Automatic wire placement thickening curved surface modeling method based on fast marching method

Abstract

An automatic wire placement thickening curved surface modeling method based on a fast marching method comprises the steps that composite material design information serves as input, a mold curved surface is voxelized, and a distance field is efficiently calculated through the fast marching method; then voxel edges are traversed, an equidistant point set is obtained through linear interpolation, and an original equidistant curved surface is established by applying a triangulation algorithm; finally, the grid segmentation algorithm is used for cutting to obtain thickened and non-thickened area curved surfaces, the two curved surfaces are sewn, and the final thickened curved surface is obtained.Compared with a traditional curved surface offset method, the method has the advantages that the problem of layering surface modeling under the condition that the thickness of a large-curvature component is increased is effectively solved, controllable high calculation precision is achieved, and the calculation process is more stable.

Description

Technical Field

[0001] The present invention belongs to the technical field of automatic wire placement of composite materials, and in particular relates to an automatic wire placement thickening surface modeling method based on a rapid marching method. Background Art

[0002] As an advanced automated manufacturing process for composite materials, automatic placement has the advantages of strong designability and adaptability to complex surfaces. Therefore, it is widely used in the manufacture of complex composite components with large curvature.

[0003] Thickness, as one of the core design parameters of composite components, has a decisive influence on component performance. In applications requiring high load-bearing capacity, composite components are typically designed with a greater thickness to ensure the structure possesses sufficient mechanical properties. Furthermore, in fields such as aerospace that pursue lightweighting, designers often optimize the thickness distribution of components to reduce the energy consumption of aircraft. Shape changes caused by thickness are particularly prominent on complex curved surfaces with large curvature, posing a challenge to the high-quality automated placement and manufacturing of complex components with large curvature.

[0004] In the automatic placement process, some existing path planning methods take into account the impact of layer thickness on the wire placement trajectory, such as the patent application entitled "A method for designing an automatic wire placement trajectory for laying the edge of a thick layer of product" (publication number CN119427787A). The basic principle is to offset the mold surface along its normal direction by a certain height, and then perform path planning on the thickened surface to compensate for the normal deviation of the path caused by the accumulation of material thickness. However, for components with negative curvature characteristics, the offset along the normal direction may cause software errors due to self-intersection of the surface, or produce thickening results that are not strictly equal in thickness. Using a non-strictly equal in thickness surface for path planning may cause over-pressure or air pressure on the roller during the placement process, or the wire length may not match the movement distance of the roller, resulting in placement defects such as bridging and wrinkling, thereby reducing the mechanical properties of the component. Summary of the Invention

[0005] In order to overcome the shortcomings of the above-mentioned prior art, the purpose of the present invention is to provide an automatic wire laying and thickening surface modeling method based on the rapid marching method, which can accurately generate strictly equal-thickness laying surfaces, properly handle the problem of surface self-intersection, reduce the occurrence of thickness-related laying defects, and has the advantages of high-precision and high-stability thickening.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is as follows:

[0007] A method for modeling an automatic wire placement thickening surface based on a rapid marching method comprises the following steps:

[0008] 1) Reading design information: The design information of the mold surface, material thickness, and ply boundary is obtained from the composite material design file. The mold surface is discretized and stored in a triangular mesh format. The data structure includes mesh vertices and the three vertex indices corresponding to each triangle face. The ply boundary is stored as a set of discrete points.

[0009] 2) Mold surface voxelization and distance field calculation: The mold surface is voxelized, and the side length of the voxel unit is determined according to the following formula:

[0010]

[0011] Where a represents the size of the voxel unit, R represents the radius of curvature of the surface, and h represents the thickness of the single layer;

[0012] The distance field is calculated in the voxel grid using the fast marching method. The fast marching method uses the vertex of the voxel unit occupied by the surface as the wave source point, selects the adjacent unknown distance points in turn, solves the eikonal equation, and obtains the distance value. The method is gradually expanded until all vertices are calculated.

