Collimating multiple laser beams for additive manufacturing

Abstract

A method for automatically adjusting scanning optical devices (17A, 17B) is disclosed, the method comprising the following steps: irradiating (step 101) an object area (3A) of a layer of powdered material (5) arranged on a building platform (11) with at least one irradiation beam (15B); irradiating (step 103) a calibration area (21A) of the layer of powdered material (5) with at least one irradiation beam (15B); directing (step 105A) the first irradiation beam (15A) onto an intermediate top surface (31) with a first scanning optical device (17A), thereby melting a first calibration pattern (37) in the intermediate top surface (31); directing (step 105A) the second irradiation beam (15A) onto an intermediate top surface (31) with a second scanning optical device (17B); and The beam (15B) is directed (step 105B) onto the intermediate top surface (31), thereby melting a second calibration pattern (39) in the intermediate top surface (31); acquiring (step 107) at least one image (35, 51A, 51B) of the intermediate top surface (31); using the at least one image (35, 51A, 51B) identifying (step 109) image points associated with geometric features of the calibration pattern; deriving (step 111) a spatial offset (O) between the first geometric feature and the second geometric feature based on the image points; and adjusting (step 113) at least one scanning optical device (17A) of the scanning optical devices while taking into account the spatial offset (O).

Description

Technical Field

[0001] The present disclosure generally relates to an additive manufacturing apparatus with multi-beam irradiation using multiple scanning optics and a method for automatically adjusting the scanning optics of an additive manufacturing apparatus. Background Art

[0002] In additive manufacturing, such as selective laser sintering or selective laser melting, a powdered material, such as metal powder or ceramic powder, is irradiated with electromagnetic radiation. Subsequently, thin layers of powder are provided into a chamber on a build platform to form a three-dimensional object by irradiating the respective powder layers with a radiation beam (e.g., a laser beam). Corresponding apparatuses are referred to as additive manufacturing apparatuses, 3D printing systems, selective laser sintering machines, selective laser melting machines, etc. For information on the operation of such apparatuses, see, for example, EP 2 732 890 A2.

[0003] To increase the speed of the build process or the size of the objects that can be built, multiple irradiation beams can be used within a single additive manufacturing device and on a single build platform. For example, in an additive manufacturing device using multiple laser beams, the coordinate system of the underlying scanning optics guiding each laser beam must be maintained throughout the build process to ensure the correct alignment of the scanned laser trajectory in the layer plane. Specifically, the accuracy of the alignment between the different laser beams / scanning optics is preferably within 50% of the laser beam diameter or less.

[0004] Regarding the calibration of an irradiation system for an apparatus for producing three-dimensional workpieces, WO 2019 / 158394A1 discloses irradiating a first irradiation beam and a second irradiation beam onto an irradiation plane so that the operating axes of the irradiation patterns intersect each other. Calibration is performed by determining the position of the intersection between the irradiation patterns of the first irradiation beam and the second irradiation beam. In addition, EP 2 983898 A1 discloses determining the relative position deviation between a first test pattern and a second test pattern generated by two laser beams in a target within a powder layer for calibrating the corresponding scanner. In addition, WO 2019 / 161886 A1 discloses constructing a first test structure by a first irradiation beam and a second test structure by a second irradiation beam and determining the offset between the two test structures.

[0005] The correct arrangement of multiple laser scanners / scanning optics is particularly necessary if multiple irradiation beams are used to manufacture a single object. Note that, especially with large parts and temperature fluctuations, the thermal effects on the various components and build parts may require continuous calibration of the laser scanners during manufacturing.

[0006] Therefore, it is an object of the present disclosure in particular to provide a calibration procedure which can be easily performed in an automated manner, can also be implemented in a simple manner in existing additive manufacturing devices, and can preferably be performed throughout the entire manufacturing process.

[0007] Generally speaking, the present disclosure is directed, at least in part, to improving or overcoming one or more aspects of existing systems, and more particularly to providing an efficient method for calibrating / adjusting coordinate systems associated with respective irradiation beams in a multi-beam additive manufacturing apparatus.

[0008] Thus, the present disclosure is directed, at least in part, to improving or overcoming one or more aspects of existing systems. Summary of the Invention

[0009] Some of these objects are achieved by a method for automatically adjusting first and second scanning optics as claimed in claim 1 and a device for additive manufacturing of three-dimensional objects as claimed in claim 15. Further aspects and developments are given in the dependent claims.

