Three-period minimal curved surface structure path planning method for laser powder bed melting

Abstract

A scanning path planning method for a metal laser powder bed fusion forming process comprises the steps that 1, a TPMS lattice model is built through an implicit function equation, slicing is conducted through an implicit slicing method, and a two-dimensional contour is obtained; 2, generating a multi-stage initial contour line path through contour surface constants of the uniform change implicit function equation; 3, calculating the distance between the adjacent initial isoline paths, and identifying an area with the distance exceeding a preset tolerance as an unqualified area and the rest as a qualified area; 4, retaining an initial isoline path in the qualified area, and generating a grid filling path in the unqualified area to obtain a mixed path set; 5, inputting a mixed path set, and on the premise of not changing geometric coordinates of each path section, carrying out connection sequence optimization on all the path sections to form a continuous scanning track; and 6, converting the formed continuous scanning track into a file format which can be executed by laser powder bed melting equipment. According to the method, the powder adhesion and deformation risks are remarkably reduced, and the SLM forming precision and efficiency are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of additive manufacturing technology, in particular to a three-period minimal surface structure path planning method for laser powder bed fusion (hereinafter referred to as LPBF). BACKGROUND

[0002] The laser powder bed fusion (LPBF) technology manufactures three-dimensional parts by layering metal powder melting, and the scanning path planning directly affects the density, residual stress, dimensional accuracy and surface quality of the formed parts. The three-period minimal surface (hereinafter referred to as TPMS) structure is widely used in the fields of aerospace, biomedicine, etc. due to its smooth surface, highly interconnected porous characteristics and mathematical controllability. However, the TPMS structure has complex morphology, and its slice cross section often has an "elongated and narrow" special-shaped area. The traditional path planning methods (such as grid path and contour offset path) have obvious defects in such areas: grid path: in the "elongated and narrow" area, "insufficient filling" is prone to occur, and "step effect" is generated on the contour edge, resulting in serious powder adhesion and reduced dimensional accuracy. Contour offset path: when the path converges to the center, "overfilling" or "underfilling" is prone to occur, resulting in internal hole defects and discontinuous molten pool. These problems limit the printing quality and engineering application of TPMS structure. The existing technology lacks a special path planning scheme for the geometric characteristics of TPMS, therefore, there is an urgent need for an innovative method that can achieve uniform filling and improve the forming quality. SUMMARY

[0003] In view of the problems existing in the prior art, the present application provides a scanning path planning method for metal laser powder bed fusion (LPBF) forming process, which aims to solve the problem of uneven filling in TPMS structure printing through implicit slicing, path generation, region detection and optimization connection, so as to be suitable for high-quality preparation of TPMS porous structure.

[0004] In order to achieve the above-mentioned purpose, the specific scheme of the present application is as follows:

[0005] The scanning path planning method for metal laser powder bed fusion forming process comprises the following steps:

[0006] Step 1: constructing a lattice model according to the implicit function equation of the three-period minimal surface structure, and performing implicit slicing on the implicit function equation to obtain two-dimensional contour data of each slice layer;

[0007] Step 2: in each slice layer, calculating the offset value of the constant C of the isosurface based on the preset scanning interval N as the step size, generating multiple initial isocline paths corresponding to the two-dimensional contour data in step 1, and obtaining the constant sequence C of the n scanning lines according to the n scanning lines n ;

[0008] Step 3, calculate the distance between adjacent initial contour paths in step 2, and set the area with a distance greater than 1.2 times the scanning interval N and less than 1.8 times the scanning interval N as an unqualified area, and the rest as a qualified area; the range of 1.2 times to 1.8 times the scanning interval N is a preset tolerance range;

[0009] Step 4, retaining the initial contour path in the qualified area in step 3 and generating a grid filling path in the unqualified area, the distance between parallel lines of the grid filling path is equal to N, and a mixed path set containing contour path segments and grid path segments is obtained;

[0010] Step 5, using the mixed path set obtained in step 4 as the entire input, connecting and optimizing the sequence of all path segments without changing the geometric coordinates of each path segment to form a continuous scanning trajectory;

[0011] Step 6, converting the continuous scanning trajectory formed in step 5 into a file format executable by a laser powder bed fusion device.

[0012] Further, step 1 generates an isosurface of a three-period minimal surface using an implicit function equation f(x, y, z) = C, where C is the isosurface constant; set C1 and C2 as two isosurface constants, and C1 > C2, which define the outer contour and inner contour of the lattice, respectively; the implicit slicing method uses the Marching Squares algorithm to slice the three-period minimal surface structure and the implicit function equation, and the slicing layer thickness is set to 20-100 μm.