[0013] 3) Extracting the isotropic surface as the original isopach surface: Traverse the edges of all voxel units. When two vertices of the edge cross the target distance value, obtain a point on the isotropic surface through linear interpolation and add this point to the set of isotropic points. After the traversal is completed, use the triangulation algorithm to build a mesh surface on the set of isotropic points to obtain the original isopach surface.

[0014] 4) Split and join to obtain the thickened surface: Project the current ply boundary onto the original constant-thickness surface generated in step 3) and use a mesh segmentation algorithm to split the original constant-thickness surface into two parts, retaining the area surrounded by the current ply boundary. Project the current ply boundary onto the surface before thickening to split the area outside the area surrounded by the current ply boundary. Use the stitching tool to stitch these two mesh parts into a complete thickened surface.

[0015] 5) Loop to generate the surface of each layer: set the thickened surface obtained in step 4 as the bottom surface of the next layer, read the ply boundary of the next layer, and then execute steps 1)-4) again until all boundaries are processed.

[0016] Step 2) Before the fast marching method begins, an initialization operation is first performed to set the vertex distance values ​​of all voxel units to infinity. Then, the voxel units occupied by the mold surface are traversed, the projection distance from the unit vertex to the mold surface is calculated, and the projection distance is saved to the corresponding vertex. After the initialization is completed, the fast marching method is executed in the voxel space.

[0017] Step 4) is specifically as follows: to obtain the thickened area surface, the current ply boundary is projected onto the original isopach surface, and the mesh segmentation algorithm is used to segment and extract the thickened area mesh surface M1; to obtain the non-thickened area surface, the current ply boundary is projected onto the underlying surface, and the mesh segmentation algorithm is used to obtain the mesh surface M2 outside the boundary; using the mesh stitching tool, a strip of triangular facets M3 is generated between the M1 and M2 meshes to fill the gaps between the M1 and M2 meshes caused by the height difference; using the mesh union operation, the M1, M2, and M3 meshes are spliced ​​together to obtain the final thickened surface.

[0018] Compared with the prior art, the present invention has the following beneficial effects:

[0019] (1) The thickening surface generated by the present invention satisfies a strict equal-thickness relationship with the underlying surface, and is closer to the actual ply surface than the results generated by traditional algorithms. Since most traditional thickening surface modeling methods are based on the principle of surface normal offset, it is difficult to accurately generate an equidistant surface in the tangential discontinuity area of ​​the convex surface, and there is a problem of thickness distortion. The present invention extracts the equidistant surface from the distance field, ensuring that each point on the equidistant surface strictly satisfies the equal-distance relationship with the underlying surface, and can automatically generate equal-thickness transition features in the discontinuous area of ​​the surface, which is closer to the actual laying effect. Using more accurate ply surface modeling results can improve the accuracy of path planning, thereby reducing the occurrence of laying defects.

[0020] (2) Compared with traditional methods, the thickened surface modeling method of the present invention has stronger stability for large curvature concave surfaces. Traditional surface offset methods are prone to abnormal program termination due to surface self-intersection on concave surfaces, or output results with broken surfaces, which cannot be used normally in subsequent path planning and other processes. The thickened surface generation method of the present invention can correctly handle the large thickness of concave surfaces or arbitrarily complex surfaces, effectively solve the problem of self-intersection of offset surfaces, and has stronger stability than traditional methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 Flowchart of an embodiment of the present invention.

[0022] Figure 2 (a) is the shape and boundary setting diagram of an embodiment of the present invention, (b) is the calculation result of the surface offset function using CAD software, (c) is the result of performing grid offset using grid processing software, and (d) is the calculation result using the method of the present invention. DETAILED DESCRIPTION

[0023] The present invention will be described in further detail below with reference to the embodiments and the accompanying drawings. The embodiments take honeycomb structures as an example, and the embodiments are used to illustrate the present invention, but are not used to limit the scope of the present invention.

[0024] Reference Figure 1 , an automatic wire placement thickening surface modeling method based on a rapid marching method, comprising the following steps:

[0025] 1) Reading design information: Key design information such as mold surface, material thickness, and ply boundaries are obtained from the composite material design file. The mold surface is discretized and stored in a triangular mesh format. The data structure includes mesh vertices and the three vertex indices corresponding to each triangle face. The ply boundaries are stored as a set of discrete points.

