A superalloy blisk and method of additive manufacturing thereof

Abstract

This invention discloses an additive manufacturing method for a high-temperature alloy integral bladed disk. Using electron beam current and scanning speed as variables, and other process parameters as fixed, several alloy samples are printed, and their relative densities are measured. A machine learning method is used to establish a machine learning prediction model for relative density, with relative density as the output. The machine learning model predicts the relative density, establishing an optimized process window for the corresponding alloy. Using this optimized process window, and based on the principle of achieving a denser alloy in additive manufacturing, the range of scanning speed and electron beam current variables is determined. Within the selected range, the scanning speed and electron beam current are chosen, and the scanning spacing is changed to print several alloy samples. The grain structure of each alloy sample is characterized to determine its microstructure. Analysis of the grain structure yields the optimal process parameters for the integral bladed disk. Finally, the integral bladed disk is printed using these optimal process parameters to produce an integral bladed disk with the required microstructure and performance.

Description

Technical Field

[0001] This invention belongs to the field of high-temperature alloy preparation technology, and particularly relates to a high-temperature alloy integral bladed disk and its additive manufacturing method. Background Technology

[0002] Integral bladed disks (IBDs) have gradually become core components of next-generation aero-engines, and their manufacturing technology signifies the development level of the entire aerospace manufacturing industry and even defense technology. IBDs integrate the disk and blades into a single unit, eliminating unnecessary connecting and supporting components and avoiding problems caused by improper assembly processes. This results in a significant reduction in the overall weight of the bladed disk, improved efficiency and reliability, and a significantly longer service life.

[0003] However, the integration of the bladed disk into a single unit has resulted in a very complex structure. Furthermore, the bladed disk requires the disk to have a fine equiaxed crystal structure while the blades have a coarse columnar crystal structure. The significant difference in the grain structure between the disk and the blades makes the integrated manufacturing process of the bladed disk a challenge for the industry. Summary of the Invention

[0004] The main objective of this invention is to provide a high-temperature alloy integral bladed disk and its additive manufacturing method, aiming to produce an integral bladed disk with the required microstructure and performance.

[0005] Therefore, one aspect of the present invention provides an additive manufacturing method for a high-temperature alloy integral bladed disk, comprising:

[0006] Selective electron beam laser melting was chosen, with electron beam current and scanning speed as variables and other process parameters fixed. Several alloy samples were printed and their relative densities were measured.

[0007] Using machine learning methods, a machine learning prediction model for relative density is established with relative density as the output.

[0008] The relative density is predicted by machine learning models, and the optimized process window for the corresponding alloy is established by combining the relative density contour map and surface morphology processing parameters.

[0009] By optimizing the alloy's process window, and taking the principle of making the additive manufacturing alloy more dense, the scanning speed and the range of electron beam current variables are determined.

[0010] Within a defined range of variables, the scanning speed and electron beam current were selected, and several alloy samples were printed by changing the scanning spacing.

[0011] The grain structure was determined by characterizing each alloy sample. The process parameters corresponding to the alloy sample with a fully columnar crystal structure were used as the blade printing process parameters, and the process parameters corresponding to the alloy sample with a fully equiaxed crystal structure were used as the disk printing process parameters.

[0012] The disk is printed using the disk printing process parameters determined above. Then, blades are printed directly on the disk using the blade printing process parameters determined above, to obtain an integral bladed disk.

[0013] Specifically, the electron beam laser selective melting process parameters were as follows: accelerating voltage 60KV, scanning spacing 100μm, layer thickness 50μm, spot size 150μm, substrate preheating temperature 900℃, electron beam current range 7.5-27.5mA with an interval of 2.5mA, scanning speed range 2000mm / s-8000mm / s with an interval of 1000mm / s. A total of 63 alloy samples were constructed, and their relative densities were measured.

[0014] Specifically, by optimizing the alloy's process window and prioritizing the density of the additive manufacturing alloy, two regions were determined for the scanning speed and electron beam current variable range, namely P1 and P2; among them,

[0015] The scanning speed corresponding to region P1 is 2000-3000 mm / s, and the electron beam current is 8-11 mA.

[0016] The scanning speed for the P2 region is 3800-4500 mm / s, and the electron beam current is 15.0-17.5 mA.