[0013] Further, the three-period minimal surface structure in step 1 is of the Primitive type, and its implicit function equation is specifically:

[0014] ,

[0015] In the formula, C is the equation value controlling the position of the isosurface; L x , L y , and L z represent the length of a single unit cell in the x, y, and z directions, respectively, and the isosurface parameters are set to C1 = 0.18 and C2 = -0.18 to define the outer and inner contours, respectively, and the unit size L x = L y = L z =π.

[0016] Further, the three-period minimal surface structure in step 1 is of the Gyroid type or Diamond type, and C1 and C2 are also used to define the outer contour and the inner contour, respectively.

[0017] Further, the offset value C' of the isosurface constant in step 2 is calculated as follows: ,

[0018] In the formula, C' represents the offset value of the lattice isosurface when generating the scanning line; C1 and C2 are isosurface parameters for determining the inner and outer contours of the lattice, wherein C1>C2; N represents the set scanning line spacing, and the preset scanning line spacing N is set to 0.06mm, which is matched with the laser spot diameter of 50-100μm; represents the minimum wall thickness of the current slice layer;

[0019] The isosurface constant sequence C n is calculated as follows: ,

[0020] In the formula, n represents the parameter of the isosurface corresponding to the nth scanning line; C n represents the isosurface constant sequence.

[0021] Further, step 31: taking two adjacent scanning lines as input, for each point on one of the scanning lines, calculating the Euclidean distance L between the point and the point on the other scanning line with the same X coordinate, and dividing the Euclidean distance L by to obtain a modified distance;

[0022] Step 32: when the modified distance is greater than 1.5 times the scanning line spacing N, recording the coordinates of the point and the previous point as a pair of cut-off points;

[0023] Step 33: taking all cut-off points as boundaries, dividing the slice plane into a qualified region and an unqualified region, wherein the modified distance in the qualified region is not greater than 1.5N, and there is at least one place in the unqualified region where the modified distance is greater than 1.5N.

[0024] Further, the grid filling path in step 4 adopts a layer rotation strategy, the rotation angle is 67°, and the grid filling path is generated by the Clipper library and intersects with the contour to ensure the coverage area.

[0025] Further, the optimization in step 5 adopts a traveling salesman problem algorithm for optimization, which is realized by the SciPy library of Python.

[0026] Further, the executable file format of the laser powder bed fusion equipment in step 6 is a CLI file or a JOB file, the CLI file includes layer height, contour data and filling path, and the JOB file includes laser parameters for controlling the laser powder bed fusion equipment, and the laser parameters include laser power, scanning speed and layer thickness

[0027] Advantages of the present application

[0028] 1. The scanning path planning method for the metal laser powder bed fusion forming process of the application avoids overfilling and underfilling by generating an isopleth path using the TPMS implicit function equation, which is naturally parallel to the geometric curved surface.

[0029] 2. The application is based on the minimum wall thickness The offset step is dynamically calculated to ensure that there is at least one path at the thinnest point, achieving uniform wall thickness and zero missing scans.

[0030] 3. The application has continuous molten pool, reduced internal porosity, high profile accuracy, reduced surface powder adhesion, reduced laser switch frequency, reduced heat accumulation, reduced deformation risk, improved profile accuracy and molten pool continuity, and good forming quality.

[0031] 3. The application has strong universality, and the same process is applicable to all TPMS types such as Primitive, Gyroid, Diamond, etc., and can be directly extended to gradient TPMS without the need for re-modeling.

[0032] 4. The application uses implicit slicing to skip STL, saving computing resources, optimizing TSP path order, shortening idle travel, and shortening printing time by 10-20%.

[0033] 5. The application can achieve uniform wall thickness of TPMS structure, reduce empty jump and laser switch frequency, significantly reduce powder adhesion and deformation risk, and improve SLM forming precision and efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0034] Figure 1 The method is a three-period minimal surface structure path planning method for laser powder bed fusion.

[0035] Figure 2 It is a comparison of the traditional slicing and TPMS lattice direct slicing algorithm process.

[0036] Figure 3 It is an algorithm for determining the minimum wall thickness of the lattice slice.

[0037] Figure 4 It is a path generated directly from the equation.

[0038] Figure 5 It is an algorithm for detecting areas that need to be optimized.

[0039] Figure 6 It is an optimization process of a hybrid path planning combined with a grid path.

[0040] Figure 7 It is a grid path detection and scan line optimization.

[0041] Figure 8Scan path for model 79th layer.

