Method for determining the orientation of a three-dimensional model for additive manufacturing

Abstract

The present invention relates to a method for determining the orientation of a 3D model (1) to be generated by an additive manufacturing device, the additive manufacturing device comprising: a vat for holding a photocurable material; and a platform (3) for holding a 3D object corresponding to the 3D model (1), wherein the platform is relatively movable with respect to the vat, the method comprising: a step of defining the surface geometry of the 3D model (1), wherein the surface geometry comprises a plurality of surface segments s i , where i represents an integer, A i represents i th the surface area of the surface segment s i ; characterized in that it further comprises: manually by the user or automatically by a computer program, assigning one or more weighting factors f i to the surface segments s i respectively, wherein the weighting factor f i indicates the sensitivity of each surface segment s i to the effect caused by the mechanical removal of any support structure thereon, wherein for a surface segment s i considered to be sensitive, the weighting factor f i is greater than 1, and for all other surface segments s i , f i is equal to 1; a step of defining an evaluation function #imgabs0#; wherein Θ and #imgabs1# represent the polar angle and the azimuth angle of the orientation of the 3D model with respect to the build direction respectively, P supp，i represents the probability of need to be supported by a support structure for each surface segment s i , and the summation represented by ∑ extends over all surface segments s i ; and a step of determining the orientation of the 3D model (1) relative to the platform by optimizing with respect to the polar angle and the azimuth angle respectively based on the evaluation function R to avoid or minimize as much as possible the support structure in the sensitive surface segments of the 3D model.

Description

Technical Field

[0001] The present invention relates to a method for additive manufacturing of three-dimensional objects. More specifically, the present invention relates to a method for determining the orientation of a 3D model to be generated by additive manufacturing. Background Art

[0002] In additive manufacturing, a 3D object is printed layer by layer by photo-induced curing of a liquid printing medium, i.e., a liquid photocurable resin, which is selectively cured under the influence of UV radiation. In a known variant of additive manufacturing, the 3D object is preferably pulled out upside down from the liquid photocurable material filled in a bucket by means of a platform. Other variants of additive manufacturing are also known to a person skilled in the art.

[0003] During 3D printing, a suitable support structure must be attached to the 3D object. The position on the surface of the 3D object to which the support structure must be attached depends on the geometry of the 3D object to be 3D printed and the orientation of the 3D object relative to the build direction.

[0004] In the art, it is generally known which points of a given geometry require a support structure for a given orientation of a 3D object. For example, US 2015 / 0151492A1 discloses a method for generating a support structure for a 3D object to be additive manufactured. Further reference is made to P. Alexander et al., “Part orientation and build cost determination in layered manufacturing”, Computer-Aided Design, vol. 30, no. 5, pp 343-356, 1998, Elsevier Science Ltd. In addition, EP 2922029A2 discloses a system for visualizing a three-dimensional (3D) model printed from a 3D printer.

[0005] After 3D printing, cleaning, thermal and / or photochemical post-curing of the 3D object, the support structure must be mechanically removed. The removal process is time-consuming and can lead to a change in the geometry of the printed 3D object, for example when removing support residues with a rotary tool or when small cracks are caused by the support removal process. For dental components such as dental restorations, drilling templates, dental models, etc., which have high quality requirements for shape accuracy, the results of mechanical post-processing can become critical. Summary of the Invention

[0006] It is an object of the present invention to overcome the disadvantages of the prior art and to provide a method for determining the orientation of a 3D model to be generated by an additive manufacturing device taking into account the local quality requirements of the resulting 3D object.

[0007] This object is achieved by the method defined in claim 1. The dependent claims relate to further improvements.

[0008] The present invention provides a method for determining the orientation of a 3D model to be generated by an additive manufacturing device, the additive manufacturing device comprising: a vat for holding a photocurable material; and a platform for holding a 3D object corresponding to the 3D model, wherein the platform is relatively movable with respect to the vat. The method comprises the following steps:

[0009] 1. Defining the surface geometry of the 3D model, wherein the surface geometry comprises a plurality of surface segments s i , where i represents an integer, A i represents i th the surface area of surface segment s i . According to the present invention, the surface geometry of the 3D model is preferably represented by triangulation because triangulation is convenient for calculation. Alternatively, other types of meshes with different geometries can also be used.

