How to manufacture 3D structures using 3D printing techniques

Abstract

The present invention relates to a method for producing a 3D structure by a 3D printing method and has the task of providing a solution for automatic correction of 3D printing data for a 3D structure 30 to be produced, used for the 3D printing method, in case of deviations occurring during the production of the 3D structure 31. The task is achieved in such a way that the 3D printing data are data of the target geometry 30 of the 3D structure, that after the production of the 3D structure 31, the produced 3D structure 31 is measured in three dimensions, generating three-dimensional, actually incomplete data of the actual geometry 31 of the 3D structure, which data are reflected in a model 34 having one or more unscanned areas 35, and that within the scanned areas, a point P37 on the surface of the target geometry of the 3D structure and a related point P38 on the surface of the actual geometry 31 of the 3D structure are measured in three dimensions. ’ 36, determining a further deviation 29 in an unscanned area 35 of the surface of the fabricated 3D structure 31, and generating corrected 3D printing data 38 using these determined first and further deviations 28, 29.

Description

[Technical field]

[0001] The present invention relates to a method for manufacturing a 3D structure by a 3D printing technique, which comprises preparing first layer data for individual layers of a 3D structure to be produced from 3D printing data, which is data of a target geometric shape of the 3D structure, and using this data to control the production of the 3D structure by the 3D printing technique. [Background technology]

[0002] For the production of individual or successive shaped parts, workpieces or moulds, it is known to use so-called 3D printing or so-called 3D printing methods, in which three-dimensional shaped parts or workpieces are produced in a layer-by-layer manner.

[0003] The shaping is carried out in a computer-controlled manner with one or more fluid or solid materials according to given dimensions and shapes. The instructions for the parts or workpieces (3D structures) to be printed are provided in the form of 3D printing data, for example from so-called Computer Aided Design (CAD) systems.

[0004] When printing 3D structures or 3D parts, a physical or chemical hardening or melting process is carried out on a particulate build material, also called build material. For example, in the so-called powder bed method or layer-by-layer method, build or mold materials such as plastics, synthetic resins, ceramics, unhardened deposits such as minerals or sand, metals, etc. are used as materials for such 3D printing methods. Furthermore, extrusion or contour-based methods using plastics or metals are also known.

[0005] When implementing 3D printing techniques, various manufacturing method flows are known.

[0006] However, several flows within those methods may be implemented using the steps of the following exemplary method: a) partial or total application of a particulate build material, also called particulate material or powdered build material, on the so-called build area in order to form a layer of uncured particulate material, the partial or total application of the particulate build material consisting of laying and levelling the particulate build material; b) selectively hardening the applied layer of unhardened particulate building material in predetermined partial areas, for example by selectively concentrating, printing or applying a treatment agent, such as an adhesive, for example by means of a print head or by means of a laser; c) repeating the steps of the preceding method in another layer plane in order to build a part or artefact in layers, in which it is provided that, before the new layer is applied partially or completely, the part or artefact to be built or printed in layers on the building area is lowered by one layer plane or layer thickness relative to the building area or the 3D printing device is raised by one layer plane or layer thickness, respectively, relative to the building area; d) thereafter removing the free unhardened particulate build material surrounding the produced build part or workpiece; It consists of:

[0007] Generally, particulate building materials are understood to be an aggregate of individual particles of a material or mixture of materials, each particle having a three-dimensional extent. Since the particles are considered to be mainly circular, elliptical or elongated particles, it is possible to specify the average diameter of such particles, which is usually in the range of 0.01 mm to 0.4 mm. Such particulate building materials may have the properties of a fluid.

[0008] From the prior art, various methods are known for fabricating three-dimensional structures or for laying down and applying particulate build material onto a build area for fabricating 3D structures.

[0009] From US Pat. No. 5,399,543 a method and device for applying a fluid and its use are known.

[0010] The method of applying the fluid, particularly with respect to particulate material applied onto a coating area, involves applying the fluid onto the coating area in front of the blade, as viewed in the forward movement direction of the blade, and then moving the blade over the applied fluid.

[0011] The object is to provide an apparatus, a method and the use of said apparatus, which makes it possible to achieve a distribution of the fluid material on the coating area that is as even as possible.

[0012] The solution provides that the blade performs a swirling motion in the form of a rotational movement, which causes the fluid applied on the coating area to be fluidized, so that not only can particulate material with a strong tendency to agglomerate be applied as evenly and smoothly as possible, but also the swirling can affect the thickening of the fluid.

[0013] In a preferred embodiment, it is provided that the application of fluid onto the coating area is carried out in excess. In this way, the constant movement of the reciprocating blade in the form of a rotary movement homogenizes the excess fluid in the form of a cylinder of fluid or particulate material formed by the forward movement of the blade in front of the blade, as seen in the forward movement of the blade, thereby filling the hollow spaces possibly occurring between the individual particle agglomerates and breaking up larger agglomerates of the particulate material by the movement of the cylinder.

[0014] From the field of manufacturing, the term "dimensional stability" is known, which means that the actual dimensions of a workpiece should lie within a defined tolerance deviation from a specified nominal dimension.

[0015] In the prior art, during the production of 3D structures, measures are taken to achieve their dimensional stability or to reduce the deviations between the 3D structure to be manufactured and the 3D structure produced by 3D printing.

[0016] From US Pat. No. 5,399,633 a method is known for scanning a second part of a printed 3D structure to modify 3D printing data relating to a first part of the 3D structure, in which a modified pre-distorted 3D specification for the first part is generated based on the scan of the second part and taking into account the printing properties.

[0017] A drawback of that known prior art technique is the lack of quality control of the fabricated 3D structures, which is typically based on deviations from given dimensions.

[0018] For example, when performing quality control of a produced 3D structure by measuring the produced 3D structure, detected deviations from a given dimension of the 3D structure can be corrected, for example, by mechanically readjusting a component or a group of components of the 3D printer.

[0019] However, such mechanical readjustments are often burdensome, since it is necessary, for example, to partially disassemble the 3D printer in order to reach the component or group of components to be adjusted, and furthermore, such readjustments also cause the 3D printer to stop, i.e., to interrupt the production of the 3D structure by the 3D printer.

[0020] This is particularly disadvantageous in areas where very narrow tolerances are defined during the manufacture of 3D structures, for example the range of such tolerances being +0.3 mm to -0.3 mm for a given maximum deviation. In order to comply with a given narrow tolerance, the manufactured 3D structures therefore need to be oversized or undersized by up to 0.3 mm in a first dimension, such as for example the length.

[0021] As a result, the prior art provides inadequate and / or burdensome approaches to achieving adequate quality control or quality assurance during the fabrication of 3D structures.

[0022] Therefore, there is a need to improve upon the known prior art and therefore to improve the methods for producing 3D structures by 3D printing techniques. [Prior art documents] [Patent documents]

[0023] [Patent Document 1] German Patent No. 10117875 [Patent Document 2] U.S. Pat. No. 1,076,9324 Summary of the Invention [Problem to be solved by the invention]

[0024] The object of the present invention is to provide a method for producing a 3D structure by a 3D printing method, in which, if deviations occur during the production of the 3D structure, an automatic correction of the 3D printing data used in the 3D printing method is performed with respect to the 3D structure to be produced.

[0025] In particular, in this way, the deviations that occur when manufacturing 3D structures by 3D printing techniques should be reduced.

[0026] The automatic correction of the 3D printing data used in this 3D printing method should also be performed in areas on the surface of the 3D structure that cannot be reached by a 3D scan of the surface. [Means for solving the problem]

[0027] This problem is solved by a method for producing a 3D structure by 3D printing having the features of independent claim 1. Improvements are presented in the dependent claims.

[0028] All methods of producing the geometric shape of the part by applying a fluid or pasty medium in layers on a substrate and then solidifying correspond to such 3D printing methods. Examples of such methods include, but are not limited to, stereolithography, laser melting, electron beam melting, binder jetting, material jetting, build-up welding, digital light processing, and filament fusion manufacturing. It is defined that the method of producing a 3D structure by this 3D printing method is, for example, a binder jetting method based on a powder bed, and that the 3D structure is produced by known conventional techniques.

[0029] In the following, even if the method for manufacturing a 3D structure by this 3D printing technique is described only with the example of a binder jetting technique, this is not intended to limit the method to such a 3D printing technique.

[0030] It is well known that in such 3D printing methods, 3D printing data of the target geometry of the 3D structure is converted into specific instructions for the production of the individual layers by the 3D printing method, also called layer data, which are used to control the production of the 3D structure by the 3D printing method.

[0031] After a desired 3D structure is fabricated on the build area of ​​a 3D printer by an additive manufacturing technique, such as a binder jetting technique, the loose particulate build material that has not hardened is removed, thereby exposing the 3D structure.

