properties and preparing high-quality AM titanium alloy parts.

[0003] With the increasing complexity of titanium alloy compositions in laser additive manufacturing, the design space for compositions involving alloying elements such as Al, V, Cu, Sn, Fe, Ni, Cr, Mo, and Zr has become vast. Furthermore, grain morphology is simultaneously influenced by different solidification conditions, making equiaxed composition design a complex problem. Existing research mostly explores grain variation patterns by preparing large-sized deposited samples, which is inefficient and relies on single solidification conditions and alloying element variables, failing to quickly clarify the influence of one or more alloying elements on grain morphology under different solidification conditions. Therefore, to further improve the efficiency of composition design for laser additive manufacturing of titanium alloys and save on new material development costs, how to efficiently and accurately explore alloy compositions that satisfy equiaxed microstructures has become an urgent problem to be solved in materials research. Summary of the Invention

[0004] The technical problem to be solved by this invention is to address the shortcomings of the prior art by providing a method for screening the grain morphology of titanium alloys manufactured by laser additive manufacturing. This method involves determining the additive elements and their range in the laser additive manufacturing of titanium alloys, constructing a dataset of the solidification behavior of laser micro-area metallurgical samples of the titanium alloys to be manufactured by laser additive manufacturing, preparing laser micro-area metallurgical samples under different laser irradiation process schemes, collecting data on the grain morphology and solidification temperature range ΔT of the laser micro-area metallurgical samples corresponding to different thermal behavior conditions, removing outliers from the dataset, and performing machine learning. The composition and content of the laser micro-area metallurgical sample, the solidification temperature range ΔT, the temperature gradient G, and the solidification rate V are used as the input layer nodes of the machine learning model, and the grain length a, width b, and aspect ratio c are used as the output layer nodes of the machine learning model. This method predicts the grain morphology of the titanium alloys to be manufactured by laser additive manufacturing, and screens out alloy compositions that satisfy the equiaxed structure requirement. This method has a wider range of applications, is highly efficient and accurate in verification, and is easy to promote and use.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a method for grain morphology screening of titanium alloys in laser additive manufacturing, characterized in that the method includes the following steps:

[0006] Step 1: Determine the additive elements and addition range for laser additive manufacturing of titanium alloys: The additive elements for laser additive manufacturing of titanium alloys include one or more of Al, V, Cu, Sn, Fe, Ni, Cr, Mo and Zr, and the addition range of each additive element for laser additive manufacturing of titanium alloys is 1 to 10 wt.%.

[0007] Step 2: Construct a dataset of solidification behavior of laser micro-area metallurgical samples of titanium alloys to be laser-added, the specific process of which is as follows:

[0008] Step 201: Use a computer and finite element calculation software to establish a geometric model of the powder bed composition of the metallurgical sample to be laser-treated in a micro-area.

[0009] Step 202: Calculate the thermophysical parameters of the metallurgical sample to be laser-treated in step 201, and assign the thermophysical parameters to the geometric model of the powder bed of the metallurgical sample to be laser-treated.

[0010] Step 203: Assign boundary conditions and initial temperature to the powder bed geometric model of the metallurgical sample to be laser-treated in the micro-area;

[0011] Step 204: Set up multiple different laser irradiation process schemes, wherein the laser irradiation process scheme includes laser power P and laser irradiation time t; and simultaneously set the absorption rate of the laser to the metallurgical sample in the micro-area to be lasered.

[0012] Step 205: Use finite element analysis software to generate a mesh and obtain the temperature change information at the end of laser treatment under different laser treatment process schemes, and obtain the temperature gradient G;

[0013] Step 206: By calculating the difference in the isotherm profile distance at different times, the solidification rate V is obtained, and a solidification behavior dataset of the micro-area metallurgical sample to be laser-treated is constructed. The solidification behavior dataset includes multiple sets of solidification behavior data, and each set of solidification behavior data includes the temperature gradient G and the solidification rate V.