[0017] Specifically, the blade printing process parameters are: electron beam current 10mA, scanning speed 2.5m / s, scanning spacing 50μm, and energy density 96J / m³. 3 The accelerating voltage is 60KV, the layer thickness is 50μm, the spot size is 150μm, and the substrate preheating temperature is 900℃.

[0018] Specifically, the disk printing process parameters are: electron beam current 16.8mA, scanning speed 4.2m / s, scanning spacing 50μm, and energy density 96J / m³. 3 The accelerating voltage is 60KV, the layer thickness is 50μm, the spot size is 150μm, and the substrate preheating temperature is 900℃.

[0019] Specifically, the alloy is IN718.

[0020] Specifically, a machine learning model with relative density as the output is constructed using support vector machine regression and Gaussian process regression.

[0021] Specifically, the grain structure of the alloy samples was determined by SEM characterization.

[0022] In another aspect, the present invention provides a high-temperature alloy integral bladed disk manufactured using the above-described additive manufacturing method.

[0023] Principles and advantages

[0024] Using electron beam current and scanning speed as two key parameters and the relative density of the alloy sample as the output, an optimal process window is obtained by combining machine learning methods. Then, based on the principle of achieving a denser alloy in additive manufacturing, the range of scanning speed and electron beam current variables is determined within the optimal process window. Within the determined variable range, the scanning speed and electron beam current magnitude are selected. By changing the scanning spacing of the key parameter, several alloy samples are printed. By analyzing the grain structure of each alloy sample, the process parameters corresponding to obtaining samples with a fully columnar crystal structure and a fully equiaxed crystal structure are determined. The disk body and blades of the integral bladed disk are printed using these process parameters. Machine learning methods are used to adjust the relevant parameters of the additive manufacturing process, thereby obtaining an integral bladed disk with an ideal microstructure. This method has the advantages of low workload, low cost, and high reliability. Using the relative density of the alloy sample as the output, the additive manufacturing of high-temperature alloys achieves density, and its mechanical properties are superior to those of forgings without the need for hot isostatic pressing. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a flowchart of the process parameters for determining the fabrication of an integral bladed disk provided by an embodiment of the present invention;

[0027] Figure 2 This is a schematic diagram of the optimized process window involved in an embodiment of the present invention;

[0028] Figure 3 These are the inverse pole figure and pole figure of the middle of the vertical cross section of the sample in region P1 involved in the embodiments of the present invention;

[0029] Figure 4 These are the inverse pole figure and pole figure of the middle of the vertical cross section of the sample in region P2 involved in the embodiments of the present invention;

[0030] Figure 5 This is a schematic diagram of a single molten pool of the P1 region sample and the P2 region sample involved in the embodiments of the present invention;

[0031] Figure 6 This is a schematic diagram of two adjacent molten pools of the P1 region sample and the P2 region sample involved in the embodiments of the present invention;

[0032] in: Figure 3 In the middle, (a) and (d) represent the inverse pole figure and pole figure of the sample when the scanning interval is 100 μm, respectively; (b) and (e) represent the inverse pole figure and pole figure of the sample when the scanning interval is 75 μm, respectively; and (c) and (f) represent the inverse pole figure and pole figure of the sample when the scanning interval is 50 μm, respectively.

[0033] Figure 4 In the middle, (a) and (d) represent the inverse pole figure and pole figure of the sample when the scanning interval is 100 μm, respectively; (b) and (e) represent the inverse pole figure and pole figure of the sample when the scanning interval is 75 μm, respectively; and (c) and (f) represent the inverse pole figure and pole figure of the sample when the scanning interval is 50 μm, respectively.

[0034] Figure 6 (a) represents a schematic diagram of two adjacent molten pools when the scanning interval of the sample in region P1 is 100 μm; (b) represents a schematic diagram of two adjacent molten pools when the scanning interval of the sample in region P2 is 100 μm; (c) represents a schematic diagram of two adjacent molten pools when the scanning interval of the sample in region P1 is 50 μm; and (d) represents a schematic diagram of two adjacent molten pools when the scanning interval of the sample in region P2 is 50 μm. Detailed Implementation

[0035] The technical solutions of 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.