[0032] Due to the 3D printing itself and the required post-processing of subsequent curing or particulate build material residues mechanically and / or with airflows from the produced 3D structure, the desired target geometry of the 3D structure given by the 3D printing data may not match the actual geometry of the produced 3D structure, such that the 3D structure is dimensionally stable as long as the deviation of the actual geometry from the target geometry does not exceed a pre-determined tolerance, otherwise the 3D structure is not dimensionally stable.

[0033] Such deviations are also referred to as distortion or shrinkage. Such distortions can also be caused by handling or transportation of the fabricated 3D structures.

[0034] Such distortions can be divided into two groups: so-called deterministic distortions and so-called non-deterministic distortions. Deterministic distortions are characterized by the fact that they occur again when multiple 3D structures are produced with the same 3D printing data or printing parameters, and are therefore reproducible.

[0035] Examples of such deterministic distortions include: a) material-induced distortion due to deviations in grain size and granularity of the particulate building material; b) Equipment induced distortions due to deviations in manufacturing tolerances, machine vibrations or mechanical distortions within the 3D printer; c) distortions resulting from the manufacturing process, such as deviations between the target and actual speeds of the operating means of the 3D printer, such as the print head, the application parts or the smoothing means, deviations in the exact nozzle actuation time of the print head, deviations in the amount of binder released from the nozzles of the print head, etc. d) distortions due to diffusion processes, for example in microwave or sintering furnaces, during hardening of three-dimensional structures; e) distortion due to cleaning, post-processing and / or transport processes; It is.

[0036] Non-deterministic distortion is a) Distortions that occur randomly and are not repeatable, rather than in a stable and systematic way; It is.

[0037] Due to this deterministic distortion, the process of fabricating the 3D structure must proceed for each new geometry or 3D structure. To avoid the burdensome mechanical adjustment of the components and component groups of the 3D printer, there are methods for manually adapting the 3D printing data to achieve the desired geometry.

[0038] In this connection, a detection of the produced 3D structure must be performed at least once, for example by 3D scanning, to provide data on the geometry of the produced 3D structure. Other suitable methods can be employed to detect the dimensions of the produced 3D structure.

[0039] A problem with measuring such fabricated 3D structures, for example by means of 3D scanning methods, is that the 3D scanning methods sometimes are not able to provide scan data for all areas of the fabricated 3D structure, for example in the form of a so-called point cloud, such a point cloud consisting of three-dimensional measurement points generated by the 3D scanning method, which describe the surface of the 3D structure.

[0040] The lack of scan data describing the surface of a 3D structure is due to areas of the fabricated 3D structure that cannot be reached by the sensors of the 3D scanning device, because these areas are behind areas that are "invisible" to the sensors of the 3D scanning device. Such areas are also referred to as recesses, dead areas, blind spots or unscanned areas in the following.

[0041] An unscanned area can also be considered as an area where very few measurement points were generated during the 3D scan. A too low measurement point density can provide only limited or no information regarding possible deviations between the actual model of the 3D structure and the target model of that 3D structure.

[0042] Therefore, a model generated by a 3D scanning technique of the surface of a manufactured 3D structure in the form of scanned actual geometric data does not completely reflect the manufactured 3D structure and has so-called unscanned areas.

[0043] In the prior art, such unscanned areas can be filled in, for example, by interpolation, so that, for example, a surface is formed over the open area. Since 3D scanning methods generate only a few measurement points in or near the unscanned areas, which are usually only present at the periphery of the unscanned areas, an exact reproduction of the exact outer contour of the surface of the 3D structure cannot be made in these unscanned areas. If only a few measurement points are generated in the unscanned areas, the reproduction of the surface or contour of the 3D structure is made partially in a stepped or terraced manner and therefore does not correspond to the exact progression of the surface or outer contour of the 3D structure.

[0044] In less complex three-dimensional structures, the process of filling such regions or surfaces is to approximate as closely as possible the actual geometry of the surface of the fabricated three-dimensional structure. In applications where a small tolerance is given between the target geometry of the 3D structure and the actual geometry of the 3D structure, the techniques for filling such regions or surfaces are often too inaccurate.

[0045] In the method for producing a 3D structure by the 3D printing method of the present invention, a) for a 3D structure to be produced, first layer data for individual layers of the 3D structure to be produced is prepared from 3D printing data, which is data of a target geometric shape of the 3D structure, and the production of the 3D structure by a 3D printing method is controlled using this data; b) the produced 3D structure is measured in three dimensions, in which case in reality incomplete 3D data of the scanned real geometry of the 3D structure of the surface of the produced 3D structure is generated, which is reflected in the model and has one or more unscanned areas; c) in a projection step, a first deviation between corresponding points P and P' in the scanning area is determined, the point P being on the surface of the target geometry of the 3D structure, and the corresponding point P' being on the model of the surface of the scanned actual geometry of the 3D structure, respectively; d) These identified first deviations are then assigned values ​​corresponding to the corresponding points P1, P2, P3, ..., P of the data of the target geometry of the 3D structure. n The deviation data is saved as e) in the interpolation step, the data of the target geometric shape of the 3D structure and the first deviation data already identified are used to interpolate another deviation belonging to the unscanned area of ​​the surface of the manufactured 3D structure, and the value of the generated another deviation data for the corresponding point P of the data of the target geometric shape of the 3D structure is stored to generate such perfect deviation data; f) generating corrected 3D printing data in a deformation step, in which points P of the data of the target geometry of the 3D structure are displaced in the opposite direction by an absolute value according to the first or another deviation of the stored values, depending on the first or another deviation determined at each point P, and g) using the corrected 3D printing data to control the production of a 3D structure by a subsequent 3D printing method; is prescribed.

[0046] As is known from the prior art, data of the target geometry of a 3D structure are converted into data for the individual layers of the 3D structure to be produced, and these data are used to control the production of the 3D structure by a 3D printer.

[0047] The 3D structure thus produced may then be removed from the print bed of the 3D printer, post-processed, and cleaned, which may also include a process for curing the 3D structure.

[0048] In the measuring step, the produced 3D structure is measured in three dimensions. This measuring process can be carried out, for example, using a corresponding laser scanning device. In the result of such a measurement, three-dimensional, in fact, often incomplete data of the actual geometry of the produced 3D structure is obtained as a model of the actual geometry. These data are incomplete because data is missing for unscanned areas that were not reached during the three-dimensional measurement.

[0049] Next, in a projection process, a first deviation between a corresponding point P' on the surface of the model of the actual geometry of the three-dimensional structure and a corresponding point P on the surface of the target geometry of the three-dimensional structure within the scanning area is determined.

[0050] The determination of the absolute value of these first deviations can be performed for each point P' present in the 3D printing data on the surface of the model of the actual geometry of the 3D structure and the corresponding point P on the surface of the target geometry of the 3D structure. For example, if a point P' identified by laser scanning on the surface of the model of the actual geometry of the produced three-dimensional structure is below the surface of the target geometry of the three-dimensional structure to be produced, the deviation at this point P is undersized.

[0051] If a point P' identified by laser scanning on the surface of the model of the actual geometry of the fabricated 3D structure lies above the surface of the target geometry of the 3D structure to be fabricated, the deviation at that point P is oversized.

[0052] If a point P' identified by laser scanning on the surface of a model of the actual geometric shape of the manufactured three-dimensional structure is on the surface of the target geometric shape of the three-dimensional structure to be manufactured, no deviation occurs at this point P.

[0053] For example, when identifying a deviation, such as a first deviation, corresponding points P′ and P on the model of the actual geometry of the fabricated 3D structure and the target geometry of the to-be-fabricated 3D structure are compared pairwise and overlap each other if no deviation occurs between the surfaces of the actual and target geometries of the 3D structure.

[0054] For each three-dimensional point P in the target geometry of the 3D structure, for which a corresponding point P' has been identified in the scanning process in the model of the scanned actual geometry data, a first deviation between the actual geometry of the 3D structure and the target geometry of the 3D structure can be or is identified and stored in the projection step, so that in unscanned areas the first deviation cannot be determined, because the model of the scanned actual geometry data does not provide the point P' in those areas.

[0055] Alternatively, the determination of these first deviations can be performed for a given area on the surface of the 3D structure, and such deviations can also be average values ​​for a given area on the surface of the 3D structure.

[0056] The first deviation thus identified has its value stored as deviation data of the corresponding point P of the target geometric data of the 3D structure, so that it can be used in another flow of the method.

[0057] The first deviations identified in the scanned area are used in a process of identifying further deviations or further deviation data in unscanned areas of the surface of the model of the fabricated 3D structure, where if the first deviations identified at the periphery of the unscanned area are known, it can be assumed that the first deviations identified at the periphery of the unscanned area continue further in the unscanned area with the same absolute value or with the same magnitude.