[0014] Step 3: Prepare laser micro-area metallurgical samples under different laser irradiation process schemes. The laser micro-area metallurgical samples use the same composition as in Step 1. The specific preparation process for laser micro-area metallurgical samples under different laser irradiation process schemes is as follows: According to the laser power P and laser irradiation time t set under the laser irradiation process scheme, a series of micro-area fixed-point heating is performed through laser irradiation to obtain laser micro-area metallurgical samples corresponding to different thermal behavior conditions. The laser irradiation process scheme includes the laser power P and the laser irradiation time t.

[0015] Step 4: Collect grain morphology data for laser micro-area metallurgical samples under different thermal behavior conditions: Measure the grain length a and width b of the laser micro-area alloyed sample perpendicular to the powder bed plane in Step 3, and calculate the aspect ratio c, where c = a / b.

[0016] Step 5: Collect the solidification temperature range ΔT under non-equilibrium solidification conditions: Calculate the Schei L solidification curves corresponding to different titanium alloy compositions using the Thermo-Calc software TTTi3 database, obtain the temperatures at the start and end of solidification, and calculate the solidification temperature range ΔT.

[0017] Step 6: Data Denoising: Based on the laser micro-area metallurgical sample results from Step 3, delete data with excessive porosity and poor fusion caused by G and V not being within the corresponding threshold range or by compositional characteristics.

[0018] Step 7, Machine Learning: Select a machine learning model, using the composition and content of the laser micro-area metallurgical sample, the solidification temperature range ΔT, the temperature gradient G, and the solidification rate V as the input layer nodes of the machine learning model, and the grain length a, width b, and aspect ratio c as the output layer nodes of the machine learning model, and train the machine learning model.

[0019] Step 8: Grain morphology screening of titanium alloys to be laser-added: Input the composition and content of the alloy sample to be laser-added, the solidification temperature range ΔT, the temperature gradient G, and the solidification rate V into the trained machine learning model to screen out alloy compositions that meet the equiaxed structure requirements of the titanium alloy to be laser-added.

[0020] The above-mentioned method for screening grain morphology of titanium alloys in laser additive manufacturing is characterized in that: the finite element calculation software includes Comsol or Ansys finite element calculation software.

[0021] The above-mentioned method for screening grain morphology of titanium alloys by laser additive manufacturing is characterized in that: in step 202, the thermophysical parameters of the laser micro-region metallurgical sample in step 201 are calculated using Jmat-Pro or Thermo-Calc software, and the thermophysical parameters include the density, specific heat capacity, thermal conductivity and enthalpy of the laser micro-region metallurgical sample composition.

[0022] The above-mentioned method for screening the grain morphology of titanium alloys by laser additive manufacturing is characterized in that the boundary conditions of the geometric model of the laser micro-area metallurgical sample powder bed include thermal conductivity, heat transfer coefficient and radiation coefficient.

[0023] The above-mentioned method for screening grain morphology of laser additive manufacturing titanium alloys is characterized in that: the machine learning model includes a multinomial regression model, a K-nearest neighbor regression model, a linear regression model, an artificial neural network model, and a random forest regression model.

[0024] Compared with the prior art, the present invention has the following advantages:

[0025] 1. This invention determines the additive elements and addition ranges for laser additive manufacturing of titanium alloys, constructs a dataset of solidification behavior of laser micro-area metallurgical samples of the titanium alloys to be laser additively manufactured, prepares laser micro-area metallurgical samples under different laser irradiation process schemes, and collects grain morphology and solidification temperature range ΔT for laser micro-area metallurgical samples corresponding to different thermal behavior conditions. Abnormal data in the dataset is removed, and machine learning is performed. The composition and content of the laser micro-area metallurgical sample, the solidification temperature range ΔT, the temperature gradient G, and the solidification rate V are used as the input layer nodes of the machine learning model, and the grain length a, width b, and aspect ratio c are used as the output layer nodes of the machine learning model. This predicts the grain morphology of the titanium alloy to be laser additively manufactured, and selects alloy compositions that satisfy the equiaxed structure of the titanium alloy to be laser additively manufactured. This invention has a wider range of applications, is highly efficient and accurate in verification, and is easy to promote and use.

