Self-adaptive support structure generation method for pull-up photocuring 3D printing

Abstract

According to the self-adaptive supporting structure generation method for pull-up type photocuring 3D printing, external force serves as a constraint condition to be included in the supporting generation process, and therefore the number and the positions of supports are optimized; layers with similar features are gathered together to significantly reduce the time cost of generating the support and reduce the number of simulations. According to the method, multiple factors such as model geometrical characteristics, material attributes, the printing direction and external force distribution are comprehensively considered to generate the optimal supporting structure, the use of the supporting structure can be minimized while successful printing is ensured, and therefore the efficiency and quality of 3D printing are improved.

Description

Technical Field

[0001] The present invention belongs to the field of additive manufacturing, and in particular relates to a method for generating an adaptive support structure for top-pull light-curing 3D printing. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] With the rapid development of additive manufacturing technology, 3D printing has become a revolutionary technology in the manufacturing industry. This technology enables designers to quickly create complex CAD models that are difficult to achieve with traditional methods. In particular, light-curing 3D printing technology is in a leading position in terms of material diversity, manufacturing accuracy and speed, and can print a variety of materials including elastic materials, ceramic materials, rigid materials and metal materials.

[0004] However, the widespread application of 3D printing technology faces a key challenge: the generation of adaptive support structures. Support structures are an important factor in ensuring the successful printing of 3D models. They not only connect the printed parts to the printing platform, but also need to resist external forces that may cause printing failure. Especially in top-pull light-curing 3D printing, release force becomes a factor that cannot be ignored. However, existing mechanically guided support optimization methods mainly focus on gravity or residual stress, and there is still a lack of support optimization methods for this release force.

[0005] The design and optimization of support structures face two main constraints: external force constraints and overhang constraints. When the external force exceeds the bearing capacity of the support structure, it may cause the support structure to break or deform, which in turn causes printing failure. At the same time, if the overhanging areas are not properly supported, these areas may sag or collapse during the printing process. Therefore, developing an algorithm that can minimize the number of support structures while satisfying these two constraints has become an important research direction in the development of 3D printing technology.

[0006] Currently, designers often need to make a lot of manual adjustments to ensure the effectiveness of the support structure. However, even after adjustments, the support structure is still at risk of breaking during the actual printing process, which may lead to serious component defects or even failure of the entire printing task. In order to increase the success rate of printing, designers often add too many support structures. However, this practice not only increases the complexity and time cost of the post-processing process, but may also damage the surface of the printed object during the removal process, thereby affecting the quality and aesthetics of the final product. Summary of the invention

[0007] In order to solve the above problems, the present invention proposes an adaptive support structure generation method for top-pull light-curing 3D printing. The present invention comprehensively considers multiple factors such as model geometric characteristics, material properties, printing direction, and external force distribution to generate an optimal support structure. It can minimize the use of support structure while ensuring successful printing, thereby improving the efficiency and quality of 3D printing.

[0008] According to some embodiments, the present invention adopts the following technical solutions:

[0009] A method for generating an adaptive support structure for top-pull light-curing 3D printing, comprising the following steps:

[0010] Build a 3D model, slice the 3D model to obtain a series of 2D cross-sectional images, preprocess them, cluster the preprocessed slice images, add constraints and select representative layers;

[0011] According to the position of the representative layer, the 3D model with added support is segmented and mechanical simulation is performed to determine the stress field area and the force condition of each support point of the model after segmentation;

[0012] Distinguish strong overhang areas and weak overhang areas based on the stress field obtained from mechanical simulation;

[0013] According to the stress conditions of each supporting point, the force of the supporting point is converted into the force field of the overhanging area, and each supporting point is optimized to obtain an optimized solution.

[0014] As an optional implementation, the process of preprocessing the slice image includes: downsampling the original slice image, binarizing the downsampled image to convert it into an image containing only black and white colors, and normalizing the binarized image to ensure that all pixel values ​​in the image are strictly 0 or 1.

[0015] As an optional implementation, the preprocessed slice images are clustered. The process of adding constraints includes: clustering the preprocessed slice images using a clustering algorithm, using relative distance as a similarity measure in the clustering process, introducing continuity constraints, determining a distance threshold, starting from the first layer, clustering layer by layer, for each layer, calculating its distance from the last layer in the current class, if the distance is less than the distance threshold, adding the layer to the current class; otherwise, starting a new class; repeating the iterative clustering process layer by layer until all layers are classified.