[0042] Figure 9 Scan path for model 72nd layer. DETAILED DESCRIPTION

[0043] The present application will be further explained and described with reference to the drawings and specific embodiments, it should be noted that the specific embodiments are not used to limit the scope of the present application.

[0044] As Figures 1 to 9 shown, the scanning path planning method for metal laser powder bed fusion forming process provided by the specific embodiment takes the TPMS structure of the Primitive type as an example, and uses 316L stainless steel powder to print on the Han's Laser M100 device, and specifically includes the following steps:

[0045] Step 1, according to the implicit function equation of the target TPMS structure, a lattice model is constructed, which is different from the traditional STL grid modeling. In this embodiment, the implicit slicing method is directly used for slicing the implicit function equation to obtain the two-dimensional contour data of each slice layer, avoiding the calculation burden and precision loss caused by the traditional STL file slicing; the specific steps are as follows:

[0046] An isosurface of a three-period minimal surface is generated by using an implicit function equation f(x, y, z) =C, wherein C is an isosurface constant; by setting an isosurface constant C1 corresponding to the outer contour and an isosurface constant C2 corresponding to the inner contour, wherein C1>C2, the implicit slicing method uses the Marching Squares algorithm to slice the three-period minimal surface structure and the implicit function equation, avoiding the generation of intermediate grid files, and the slice layer thickness is set to 20-100μm to adapt to the process parameters of laser powder bed fusion.

[0047] This embodiment takes the TPMS structure of the Primitive type as an example, and the implicit function equation thereof is specifically as follows:

[0048] (1),

[0049] In the formula, C is the equation value for controlling the position of the isosurface; L x , L y and L z represent the length of a single unit cell in the x, y and z directions respectively, the isosurface parameter is set to C1=0.18 to define the outer contour, C2=-0.18 to define the inner contour, and the unit size L x = L y = L z= π, and the designed lattice size is 3.144 mm. The lattice wall thickness and porosity are defined. By adjusting C1 and C2, the volume fraction of the lattice is controlled. The implicit slicing method is used to slice the mathematical equation directly, avoiding the generation of STL intermediate files. The Marching Squares algorithm is used to implement it in the Python environment: the lattice model domain is discretized into a grid with a resolution of 0.1 mm, the relationship between each grid point and the isosurface is calculated by interpolation, and the slicing contour is extracted by linear interpolation. The slicing layer thickness is set to 30 μm to adapt to the typical parameters of the SLM process. As shown in FIG. 3, the traditional slicing algorithm is compared with the direct slicing algorithm. Figure 2

[0050] If the TPMS structure is Gyroid or Diamond, C1 and C2 are also used to define the outer contour and the inner contour, respectively.

[0051] Step 2, first call Figure 3 algorithm to process the lattice contour composed of the outer contour generated by the isosurface constant C1 and the inner contour generated by the isosurface constant C2 for the current slicing layer, and calculate the minimum wall thickness : determine whether there is a containment relationship between the outer contour and the inner contour; if there is, match them in pairs; if there is not, divide the two types of contours into two groups according to the X coordinate and then match them. Then, calculate the distance between each pair of points on the matched contour, record the minimum distance and the corresponding point coordinates; after traversing all the points, output the minimum value of and the corresponding coordinates. Based on the preset scanning pitch N = 0.06 mm (matching the laser spot diameter of 50 μm), the isosurface constant sequence C n is generated in the following way:

[0052] In step 1, C1, C2 and of each slicing layer are obtained, and based on the preset scanning pitch N, the offset value of the isosurface constant C is calculated as the step size , generating multiple initial isosurface line paths corresponding to the two-dimensional contour data described in step 1, starting from C1 and C2 respectively, and uniformly offsetting inward along the wall thickness direction, and obtaining the isosurface constant sequence C n of n scanning lines through different scanning lines, ensuring that the initial isosurface line path fits the TPMS geometric morphology.

[0053] The formula of the isosurface constant sequence C n is as follows:

[0054] (2),

[0055] (3),

[0056] ​In formula (2) and formula (3), C' represents the offset value of the lattice equivalent surface when generating the scanning line; C1 and C2 are equivalent surface parameters for determining the inner and outer contours of the lattice, wherein C1 > C2; N represents the set scanning line spacing, and the preset scanning line spacing N is set to 0.06 mm, which is matched with the laser spot diameter of 50-100 μm; n represents the parameter of the equivalent surface corresponding to the nth scanning line; C n represents an equivalent surface constant sequence; represents the minimum wall thickness of the current slice layer; this ensures that the path is naturally offset based on the TPMS topography and fits the profile height. The generated initial contour line path is as shown in Figure 4 , but there is a gap in the partial area that needs to be optimized.