[0010] 2. Manually by the user or automatically by a computer program, assigning one or more weighting factors f i to the surface segments s i respectively, wherein the weighting factor f i indicates the sensitivity of each surface segment s i to the effect caused by the removal of any support structure thereon. Preferably, for surface segments s i considered sensitive to the post-treatment of the removal of any support structure, the weighting factor f i is greater than 1, and for all other surface segments s i , f i is equal to 1.

[0011] 3. Defining an evaluation function

[0012]

[0013] where θ and respectively represent the polar angle and the azimuthal orientation angle of the 3D object with respect to the build direction, p supp，i represents the probability of need for each surface segment s i to be supported by a support structure, and the summation represented by ∑ extends over all surface segments s i . The evaluation function R is normalized by the total surface area and has a negative overall sign. Alternative normalizations can be selected. For example, normalization by the sum of all support probabilities can provide decoupling of the avoidance of support structures on sensitive surface segments from the reduction in the total number of support structures.

[0014] 4. Based on the evaluation function R, the orientation of the 3D model (1) relative to the platform is determined by optimizing with respect to the polar angle and the azimuth angle respectively to avoid or minimize as much as possible the support structure attached to the sensitive surface section of the 3D model.

[0015] The main advantageous effects of the present invention are that by determining the orientation of the 3D model to be printed, the position of the 3D model to which the support structure must be attached can be influenced. More specifically, by optimizing the evaluation function R to determine the orientation, the support structure in the sensitive surface section of the 3D model can be avoided or reduced, but instead positioned in other insensitive areas. Thereby, surface artifacts that may occur due to mechanical post-processing in the sensitive surface section can be avoided or reduced, and the need to invest additional manual work for precise removal of the support structure can be avoided or minimized as much as possible. Therefore, the user can be spared the damage of accepting manually reworked surfaces. Another main advantageous effect of the present invention is that by assigning the weighting function f, the removal process of any support structure itself can also be affected. In the case of dental components, when assigning the weighting factor f, the surface sections of the 3D model to be printed that require less sensitivity compared to other parts, or the surface sections where the support structure can be removed more easily and / or quickly compared to other parts, or the surface sections where less damage is expected from manual post-processing compared to other parts can be easily identified and taken into account. Another main advantageous effect of the present invention is that due to the evaluation function R based on the weighting factor f, the method steps provide an automatic algorithm for avoiding attachment of the support structure at such positions where there is an increased need for protection against the negative impact from mechanical post-processing compared to other positions of the 3D model to be printed. Thereby, the user does not need to manually orient the 3D object and / or place the support structure. Manual manipulation of the placement of the support structure requires the user's experience and is only possible to a limited extent, and thus may increase the risk of misprinting. Therefore, another main advantageous effect of the present invention is that the risk of misprinting can be reduced.

[0016] According to the present invention, the orientation of the 3D model to be printed is determined by optimizing the evaluation function R. Therefore, at the optimized orientation of the 3D model, the evaluation function R must have an extreme value. Using the negative total algebraic sign of the evaluation function R, the case where the weighting factor f of the sensitive surface section is >1 and the others are 1, and the positive probability value p defined as above supp，i , the best orientation is given by the maximum value of the evaluation function R. Therefore, the present invention provides an accurate algorithmic solution to find the optimized orientation regarding the protection of the surface section that is sensitive to the negative impact of mechanically removing the support structure attached thereto.