[0016] As a further implementation, the process of introducing the continuity constraint is that if the image indexes between two classes are not continuous, the distance between them is defined as being greater than a preset distance threshold;

[0017] The process of determining the distance threshold is to analyze the distance distribution between the slice images, calculate the mean and standard deviation of the distance, and adaptively determine the optimal distance threshold.

[0018] As an optional implementation, the process of selecting a representative layer includes using a prediction model based on a multi-layer perceptron to predict the Z-direction pulling force required to destroy the vacuum force of the layer based on the processed slice image, calculating the total external force of each layer, and selecting the layer with the largest force as the representative layer of this class.

[0019] As an optional implementation, voxel grids are used to divide the model during mechanical simulation.

[0020] As an optional implementation, the process of distinguishing strong overhang areas from weak overhang areas based on the stress field obtained from mechanical simulation includes: setting a stress threshold, distinguishing strong overhang areas from weak overhang areas based on the comparison between the stress threshold and the force magnitude, and for areas where the stress value exceeds the predetermined threshold, it is a strong overhang area and the support points are increased; for areas where the stress value is less than the predetermined threshold, it is a weak overhang area and the support points are reduced.

[0021] As a further implementation method, a transition zone is provided between the strong and weak overhanging areas so that the support density changes gradually.

[0022] As an optional implementation method, according to the force conditions of each support point, the process of converting the force of the support point into the force field of the overhanging area includes selecting the maximum force encountered by each support point in all slice models as the force of the support point, and using the Gaussian distribution formula to convert the force of the support point into the force field of the overhanging area, and in the process of conversion, adaptively adjusting the expansion coefficient of the support force of each support point mapped to the Gaussian field.

[0023] As an optional implementation method, the process of optimizing each support point includes: constructing a total energy function: E total =E f +λ k E o , the total energy function is minimized by the gradient descent method;

[0024] Where E f Represents the energy caused by the force, E o represents the energy term for overhang constraint optimization, E f It is used to optimize the position of the support points so that the forces borne by all the support points are as similar as possible, thereby minimizing the number of support points. o The optimization goal is to ensure that the support point is the center of the circle and the self-support The circle with a radius of λ can completely cover the strong overhang area, and the calculation process is to calculate the sum of the product of the farthest distance from the support point and its weight in the Voronoi area of ​​each support point;k It is a decreasing coefficient, which decreases gradually with the increase of the number of iterations.

[0025] As an alternative embodiment,

[0026] Where χ(p) represents the density at point p

[0027]

[0028] When the distance is less than the maximum overhang distance d self-support When , it indicates that the support of the current area satisfies the overhang constraint, ω i If it is set to 0, the area will be optimized only according to the force constraint, and the distribution of support points will not be affected by the overhang constraint. Otherwise, it is set to 1, where P is infinite, that is, the infinite norm is taken for ||p-pi||.

[0029] As an optional implementation, a minimum spanning tree algorithm is used to construct the support structure, and support points whose distances are less than a predetermined value are connected by trusses.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] The present invention innovatively proposes an adaptive support structure generation method for top-up light-curing 3D printing, which can automatically optimize the number and distribution of supports according to external force requirements, realize the automation of support structure generation, and significantly improve the efficiency and quality of 3D printing.

[0032] The present invention achieves a significant reduction in the number of simulations (by comparison, at least 1-2 orders of magnitude) through an innovative clustering method, greatly shortens the time required for simulation, and improves the efficiency of support structure generation.

[0033] The present invention introduces the concept of weak overhang areas, and optimizes the number and layout of supports to meet overhang and force constraints, thereby achieving adaptive adjustment of the support structure, reducing the number of required supports, minimizing damage to the model surface, and reducing manual post-processing time, significantly improving the printing success rate and model quality.

[0034] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0036] Figure 1It is a schematic diagram of the process of distinguishing strong overhang areas from weak overhang areas;

[0037] Figure 2 It is a schematic diagram of adding additional supports in the weak overhang area;

[0038] Figure 3 is an overall flow chart of an embodiment;

[0039] Figure 4 is a schematic diagram of a maximum overhang distance and a maximum overhang angle experiment of an embodiment;

[0040] Figure 5 is a schematic diagram of boundary conditions of a simulation of an embodiment;

[0041] Figure 6 is an iterative force result diagram of an embodiment;

[0042] Figure 7 This is a comparison of the printing results of an embodiment with Materialise Magics e-Stage when the overhang angle is set to 30°. DETAILED DESCRIPTION

[0043] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0044] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0045] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0046] In the absence of conflict, the embodiments in this application and the features in the embodiments may be combined with each other.