[0036] This application uses IN718 alloy as an example to describe in detail the additive manufacturing method of integral bladed disks. The average particle size of the powder is approximately 35 μm, and the chemical composition of the IN718 alloy (by mass percentage) is as follows:

[0037] Ni: 50-55%, Cr: 17-21%, Nb: 4.75-5.5%, Mo: 2.8-3.3%, Ti: 0.65-1.15%, Al: 0.2-0.8%, Cu: ≤0.3%, Co: ≤1%, C: ≤0.08%, Mn: ≤0.35%, S: ≤0.015%, P: ≤0.015%, balance Fe.

[0038] See Figure 1 The specific preparation steps of the additive manufacturing method for integral bladed disks are as follows:

[0039] Step 1: Electron beam laser selective melting process parameters: accelerating voltage 60KV, scanning spacing 100μm, layer thickness 50μm, spot size 150μm, substrate preheating temperature 900℃, electron beam current range 7.5-27.5mA, spacing 2.5mA, scanning speed range 2000mm / s-8000mm / s, spacing 1000mm / s. A total of 63 samples were constructed, and their relative densities were measured.

[0040] Step 2: Construct a machine learning model with relative density as the output using Support Vector Machine Regression (SVR) and Gaussian Process Regression (GPR).

[0041] Step 3: Using a machine learning model, predict the relative density of IN718 alloy. Combine this with relative density contour plots and surface morphology processing parameters to establish an optimized process window for electron beam laser selective melting forming of IN718 alloy. Specifically, as follows... Figure 2 As shown.

[0042] Step 4: In the process window optimization, the alloy density is higher near P1 and P2, resulting in a denser alloy. Select process parameters within this range. The parameter range near P1 is: scanning speed: 2000-3000 mm / s, electron beam current: 8-11 mA; the parameter range near P2 is: scanning speed: 3800-4500 mm / s, electron beam current: 15.0 mA-17.5 mA.

[0043] Step 5: Within the defined variable range, i.e., regions P1 and P2, select the scanning speed and electron beam current respectively, and print several alloy samples by changing the scanning spacing.

[0044] Step 6: Characterize each alloy sample to determine the grain structure. Use the process parameters corresponding to the alloy sample with a fully columnar crystal structure as the blade printing process parameters, and use the process parameters corresponding to the alloy sample with a fully equiaxed crystal structure as the disk printing process parameters, thereby obtaining the optimal process parameters for the overall bladed disk preparation.

[0045] The disk is printed using the disk printing process parameters determined above. Then, blades are printed directly on the disk using the blade printing process parameters determined above, to obtain an integral bladed disk.

[0046] This application uses electron beam current and scanning speed as two key parameters, with the relative density of the alloy sample as the output, and combines machine learning methods to obtain the optimal process window. Then, based on the principle of achieving a denser alloy in additive manufacturing, the scanning speed and electron beam current variable range are determined within the optimal process window. Within the determined variable range, the scanning speed and electron beam current magnitude are selected. By changing the scanning spacing, several alloy samples are printed. By analyzing the grain structure of each alloy sample, the process parameters corresponding to obtaining samples with a fully columnar crystal structure and a fully equiaxed crystal structure are determined. The disk body and blades of the integral bladed disk are printed using these process parameters. Machine learning methods are used to adjust the relevant parameters of the additive manufacturing process, thereby obtaining an integral bladed disk with an ideal microstructure. This method has the advantages of low workload, low cost, and high reliability. Using the relative density of the alloy sample as the output, the additive manufacturing of high-temperature alloys is dense, and its mechanical properties are superior to those of forgings without the need for hot isostatic pressing.

[0047] Specifically, the processes in steps 5 and 6 are as follows:

[0048] In regions P1 and P2, a set of parameters is selected respectively. For example, in region P1, the electron beam selective melting process is selected with an electron beam current of 10mA and a scanning speed of 2.5m / s, and the other parameters remain unchanged as shown in step 1. In region P2, the electron beam selective melting process is selected with an electron beam current of 16.8mA and a scanning speed of 4.2m / s, and the other parameters remain unchanged as shown in step 1. Only the scanning spacing is changed, so that the scanning spacing parameter is reduced from 100μm to 50μm, with an interval of 25 micrometers, to form a high-temperature alloy sample.