[0010] In a first aspect, the present disclosure relates to a method for automatically adjusting a first scanning optic associated with a first radiation beam and a second scanning optic associated with a second radiation beam during additive manufacturing of a target object. The method comprises the following steps:

[0011] irradiating an object region of a layer of powdered material disposed on a build platform with at least one of the first radiation beam and the second radiation beam, wherein the object region is associated with a cross-section of the target object;

[0012] irradiating a calibration region of the layer of powdered material with at least one of the first radiation beam and the second radiation beam, wherein the calibration region is associated with a cross-section of a calibration object to form an intermediate top surface of the calibration object;

[0013] directing the first irradiation beam onto the intermediate top surface with the first scanning optics to melt a first calibration pattern having first geometric features into the intermediate top surface;

[0014] directing the second irradiation beam onto the intermediate top surface with the second scanning optics to melt a second calibration pattern having second geometric features into the intermediate top surface;

[0015] acquiring at least one image of the intermediate top surface, the at least one image having the first calibration pattern and the second calibration pattern;

[0016] identifying image points associated with the first geometric feature and the second geometric feature using the at least one image;

[0017] deriving a spatial offset between the positions of the first geometric feature and the second geometric feature within the layer of powdered material based on the image points; and

[0018] At least one of the first scanning optics and the second scanning optics is adjusted taking into account the spatial offset.

[0019] In another aspect, an apparatus for additively manufacturing a three-dimensional object from powdered material comprises:

[0020] an object forming chamber having a work surface and a building platform, wherein the building platform is adjustable in a vertical position relative to the building platform opening for layer-by-layer fabrication of three-dimensional objects and calibration objects on the building platform;

[0021] a calibration camera configured to acquire an image of at least one calibration area of ​​the work surface associated with the calibration object; and

[0022] an irradiation system comprising at least two scanning optics configured to direct respective irradiation beams onto a layer of powdered material distributed on top of the building platform; and

[0023] A controller having at least one microprocessor and at least one memory device storing instructions that, when executed by the at least one microprocessor, are operable to cause the device to control the vertical position of the build platform, receive image data from the calibration camera, control the irradiation system, and perform the method as described above.

[0024] Further embodiments of the above-mentioned aspects are disclosed in the dependent claims, which are incorporated herein by reference.

[0025] For example, in some embodiments of the method, when forming the intermediate top surface of the calibration object, the calibration area may be irradiated in a plurality of portions, wherein:

[0026] - at least one of the plurality of sections is manufactured to have a corrugated surface; and / or

[0027] - irradiating a first portion of the plurality of portions by directing at least one of the first and second irradiation beams along scan vectors extending parallel to each other or extending less than 10° out of parallel relation.

[0028] In some embodiments, the method may further include the following steps:

[0029] Directing an illumination beam onto the intermediate top surface having the first calibration pattern and the second calibration pattern, wherein the incident direction of the illumination beam has:

[0030] an azimuth component in the range of 45° to 90° relative to a scan vector of one of the portions of the calibration area, and

[0031] An inclination component, the inclination component relative to the normal direction of the powdery material layer is in the range of 25° to 80°, preferably in the range of 40° to 60°, and wherein,

[0032] When the illumination beam is directed onto the intermediate top surface, at least one image of the intermediate top surface is acquired having the first calibration pattern and the second calibration pattern.

[0033] In some embodiments, for each orientation of the parallel scan vectors, directing an illumination beam onto the intermediate top surface and acquiring an image, and the method may further include:

[0034] generating a superposition of images acquired for each configuration of the scan vector, and

[0035] Based on the superposition of these images, the image points of the first feature and the second feature are identified.

[0036] In some embodiments, the method may further include the following steps:

[0037] For each of the first calibration pattern and the second calibration pattern, a plurality of straight line segments having flat surface areas are melted, wherein,

[0038] For each of the first calibration pattern and the second calibration pattern, at least one of the plurality of straight line segments extends in the first portion and is oriented at a first intersection angle in the range of 45° to 90°, preferably in the range of 60° to 90° or 60° to 80°, relative to a scan vector for the first portion.