[0058] Since the first deviations determined at the periphery of the unscanned area are known, the determination of further deviations or further deviation data in the unscanned area can be performed in such a way that the information is successively carried out from known points in the scanned area to points in the unscanned area by means of assumed small partial planes, thus realizing a gradual completion of the missing further deviations in the unscanned area, where these points are at the vertices of assumed small partial areas, which are for example triangular partial planes used in a triangulation process of the surface of the target structure to be produced.

[0059] Such partial planes filling the unscanned area can be n-sided or free-form planes with vertices, in particular triangles. In the following, the method is advantageously described based on target geometric shapes described by triangular partial planes, without this being intended to limit the method to these partial planes.

[0060] Here, both the first deviation data already determined and further deviation data to be determined for the unscanned regions are determined and saved at the corner points P of the partial plane. When using free-form planes, the deviation data at correspondingly determinable points are saved.

[0061] It is now defined that the identification of further deviations within the unscanned area begins at the periphery of the unscanned area and continues this process towards the assumed centre of the unscanned area.

[0062] Further, small sub-planes of the target geometry, for example triangular sub-planes, whose vertices are used to identify the deviations in the unscanned region, are defined. The repeated identification of deviations using these triangular sub-planes is performed starting from the periphery of the unscanned region until all deviations have been identified and stored for each point in the unscanned region.

[0063] An example process for identifying this further deviation or further deviation data is described in detail below.

[0064] A prerequisite for determining further deviations is that the surface of the target part, i.e. the surface of the target geometry of the 3D structure, is already present in triangulated form. Such triangulation is carried out in a transformation process, in which the surface of the target geometry of the 3D structure to be produced is completely covered or represented, for example, by a number of triangular partial planes, the corners of which are respectively denoted by P1, P2, P3, ..., P n Therefore, points P1, P2, P3, ..., P at the corners of the triangular partial plane are assigned. n Data on neighborhood relationships is obtained and then used.

[0065] On the surface of the target geometric shape, triangular partial planes lying on the periphery of the unscanned area are determined, in which the first deviation data for two corners have already been determined in the projection process. In this example, it can be a triangular partial plane of a triangle with vertices P1, P2, P3, in which the first deviation data at the vertices P1 and P2 are known, because in that case the corresponding points P1 and P2, respectively, are determined. ’ Between P2 and P2 ’ However, it is not possible to determine the deviation between corresponding points P and P' in the unscanned regions, since in those regions the data for point P' is missing.

[0066] When determining or identifying the further deviation, the starting point is that the deviation value or absolute value determined at the periphery of the unscanned area continues into the unscanned area with the same value or absolute value for the further deviation to be determined. In this example, the further deviation for the vertex P3 can be determined using a mathematical function, for example, the arithmetic mean value of the known first deviations at the points P1 and P2 is determined. The first and further deviations determined for the points P1, P2, P3 have a value or absolute value for the determined deviation, with the direction of the deviation running along the vertex normal corresponding to the points P1, P2 or P3 of the triangular partial plane, respectively.

[0067] Using a mathematical function such as an arithmetic mean of the known first deviations for these points P1 and P2 to determine the other deviation for the vertex P3 is one example, although in practice other functions or calculation rules may be stored other than the mean mean, for example taking into account a more complex collection of sub-areas consisting of multiple points, weighting by priority when the first deviations differ significantly, or a degree of uncertainty in determining the other deviation based on the distance to the determined first deviation.

[0068] Area, angle or distance weighted average calculations can also be used as functions, so that not only neighbourhood relationships but also existing geometrical circumstances are included in the calculation, e.g. when weighting by distance, closer points have a greater influence than more distant points.

[0069] After determining another deviation with respect to the vertex P3, another triangular partial plane on the surface of the target geometry, for example on the periphery of the unscanned area, is determined, for which two vertices have already been determined the first deviation data in the projection step. In this example, it can be the triangle P2, P4, P5, for which the first deviation with respect to the vertices P2 and P4 is known, because there the corresponding points P2 and P2, respectively, are located at the periphery of the unscanned area. ’ Between P4 and P4 ’ The determination of this further deviation for vertex P5 is again performed, for example, by calculating the arithmetic mean value of the known first deviations for points P2 and P4.

[0070] The determination of the further deviations for the corresponding vertices of the further triangular partial planes can be continued, for example, along the periphery of the unscanned area until the first sequence of triangular partial planes has been used to determine the further deviations along the entire periphery area. nThe process of identifying the further deviation based on the known first deviation with respect to is an iterative process that uses the triangular sub-planes created during the triangulation of the surface of the target geometric shape as triangular sub-planes and does not use the triangular sub-planes created during the identification of the further deviation.

[0071] For example, if a first sequence of further deviations is identified near the periphery of the unscanned region, then the identification of further deviations, for example with a second sequence, may continue in areas further away from the periphery of the unscanned region.

[0072] Continuing to identify these further deviations can be done using another triangular partial plane having vertices P3, P5, and P6 in which further deviations relative to vertices P3 and P5 have been identified, as described above when identifying further deviations relative to vertices P3 and P5.

[0073] The determination of the further deviation with respect to vertex P6 can be performed in the manner described above, for example by an arithmetic mean value calculation.

[0074] The continuation of the identification of another deviation with this other sequence is one example, this other sequence being further away from the periphery of the unscanned area than the first sequence.

[0075] Points P7, P8, P9,..., P in the unscanned region n The process of identifying further deviations for continues iteratively, for example for the entire unscanned area, until further deviations are generated.

[0076] Points P1, P2, P3, ..., P n The determination of the different deviations with respect to is carried out independently of the progression of the surface of the produced 3D structure, which is a particular advantage of the method.

[0077] In this approach, steps may occur between adjacent points P, especially in the region of the supposed center of the unscanned area, during the subsequent deformation, for example due to different first deviations at different peripheries of the unscanned area. These steps are caused, for example, by the fact that the determined further deviation values ​​or absolute values ​​between adjacent points P are too different.

[0078] It is provided that smoothing of further deviations identified in the unscanned areas is carried out.

[0079] Instead, it is provided that a smoothing of the first deviation and the further deviation takes place.

[0080] During this smoothing, the corresponding points P1, P2, P3, ..., P are smoothed using a weighted average or other value, such as a smoothing rule that takes into account the distance, angle, and geometric shape of the corresponding deviation values. n By correcting each identified deviation value with respect to , the jumps between such adjacent deviation values ​​are reduced, thereby mitigating steps that would otherwise occur in subsequent deformation steps.

[0081] For this purpose, for example, when calculating the weighted average value for point P1, the deviation value for point P1 itself and the deviation values ​​for adjacent points P2, P3, ..., P2, which are also called adjacent points, immediately surrounding point P1 are used. n The weighted average value for point P1 is determined by processing a plurality of deviation values ​​for each of the points P2, P3, and P4. Therefore, when a triangular partial plane is used, at least the deviation values ​​for three adjacent points P2, P3, and P4 and the deviation value of point P1 itself are used in calculating the weighted average value.

[0082] Further, for example, it is provided that surface normals associated with triangular partial planes are calculated for each triangle and converted to vertex normals associated with the vertices defining that triangle, and these normals are used in further flows of the method.

[0083] A surface normal is a normalized, outward-pointing vector perpendicular to the surface of a triangular subplane.

[0084] For each point, the vertex normal is calculated by angle- and area-weighted addition of the surface normals associated with adjacent or neighboring triangles, followed by normalization.

[0085] Here, based on the method, each point P1, P2, P3, ..., P of a target model or target geometric shape of a 3D structure is n , the first or another deviation value or absolute value and the vertex normal are obtained, where the vector resulting from the multiplication of the deviation value and the vertex normal is used to map, in direction and absolute value, each point P of the target geometry of the 3D structure to its actual position P in the actual model produced, i.e., in the actual geometry of the 3D structure. ’ This corresponds to a displacement vector that displaces the object.

[0086] This displacement vector is the set of points P1, P2, P3, ..., P n If the displacement vector is pointing outward at the point, it means that the fabricated three-dimensional structure is oversized for the given target structure data at that point. If the length of this displacement vector is zero, it means that the fabricated three-dimensional structure has no deviation for the given target structure data at that point. If the length of this displacement vector is within the tolerance range, the deviation does not impair the dimensional stability. If the displacement vector is pointing inward, it means that the fabricated three-dimensional structure is undersized for the given target structure data at that point.

[0087] It is also provided that from a number of produced 3D structures, data on the actual geometry of the 3D structure can be determined by the model, and a comparison is made between the data on the desired target geometry of the 3D structure and the data on a certain number of actual geometries of the 3D structure, in such a way that a deterministic deviation, i.e. a mean value generated using statistical methods, is calculated for the deviations determined at different points.

[0088] In this way, for example, on the occurrence of one first or further deviation, or one too large deviation, a different error procedure can be initiated than on the occurrence of a systematic deviation, this also applies for the occurrence of a certain number of first or further deviations.