[0026] This embodiment reads the honeycomb digital model and ply boundary information from the CATIAPart document. The shape and boundary settings are as follows: Figure 2 As shown in (a), the target thickness of the thickening is set to 50 mm. In the case of a layer thickness of 0.2 mm, this is equivalent to laying down about 250 layers of material;

[0027] 2) Mold surface voxelization and distance field calculation: First, the mold surface is voxelized, that is, it is covered with a set of equal-sized cubic grids. The side length of the voxel unit is determined according to the following formula:

[0028]

[0029] Where a represents the size of the voxel unit, R represents the curvature radius of the surface, and h represents the thickness of a single layer. The voxel side length calculated using the above formula can ensure that the calculation error is less than the single layer thickness h. The measurement results of this embodiment show that the minimum R corner radius of the honeycomb surface is 5 mm, and the layer thickness h is set to 0.2 mm. It is calculated that the voxel size a of the honeycomb should be less than 2.8 mm.

[0030] Then, the distance field is calculated in the voxel grid using the fast marching method. The fast marching method is an efficient solution to interface evolution problems and has been widely used in the field of wave optics. Before the fast marching method starts, an initialization operation is first performed. The vertex distance values ​​of all voxel units are set to infinity. Then, the voxel units occupied by the mold surface are traversed, the projection distance from the unit vertex to the mold surface is calculated, and the projection distance is saved to the corresponding vertex. After the initialization is completed, the fast marching method is executed in the voxel space. The fast marching method uses the vertex of the voxel unit occupied by the surface as the wave source point, and selects the adjacent unknown distance points in turn to solve the eikonal equation to obtain the distance value, and gradually expands until all vertices are calculated.

[0031] This embodiment uses a voxelization tool to voxelize the mold surface at a granularity of 2.5 mm to obtain a spatial voxel unit grid, which constitutes the solution domain of the fast marching method. The voxel units occupied by the surface are separated, and the voxel units are traversed, and their vertices are projected onto the surface. The projection distance is assigned to the voxel vertices as the initial condition of the fast marching method. The initial distance field is then propagated to the entire voxel space using the fast marching method.

[0032] 3) Extracting the equidistant surface as the original isopach surface: Traverse all voxel units and traverse each edge within each voxel unit to determine whether the two vertices of the current edge are on different sides of the target distance value. If so, use linear interpolation to obtain a target distance point and add it to the equidistant point set. After the traversal is completed, triangulate the points in the equidistant point set to obtain the original isopach surface. At this time, the isopach surface is generated within the entire range of the mold surface and needs to be further segmented according to the ply boundaries to retain the thickened surface in the required area.

[0033] 4) Split and join to obtain the thickened area: The original isopach surface covers all areas of the underlying surface, but only the area surrounded by the current ply boundary is required. Therefore, the original isopach surface needs to be segmented. In order to obtain the thickened area surface, the current ply boundary is projected onto the original isopach surface, and the mesh segmentation algorithm is used to segment and extract the thickened area mesh surface M1; in order to obtain the non-thickened area surface, the current ply boundary is projected onto the underlying surface, and the mesh segmentation algorithm is used to obtain the mesh surface M2 outside the boundary; using the mesh stitching tool, a strip of triangular facets M3 is generated between the M1 and M2 meshes to fill the gaps caused by the height difference between the M1 and M2 meshes; using the mesh union operation, the M1, M2, and M3 meshes are spliced ​​together to obtain the final thickened surface;

[0034] 5) Loop through the ply surfaces: Set the thickened surface obtained in step 4 as the base surface of the next layer, read the ply boundaries of the next layer, and then repeat steps 1–4 until all boundaries are processed. After each boundary is processed, export the thickened surface of that layer to disk for use in subsequent path planning or finite element analysis.