[0026] 2. This invention establishes the relationship between the ΔT, G, and V values ​​of materials and the original grain morphology of laser additive manufacturing of titanium alloys. It can provide database support for material screening as needed, enabling rapid screening of the composition of laser additive manufacturing titanium alloys. The process is precise and controllable, and it makes full use of metal powder, with the advantages of high material utilization and short overall sample preparation cycle. The composition-grain morphology relationship of laser additive manufacturing titanium alloys established by machine learning has broad application prospects in the field of materials genome engineering. It can be used to guide the design and screening of additive manufacturing compositions of various metals, including but not limited to titanium alloys, aluminum alloys, high-entropy alloys, steel, and high-temperature alloys. It can be promoted and used in various additive manufacturing alloy systems, which will promote the research and development of new materials.

[0027] 3. In this invention, while collecting the solidification temperature range ΔT, the Scheil solidification curve of the alloy can be obtained, which is of guiding significance for studying the non-equilibrium solidification process of the alloy and the decrease in solids line caused by solute segregation, as well as clarifying the composition of the liquid phase that solidifies between dendrites.

[0028] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0029] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0030] like Figure 1 As shown, the present invention provides a method for grain morphology screening of titanium alloys in laser additive manufacturing, comprising the following steps:

[0031] Step 1: Determine the additive elements and addition range for laser additive manufacturing of titanium alloys: The additive elements for laser additive manufacturing of titanium alloys include one or more of Al, V, Cu, Sn, Fe, Ni, Cr, Mo and Zr, and the addition range of each additive element for laser additive manufacturing of titanium alloys is 1 to 10 wt.%.

[0032] Step 2: Construct a dataset of solidification behavior of laser micro-area metallurgical samples of titanium alloys to be laser-added, the specific process of which is as follows:

[0033] Step 201: Use a computer and finite element calculation software to establish a geometric model of the powder bed composition of the metallurgical sample to be laser-treated in a micro-area.

[0034] Step 202: Calculate the thermophysical parameters of the metallurgical sample to be laser-treated in step 201, and assign the thermophysical parameters to the geometric model of the powder bed of the metallurgical sample to be laser-treated.

[0035] Step 203: Assign boundary conditions and initial temperature to the powder bed geometric model of the metallurgical sample to be laser-treated in the micro-area;

[0036] Step 204: Set up multiple different laser irradiation process schemes, wherein the laser irradiation process scheme includes laser power P and laser irradiation time t; and simultaneously set the absorption rate of the laser to the metallurgical sample in the micro-area to be lasered.

[0037] Step 205: Use finite element analysis software to generate a mesh and obtain the temperature change information at the end of laser treatment under different laser treatment process schemes, and obtain the temperature gradient G;

[0038] Step 206: By calculating the difference in the isotherm profile distance at different times, the solidification rate V is obtained, and a solidification behavior dataset of the micro-area metallurgical sample to be laser-treated is constructed. The solidification behavior dataset includes multiple sets of solidification behavior data, and each set of solidification behavior data includes the temperature gradient G and the solidification rate V.

[0039] Step 3: Prepare laser micro-area metallurgical samples under different laser irradiation process schemes. The laser micro-area metallurgical samples use the same composition as in Step 1. The specific preparation process for laser micro-area metallurgical samples under different laser irradiation process schemes is as follows: According to the laser power P and laser irradiation time t set under the laser irradiation process scheme, a series of micro-area fixed-point heating is performed through laser irradiation to obtain laser micro-area metallurgical samples corresponding to different thermal behavior conditions. The laser irradiation process scheme includes the laser power P and the laser irradiation time t.

[0040] Step 4: Collect grain morphology data for laser micro-area metallurgical samples under different thermal behavior conditions: Measure the grain length a and width b of the laser micro-area alloyed sample perpendicular to the powder bed plane in Step 3, and calculate the aspect ratio c, where c = a / b.

[0041] Step 5: Collect the solidification temperature range ΔT under non-equilibrium solidification conditions: Calculate the Schei L solidification curves corresponding to different titanium alloy compositions using the Thermo-Calc software TTTi3 database, obtain the temperatures at the start and end of solidification, and calculate the solidification temperature range ΔT.

[0042] Step 6: Data Denoising: Based on the laser micro-area metallurgical sample results from Step 3, delete data with excessive porosity and poor fusion caused by G and V not being within the corresponding threshold range or by compositional characteristics.