[0039] In some embodiments, a second portion of the plurality of portions may be irradiated by directing at least one of the first irradiation beam and the second irradiation beam along scan vectors extending parallel to each other or extending less than 10° from parallel, and

[0040] The scanning vector of the second portion differs in orientation from the scanning vector of the first portion by an angle in the range of 45° to 90°, preferably by an angle in the range of 80° to 90°.

[0041] In some embodiments, for each of the first calibration pattern and the second calibration pattern, at least one of the plurality of straight line segments may extend in the second portion and be oriented at an intersection angle in the range of 45° to 90°, preferably in the range of 60° to 90° or 60° to 80°, relative to a scan vector for the second portion.

[0042] In some embodiments of the method, the calibration area may comprise two opposing portion pairs, each portion pair being irradiated by directing at least one of the first irradiation beam and the second irradiation beam along a scan vector, wherein the scan vectors of the two portion pairs are oriented relative to each other at an angle in the range of 45° to 90°, preferably in the range of 80° to 90°,

[0043] The first calibration pattern may include a pair of straight line segments that intersect at a first reference point defining the first geometric feature, each of the straight line segments may extend sufficiently within one of a pair of opposing portions of the calibration area to enable estimation of the orientation of the straight line segment (e.g., within 20% or more for small calibration objects),

[0044] the second calibration pattern may include a pair of straight line segments that intersect at a second reference point defining the second geometric feature, and each of the straight line segments extends sufficiently within one of the pair of opposing portions of the calibration area to enable estimation of the orientation of the straight line segment (e.g., for small calibration objects, to within 20% or more),

[0045] The spatial offset is the difference between the positions of the first reference point and the second reference point,

[0046] These straight line segments are respectively generated by moving the corresponding irradiation beam along a linear scanning vector, wherein the orientation of the linear scanning vector of the first calibration pattern is inclined relative to the orientation of the linear scanning vector of the second calibration pattern, and preferably the linear scanning vector of the first calibration pattern and the linear scanning vector of the second calibration pattern intersect at an angle in the range of 80° to 90°.

[0047] In an exemplary embodiment, a line segment having a length of, for example, 8 mm may be created on a cylinder having a diameter of 36 mm.

[0048] In some embodiments of the method, the calibration area may comprise at least two portions, and the at least two portions are irradiated by directing at least one of the first irradiation beam and the second irradiation beam along a scan vector, wherein the scan vector in one of the at least two portions is oriented at an angle in the range of 45° to 90°, preferably in the range of 80° to 90°, relative to the scan vector in the other of the at least two portions, and

[0049] The first calibration pattern and the second calibration pattern may each comprise straight line segments produced by directing a respective irradiation beam along a linear scan vector, wherein the orientation of the linear scan vector of the first calibration pattern is tilted relative to the orientation of the linear scan vector of the second calibration pattern, and

[0050] These straight line segments can be associated with linear extensions that intersect at a first reference point defining the first geometric feature or a second reference point defining the second geometric feature, respectively, wherein these linear extensions preferably intersect at an angle in the range of 80° to 90° for the first calibration pattern and the second calibration pattern, respectively.

[0051] In some embodiments of the method, the first calibration pattern and the second calibration pattern may each include straight line segments, and the method may further include:

[0052] In the at least one image, deriving straight lines extending through the image of the straight line segments and identifying image points at intersections of the straight lines, and

[0053] The spatial offset in the powdery material layer is determined from the distance between the image points in the image.

[0054] In some embodiments of the method, adjusting the first scanning optics or the second scanning optics may include setting a beam path through the respective scanning optics or a zero point associated with the respective scanning optics for directing the laser beam across the powder bed according to a machine coordinate system; and / or

[0055] A first layer associated with the calibration object can be attached to the build platform, and the calibration object can preferably be positioned at a border region of the build platform. Additionally or alternatively, the first and second calibration patterns can be melted into corresponding intermediate top surfaces of the calibration object as multiple layers of powdered material are applied to produce the target object, so that adjustments of the first and second scanning optics can be repeatedly performed throughout additive manufacturing.

[0056] In some embodiments of the apparatus, the apparatus may further comprise:

[0057] an illuminator positioned to direct an illumination beam onto the calibration area, and

[0058] The at least one storage device may further store instructions that, when executed by the at least one microprocessor, are operable to activate the illuminator to illuminate the calibration area and control the calibration camera to acquire an image when the illuminator is activated.