[0047] As described in the background technology, the present invention proposes an adaptive support structure generation method for top-pull light-curing 3D printing, aiming to solve the following key technical problems:

[0048] (1) Minimize the impact of the support structure on the model surface while satisfying the force constraints and overhang constraints.

[0049] (2) Reduce the computational complexity and time cost of support structure generation.

[0050] (3) Optimize the distribution of support structures to adapt to the geometric characteristics and stress conditions of complex models.

[0051] (4) Solve the problem of insufficient support caused by overly strict definition of overhanging area in traditional methods.

[0052] Specifically, the adaptive support structure generation method for pull-up light-stereolithography 3D printing incorporates external forces as constraints into the support generation process, thereby optimizing the number and position of supports. At the same time, an efficient clustering method is used to cluster layers with similar characteristics together to significantly reduce the time cost of generating supports and reduce the number of simulations.

[0053] like Figure 3 As shown, the specific steps include:

[0054] Step (1): construct a 3D model, slice the 3D model to obtain a series of 2D cross-sectional images, preprocess the images, cluster the preprocessed slice images, add constraints and select representative layers;

[0055] Step (2): According to the position of the representative layer, the 3D model with added support is segmented and mechanical simulation is performed to determine the stress field area and the force condition of each support point of the model after segmentation;

[0056] Step (3): distinguishing the strong overhang area and the weak overhang area according to the stress field obtained by mechanical simulation;

[0057] Step (4): According to the stress conditions of each support point, the force of the support point is converted into the force field of the overhanging area, and each support point is optimized to obtain an optimized solution.

[0058] The details of each step are as follows:

[0059] Step (1) mainly includes the following steps:

[0060] The main goal of this step is to reduce the number of layers that need to be simulated mechanically by clustering methods, while ensuring that the selected representative layers can accurately reflect the mechanical properties of the entire model. It specifically includes the following sub-steps:

[0061] Step (1-1): Model slicing and clustering preprocessing:

[0062] First, the 3D model is sliced ​​along the printing direction to obtain a series of two-dimensional cross-sectional images. In order to improve the efficiency and accuracy of subsequent clustering, these slice images need to be preprocessed:

[0063] a) Downsampling: Downsample the original slice image to reduce its resolution. This step can significantly reduce the number of pixels in each image, thereby reducing the dimension of the feature vector during clustering. The degree of downsampling needs to strike a balance between retaining sufficient details and reducing the amount of computation.

[0064] b) Binarization: Binarize the downsampled image and convert it into an image containing only black and white. This step can further reduce the complexity of the information while retaining the key geometric features of the model section.

[0065] c) Normalization: Normalize the binarized image to ensure that all pixel values ​​in the image are strictly 0 or 1. This step can eliminate small numerical errors that may be introduced during image processing and provide a consistent basis for subsequent distance calculations.

[0066] Step (1-2): Adaptive clustering:

[0067] The clustering algorithm is used to cluster the preprocessed slice images. The key to this step is to define a suitable distance function and automatically determine the distance threshold:

[0068] a) Relative distance function: In order to accommodate both large and small cross sections that may exist in the model, we use relative distance as the similarity measure. The definition of relative distance is shown in formula (1):

[0069]

[0070] d(i,j) represents the Euclidean distance between image i and image j, and |i| represents the number of non-zero pixels in image i. This definition of relative distance can effectively handle cross-sections of different scales, making the clustering process more robust to changes in the overall shape of the model.

[0071] b) Continuity constraint: To ensure the continuity of slice image indexes for each class, a continuity constraint is introduced in the clustering process. Specifically, if the image indexes between two classes are not continuous, the distance between them is defined as greater than a preset distance threshold. This avoids merging non-adjacent images into the same class and ensures that the images within each class are continuous.

[0072] c) Automatic threshold determination: The method proposed by Rahmah et al. (Rahmah N, Sitanggang IS. Determination of optimal epsilon (eps) value on dbscan algorithm to clustering data on peatland hotspots in sumatra [C] / / IOP conference series: earth and environmental science. IoP Publishing, 2016, 31 (1): 01-2012) is used to automatically calculate the clustering distance threshold ∈ d This method analyzes the distance distribution between slice images, calculates the mean and standard deviation of the distance, and then uses an adaptive formula to determine the optimal distance threshold. This method can automatically adjust the strictness of clustering according to the characteristics of different models and improve the adaptability of the algorithm.