[0089] In the deformation process, the point P for generating the corrected 3D printing data can be displaced in the opposite direction to the specified deviation or displacement vector, for example, by their absolute value. In one example, the point P1, which is specified as a deviation of 0.2 mm undersize, can be displaced outward in the opposite direction to the specified deviation by an absolute value of 0.2 mm. In this way, the deterministic distortion is corrected.

[0090] Conversely, for example, an oversize of 0.1 mm at a given point P2 on the surface of a produced three-dimensional structure can be eliminated by specifying that the corrected 3D printing data for that point will produce the next three-dimensional structure with an undersize of 0.1 mm, thus also correcting for deterministic distortions.

[0091] Instead, during the transformation, points P1, P2, P3, ..., P n The displacement can be made in the opposite direction to the determined deviation by a partial or increased absolute amount of the determined absolute value of the deviation. These partial amounts can be application specific and based on a defined area or existing geometry of the 3D structure. As such, the displacement can be in steps, e.g., 50%, 75%, 90%, 110%, 125% or 150% of the determined absolute value of the first or another deviation.

[0092] Furthermore, in the deformation process, the points P1, P2, P3, ..., P nis defined as being displaced by the product obtained from the value stored for the first or another deviation as a first factor and a second factor lying in the range of 0.3 to 1.7. In this way, the deformation can be adapted continuously within said range.

[0093] These corrected 3D printing data also provide data for the individual layers of the 3D structure to be manufactured, which are used to control the production of the 3D structure by a 3D printing method, such data being also called layer data.

[0094] Alternatively, it can be provided that the first and further identified deviations are used to correct the 3D printing data of the individual layers already present for the first print. In this way, corrected 3D printing data of the individual layers of the 3D structure to be manufactured are generated, which data can be directly used for the production of the 3D structure by a subsequent 3D printing method. Thus, it is omitted to convert these corrected 3D printing data into data of the individual layers of the 3D structure to be manufactured.

[0095] The thus corrected 3D printing data is then used to control the production of another 3D structure, so that these corrected 3D printing data take into account deformations or deviations of the actual geometry of the produced 3D structure relative to a given target geometry of the 3D structure to be produced. In this way, for example, a first or another identified deviation, such as an undersize of 0.2 mm at a given point on the surface of the produced 3D structure, can be eliminated by the corrected 3D printing data for that point prescribing the subsequent production of a 3D structure with an oversize of 0.2 mm.

[0096] In this example, for example, the deviation identified at point P of the given target geometry of the three-dimensional structure to be produced is stored such that its direction is towards the vertex normal and its absolute value is 0.2 mm. If it is undersized, the direction of the identified deviation is towards the inside or inner side of the three-dimensional structure to be produced or the given target geometry of the three-dimensional structure to be produced. If it is oversized, the direction of the identified deviation is towards the outside or outer side of the three-dimensional structure to be produced or the given target geometry of the three-dimensional structure to be produced.

[0097] The method provides that the control of the 3D printing method is performed to correct the determined deviation or deterministic distortion based on the corresponding points, including determining first and further deviation data between the actual geometric shape of the 3D structure to be produced and the data of the target geometric shape of the 3D structure to be produced over the entire surface or outer contour of the 3D structure. The determination of the deviation or deviation data for points in the unscanned area by the method can realize the determination of the deviation or deterministic distortion even for the unscanned area that is not detectable.

[0098] If the process of determining the different deviation data in the unscanned areas results in perfect deviation data, then a so-called deformed 3D structure is produced, which the corrected 3D printing data represents. The manufacturing of the 3D structure using these corrected 3D printing data corrects the deterministic distortions caused by the material, process and equipment, and thus the conclusion is reached that a 3D structure is produced that corresponds as closely as possible to the target geometric data of the given 3D structure, i.e. the 3D printing data obtained from the CAD system.

[0099] The creation of the deformed 3D structure or corrected 3D printing data comprises the following steps:

[0100] In a first step, surface normals of the triangulated target part are calculated from the initial given 3D printing data, which is the data of the target geometry of the 3D structure. Such surface normals are determined, for example, with respect to small triangular part planes that reproduce the surface of the 3D structure. The reproduction of the surface of the three-dimensional structure can also be performed with part planes other than triangles.

[0101] From these surface normals, the vertex normals are calculated.

[0102] Here, using the vertex normal, for example, for each triangular partial plane, a displacement direction is obtained, and a predetermined deviation is assigned along this direction. n can be displaced along its vertex normal.

[0103] The magnitude of the displacement along the vertex normal is the magnitude of the displacement along the vertex normal at the points P1, P2, P3, ..., P n This displacement is obtained from the deviation values ​​respectively specified at the points P1, P2, P3, ..., P on the surface of the target geometric shape of the three-dimensional structure to be manufactured, through which the vertex normal passes. n This vector can be understood as having its origin at . This vector points parallel to the associated vertex normal in one of two possible directions. n If the deviation determined for a point on the surface of the target geometry is less than zero, the vector points away from the 3D structure. If the deviation determined for a point on the surface of the target geometry is less than zero, the vector points into the 3D structure.

[0104] During the generation of a deformed 3D structure or corrected 3D printing data, called deformation or deformation process for short, a point on the surface of the 3D structure to be produced, which is contained in the data of the target geometry of the 3D structure, is displaced along the associated vertex normal, in the opposite direction to the determined deviation, by an absolute value that depends on the first or another deviation determined at this point.

[0105] The points P1, P2, P3, ..., P on the surface of the target geometric shape of this 3D structure n The deformation is performed on a predetermined number of points, on selected points, or on all points, thus generating a deformed 3D structure or corrected 3D printing data.

[0106] The corrected 3D printing data of these deformed three-dimensional structures is sliced ​​into layer data as usual, thus generating geometric or layer data for controlling the 3D printing process.

[0107] In an alternative variant of the method for producing a 3D structure by this 3D printing method, instead of deforming the target part, the points P1, P2, P3, ..., P4 with deviation data to the coordinate system of the layer data already obtained in the first printing are n In this way, corrected layer data are obtained, and these data are used to control the production of the 3D structure on a 3D printer.

[0108] It is also provided that the corrected 3D printing data of the 3D structure to be manufactured or the corrected 3D printing data of the individual layers are generated according to the method, regardless of the magnitude of the predetermined deviations, in this way even the smallest detected deviations are processed during the transformation and act on the corrected 3D printing data to be generated.

[0109] The program for implementing the method for manufacturing a 3D structure by the 3D printing method is executed, for example, in a control unit for processing a print job, a CAD computer, or a central control unit of a 3D printer. This central control unit controls the flow for manufacturing the 3D structure based on 3D data of the dimensions of the 3D structure to be manufactured that is delivered thereto. Such data is generated, for example, by a computer-aided design system and delivered to the central control unit.

[0110] This central control unit thus provides or generates parameters for driving the 3D printer, such as, for example, a parameter of the nozzle actuation time or a parameter of the speed of movement of the operating means of the 3D printer over the build area. Thus, for example, the parameter of the nozzle actuation time can be adjusted by this central control unit. This parameter of the nozzle actuation time can be displaced in time by the central control unit with respect to a given value of the actuation time, so that this displaced actuation time is located before or after the given value of the actuation time. The direction of this displacement depends on the direction of the determined deviation of these dimensions.

[0111] Whereas in known prior art, usually only a small number of assumed representative measurement points on the 3D structure, e.g. 30, are defined and used for quality assurance, the method according to the invention significantly improves the quality control situation.

[0112] The method solves this problem by complementing imperfect measurements, combining multiple measurements of a single fabricated part, and comparing measurements of different similar parts when printing the same geometry multiple times.

[0113] This comparison is made possible by storing the identified deviations with respect to a point P of the target geometry of the 3D structure. In this way, the identified deviations for a number of produced 3D structures are always assigned or stored to the same point P of the target geometry of the 3D structure, so that the target geometry of the 3D structure serves, as it were, as a common base of reference for the different identified deviations, so that they can be compared.

[0114] This interpolation process, i.e. the generation of separate deviation data for unscanned areas, makes it possible to process a comparable number of measurement points in every measurement, which qualitatively improves both the comparison technique described above and the deformation process described herein.

[0115] The advantage of the method is that it is possible to store the first and further deviations determined from different measurements of the fabricated 3D structure at the relevant points P of the data of the target geometry of the 3D structure and thus to perform an integration of the different measurements on the 3D structure. In this way, it is possible to detect the areas of the 3D structure where the fabricated 3D structure is present or located during the first three-dimensional scan.

[0116] Furthermore, in order to be able to distinguish deterministic distortions from non-deterministic distortions by statistical methods, it is possible to fabricate multiple 3D structures and determine the interpolated deviation data for all fabricated 3D structures by the techniques described herein.

[0117] The method also provides for measuring the subsequently produced 3D structure in three dimensions using the corrected 3D printing data in order to determine a first and further deviation or perfect deviation between the target geometry of the 3D structure to be produced and the actual geometry of the subsequently produced 3D structure, the perfect deviation or perfect deviation data thus obtained being then used to transform the already corrected 3D printing data.