[0059] In some embodiments of the device, the illuminator can be configured to emit a light beam onto the intermediate top surface of the calibration object with an incident direction in the range of 25° to 80°, preferably in the range of 40° to 60°, relative to the normal direction of the powdered material layer.

[0060] In some embodiments of the apparatus, at least one of the following is included:

[0061] The calibration camera may be positioned vertically above the calibration area,

[0062] The calibration camera can be configured to image the entire work surface.

[0063] The calibration area can be located at the edge of the building platform opening, or

[0064] The apparatus may further include a powder system configured to distribute powdered material onto the work surface using a powder deposition tool to provide the powdered material layer by layer onto the build platform during additive manufacturing. The first scanning optics and the second scanning optics are preferably adjusted so that when the calibration procedure is repeated in a subsequent layer of powdered material, the first feature / reference point and the second feature / reference point differ in position by less than a threshold. The threshold may optionally be set to be equal to or less than a laser beam spot size, for example, 50% of the laser beam spot size on the layer of powdered material.

[0065] Another aspect relates to a computer program product comprising instructions for causing the apparatus for additive manufacturing disclosed herein to perform the steps of the method for automatic adjustment.

[0066] In some embodiments, when an additive manufacturing instruction protocol including irradiation instructions for manufacturing a target object is read into a control unit, the control unit adds irradiation instructions for manufacturing a calibration object and calibration instructions for performing image analysis and offset determination layer by layer.

[0067] Another aspect relates to a method for marking and analyzing an intermediate layer, in particular for optically calibrating one or more scanning optical devices. The method comprises:

[0068] generating an intermediate layer having a diffusely scattering and / or rough surface, wherein the intermediate layer is generated by additive manufacturing based on a first irradiation strategy;

[0069] marking the intervening layer at the top surface using a laser beam for additive manufacturing based on a second irradiation strategy;

[0070] illuminating the top surface with a light beam at an angle that produces increased scattering; and

[0071] An image of the top surface is captured in which the imaging signal / detected light of areas produced only with the first irradiation strategy (eg for diffuse scattering and / or rough surfaces) is enhanced relative to the imaging signal of areas marked with the second irradiation strategy.

[0072] An advantage of the calibration disclosed herein is that any part / object can be produced with multiple laser beams without any quality drawbacks due to misadjustments of the laser beams, in particular of the scanning optics.

[0073] Other features and aspects of the present disclosure will become apparent from the following description and accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] The accompanying drawings, which are incorporated herein and constitute a part of the specification, illustrate exemplary embodiments of the present disclosure and, together with the description, serve to explain the principles of the present disclosure. In the drawings:

[0075] Figure 1 is a schematic diagram of an exemplary additive manufacturing apparatus;

[0076] Figure 2 is a schematic image of a top surface of a calibration structure having two pairs of cross lines as a calibration pattern;

[0077] Figure 3 is a flow chart illustrating an exemplary calibration procedure during additive manufacturing;

[0078] Figure 4A is a schematic diagram of an exemplary dual-beam scanning procedure for preparing an intermediate calibration layer having different reflective properties in different portions of a corresponding intermediate top surface;

[0079] Figure 4B The first laser beam is applied to Figure 4A Schematic diagram of the first pair of cross lines on the prepared calibration layer shown;

[0080] Figure 4C It is a schematic diagram of using scan vectors to generate cross lines;

[0081] Figure 4D is a schematic diagram of calibrating multiple scanning optical devices;

[0082] Figure 4E A second laser beam is applied to Figure 4B A schematic diagram of a second pair of crosshairs on the calibration layer shown;

[0083] Figure 5A It passes through Figure 4A a schematic cross-sectional view of the calibration layer of the indicated portion II or III for illustrating that the light of the illuminator is partially redirected onto the calibration camera;

[0084] Figure 5B is a schematic cross-sectional view of a calibration layer along the lines of a calibration pattern to illustrate specular reflection of light from an illuminator away from a camera;

[0085] Figure 5C is Figure 5B Images of the top surface acquired by the calibrated camera under the lighting conditions shown;

[0086] Figure 6A and Figure 6B An image of an intermediate calibration layer is shown, where two pairs of intersecting scan lines are illuminated from the side by corresponding illuminators;

[0087] Figure 7A and Figure 7B Schematic diagrams of alternative configurations of the calibration object and another calibration procedure, respectively;