[0073] d) Iterative clustering process: Starting from the first layer, clustering is performed layer by layer. For each layer, its distance to the last layer in the current class is calculated. If the distance is less than a threshold ∈ d , then add the layer to the current class; otherwise, start a new class. This process continues until all layers are classified.

[0074] Step (1-3): Select the representative layer:

[0075] In each cluster, a representative layer needs to be selected for subsequent mechanical simulation. The principle of selecting the representative layer is to ensure that it can represent the most unfavorable mechanical situation in the cluster:

[0076] a) Calculate external force: It mainly consists of two parts: gravity and release force.

[0077] The gravity calculation is relatively simple and can be directly calculated based on the volume and material density of each layer.

[0078] The release force is calculated using a prediction model based on a multi-layer perceptron (MLP). The model uses the slice image as input and predicts the Z-direction pulling force required to break the vacuum force of the layer. This method greatly improves the calculation efficiency compared to directly using the 3D model for simulation.

[0079] b) Select the layer with the largest stress: After calculating the total external force (gravity + release force) of each layer, select the layer with the largest stress as the representative layer of this type. This ensures that the support structure is designed under the most unfavorable conditions.

[0080] In this embodiment, step (2) mainly includes the following steps:

[0081] Step (2-1): By clustering similar layers and selecting representative layers, we determine the specific slice layers that need to be simulated. According to the position of the representative layer, the model with added support is sliced ​​and simulated;

[0082] Step (2-2): Dynamic simulation is used based on the simulation results to improve the accuracy and reliability of the simulation, so as to better simulate the dynamic behavior in the actual production process and ensure that the simulation results can accurately reflect the stress and deformation in actual operation.

[0083] In this embodiment, the boundary conditions of the simulation are as follows: Figure 5 As shown, a uniform velocity in the positive direction of the Z axis is applied to the top end point of the support, while the bottom surface of the PDMS material of the material box is fixed.

[0084] Step (2-3): In terms of meshing, although adaptive meshing can provide more sophisticated mesh management, when adding support structures to the segmented model, it may introduce mesh self-intersection problems, causing the mesh to become non-manifold. In order to solve this problem and ensure the stability and efficiency of the simulation, we use voxel meshing to divide the model.

[0085] a) The use of voxel grids is due to their fast meshing speed, which can greatly improve simulation efficiency.

[0086] b) When iteratively refining the mesh in support structures and key areas, there is no need to re-mesh the entire model, which simplifies the mesh update process.

[0087] The step (3) mainly comprises the following steps:

[0088] In this step, the strong overhang area and the weak overhang area are distinguished according to the previously calculated stress field. Figure 1 This process is crucial for subsequent support optimization because it determines the support strength and density required in different areas. The following is a detailed implementation method:

[0089] Strong overhang areas are those areas where the overhang is large and support structures must be added and overhang constraints need to be met to prevent printing failure. Its definition is the same as the overhang area in traditional methods. These areas are usually at a very small angle to the printing platform, or the overhang length exceeds the self-supporting capacity of the material.

[0090] Weak overhang areas are those areas where the support structure can be expanded. When the support structure in the strong overhang area is not sufficient to meet the force constraints, additional supports can be added in the weak overhang area to improve the overall strength and stability of the support structure, such as Figure 2 as shown in (a) and (b).

[0091] It is worth noting that the weak overhang regions do not need to satisfy the overhang constraints because their main role is to strengthen the supporting structure rather than prevent the collapse of the overhang region.

[0092] Step (3-1): Determine a stress threshold to distinguish between strong and weak overhang areas.

[0093] a) For areas where the force exceeds the predetermined threshold, add support points to prevent breakage;

[0094] b) For areas with less stress, reduce the number of support points to save material and reduce damage to the model surface.

[0095] Step (3-2): Divide the model surface into strong and weak overhang areas:

[0096] a) Strong overhang areas: areas where stress values ​​are higher than the threshold, these areas require stronger support.

[0097] b) Weak overhang areas: areas where stress values ​​are below the threshold, these areas can use lighter supports or no support.

[0098] c) Transition zone: Set a transition zone between the strong and weak overhang areas to make the support density change gradually and avoid sudden changes.

[0099] The step (4) mainly comprises the following steps:

[0100] In this step, based on the previous analysis results, the support structure is fully optimized to balance the force requirements and overhang constraints. This process is the core of the algorithm and directly affects the final printing quality and efficiency. The following is a detailed optimization strategy:

[0101] Step (4-1): The force conditions of each support point of each sliced ​​model are obtained through simulation analysis. In order to ensure the integrity of the support structure during the entire printing process, we select the maximum force encountered by each support point in all sliced ​​models as the force of the support point.