[0043] Step 7, Machine Learning: Select a machine learning model, using the composition and content of the laser micro-area metallurgical sample, the solidification temperature range ΔT, the temperature gradient G, and the solidification rate V as the input layer nodes of the machine learning model, and the grain length a, width b, and aspect ratio c as the output layer nodes of the machine learning model, and train the machine learning model.

[0044] Step 8: Grain morphology screening of titanium alloys to be laser-added: Input the composition and content of the alloy sample to be laser-added, the solidification temperature range ΔT, the temperature gradient G, and the solidification rate V into the trained machine learning model to screen out alloy compositions that meet the equiaxed structure requirements of the titanium alloy to be laser-added.

[0045] In this embodiment, the finite element calculation software includes Comsol or Ansys finite element calculation software.

[0046] In this embodiment, in step 202, the thermal properties of the laser micro-region metallurgical sample in step 201 are calculated using Jmat-Pro or Thermo-Calc software. The thermal properties include the density, specific heat capacity, thermal conductivity and enthalpy of the laser micro-region metallurgical sample.

[0047] In this embodiment, the boundary conditions of the geometric model of the laser micro-area metallurgical sample powder bed include thermal conductivity, heat transfer coefficient, and radiation coefficient.

[0048] In this embodiment, the machine learning model includes a multinomial regression model, a K-nearest neighbor regression model, a linear regression model, an artificial neural network model, and a random forest regression model.

[0049] In this invention, by determining the additive elements and their range in laser additive manufacturing of titanium alloys, a dataset of solidification behavior of laser micro-area metallurgical samples for laser additive manufacturing of titanium alloys is constructed. Laser micro-area metallurgical samples under different laser irradiation process schemes are prepared. Grain morphology and solidification temperature range ΔT are collected for laser micro-area metallurgical samples corresponding to different thermal behavior conditions. Outliers in the dataset are removed, and machine learning is performed. The composition and content of the laser micro-area metallurgical sample, solidification temperature range ΔT, temperature gradient G, and solidification rate V are used as input layer nodes of the machine learning model, and grain length a, width b, and aspect ratio c are used as output layer nodes. This predicts the grain morphology of the titanium alloy to be laser additive manufactured, and selects alloy compositions that satisfy the equiaxed structure. This method has a wider application range and is highly efficient and accurate in verification. It also establishes the material ΔT, G, and V values ​​and their relationship with the laser additive manufacturing process of titanium alloys. The relationship between initial grain morphology can provide database support for material screening as needed, enabling rapid compositional screening of laser additive manufacturing titanium alloys. The process is precise and controllable, while making full use of metal powder, resulting in high material utilization and a short overall sample preparation cycle. The composition-grain morphology relationship of laser additive manufacturing titanium alloys established by machine learning has broad application prospects in the field of materials genome engineering. It can be used to guide the design and screening of additive manufacturing compositions for various metals, including but not limited to titanium alloys, aluminum alloys, high-entropy alloys, steel, and high-temperature alloys. It can be widely used in various additive manufacturing alloy systems, which will promote the research and development of new materials. While collecting the solidification temperature range ΔT, the Scheil solidification curve of the alloy can be obtained, which is of guiding significance for studying the non-equilibrium solidification process of the alloy and the solidus line reduction caused by solute segregation, as well as clarifying the composition of the liquid phase that solidifies between dendrites.