[0088] Figure 8A and Figure 8B is a schematic diagram of repeated calibration and readjustment of an irradiation system during additive manufacturing and the resulting offset reduction;

[0089] Figure 9A and Figure 9B Schematic diagram of an exemplary additive manufacturing setup with two powder bed cameras and an image of the powder bed, respectively;

[0090] Figure 10A and Figure 10B are two images displaced relative to each other;

[0091] Figure 11A and Figure 11B are used for Figure 10A and Figure 10B a graph with correlation curves and corresponding correlation maps for an image of ;

[0092] Figure 12A and Figure 12B are the superimposed image and the resulting composite image, respectively; and

[0093] Figure 13 is a superimposed image where stitching is applied to the corners of the image. DETAILED DESCRIPTION

[0094] The following is a detailed description of exemplary embodiments of the present disclosure. The exemplary embodiments described herein and illustrated in the accompanying drawings are intended to teach the principles of the present disclosure, thereby enabling one of ordinary skill in the art to implement and use the disclosure in many different environments and for many different applications. Therefore, the exemplary embodiments are not intended to, and should not be construed as, limiting the scope of patent protection. Rather, the scope of patent protection is defined by the appended claims.

[0095] The present disclosure is partly based on the recognition that correct adjustment of a multi-beam irradiation system can be ensured by additionally generating a calibration structure during the additive manufacturing process (i.e., simultaneously with and preferably separately from the generation of the 3D object). The calibration structure comprises specific intermediate calibration layers, which are used to analyze the calibration state of the multi-beam irradiation system. The intermediate calibration layers can comprise specific calibration patterns, which are each generated with one of the irradiation systems and each define a reference point for this irradiation system. Furthermore, it has been recognized that the detection of the calibration pattern can be improved by forming the calibration pattern on a specially prepared intermediate calibration layer, which improves the image acquisition, in particular the contrast, under correspondingly set lighting conditions.

[0096] The following combination Figure 1 Describe the additive manufacturing device. Figures 2 to 6B , an exemplary implementation of the concept proposed herein for calibrating multiple radiation sources is described for additive manufacturing based on multiple laser beams. Figure 7A and Figure 7B , describes alternative geometrical implementations and further uses of the imaging concepts described herein. Figure 8A and Figure 8B Exemplary implementations of concepts for calibrating multiple scanning optics into an additive manufacturing process are described.

[0097] Figure 1 There is schematically shown an apparatus 1 for additively manufacturing a three-dimensional object 3 (also referred to herein as a target object) from a powdered material 5. The apparatus 1 comprises an object forming chamber 7 within a housing providing the required environment for the manufacturing process. Figure 1 Schematically shown are a top cover 9A and a work surface 9B defining the chamber 7. The chamber 7 is further defined by side walls and a rear wall (not shown). A front door (also not shown) may provide access to the chamber 7.

[0098] The additive manufacturing process is performed on work surface 9B, in particular on build platform 11. Build platform 11 can be lowered in the Z direction to successively provide new layers of powdered material for the layer-by-layer manufacturing process. The powdered material can be evenly spread onto build platform 11 by powder deposition tool 13 to form a powder bed. Powder deposition tool 13 extends in the Y direction and can be moved in the X direction across work surface 9B to build platform opening 11A, where the powder bed is formed on top of build platform 11. Thus, the layers of powdered material 5 extend within a layer plane (here, the XY plane).

[0099] The powder bed is prepared for irradiation with laser radiation so that a powder melting process is initiated by absorption of the laser radiation, followed by a solidification process within the corresponding cross section of the object 3 .

[0100] To this end, the powdered material is irradiated with a plurality of laser beams 15A, 15B, emitted by one or more laser sources and directed to specific locations on the powder bed. Each laser beam 15A, 15B is directed by scanning optics 17A, 17B (also referred to as a scanning system or scanner), which are arranged relative to a machine coordinate system extending in the work surface 9B. The scanning optics 17A, 17B may include lenses and mirrors, and in particular scanning mirrors, whose orientation is adjustable in space, so as to reflect the laser beam in a specific direction to a desired point in the machine coordinate system. The orientation of the scanning mirrors can be set by microactuators, for example, which are used to orient the scanning mirrors, for example, in 3D space. In principle, the scanning field achieved by the scanning optics can be associated with the machine coordinate system in the layer plane (XY plane). Each scanning optic 17A, 17B is calibrated over its respective scanning field (the scanning field does not necessarily have to be 100% of the entire powder bed). Calibration of the scanning field is typically performed by software / digital calibration of the parameters used to drive the scanning optics 17A, 17B. For example, by adjusting the scanning mirrors in 3D space, the parameters allow setting the XY offset of the zero point relative to the powder bed and the rotation angle of the scanning movement (e.g., the scanning movement in one or more cardinal directions on the powder bed, such as the XY directions). In addition, with respect to the scanning process, the amplitude and / or speed of the scanning movement can be set.