[0118] Thus, new improvements in dimensional stability are obtained. The method can be iterative. One approach is to perform the iterations described here not immediately in sequence for each produced 3D structure, but at intervals in time or only after the production of a given number of 3D structures, to deal with possible mechanical changes or subsequent processes in the 3D printer. Instead, knowledge of the need for further iterations is obtained by evaluating the dimensional stability of the produced 3D structures over time.

[0119] This allows for repeatable improvements in precision or dimensional stability, as well as a self-regulating matching process during mass production.

[0120] The previously described features and advantages of the present invention will be better understood and appreciated after careful consideration of the following detailed description of the presently advantageous, non-limiting embodiments thereof in conjunction with the accompanying drawings, in which: [Brief description of the drawings]

[0121] [Figure 1] Example of a flow chart for a method according to the invention [Diagram 2] FIG. 1 is an example of a flowchart for a process of imputing missing deviation data related to unscanned areas of the surface of a fabricated 3D structure. [Diagram 3] Example of a flow chart for the smoothing process [Figure 4] Schematic of the 3D structure to be produced using the 3D printing method, i.e. the target geometry of the 3D structure [Diagram 5] 3D structures produced using 3D printing techniques, i.e. schematic of the actual geometry of the 3D structure [Figure 6] Schematic representation of a model of the scanned real geometric data of a 3D structure provided by the measurement process. [Figure 7] A cross-section of the model illustrated in FIG. 6 or the scanned data of the actual geometry. [Figure 8]FIG. 2 is a schematic diagram visualizing a projection step 9 of projecting the actual geometry data and the target geometry data onto one another to generate first deviation data; [Figure 9] Schematic diagram of an example of identifying different deviations for unscanned regions [Figure 10] Schematic diagram of the process of identifying additional deviations based on a known first deviation. [Figure 11] Schematic of the process continuing to identify different deviations in different sequences. [Figure 12] Schematic of the process of generating and interpolating deviations for the entire unscanned area [Figure 13a] An excerpt from the process of transforming and generating corrected 3D printing data during the transformation process [Figure 13b] An excerpt from the process of transforming and generating corrected 3D printing data during the transformation process [Figure 14] Illustration of perfect transformation or perfect generation of corrected 3D printing data DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0122] FIG. 1 shows an example of a flow chart of a method according to the invention.

[0123] The method starts with a preparation step 1 of 3D printing data, which is data of a desired target geometry 30 of a 3D structure to be produced. Such 3D printing data can be provided, for example, from a computer-aided design system.

[0124] In the conversion step 2, these 3D printing data, i.e., data of the target geometric shape 30 of the 3D structure, are first converted into partial planes, such as small triangular partial planes that describe the outer contour of the 3D structure 30 to be produced. That is, for example, a triangular partial plane is used to perform triangulation that completely covers the surface of the target geometric shape 30 of the 3D structure to be produced.

[0125] For the partial planes present in these triangles, the associated surface normals are determined for each triangular surface, and based thereon, the associated vertex normals are calculated for each vertex, which are used in the further flow of the method. The data of the target geometry 30 of the 3D structure obtained in these transformed forms are then transformed between the corresponding points, i.e., the points P on the surface of the actual geometry 31 of the 3D structure. ’ , P37 on the surface of the target geometric shape 30 of the 3D structure. n are placed to match the vertices of the triangular subplane, and the generated vertex normals are aligned to these points P1, P2, P3, ..., P n It is defined as extending through.

[0126] In a first slicer step 3, the 3D printing data of the target geometry 30 of the 3D structure is converted into specific instructions for producing the individual layers by a 3D printing method, after which first layer data 4 of the target geometry 30 of the 3D structure is obtained.

[0127] In a first printing step 5, a 3D printer uses the prepared layer data 4 to produce a 3D structure by 3D printing techniques.

[0128] After the fabrication of the 3D structure in the 3D printer 31 is completed, the structure is removed from the print bed of the 3D printer and cleaned in a post-processing step 6. This post-processing step 6 may also include processes of curing, post-processing and transportation of the 3D structure 31. Post-processing is understood to be, for example, the removal and cleaning of the support structure of the fabricated 3D structure 31.

[0129] The produced 3D structure 31 is measured in three dimensions in a measurement step 7. This measurement process can be carried out, for example, by means of a corresponding laser scanning device (3D scanning device). As a result of such measurements, three-dimensional, actually mostly incomplete data 34 of the real geometry of the produced 3D structure 31 are obtained. These data are incomplete because data are missing for unscanned areas, which are caused, for example, by recesses. Such unscanned areas 35 arise in measurement step 7 when parts of the surface of the produced real structure 31 are not reachable by the sensor of the 3D scanning device or are "invisible" to the 3D scanning device. The scanned data 34 of the real geometry 31 correspond to a model 34 which optically represents these data 34 and can therefore be realized.

[0130] Optionally, before continuing with the method, one can proceed with another first printing step 5, another post-processing step 6 and another measurement step 7 to fabricate and measure multiple 3D structures 31.

[0131] At the end of the measurement step 7, scanned actual geometric data 34 of the produced three-dimensional structure 31 are obtained as a measurement point cloud 8. These three-dimensional data of the measurement point cloud 8 are, for example, the vertices P1, P2, P3, ..., P4 of the partial planes used in the triangulation process of the surface of the target structure to be produced in order to describe the outer contour of the three-dimensional structure to be produced. n is assigned to.

[0132] In the projection step 9, the data of the actual geometric shape 31 of the produced three-dimensional structure and the data of the target geometric shape 30 of the three-dimensional structure to be produced are projected onto each other or compared with each other. Therefore, the corresponding points P37 and P38 in the scanning area are projected onto each other or compared with each other. ’ 36 ’ Based on this, the deviation between the actual geometry 31 of the three-dimensional structure and the target geometry 30 of the three-dimensional structure can be determined point by point. n37 are respectively located on the surface of the target geometric shape 30 of the three-dimensional structure, and the points P1 ’ , P2 ’ , P3 ’ ,..., P k ’ 36 are respectively located on the surface of the actual geometric shape 31 of the three-dimensional structure. In this case, the relational expression k < n holds, which means that the number of measurement points recorded during measurement for which the points P ’ 36 can be assigned is less than the number of points P37 on the target geometric shape 30.

[0133] Therefore, these determined first deviations 28 or deviation data 28 respectively represent the intervals between the points P37 in the scanning area and the related points P ’ 36. These deviation data 28 store the values specified for each of the points P1, P2, P3,..., P n on the surface of the target geometric shape 30 of the three-dimensional structure.

[0134] The data 34 of the scanned actual geometric shape 31 of the fabricated three-dimensional structure generated by this three-dimensional scan is incomplete, that is, it has an unscanned area 35. Therefore, in the next complementing step 10, based on this method, a second deviation 29 or second deviation data 29 regarding the unscanned area 35 is generated or interpolated. In this complementing step 10 based on this method, the first deviation 28 specified within the scanning area in the projection step 9 is used, in particular, the first deviation 28 confirmed between the corresponding points P37 and P ’ 36 at the periphery of the unscanned area 35.

[0135] After the complementing step 10, complete deviation data consisting of all the deviation data 28 and 29 between the 3D structure 30 to be fabricated and the fabricated 3D structure 31, that is, both the first deviation data 28 specified within the scanning area and the second deviation data 29 specified within the unscanned area 35, is obtained.

[0136] This interpolation step 10 can optionally be followed by a smoothing step 11, in which the now complete deviation data, i.e. the first deviation data 28 of the scanned area and the second deviation data 29 of the unscanned area, are smoothed or post-processed, for example to correct the deviation 28 or 29 at any point P using a weighted average value for this deviation 28 or 29.

[0137] After this, smoothed perfect deviation data 12 is obtained that describes all differences or all deviations between the target geometry 30 of the 3D structure and the actual geometry 31 of the 3D structure for all areas of the surface of the actual geometry 31 of the 3D structure.

[0138] Optionally, in a next step a deterministic detection of the distortion 13 can be performed based on evaluation or analysis of multiple results of multiple 3D scans of multiple fabricated 3D structures 31. In this way, for example, the average value of the deviations occurring at points on the surface of the 3D structure can be determined.

[0139] As a result of this optional detection 13, the perfect deviation data determined in step 12 or a further corrected perfect deviation data 14 is then obtained and used in another flow of the method.