[0017] Generally, when a surface section with a high degree of sensitivity is arranged to be oriented away from the construction platform or at a small angle with respect to the construction direction, the support structure to be positioned on such a surface section can be avoided or minimized as much as possible. This principle can be implemented in the present invention through several different evaluation functions R, which can be used to determine the optimized orientation of the 3D model. In one embodiment, the need probability p to be supported by the support structure of each surface section s is estimated by projecting the surface section onto a plane parallel to the construction platform i The need probability p to be supported by the support structure supp，i :

[0018]

[0019] where represents the normal vector of the i th surface section s i , θ and represent the polar angle and azimuth angle of the 3D model orientation respectively, represents the unit vector perpendicular to the platform, i.e., in the construction direction, and "·" represents the scalar product. Using this estimate of the support probability, the evaluation function is given by

[0020]

[0021] In an alternative embodiment,

[0022]

[0023] and

[0024]

[0025] where max(a, b) represents the maximum value of two values a and b and n is a real-valued exponent

[0026] The present invention also provides a computer program having computer-readable code for causing a CAM module to perform the method steps. The present invention also provides a computer-readable memory storing the computer program

[0027] In the present invention, the computer program provides means for performing the allocation step, where the weighting factor f iis assigned to each surface section. In a first embodiment, the assignment is performed manually by a user by separately marking one or more surface sections on a display of a 3D model with a desired weighting factor f. In another embodiment, the assignment of the weighting factor is performed by CAD software used to design the 3D model to be printed. During the design process, the functions of different surface regions of the model may be known, and the assigned weighting factor can be selected according to the function of the corresponding surface section to which the region corresponds. For example, in a dental application, regions of a dental prosthesis related to a proper fit of the prosthesis can be identified during the design process and assigned a high degree of sensitivity to the effects from removing any support structures thereon.

[0028] In a preferred embodiment, a computer program utilizes a neural network to automate the assignment step, where the weighting factor f i is assigned to each surface section. In such an embodiment, the neural network can be used to classify characteristic regions of a 3D model based on the local surface geometry of the 3D model. For example, if the 3D model represents a dental working model, suitable characteristic region labels can be teeth, gums, alveolar bone, etc. Based on this classification, a suitable weighting factor f i can be assigned to the surface sections belonging to a specific characteristic region. In the previous example of a dental working model, a possible suitable assignment would be a large factor corresponding to a high sensitivity for sections belonging to the "tooth" region and a factor f i = 1 corresponding to no sensitivity for sections belonging to the "alveolar bone" region. For example, a 3D model can be used to train the neural network, where the characteristic regions have been identified either manually by the user or by a CAD software program. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In the following description, other aspects and advantageous effects of the present invention will be described in more detail by using exemplary embodiments and by referring to the accompanying drawings, where

[0030] Figure 1 is a schematic vertical cross-sectional view of a hemisphere as a 3D model according to an embodiment, where the curved surface indicated by the bold portion has been assigned a weighting factor f > 1 that is considered sensitive to post-processing;

[0031] Figure 2 is a schematic vertical cross-sectional view of a hemisphere as a 3D model according to an alternative embodiment, where the flat surface indicated by the bold portion has been assigned a weighting factor f > 1 that is considered sensitive to post-processing.

[0032] The reference numerals shown in the drawings represent the elements listed below and will be referred to in the subsequent description of the exemplary embodiments:

[0033] 1. 3D model

[0034] i: integer

[0035] s: surface section

[0036] s i : i th surface section

[0037] f: weighting factor,

[0038] f i : assigned to i th weighting factor of the surface section

[0039] A i : i th surface area of the surface part

[0040] evaluation function

[0041] θ: polar angle

[0042] azimuth angle

[0043] i th normal vector of the surface section

[0044] unit vector in the vertical direction

[0045] Figure 1 is a schematic vertical cross-sectional view of a hemisphere of the 3D model (1), the orientation of which has been determined by a method according to an embodiment of the present invention.

[0046] In a defining step of the method, the surface geometry of the 3D model (1) is first defined. The surface geometry includes a plurality of surface sections s i , where i is an integer. The surface geometry of the 3D model (1) is preferably represented by triangulation, where i th the surface section is a triangle (not shown) having a surface area A i . In another defining step, an evaluation function is defined such that

[0047]

[0048] the evaluation function R depends on the weighting factor f i and p supp，i , the weighting factor respectively indicating the degree of sensitivity of the surface section s i to the effect of removing any support structure, p supp，i indicating the probability that the surface section will need to be supported by a support structure. The summation represented by ∑ extends over all surface sections s i .