[0050] The above description is merely a preferred embodiment of the present invention and does not constitute any limitation on the present invention. Any simple modifications, alterations, or equivalent structural changes made to the above embodiments based on the technical essence of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for screening grain morphology in laser additive manufacturing of titanium alloys, characterized in that, The method includes the following steps: Step 1: Determine the additive elements and addition range for laser additive manufacturing of titanium alloys: The additive elements for laser additive manufacturing of titanium alloys include one or more of Al, V, Cu, Sn, Fe, Ni, Cr, Mo and Zr, and the addition range of each additive element for laser additive manufacturing of titanium alloys is 1 to 10 wt.%. Step 2: Construct a dataset of solidification behavior of laser micro-area metallurgical samples of titanium alloys to be laser-added, the specific process of which is as follows: Step 201: Use a computer and finite element calculation software to establish a geometric model of the powder bed composition of the metallurgical sample to be laser-treated in a micro-area. Step 202: Calculate the thermophysical parameters of the metallurgical sample to be laser-treated in step 201, and assign the thermophysical parameters to the geometric model of the powder bed of the metallurgical sample to be laser-treated. Step 203: Assign boundary conditions and initial temperature to the powder bed geometric model of the metallurgical sample to be laser-treated in the micro-area; Step 204: Set up multiple different laser irradiation process schemes, wherein the laser irradiation process scheme includes laser power P and laser irradiation time t; and simultaneously set the absorption rate of the laser to the metallurgical sample in the micro-area to be lasered. Step 205: Use finite element analysis software to generate a mesh and obtain the temperature change information at the end of laser treatment under different laser treatment process schemes, and obtain the temperature gradient G; Step 206: By calculating the difference in the isotherm profile distance at different times, the solidification rate V is obtained, and a solidification behavior dataset of the micro-area metallurgical sample to be laser-treated is constructed. The solidification behavior dataset includes multiple sets of solidification behavior data, and each set of solidification behavior data includes the temperature gradient G and the solidification rate V. Step 3: Prepare laser micro-area metallurgical samples under different laser irradiation process schemes. The laser micro-area metallurgical samples use the same composition as in Step 1. The specific preparation process for laser micro-area metallurgical samples under different laser irradiation process schemes is as follows: According to the laser power P and laser irradiation time t set under the laser irradiation process scheme, a series of micro-area fixed-point heating is performed through laser irradiation to obtain laser micro-area metallurgical samples corresponding to different thermal behavior conditions. The laser irradiation process scheme includes the laser power P and the laser irradiation time t. Step 4: Collect grain morphology data for laser micro-area metallurgical samples under different thermal behavior conditions: Measure the grain length a and width b of the laser micro-area alloyed sample perpendicular to the powder bed plane in Step 3, and calculate the aspect ratio c, where c = a / b. Step 5: Collect the solidification temperature range ΔT under non-equilibrium solidification conditions: Calculate the Scheil solidification curves corresponding to different titanium alloy compositions using the Thermo-Calc software TTTi3 database, obtain the temperatures at the start and end of solidification, and calculate the solidification temperature range ΔT. Step 6: Data Denoising: Based on the laser micro-area metallurgical sample results from Step 3, delete data with excessive porosity and poor fusion caused by G and V not being within the corresponding threshold range or by compositional characteristics. Step 7, Machine Learning: Select a machine learning model, using the composition and content of the laser micro-area metallurgical sample, the solidification temperature range ΔT, the temperature gradient G, and the solidification rate V as the input layer nodes of the machine learning model, and the grain length a, width b, and aspect ratio c as the output layer nodes of the machine learning model, and train the machine learning model. Step 8: Grain morphology screening of titanium alloys to be laser-added: Input the composition and content of the alloy sample to be laser-added, the solidification temperature range ΔT, the temperature gradient G, and the solidification rate V into the trained machine learning model to screen out alloy compositions that meet the equiaxed structure requirements of the titanium alloy to be laser-added.

2. The method for grain morphology screening of laser additive manufacturing titanium alloys according to claim 1, characterized in that: The finite element calculation software includes Comsol or Ansys finite element calculation software.

3. The method for grain morphology screening of laser additive manufacturing titanium alloys according to claim 1, characterized in that: In step 202, the thermophysical parameters of the laser micro-region metallurgical sample in step 201 are calculated using Jmat-Pro or Thermo-Calc software. The thermophysical parameters include the density, specific heat capacity, thermal conductivity and enthalpy of the laser micro-region metallurgical sample.

4. The method for grain morphology screening of laser additive manufacturing titanium alloys according to claim 1, characterized in that: The boundary conditions of the geometric model of the laser micro-area metallurgical sample powder bed include thermal conductivity, heat transfer coefficient, and radiation coefficient.

5. The method for grain morphology screening of laser additive manufacturing titanium alloys according to claim 1, characterized in that: The machine learning models include multinomial regression, K-nearest neighbor regression, linear regression, artificial neural network, and random forest regression.

Images