[0049] Through analysis of SEM and EBSD images of the samples, specifically as follows: Figure 3 As shown in Figure 4, it was determined that at a scanning interval of 50 μm, the process parameters in region P1 yielded fully columnar crystals, and region P2 yielded fully equiaxed crystals, thus achieving grain structure control of the nickel-based superalloy IN718.

[0050] To obtain a fully columnar crystal structure, the electron beam selective melting process should be selected with an electron beam current of 10 mA, a scanning speed of 2.5 m / s, a scanning spacing of 50 μm, and an energy density of 96 J / m². 3 The accelerating voltage is 60KV, the layer thickness is 50μm, the spot size is 150μm, and the substrate preheating temperature is 900℃, which are the blade printing process parameters.

[0051] To obtain a fully equiaxed grain structure, the electron beam selective melting process should be selected with an electron beam current of 16.8 mA, a scanning speed of 4.2 m / s, a scanning spacing of 50 μm, and an energy density of 96 J / m². 3 The acceleration voltage is 60KV, the layer thickness is 50μm, the spot size is 150μm, and the substrate preheating temperature is 900℃, which are the process parameters for disk printing.

[0052] If the above steps still do not yield a fully columnar or fully equiaxed grain structure, then select another set of parameters in regions P1 and P2 respectively and repeat the above steps until a fully columnar or fully equiaxed grain structure is obtained.

[0053] Using the aforementioned blade and disk printing process parameters, a monolithic disk was printed. Tensile specimens of the blade and disk were then cut from the monolithic disk. The relative density of both the blade and disk tensile specimens was ≥99.6%. Tensile tests were conducted on the blade and disk tensile specimens, and the properties are as follows:

[0054] At room temperature, the blade tensile specimen has σb≥1280MPa, σ0.2≥1120MPa, and A≥30%; at 650℃, the blade tensile specimen has σb≥1000MPa, σ900≥900MPa, and A≥28%.

[0055] At room temperature, the σ of the disc tensile specimen b ≥1380MPa, σ0.2≥1200MPa, A≥24%; at 650℃, the tensile specimen of the disc has σb≥1110MPa, σ0.2≥900MPa, and A≥15%;

[0056] The data above shows that the obtained integral bladed disk has a dense structure, and its high-temperature mechanical properties are significantly better than those of forged high-temperature alloys without the need for hot isostatic pressing.

[0057] Considering that electron beam current and scanning speed are the main parameters affecting density in the SEBM process, and that their effects are more significant and have a wider range of variation, this application selects these two parameters as variables, takes the relative density of the alloy sample as the output, and uses machine learning methods to obtain the optimal process window. Then, based on the selection principle of relatively dense alloys, the range of electron beam current and scanning speed is determined within the optimal process window. Subsequently, within the selected range, the optimal process parameters for both columnar and equiaxed crystal structures are obtained by changing another key parameter, the scanning spacing.

[0058] The reason why columnar and equiaxed crystal structures can be obtained by adjusting the scanning interval in this application is as follows:

[0059] The inventors discovered that, at the same energy density, a lower scan rate results in a smaller aspect ratio and a shorter longitudinal length in the molten pool. The P2 region, with its higher scan rate, produces a shallower, wider, and longer molten pool, while the P1 region produces a deeper, narrower, and shorter molten pool. Figure 5As shown, the shallower molten pool in region P2 facilitates rapid cooling and uniform cooling across all areas. During crystallization, the growth direction of crystal nuclei tends to be disordered, resulting in the formation of equiaxed crystals. In region P1, the molten pool is deeper, and the undercooling is different between the internal and surface regions. Crystals tend to grow from the pre-crystallized region into the liquid, thus easily forming columnar crystal regions.

[0060] like Figure 6 As shown, when the scanning spacing is 100 μm, there is a certain overlap and void region between the two molten pools, indicated by dashed circles. The sample in region P1 has a smaller overlap and a larger void region. The void region may be smaller than the powder particle size, thus preventing fusion defects. However, the void region may form aggregated equiaxed crystals because these equiaxed crystals retain the original grains of the incompletely melted powder. Therefore, reducing the molten pool spacing (scanning spacing) in region P1 can effectively reduce molten pool overlap and the equiaxed crystal content in region P1, resulting in a fully columnar crystal structure. Equiaxed crystals easily form in the molten pool of region P2. Adjusting the scanning spacing (molten pool spacing) makes the solidification temperature gradient more uniform, which is beneficial for the formation of equiaxed crystals. Therefore, the scanning spacing is selected for optimization.