[0057] Step 3, calculate the spacing between adjacent initial contour line paths in step 2, and set the area with a spacing greater than 1.2 times the scanning spacing N and less than 1.8 times the scanning spacing N as an unqualified area, and the remaining area as a qualified area; the range of 1.2 times to 1.8 times the scanning spacing N is a preset tolerance range;

[0058] Input: all initial contour line paths in the current slice layer, the path direction is uniform clockwise, and the scanning spacing N = 0.06 mm.

[0059] As shown in Figure 5 , the specific steps of the spacing between adjacent initial contour line paths are as follows:

[0060] 1. Extract two adjacent paths A and B in order; the path A is one scanning line, and the path B is another scanning line;

[0061] 2. Determine whether the path A is contained in the path B:

[0062] If not, find the nearest point on the other path B using the starting point of the path A, and continue to the next step;

[0063] If yes, continue to the next step.

[0064] 3. Perform the following operations on each point on the path A:

[0065] a) Find the corresponding point with the same X coordinate on the path B as the point;

[0066] b) Calculate the Euclidean distance L between the two points;

[0067] c) Correct the Euclidean distance: ;

[0068] d) Check whether the corrected Euclidean distance L is greater than 1.5 N (i.e., 0.09 mm):

[0069] If greater, record the coordinates of this point and its previous point, and record the IDs of paths A and B;

[0070] If not greater, continue to the next point;

[0071] 4. Until all points of path A are processed, return to step 1 to continue the next pair of adjacent paths;

[0072] 5. After all pairs of adjacent paths are processed, output the recorded coordinates of the cut-off points and the IDs of paths A and B corresponding to them.

[0073] Step 4: In the eligible area described in step 3, keep the initial contour line paths, which are parallel to the contour, to ensure edge smoothness and uniformity of filling; in the unqualified area, use grid path filling to generate grid filling paths, which use interlayer rotation strategy with a rotation angle of 67° to reduce residual stress and improve the continuity of the molten pool, with a scanning interval of 0.06 mm. The grid filling paths are generated by the Clipper library and intersected with the contour to ensure coverage. The parallel line distance of the grid filling paths is equal to N, resulting in a mixed path set containing both contour line segments and grid path segments; this mixed strategy combines the geometric adaptability of TPMS paths and the robustness of grid paths, and the optimization process is as shown in Figure 6 .

[0074] Step 5: Take the mixed path set obtained in step 4 as the total input, and on the premise of not changing the geometric coordinates of each path segment, use the Traveling Salesman Problem algorithm (TSP) to optimize the connection order of all path segments to form a continuous scanning trajectory; the optimization is realized by the SciPy library of Python, which is used to calculate the shortest connection path to reduce empty travel and the number of laser switchings.

[0075] Assuming that the endpoints of the path segments are considered as cities, and the connection distance between path segments is the weight. The TSP solver in the SciPy library of Python is used to calculate the shortest connection path to reduce empty travel and the number of laser switchings. The paths generated directly from the TPMS implicit function equation are deleted in the unqualified area; the contour lines in the unqualified area are offset, and the offset area is filled with grid paths; the TSP algorithm globally optimizes the paths, and the optimized paths are as shown in Figure 7 . The continuity of the optimized paths is improved, the printing time is reduced by about 15%, and the molten pool continuity is improved. The filling is more uniform.

[0076] Step 6, path format conversion: converting the continuous scanning trajectory formed in step 5 into a CLI file or a JOB file format executable by a laser powder bed fusion device. The CLI file contains layer height, contour data and filling path, and the JOB file includes laser parameters for controlling the laser powder bed fusion device, the laser parameters including laser power, scanning speed and layer thickness, the laser power being preferably 75 W, the scanning speed being preferably 600 mm / s, and the layer thickness being preferably 30 μm.

[0077] Laser powder bed fusion device compatibility: adapted to Han's Laser M100 system, and the path file is imported through the device software.

[0078] Print verification: 316L gas atomized powder (particle size 10-40 μm) is used to print in an argon atmosphere (oxygen content <0.01%).

[0079] Comparing the experimental results of the contour offset path, the grid path segment and the improved path of the 79th layer and the 72nd layer of the lattice model, it is shown that the new path reduces powder adhesion and internal porosity, and realizes uniform distribution of filling lines in long and narrow areas. This eliminates the uneven areas in the contour path and the loss of filling lines in the grid path segment, thereby improving the printing quality of the TPMS lattice. As shown in Figure 8 and Figure 9 .

[0080] The method of the above embodiment can help to realize uniform filling and improve the forming quality of a three-period minimal surface.