[0035] Since this embodiment only requires the uniform thickness thickening result of a single boundary-enclosed area, all boundaries have been processed after one thickening pass, and there is no need to repeat the thickening surface generation step. The final calculation result of the thickened surface is as follows: Figure 2 As shown in (d).

[0036] As a comparison with the method in this paper, Figure 2 (b) shows the calculation results using the surface offset function of the CAD software. Figure 2(c) shows the result of performing a mesh offset using mesh processing software. As can be seen, due to the complexity of the surface, the CAD software offset error occurred, making this method unsuitable for generating equidistant surfaces. While the mesh offset method can produce offset results, the offset surface is broken at corners due to self-intersection. Furthermore, the offset mesh becomes coarse at edges, which means that the surface accuracy is reduced. In contrast, the proposed method can correctly generate thickened surfaces. Furthermore, since the accuracy of the thickened surface is only related to the size of the voxel unit, the accuracy of the thickened surface is not affected by the surface curvature and thickness, resulting in a smoother thickening result.

[0037] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Any simple modification or equivalent change made to the above embodiment based on the technical essence of the present invention shall fall within the scope of protection of the present invention.

Claims

1. A method for modeling a surface of an automatic wire laying and thickening process based on a rapid marching method, characterized in that: The following steps are involved: 1) Reading design information: The design information of the mold surface, material thickness, and ply boundary is obtained from the composite material design file. The mold surface is discretized and stored in a triangular mesh format. The data structure includes mesh vertices and the three vertex indices corresponding to each triangle face. The ply boundary is stored as a set of discrete points. 2) Mold surface voxelization and distance field calculation: The mold surface is voxelized, and the side length of the voxel unit is determined according to the following formula: Where a represents the size of the voxel unit, R represents the radius of curvature of the surface, and h represents the thickness of the single layer; The distance field is calculated in the voxel grid using the fast marching method. The fast marching method uses the vertex of the voxel unit occupied by the surface as the wave source point, selects the adjacent unknown distance points in turn, solves the eikonal equation, and obtains the distance value. The method is gradually expanded until all vertices are calculated. 3) Extracting the isotropic surface as the original isopach surface: Traverse the edges of all voxel units. When two vertices of the edge cross the target distance value, obtain a point on the isotropic surface through linear interpolation and add the point to the set of isotropic points. After the traversal is completed, a triangulation algorithm is used to build a mesh surface on the set of equidistant points, thereby obtaining the original equal-thickness surface; 4) Split and join to obtain the thickened surface: Project the current ply boundary onto the original constant-thickness surface generated in step 3) and use a mesh segmentation algorithm to split the original constant-thickness surface into two parts, retaining the area surrounded by the current ply boundary. Project the current ply boundary onto the surface before thickening to split the area outside the area surrounded by the current ply boundary. Use the stitching tool to stitch these two mesh parts into a complete thickened surface. 5) Loop to generate the surface of each layer: set the thickened surface obtained in step 4 as the bottom surface of the next layer, read the ply boundary of the next layer, and then execute steps 1)-4) again until all boundaries are processed.

2. The method according to claim 1, wherein: Step 2) Before the fast marching method begins, an initialization operation is performed to set the vertex distance values ​​of all voxel units to infinity. Then, the voxel units occupied by the mold surface are traversed, the projection distance from the unit vertex to the mold surface is calculated, and the projection distance is saved to the corresponding vertex; After initialization, fast marching is performed in voxel space.

3. The method according to claim 1, characterized in that Step 4) is specifically as follows: to obtain the thickened area surface, the current ply boundary is projected onto the original isopach surface, and the mesh segmentation algorithm is used to segment and extract the thickened area mesh surface M1; to obtain the non-thickened area surface, the current ply boundary is projected onto the underlying surface, and the mesh segmentation algorithm is used to obtain the mesh surface M2 outside the boundary; using the mesh stitching tool, a strip of triangular facets M3 is generated between the M1 and M2 meshes to fill the gaps between the M1 and M2 meshes caused by the height difference; using the mesh union operation, the M1, M2, and M3 meshes are spliced ​​together to obtain the final thickened surface.

Images