[0102] Step (4-2): After obtaining the maximum force at each support point, convert the support force into the force field of the overhanging area to further optimize the position and number of supports.

[0103] a) Use the Gaussian distribution formula to transform the force at the support point into the force field in the overhanging area, as shown in formula (2):

[0104]

[0105] Among them, F si is the force on the ith support, where μ xi is the x value of the support point, μ yi is the y value of the support point, is the expansion coefficient of the support force of the ith support mapped to the Gaussian field, σ x represents the standard deviation of the Gaussian distribution in the x direction, σ y represents the standard deviation of the Gaussian distribution in the y direction, d self-support Represents the self-supporting distance, and π is the pi constant. Adaptively adjust The value of can be used to expand the Gaussian field range of the support mapping. In this way, the nearby weak overhang area can be included in the force field, as shown in formula (3):

[0106]

[0107] where F break is the stress threshold for support fracture.

[0108] b) An innovative total energy function is proposed to optimize the distribution of support points, which is defined as shown in formula (4):

[0109] E total =λ k E f +E o (4)

[0110] Where E f Represents the energy caused by the force, E o represents the energy term for overhang constraint optimization. E f The purpose is to optimize the position of the support points so that the forces borne by all support points are as similar as possible, thereby minimizing the number of support points. The specific expression is shown in formula (5):

[0111]

[0112] Where χ(p) represents the density at point p; E o The optimization goal is to ensure that the support point is the center of the circle and the self-support The circle with a radius of can completely cover the strong overhang area, which is specifically expressed as shown in formula (6):

[0113]

[0114] This energy term calculates the maximum distance from the support point and its weight ω in the Voronoi region of each support point. i The sum of the products of . Figure 4 As shown, when the distance is less than the maximum overhang distance d self-support , indicating that the support of the current area satisfies the overhang constraint, so ω iSet to 0, so that the area will be optimized only according to the force constraint, and the distribution of support points will not be affected by the overhang constraint. Where P is infinite, that is, the infinite norm is taken for ||p-pi||, ω i It is expressed as shown in formula (7):

[0115]

[0116] In order to ensure the dynamic balance force constraint and overhang constraint during the optimization process, a decreasing coefficient λ is introduced. k , as shown in formula (8):

[0117] λ k =λ k-1 *θ k (8)

[0118] where θ k The value of is less than 1, ensuring that λ k It will gradually decrease with the increase of iteration times. To ensure that the force field constraint is the main optimization target in the initial stage, it is helpful to determine the reasonable initial support position according to the force distribution. In the later stage, for the area that does not meet the overhang constraint, the optimization focus shifts to the overhang constraint.

[0119] Step (4-3): Minimize the energy by gradient descent. Use the minimum spanning tree algorithm to build the support structure, and connect the support points that are close to each other through trusses to enhance the overall stability of the structure and prevent breakage caused by mechanical shaking during the printing process.

[0120] In order to reduce the contact area between the model and the support, the cross-sectional area of ​​the contact area is set smaller than the cross-sectional area of ​​the other parts of the support.

[0121] As an implementation, Figure 6 The top picture shows the number of supports, support force distribution, maximum force, and 3D support results of the dragon model during the iteration process. The bottom picture shows the number of supports and maximum force of the cat, armadillo, sofa, and semi-hollow skeleton model. In the figure, the black line indicates the maximum force during the iteration process, and the blue line shows the number of supports. The specific values ​​are represented in black and blue fonts respectively. The red dotted line represents the maximum force threshold of 1.2N.

[0122] Figure 7The figure below compares the printing results of this method with those of Materialise Magics e-Stage when the overhang angle is set to 30°. The rendering and printing results of the supports generated by Materialise Magics e-Stage show that only the hollow cat model was successfully printed, and the supports of the armadillo model broke during the printing process. The black dotted box shows an enlarged view of the broken supports of the armadillo model. The figure on the right shows the rendering and printing results of the adaptive supports of the method proposed in this embodiment, all of which were successfully printed.

[0123] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0124] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0125] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0126] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0127] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made by those skilled in the art within the spirit and principle of the present invention without creative labor shall be included in the protection scope of the present invention.