[0140] In the deformation step 15, corrected 3D printing data is generated, which includes deformations or deviations of the actual geometry 31 of the produced 3D structure with respect to a given target geometry 30 of the 3D structure to be produced. In order to eliminate or at least minimize the deterministic distortions that occur as a result of the production of another 3D structure based on this corrected 3D printing data 38, a deformation of the existing target model of this 3D structure is performed in the opposite direction to the identified deviations. This "opposite direction to the identified deviations 28, 29" is such that a specified oversize of, for example, 0.2 mm at a given point P1 37 on the surface of the 3D structure is compensated by an undersize of 0.2 mm at the given point P1 37. In this deformation, the data of the target geometry 30 of the three-dimensional structure at the points P1, P2, P3, ..., P4, ..., P5, ..., P6, ..., P7, ..., P8, ..., P9, ..., P10, ..., P11, ..., P12, ..., P13, ..., P14, ..., P15, ..., P16, ..., P17, ..., P18, ..., P19, ..., P20, ..., P21, ..., P22, ..., P23, ..., P24, ..., P25, ..., P26, ..., P27, ..., P28, ..., P29, ..., P30, ..., P31, ..., P29, ..., P32, ..., P33, ..., P34, ..., P35, ..., P36, ..., P37, ..., P38, ..., P39, ..., P40, ..., P41, ..., P42, ..., P43, ..., P44, ..., P45, ..., P46, ..., P47, ..., P48, ..., P49, n For each point P1, P2, P3, ..., P n In response to the deviations 28, 29 determined in, the displacement is made in the opposite direction to the determined deviations 28, 29 by the absolute value of the determined deviations 28, 29.

[0141] In the following, the 3D printing data generated in this deformation step 15 appears as deformation data 16 which represents the corrected 3D printing data 38.

[0142] In the next second slicer step 17, the deformation data 16 of the 3D structure or the corrected 3D printing data 38 are converted into specific instructions for producing an individual layer by a 3D printing method, after which another layer data 18 for producing another 3D structure is obtained.

[0143] In another printing process 19, a 3D printer uses the prepared layer data 18 to produce another 3D structure by a 3D printing technique.

[0144] After producing another 3D structure in another printing step 19, the method can proceed anew and continue in the measuring step 7.

[0145] Alternatively, in the flow of the method, it can be provided that corrected 3D printing data for the individual layers are generated in an alternative deformation step 20, including deformations or deviations of the actual geometry 31 of the produced 3D structure relative to a given target geometry 30 of the 3D structure to be produced.

[0146] The smoothed perfect deviation data 12 or the corrected deviation data 14 are used, for example, to generate so-called layer deformation data 21 from the obtained first layer data 4 using the known deviations 28, 29 for each layer.

[0147] Similarly, deformations of the first layer data 4 are performed in the opposite direction to the identified deviations 28, 29 in order to eliminate or at least minimize deterministic distortions that arise as a result of producing another 3D structure based on these corrected 3D printing data 38.

[0148] In a further flow of the method, the layer deformation data 21 is converted into second layer data 18 to produce another 3D structure in another printing step 19.

[0149] FIG. 2 illustrates an example of a flow chart of a process according to the present method for complementing missing further deviation data 29 related to unscanned areas 35 of the surface of a fabricated 3D structure 31 in the complementation step 10 using data of a target geometry 30 of the 3D structure and a first deviation 28 already identified.

[0150] The input data 22 for the interpolation step 10 are the surface data of the target geometry 30 of the 3D structure, which have already been processed in the transformation step 2, so that the surface of the target geometry 30 of the 3D structure is reconstructed or triangulated using small partial planes, for example triangles, whose vertices are represented by points P1, P2, P3, ..., P n The first deviation values ​​or first deviation data 28 already determined in the projection step 9 for the scanned area also belong to the input data 22 .

[0151] The first deviation data 28 already determined for the scanned area, in particular the points P1, P2, P3, ..., P on the periphery of the unscanned area 35 n Starting from the deviation data 28 specified in 37, points P1, P2, P3, ..., P n An interpolation 23 is then performed to determine stepwise another deviation data 29 for the point P337 located in the unscanned region 35. The process of the interpolation 23 uses the known first deviation data 28 for, for example, two points P137 and P237 of a triangular partial plane located in the scanned region, in order to identify another deviation value 29 for the point P337 located in the unscanned region 35. In this example, in the step of the interpolation 23, the another deviation value for the point P337 can be determined in such a way that the arithmetic mean value of the known first deviations 28 for the points P137 and P237 is determined.

[0152] This interpolation 23 may, for example, be performed along the periphery of the unscanned area 35 by interpolating points P1, P2, P3, . . . , P n 37, all points P1, P2, P3, . . . , P4 of the first sequence lying near the periphery of the unscanned area 35 are n 37, the process continues until the further deviation values ​​29 associated therewith are determined. The further deviation values ​​or further deviation data 29 thus determined also have their absolute values ​​corresponding to the associated points P1, P2, P3, ..., P4 of the data of the target geometric shape 30 of the three-dimensional structure. n Saved regarding 37.

[0153] Next, a second sequence of assumed points P1, P2, P3, ..., P4 having a greater distance to the periphery of the unscanned area 35 is n 37, etc., can be interpolated 23. By this interpolation 23, all points P1, P2, P3, ..., P n , P3, . . . , P4, . . . n Regarding 37, the additional deviation data 29 is interpolated from the outside to the inside.

[0154] In this way, points P1, P2, P3, . . . , P n The further deviation data 29 generated with respect to 37 is the output data 24 of the interpolation process 10 .

[0155] 3 illustrates an example of a flow for the smoothing step 11. When determining a deviation 28 in a scanned area, or when determining another deviation 29 in an unscanned area 35 or at the transition between the scanned and unscanned areas 35, sudden changes in the values ​​of the deviation data 28, 29 for adjacent points P can occur.

[0156] The smoothing input data 25 provided to this optional smoothing step 11 is a set of points P1, P2, P3, ..., P4 in scanned and unscanned regions 35 of the surface of a target geometric shape 30 of a three-dimensional structure. n 37 are the perfect deviation data obtained after imputation step 10.

[0157] In this smoothing step 11, steps that would otherwise occur in the subsequent deformation step 15 are removed by removing the corresponding points P1, P2, P3, ..., P n By correcting the deviation values ​​28, 29 identified for 37 by a weighted average operation 26 on the respective corresponding deviation values, such jumps between adjacent deviation data 28, 29 values ​​are mitigated by reducing them.

[0158] For this purpose, for example, the weighted average value for the point P137 is calculated by calculating the weighted average value 26 for the point P137 using the deviation value 28 or 29 for the point P137 itself and the deviation values ​​29 for the adjacent points P2, P3, P4, ..., P n Therefore, when a triangular partial plane is used, at least the deviation values ​​28, 29 of the three adjacent points P2, P3 and P4 and the deviation value of the point P1 37 itself are used in this weighted average calculation 26.

[0159] This smoothing step 11 is carried out by applying a given number of points P1, P2, P3, ..., P n37, points P1, P2, P3, ..., P4 defined with respect to a defined area on the surface of the target geometric shape 30 of the three-dimensional structure are n 37 or all points P1, P2, P3, ..., P n 37. The result of this smoothing step 11 is the smoothed output data 27, which is the points P1, P2, P3, ..., P n 37 are generated. These smoothed output data 27 are converted to smoothed perfect deviation data 12 in the method flow.

[0160] 4 illustrates a 3D structure 30 to be produced by means of a 3D printing method, i.e. a target geometry of the 3D structure to be produced 30. For this target geometry of the 3D structure 30 data of the target geometry of the 3D structure 30 are obtained and these data are converted, for example in a first slicer step 3, into specific instructions for producing an individual layer by means of the 3D printing method, i.e. first layer data 4 of the target geometry of the 3D structure 30.

[0161] In the next first printing step 5, a 3D printer produces a 3D structure by 3D printing using the provided layer data 4. Although the steps of the method of Fig. 1 are not shown in Fig. 4 and the following figures, for a better understanding, the steps of the method of Fig. 1 are designated by their numbers with respect to Fig. 4 and the following figures.

[0162] FIG. 5 illustrates a 3D structure 31 produced by a 3D printing method, i.e., for example, the actual geometry 31 of the 3D structure after a first printing step 5, a curing and post-treatment step 6. This produced 3D structure 31 corresponds to the data of the actual geometry 31 of the 3D structure. To simplify the drawing, a rectangular parallelepiped-shaped object is selected as the 3D structure. In reality, such a 3D structure 31 may have other shapes. Even if there is no unscanned area during 3D scanning of such a rectangular parallelepiped-shaped object, this is assumed here as an example to easily explain the method.

[0163] 5, for example, has two deviations 32, 33 with respect to a given target geometry 30. In this example, the deviations are a concave deviation 32 and a convex deviation 33 displayed on the surface of the fabricated three-dimensional structure 31.