[0049] As Figure 1 shown, the entire surface represented by the bold portion has been assigned a weighting factor f > 1 that is considered sensitive to post-processing. The flat surface has been assigned a weighting factor f = 1. In the determination step, the orientation of the 3D model (1) relative to the build direction is determined based on the evaluation function R with the assigned weighting factor f. In Figure 1 , the hemisphere, i.e., the 3D model (1), is oriented in the optimized direction where R is a maximum. The 3D model (1) can be generated by an additive manufacturing device (not shown). The additive manufacturing device has a bucket for holding a photocurable material and a platform for holding a 3D object corresponding to the 3D model (1). The platform is relatively movable with respect to the bucket. When the 3D model (1) is generated using the additive manufacturing device, all support structures will be located on the flat surface and will thus protect the curved surface from the effects caused by the mechanical removal of the support structures.

[0050] Figure 2 is a schematic vertical cross-sectional view of a hemisphere as the 3D model (1) according to an alternative embodiment. In this alternative embodiment, the curved surface indicated by the bold portion has been assigned a weighting factor f = 1. The flat surface has been assigned a weighting factor f > 1, which is considered sensitive to the effects resulting from the mechanical removal of the support structures. In Figure 2 , the hemisphere, i.e., the 3D model (1), is oriented in the optimized direction where R is a maximum. When the 3D model (1) is generated using the additive manufacturing device, all support structures will be located on the curved surface and will thus protect the flat surface from mechanical post-processing.

[0051] The above two exemplary embodiments in Figure 1 and Figure 2 have been selected to illustrate the present invention by using a relatively simple 3D model (1). Of course, the method can be easily applied to more complex geometries, such as dental restorations, etc.

Claims

1. A method for determining the orientation of a 3D model (1) to be generated by an additive manufacturing device, the additive manufacturing device comprising: A bucket for holding a photocurable material; and a platform for holding a 3D object corresponding to a 3D model (1), wherein the platform is relatively movable with respect to the bucket, and the method includes: Steps for defining the surface geometry of a 3D model (1), wherein the surface geometry includes a plurality of surface segments s i , where i represents an integer, A i represents i th the surface area of the surface segment s i ; Characterized in that it further includes: The step of manually or automatically by a computer program assigning one or more weighting factors f i to the surface segments s i respectively, where the weighting factor f i indicates the sensitivity of each surface segment s i to the effect caused by the mechanical removal of any support structure thereon, wherein for the surface segments s i considered to be sensitive to the effect caused by the mechanical removal of any support structure, the weighting factor f i is greater than 1, and for all other surface segments s i , f i is equal to 1; The step of defining an evaluation function ; where θ and respectively represent the polar angle and the azimuth angle of the 3D object with respect to the building direction, and p supp,i represents the required probability indicating that each surface section s i is to be supported by the support structure, and the summation represented by ∑ extends over all surface sections s i ; and Based on the evaluation function R, by respectively with respect to the polar angle and the azimuthal angle θ and The step of optimizing to determine the orientation of the 3D model (1) relative to the building direction to avoid or minimize as much as possible the need for a support structure in the sensitive surface sections of the 3D model.

2. The method according to claim 1, wherein In the evaluation function in which represents i th the surface section s i is the normal vector of, represents the unit vector in the vertical direction perpendicular to the platform and pointing in the construction direction, and "·" represents the scalar product.

3. The method according to claim 1 or 2, characterized in that The surface geometry of the 3D model (1) is represented by triangulation, where i th surface section s i is a triangle with a surface area A i ​ 4. The method according to any one of claims 1 to 2, characterized in that, The assignment step is manually performed by the user by marking one or more surface segments s on the display of the 3D model (1) with a weighting factor f i respectively i to perform the assignment step manually.

5. The method according to any one of claims 1 to 2, characterized in that, The allocation step is performed by a computer program, the computer program including a neural network that has been trained to classify regions of the 3D model based on local surface geometry of the 3D model and to assign a weighting factor f i to surface segments.

6. A computer program product comprising computer-readable code for causing a CAD module to perform the method steps according to any one of claims 1 to 5.

7. A computer-readable memory storing computer-readable code for causing a CAD module to perform the method steps according to any one of claims 1 to 5.

Images