[0061] The temperature gradient deviation angle between the narrower and deeper molten pools in the P1 region sample is smaller. After multiple molten pools are stacked, a spatial temperature gradient almost parallel to the printing direction is formed in the thickness of one layer, similar to a directional solidification process. Due to the higher scanning speed, the temperature gradient in the molten pool of the P2 region sample produces a larger deviation angle towards the scanning direction between different molten pools. This deviation alternates with the change of scanning direction, thus forming a temperature gradient with different directions but uniform in space. The uniform temperature gradient results in a wide compositional supercooled zone in the liquid formed at the solid-liquid interface front of the unsolidified area. The entire liquid can nucleate, and dendrites grow uniformly in all directions, resulting in better equiaxed grains.

[0062] The above embodiments are merely illustrative examples to clearly illustrate the present invention and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all embodiments here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method for additive manufacturing of a high-temperature alloy integral bladed disk, characterized in that, include: Selective electron beam laser melting was chosen, with electron beam current and scanning speed as variables and other process parameters fixed. Several alloy samples were printed and their relative densities were measured. Using machine learning methods, a machine learning prediction model for relative density is established with relative density as the output. The relative density is predicted by machine learning models, and the optimized process window for the corresponding alloy is established by combining the relative density contour map and surface morphology processing parameters. By optimizing the alloy's process window, and taking the principle of making the additive manufacturing alloy more dense, the scanning speed and the range of electron beam current variables are determined. Within a defined range of variables, the scanning speed and electron beam current were selected, and several alloy samples were printed by changing the scanning spacing. The grain structure was determined by characterizing each alloy sample. The process parameters corresponding to the alloy sample with a fully columnar crystal structure were used as the blade printing process parameters, and the process parameters corresponding to the alloy sample with a fully equiaxed crystal structure were used as the disk printing process parameters. The disk body is printed using the disk body printing process parameters determined above. Then, blades are printed directly on the disk body using the blade printing process parameters determined above, to obtain an integral bladed disk. By optimizing the alloy's process window and prioritizing a denser additive manufacturing alloy, two regions, P1 and P2, were determined for the scanning speed and electron beam current variable range; among them, The scanning speed corresponding to region P1 is 2000-3000 mm / s, and the electron beam current is 8-11 mA. The scanning speed corresponding to the P2 region is 3800-4500 mm / s, and the electron beam current is 15.0-17.5 mA; The blade printing process parameters are: electron beam current 10mA, scanning speed 2.5m / s, scanning spacing 50μm, and energy density 96J / m³. 3 The accelerating voltage is 60KV, the layer thickness is 50μm, the spot size is 150μm, and the substrate preheating temperature is 900℃. Disk printing process parameters: electron beam current 16.8mA, scanning speed 4.2m / s, scanning spacing 50μm, energy density 96J / m³. 3 The accelerating voltage is 60KV, the layer thickness is 50μm, the spot size is 150μm, and the substrate preheating temperature is 900℃. The alloy is IN718.

2. The additive manufacturing method for a high-temperature alloy integral bladed disk according to claim 1, characterized in that: The electron beam laser selective melting process parameters were as follows: accelerating voltage 60KV, scanning spacing 100μm, layer thickness 50μm, spot size 150μm, substrate preheating temperature 900℃, electron beam current range 7.5-27.5mA with an interval of 2.5mA, and scanning speed range 2000mm / s-8000mm / s with an interval of 1000mm / s. A total of 63 alloy samples were constructed, and their relative densities were measured.

3. The additive manufacturing method for a high-temperature alloy integral bladed disk according to claim 1, characterized in that: We construct a machine learning model with relative density as the output using support vector machine regression and Gaussian process regression.

4. The additive manufacturing method for a high-temperature alloy integral bladed disk according to claim 1, characterized in that: The grain structure of the alloy samples was determined by SEM characterization.

5. The additive manufacturing method for a high-temperature alloy integral bladed disk according to claim 1, characterized in that: The average particle size of the alloy is 30-40 μm.

6. A high-temperature alloy integral bladed disk, characterized in that: It is manufactured using the additive manufacturing method for high-temperature alloy integral bladed disks as described in any one of claims 1-5.

Images