[0164] The fabricated three-dimensional structure 31, illustrated in Fig. 5, is measured in three dimensions, for example, in a measurement step 7. A model 34 of the scanned data of the actual geometry 31 of the three-dimensional structure provided by this measurement step 7 is illustrated in Fig. 6. This model 34 includes points P1, P2, P3, P4, P5, P6, P7, P8, P9, P10, P11, P12, P13, P14, P15, P16, P17, P18, P19, P20, P21, P22, P23, P24, P25, P26, P27, P28, P29, P30, P31, P32, P33, P34, P35, P36, P37, P38, P39, P40, P41, P42, P43, P44, P45, P46, P47, P48, P49, P50, P51, P52, P53, P54, P55, P56, P57, P58, P59, P60, P61, P62, P63, P64, P65, P66, P67, P68, P70, P71, P72, P73, P74, P75, P76, P77, P78, P79, P80, P81, P82, P83, P84, P85, P86, P87, P88, P90, P91, P92, P93, P94, P95, P96, P97, P98, P99, P100, P99, P111, P99, P122, P132, P143, P154, P165, P170, P185, P ’ ,P2 ’ ,P2 ’ ,...,P k ’ 6, the measurement points 8 or data are made up of a small number of points P1 and P2 of the actual geometry 31 of the three-dimensional structure. ’ ,P2 ’ ,P2 ’ ,...,P k ’ Only 36 are shown as examples.

[0165] For example, in the drawing of FIG. 6, an unscanned area 35 is shown. In the measurement step 7, in this unscanned area 35, a point P1 on the surface of the fabricated 3D structure 31 is ’ ,P2 ’ ,P2 ’ ,...,P k ’ No three-dimensional data regarding 36 has been determined, for example, because such area 35 was "invisible" to the sensors of the 3D scanning device.

[0166] Figure 7 illustrates the scanned data of the actual geometry 34 along a cross section or section line AA of the model 34 illustrated in Figure 6. Figure 7 illustrates that in the unscanned areas 35 no data is available regarding the progression of the surface or outer contour of the model 34. In Figure 7 the concave deviations 32 are fully illustrated and the convex deviations 33 are at least partially visible.

[0167] Fig. 8 visualizes a projection step 9 in which data of the actual geometry 31 of the produced three-dimensional structure and data of the target geometry 30 of the three-dimensional structure to be produced are projected onto one another or compared with one another. In this projection step 9, a first deviation 28 between the actual geometry 31 of the three-dimensional structure and the target geometry 30 of the three-dimensional structure is determined at corresponding points. The data of the target geometry 30 is shown with a dashed line, whereas the data of the actual geometry 31 is shown with a solid line. Furthermore, Fig. 8 illustrates an example of an unscanned area 35.

[0168] In the projection step 9, corresponding points P37 and P38 in the scanning area 35 are projected. ’ A first deviation 28 between points P1, P2, P3, ..., P36 is determined. n 37 are on the surface of the target geometric shape 30 of the three-dimensional structure, and their associated points P1 ’ ,P2 ’ ,P3 ’ ,...,P k ’ 8, points P1 37 and P1 36 are located on the surface of the actual geometric shape 31 of the three-dimensional structure. ’ 36 is illustrated in the first deviation 28 between 36 .

[0169] The problem to be solved by this method is to find the corresponding points P37 and P38 in the unscanned area 35. ’ The first deviation 28 between 36 cannot be identified.

[0170] The determined first deviations 28 or deviation data 28 each represent, for example, only the distance between the point P1 37 and its associated point P1' 36 in the scanning area. These determined first deviations 28 are then calculated for each point P1, P2, P3, ..., P1 lying on the surface of the target geometric shape 30 of the three-dimensional structure. n The value specified for 37 is stored.

[0171] FIG. 9 illustrates an example of identifying another deviation 29 for an unscanned area 35 .

[0172] FIG. 9 illustrates a cross-section of the transition from the scanned area to the unscanned area 35 in an enlarged view.

[0173] A prerequisite for determining the further deviations 29 is that the surface of the target part is already present in a triangulated form. As explained in the transformation step 2, the surface of the target geometric shape 30 of the 3D structure to be produced is, for example, completely covered by triangular partial planes. Therefore, the vertices P1, P2, P3, ..., P4 of these triangular partial planes are completely covered by the vertices P1, P2, P3, ..., P4 of the triangular partial planes. n Data on 37 neighbourhood relationships is obtained and subsequently used.

[0174] A triangular partial plane or triangle on the surface of the target geometric shape 30, which is on the periphery within the unscanned area 35 and for which the first deviation data 28 at two vertices have already been determined in the projection step 9, is determined. In the example of FIG. 9, the corresponding points P1 37 and P1 ’ Between 36 and P237 and P2 ’ Since the two first deviations 28 between the points P1 and P2 have been identified, it is a first triangular partial plane consisting of the vertices P1, P2, and P3, in which the first deviations 28 with respect to the vertices P1 and P2 are known. However, within the unscanned area 35, the points P ’ Since data for 36 is missing, within that region 35, the corresponding points P37 and P ’ It is not possible to specify the deviation between the 36.

[0175] When determining or identifying the further deviations 29, the starting point is that the values ​​of the first deviations 28 at the periphery within the unscanned area 35 continue towards the unscanned area so as to have an equivalent value for the further deviations 29. In this example, the further deviations 29 for the vertex P3 of the first triangular partial plane can be determined by identifying the arithmetic mean value of the first deviations 28 known for the points P1 and P2. The first deviations 28 and the further deviations 29 identified for the points P1, P2 and P3 have one value for the deviations, and the direction of these deviations 28 and 29 extends along the vertex normals associated with the points P1, P2 or P3, respectively, which are not shown in the drawing of FIG.

[0176] For example, the determination of another deviation 29 for point P3 by arithmetic averaging of first deviations 28 known for points P1 and P2 is one example. In practice, other calculation rules can be stored besides the averaging, for example taking into account a more complex collection of partial planes consisting of several points P, weighting by priority when the first deviations differ significantly, or a degree of uncertainty of a given another deviation based on its distance to the determined first deviation.

[0177] After the determination of the further deviation 29 for the vertex P3, a further triangular partial plane is determined on the surface of the target geometric shape 30, which is, for example, on the periphery of the unscanned area 35 and for which the first deviation data 28 for the two vertices have already been determined in the projection step 9. In the example of FIG. 9, the corresponding points P2 37 and P2 ’ Between 36 and P437 and P4 ’ Since the two first deviations 28 between the points 36 have been determined, it is a second triangular partial plane consisting of the vertices P2, P4, P5, for which the first deviations 28 for the vertices P2 and P4 are known. The determination of another deviation 29 for the vertex P5 of this second triangular partial plane is performed by calculating the arithmetic mean value of the first deviations 28, again known for the points P2 and P4.

[0178] The corresponding vertices P6, P7, P8, ..., P nThe identification of further deviations 29 with respect to the first vertex or point P37 on the surface of the target geometry 30 of the 3D structure to be produced can be continued, for example, along the periphery of the unscanned area 35 until the first assumed sequence of assumed triangular partial planes is established along the entire peripheral area. This process of identifying further deviations 29 based on the known first deviations 28 with respect to the corresponding vertex or point P37 on the surface of the target geometry 30 of the 3D structure to be produced is illustrated by way of example in FIG. 10. The triangular partial plane illustrated in the drawing of FIG. 10 is illustrated in a diagrammatic form with a model 34 in order to visualize the iterative process of identifying further deviations 29. As already explained, the triangular partial plane is a triangular partial plane created by triangulation of the surface of the target geometry 30.

[0179] For example, if a first sequence of further deviations 29 is identified near the periphery of the unscanned area 35, then it is possible to continue identifying, for example, a second sequence of further deviations 29 in an area further away from the periphery of the unscanned area 35.

[0180] The further determination of this further deviation 29 is explained by the third triangular partial plane consisting of vertices P3, P5, and P6 shown in Fig. 9. In this third triangular partial plane, the further deviation 29 for the vertices P3 and P5 is determined as described above. Therefore, the determination of the further deviation 29 for the vertex P6 can be performed in the manner described above.

[0181] The further identification of another deviation 29 according to this other assumed sequence (which is further away from the periphery of the unscanned area 35 than the first sequence) is illustrated in Fig. 11. The triangular partial plane shown in the drawing of Fig. 11 is illustrated in diagrammatic form in the drawing according to the model 34 in order to visualize the iterative process of identifying another deviation 29 according to the other sequence.

[0182] The process of identifying further deviations 29 within the unscanned region 35 continues iteratively until further deviations 29 are generated, for example relating to the entire unscanned region 35, as visualized in FIG.

[0183] As can be seen from figures 11 and 12, the determination of the further deviations 29 by the method is carried out independently of the progression of the surface of the produced 3D structure 31, which is a particular advantage of the method.

[0184] 13a and 13b, the process of transforming and generating corrected 3D printing data proceeding in transformation step 15 is illustrated diagrammatically in two condensed diagrams.

[0185] In FIG. 13a an excerpt from FIG. 8 is shown, in which the deviations 28 between the actual geometry 31 of the 3D structure and the target geometry 30 of the 3D structure at corresponding points are determined, and these first deviations 28 are represented by corresponding points P1, P2, P3, ..., P n 37 and P1 ’ ,P2 ’ ,P3 ’ ,...,P k ’ We have already explained that the decision is made between the parties.