Claims

1. A scanning path planning method for metal laser powder bed fusion molding process, characterized in that, Includes the following steps: Step 1: Construct a lattice model based on the implicit function equation of the three-period minimal surface structure, and implicitly slice the implicit function equation to obtain the two-dimensional contour data of each slice layer. Step 2: Within each slice layer, calculate the offset value C' of the isosurface constant based on the preset scan interval N as the step size, generate multiple initial contour paths corresponding to the two-dimensional contour data described in Step 1, and obtain the isosurface constant sequence C of n scan lines based on the n scan lines. n ; Step 3: Calculate the spacing between adjacent initial contour lines in Step 2. Regions with a spacing greater than 1.2 times the scan spacing N and less than 1.8 times the scan spacing N are designated as unqualified regions, while the remaining regions are qualified regions. The range of 1.2 to 1.8 times the scan spacing N is a preset tolerance range. Step 4: In the qualified areas described in Step 3, the initial contour paths are retained, and grid-filled paths are generated in the unqualified areas. The parallel line spacing of the grid-filled paths is equal to N, resulting in a mixed path set that simultaneously contains contour path segments and grid path segments. Step 5: Using the mixed path set obtained in Step 4 as the entire input, optimize the connection order of all path segments without changing the geometric coordinates of each path segment to form a continuous scanning trajectory. Step 6: Convert the continuous scanning trajectory formed in Step 5 into a file format executable by the laser powder bed melting equipment.

2. The method according to claim 1, characterized in that, Step 1 uses the implicit function equation f(x, y, z) = C to generate the isosurface of the three-period minimal surface, where C is the isosurface constant; two isosurface constants C1 and C2 are set, with C1 > C2, to define the outer and inner contours of the lattice, respectively; the implicit slicing method uses the Marching Squares algorithm to slice the three-period minimal surface structure and the implicit function equation, with the slice layer thickness set to 20-100 μm.

3. The method according to claim 2, characterized in that, The three-period minimal surface structure described in step 1 is of the Primitive type, and its implicit function equation is as follows: ， In the formula, C is the equation value controlling the position of the isosurface; L x L y and L z These represent the lengths of individual cells in the x, y, and z directions, respectively. The isosurface parameters C1=0.18 and C2=-0.18 define the outer and inner contours, respectively, with cell size L. x = L y =L z =π.

4. The method according to claim 2, characterized in that, The three-period minimal surface structure mentioned in step 1 is of the Gyroid or Diamond type, and C1 and C2 are used to define the outer and inner contours respectively.

5. The method according to claim 1, characterized in that, The formula for the offset value C' of the isosurface constant mentioned in step 2 is as follows: ， In the formula, C' represents the offset value of the lattice isosurface when generating scan lines; C1 and C2 are isosurface parameters that determine the inner and outer contours of the lattice, where C1>C2; N represents the set scan line spacing, and the preset scan line spacing N is set to 0.06mm, which matches the laser spot diameter of 50-100μm; Indicates the minimum wall thickness of the current slice layer; The isosurface constant sequence C n The formula is as follows: ， In the formula, n represents the parameter of the isosurface corresponding to the nth scan line; C n This represents the sequence of isosurface constants.

6. The method according to claim 1, characterized in that, The method for calculating the spacing between adjacent initial contour line paths in step 3 includes the following steps: Step 31: Using two adjacent scan lines as input, for each point on one scan line, calculate the Euclidean distance L between it and points on the other scan line with the same X-coordinate, and divide the Euclidean distance L by... Obtain the corrected distance; Step 32: When the correction distance is greater than 1.5 times the scanning interval N, record the coordinates of the point and the previous point as a pair of cutoff points; Step 33: Using all cut-off points as boundaries, divide the slice plane into qualified and unqualified regions. The correction distance in the qualified region is no greater than 1.5 N, and there is at least one correction distance greater than 1.5 N in the unqualified region.

7. The method according to claim 1, characterized in that, The grid fill path described in step 4 adopts an inter-layer rotation strategy with a rotation angle of 67°, and the grid fill path is generated by the Clipper library and intersected with the contour.

8. The method according to claim 1, characterized in that, The optimization described in step 5 uses the Traveling Salesman Problem algorithm, implemented using the SciPy library in Python.

9. The method according to claim 1, characterized in that, The executable file format of the laser powder bed melting device described in step 6 is a CLI file or a JOB file. The CLI file contains layer height, contour data and fill path, and the JOB file includes laser parameters for controlling the laser powder bed melting device, including laser power, scanning speed and layer thickness.

Images