Claims

1. A method for generating an adaptive support structure for top-pull light-curing 3D printing, characterized in that: The following steps are involved: Build a 3D model, slice the 3D model, obtain a series of 2D cross-sectional images, preprocess them, cluster the preprocessed slice images, add constraints and select representative layers; According to the position of the representative layer, the 3D model with added support is segmented and mechanical simulation is performed to determine the stress field area and the force condition of each support point of the model after segmentation; Distinguish strong overhang areas and weak overhang areas based on the stress field obtained from mechanical simulation; According to the stress conditions of each supporting point, the force of the supporting point is converted into the force field of the overhanging area, and each supporting point is optimized to obtain an optimized solution.

2. The method for generating an adaptive support structure for top-pull light-curing 3D printing according to claim 1, characterized in that: The preprocessed slice images are clustered, and the process of adding constraints includes: clustering the preprocessed slice images using a clustering algorithm, using relative distance as a similarity measure in the clustering process, introducing continuity constraints, determining a distance threshold, starting from the first layer, and clustering layer by layer. For each layer, calculate its distance from the last layer in the current class. If the distance is less than the distance threshold, add the layer to the current class; otherwise, start a new class; repeat the iterative clustering process layer by layer until all layers are classified.

3. The method for generating an adaptive support structure for top-pull light-curing 3D printing according to claim 2, characterized in that: The process of continuity constraint is that if the image indexes between two classes are not continuous, the distance between them is defined as greater than a preset distance threshold; The process of determining the distance threshold is to analyze the distance distribution between the slice images, calculate the mean and standard deviation of the distance, and adaptively determine the optimal distance threshold.

4. The method for generating an adaptive support structure for top-down photocuring 3D printing according to claim 1, characterized in that: The process of selecting the representative layer includes using a prediction model based on a multi-layer perceptron to predict the Z-direction pulling force required to destroy the vacuum force of the layer based on the processed slice image, calculating the total external force of each layer, and selecting the layer with the largest force as the representative layer of this class; and using voxel grids to divide the model when performing mechanical simulation.

5. The method for generating an adaptive support structure for top-pull light-curing 3D printing according to claim 1, characterized in that: The process of distinguishing strong overhang areas from weak overhang areas according to the stress field obtained from mechanical simulation includes: setting a stress threshold, distinguishing strong overhang areas from weak overhang areas according to the comparison between the stress threshold and the force magnitude, and for areas where the stress value exceeds the predetermined threshold, it is a strong overhang area and the support points are increased; for areas where the stress value is less than the predetermined threshold, it is a weak overhang area and the support points are reduced.

6. The method for generating an adaptive support structure for top-pull light-curing 3D printing according to claim 1, characterized in that: A transition zone is set between the strong and weak overhang areas to make the support density gradually change.

7. The method for generating an adaptive support structure for top-pull light-curing 3D printing according to claim 1, characterized in that: According to the stress conditions of each support point, the process of converting the force of the support point into the force field of the overhanging area includes selecting the maximum force encountered by each support point in all slice models as the force of the support point, and using the Gaussian distribution formula to convert the force of the support point into the force field of the overhanging area. In the conversion process, the expansion coefficient of the support force mapping of each support point into the Gaussian field is adaptively adjusted.

8. The method for generating an adaptive support structure for top-pull light-curing 3D printing according to claim 1, characterized in that: The process of optimizing each support point includes: constructing the total energy function: E total =E f +λ k E o , the total energy function is minimized by the gradient descent method; Where E f Represents the energy caused by the force, E o represents the energy term for overhang constraint optimization, E f It is used to optimize the position of the support points so that the forces borne by all the support points are as similar as possible, thereby minimizing the number of support points. o The optimization goal is to ensure that the support point is the center of the circle and the self-support The circle with a radius of λ can completely cover the strong overhang area, and the calculation process is to calculate the sum of the product of the farthest distance from the support point and its weight in the Voronoi area of ​​each support point; k It is a decreasing coefficient, which decreases gradually with the increase of the number of iterations.

9. The method for generating an adaptive support structure for top-pull light-curing 3D printing according to claim 8, characterized in that: Where χ(p) represents the density at point p; When the distance is less than the maximum overhang distance d self-support When , it indicates that the support of the current area satisfies the overhang constraint, ω i If it is set to 0, the area will be optimized only according to the force constraint, and the distribution of support points will not be affected by the overhang constraint. Otherwise, it is set to 1, where P is infinite, that is, the infinite norm is taken for ||p-pi||.

10. The method for generating an adaptive support structure for top-down photocuring 3D printing according to claim 8, wherein: The minimum spanning tree algorithm is used to construct the support structure, and the support points whose distance is less than the predetermined value are connected by trusses.

Images