[0186] An example of this is shown in the drawing of FIG. 13a, for example, in which points P137 and P1 ’ These determined first deviations 28 or deviation data 28 are respectively shown as point P37 and its associated point P36 in the scan area. ’ 36, and each point P1, P2, P3, ..., P n The value specified for 37 is stored.

[0187] In the deformation process illustrated in Fig. 13b, points P1, P2, P3, ..., P n is displaced in the opposite direction to the identified deviation 28 or 29 by the absolute amount of the stored value of the deviation 28 or 29, which depends on the deviation 28 or 29. In this example of Fig. 13, the first identified deviation 28 has, for example, a value of -0.2 mm and points towards the three-dimensional structure, which corresponds to an undersizing of 0.2 mm.

[0188] Therefore, in this example, to generate corrected 3D printing data 38, point P1 is displaced outward from the surface of the target geometry 30 of the 3D structure by the absolute value of 0.2 mm specified for this first deviation 28, in the opposite direction, away from the 3D structure, as illustrated, for example, in FIG. 13b using the double arrow 39 representing the deformation.

[0189] FIG. 14 illustrates an example of a diagram in which corrected 3D printing data 38 that differs from the original data of the target geometric shape 30 is perfectly transformed or perfectly generated, thus reducing or eliminating deterministic distortions when using the corrected 3D printing data 38 to print another 3D structure.

[0190] This deformation and generation of corrected 3D printing data 38 is performed in the scanned areas with the first deviation 28 and in the unscanned areas 35 with the identified other deviation 29. In this way, deformation can be performed over the entire surface of the target geometry 30 of the 3D structure. The deformation process is also illustrated in Figure 14 by the double arrow 39.

[0191] In a first alternative, only a portion of the value specified for the deviation can be used during this transformation, for example 50%, 75% or 90% of the deviation value at point P37.

[0192] In a second alternative, during this transformation, the determined deviation value is multiplied by a coefficient, thus enabling, for example, a transformation of 110%, 125% or 150% of the absolute value of the first or another deviation determined at point P37 to be performed. [Explanation of symbols]

[0193] 1 Preparation process 2. Conversion process 3. First slicer process 4 First layer data 5. First printing process 6 Post-processing 7 Measurement process 8. Measurement points 9 Projection process 10 Complementary process 11 Smoothing process 12 Smoothed perfect deviation data 13 Detection 14. Corrected perfect deviation data 15 Transformation process 16 Deformation Data 17 Second slicer process 18 Second Layer Data 19. Another printing process 20 Alternative transformation process 21 Layer Deformation Data 22 Input Data 23 Interpolation 24 Output Data 25 Smoothing input data 26 Weighted average calculation 27 Smoothed output data 28 First deviation / First deviation data 29 Another deviation / second deviation data 30 3D structure to be produced / target geometric shape of 3D structure 31 Manufactured 3D structure / actual geometry of 3D structure 32 Concave deviation 33 Convex deviation 34 Scanned real geometry data / model 35 Unscanned Areas 36 Points / P' of the actual geometry of the 3D structure 37 Target geometric shape point of 3D structure / P 38 Corrected 3D printing data 39 Double Arrow / Transformation

Claims

1. A method for manufacturing a 3D structure by a 3D printing technique, comprising the steps of: preparing first layer data (4) for individual layers of a 3D structure (30) to be manufactured from 3D printing data, which is data of a target geometric shape (30) of the 3D structure; and using the layer data to control the manufacturing of the 3D structure by the 3D printing technique, The fabricated 3D structure (31) is measured in three dimensions, and in fact incomplete 3D data of the scanned actual geometry (31) of the 3D structure of the surface of the fabricated 3D structure (31) is generated, which has one or more unscanned areas (35) that are reflected in the model (34); In the projection step (9), corresponding points P (37) and P ’ (36) is determined, and this point P (37) is located on the surface of the target geometric shape (30) of the 3D structure and is associated with a point P ’ (36) on a model (34) of the surface of the scanned actual geometry (31) of the 3D structure, respectively; These identified first deviations (28) are then used to calculate the associated points P of the data of the target geometry (30) of the 3D structure. １ , P ２ , P ３ ,... ,P ｎ The value is stored as deviation data for (37); In the interpolation step (10), the interpolation of the other deviations (29) associated with the unscanned areas (35) of the surface of the fabricated 3D structure (31) is performed using the data of the target geometry (30) of the 3D structure and the first deviation data (28) already identified, to obtain the associated points P of the data of the target geometry (30) of the 3D structure. １ , P ２ , P ３ ,... ,P ｎ The value of the other deviation data (29) generated for (37) is stored, thus generating a complete deviation data; Corrected 3D printing data (38) is created in a deformation step (15) in which the data points P １ , P ２ , P ３ ,... ,P ｎ (37) is the point P １ , P ２ , P ３ ,... ,P ｎ and depending on the first or other deviation (28, 29) determined and stored with respect to (37), it is displaced in the opposite direction to the determined first or other deviation (28, 29) by the absolute value of the stored value, which depends on the determined first or other deviation (28, 29). These corrected 3D printing data (38) are used to control the fabrication of subsequent 3D structures by 3D printing techniques; A method characterized by:

2. 10. The method of claim 1, The method is characterized in that the unscanned areas (35) are areas of the surface of the fabricated three-dimensional structure (31) that could not generate data in the model (34) when measured in three dimensions.

3. 10. The method of claim 1, In the transformation step (2), the surface of the target geometric shape (30) of the three-dimensional structure to be produced is reproduced, i.e., triangulated, using a plurality of partial planes each having a plurality of vertices, and points P １ , P ２ , P ３ ,... ,P ｎ (37) is allocated.

4. 3. The method of claim 2, In the above-mentioned complementation step (10), in the first step, a point P on the periphery of the unscanned area (35) is １ , P ２ , P ３ ,... ,P ｎ From the plurality of first deviations (28) identified with respect to (37), a point P １ , P ２ , P ３ ,... ,P ｎ Another deviation (29) with respect to (37) is generated using the function to find the point P １ , P ２ , P ３ ,... ,P ｎ Another deviation (29) with respect to (37) is identified and in a subsequent step this function is used to find the point P １ , P ２ , P ３ ,... ,P ｎ From the plurality of first deviations (28) or other deviations (29) identified with respect to (37), a point P １ , P ２ , P ３ ,... ,P ｎ and identification of another deviation (29) with respect to (37).

5. 5. The method of claim 4, The method, wherein the function is an arithmetic average value operation, an area weighted average value operation, an angle weighted average value operation, or a distance weighted average value operation.

6. 3. The method of claim 2, Point P in the unscanned area (35) １ , P ２ , P ３ ,... ,P ｎ A method characterized in that the identification of the further deviations (29) with respect to (37) starts from the periphery of the unscanned area (35) and continues in the direction of the assumed center of the unscanned area (35).

7. 10. The method of claim 1, In the deformation step (15), point P １ , P ２ , P ３ ,... ,P ｎ The displacement of (37) is １ , P ２ , P ３ ,... ,P ｎ (37) along a vertex normal extending through

8. 10. The method of claim 1, After generating the additional deviations (29) in the interpolation step (10), in the smoothing step (11), each point P １ , P ２ , P ３ ,... ,P ｎ (37) or the selected point P １ , P ２ , P ３ ,... ,P ｎ Regarding (37), the weighted average value is calculated for each point P １ , P ２ , P ３ ,... ,P ｎ smoothing the absolute values ​​of the first and other deviations (28, 29) stored with respect to (37), thus generating smoothed perfect deviation data (12).

9. 10. The method of claim 1, In the transformation step (15), the data points P of the target geometric shape (30) of the 3D structure are １ , P ２ , P ３ ,... ,P ｎ The transformation in (37) is １ , P ２ , P ３ ,... ,P ｎ (37) or all points P １ , P ２ , P ３ ,... ,P ｎ (37).

10. 10. The method of claim 1, In the transformation step (15), the data points P of the target geometric shape (30) of the 3D structure are １ , P ２ , P ３ ,... ,P ｎ (37) is displaced by the product obtained from the absolute value stored for the first or another deviation (28, 29) as a first coefficient and a second coefficient lying in the range of 0.3 to 1.

7.

11. 10. The method of claim 1, a 3D structure subsequently produced using the corrected 3D printing data (38) is measured in three dimensions in order to newly determine by the method first and further deviations (28, 29) between a target geometry (30) of the 3D structure to be produced and an actual geometry (31) of the subsequently produced 3D structure, followed by a deformation step in which the already corrected 3D printing data (38) is at least partially deformed using the first and further deviations (28, 29) to generate further corrected 3D printing data, which are used to control the production of the 3D structure subsequently to be produced by a 3D printing method.

Images