[0101] Since the concept disclosed herein is also applicable to more than two laser beams, Figure 1 An exemplary third scanning optics 17C is indicated in dashed lines in FIG. The laser source(s), such as a fiber laser or a disk laser, may be provided on top of the top cover 9A, or may be positioned differently such that the laser beam needs to be directed to the scanning optics 17A, 17B, for example by optical fibers.

[0102] The fabrication process can be monitored using, for example, a monitoring camera 19 positioned centrally above build platform 11. Monitoring camera 19 can have a resolution of approximately 90 μm per pixel and can be configured to provide low-distortion images of work surface 9B. Providing a dedicated camera for monitoring generative growth can provide improved performance. However, as described below, a single camera can perform multiple tasks; for example, monitoring camera 19 can also be used for calibration procedures.

[0103] Each of the laser beams 15A, 15B moves across the powder bed according to a scan trajectory. For example, the scan trajectory can be composed of a plurality of linear scan vectors, i.e., linear trajectory segments. Parallel linear scan vectors can be grouped into hatches. As defined in the CAD / CAM control file for a particular manufacturing job (also referred to herein as the additive manufacturing instruction protocol), the plurality of hatches extend across each cross-section of the object 3.

[0104] To control additive manufacturing, apparatus 1 includes a controller 20 having at least one microprocessor 20A and at least one memory device 20B. The at least one memory device 20B stores instructions that, when executed by the at least one microprocessor 20A, are operable to cause apparatus 1 to control the vertical position of build platform 11, receive image data from a camera, and control the irradiation system (e.g., setting laser parameters and directing the laser beam across the powder bed). For example, a CAD / CAM control file can be loaded into the at least one memory device 20B to provide instructions to microprocessor 20A.

[0105] The CAD / CAM control file includes instructions that, when executed by at least one microprocessor 20A of the controller 20, define which laser beam is to be directed onto which portion of the cross-section. To this end, the controller 20 is connected, for example, to the scanning optics 17A, 17B to set up the corresponding drivers. Obviously, a high spatial alignment of the laser beam trajectories of the different laser beams 15A, 15B is required to ensure proper bordering of the areas irradiated by the different laser beams 15A, 15B.

[0106] Furthermore, the CAD / CAM control file may include parameter values ​​for the laser system, for lowering the build platform 11 , for the powder deposition tool 13 , and the like.

[0107] For the calibration disclosed herein, the CAD / CAM file is extended with instructions to form the calibration object 21 at the same time as the object 3 is formed. Figure 1 As shown, calibration object 21 is, for example, cylindrical with a diameter of approximately 8 to 10 mm. Calibration object 21 is positioned close to the edge of building platform 11, thereby occupying only a limited space for arranging object 3 to be built.

[0108] For calibration, the device 1 further comprises a calibration camera 23 and two illuminators 25, 27 for illuminating the working surface 9B, in particular the top layer of the calibration object 21. Schematically, Figure 1 Two illumination light beams 25A, 27A are indicated in . The illuminators 25 , 27 may be LED light sources configured to emit light beams onto the work surface 9B, in particular onto the intermediate top surface of the calibration object 21 .

[0109] The calibration proposed in this paper can be performed every few layers (i.e., intermediate layers), for example, every four layers. Calibration can also be started as needed. Figure 1 A plurality of intermediate layers 29 in the calibration object 21 are schematically shown. For example, the intermediate layers 29 are used to calibrate the offset. "Intermediate" here relates to the fact that these layers are used for calibration but are subsequently overlaid. A first intermediate layer 29_1 can be attached to the building platform 11 and allows a first calibration to be performed. The intermediate layer 29A is Figure 1 The "current" is processed and used for calibration.

[0110] For the intermediate layer 29A, Figure 1 An illustration of a circular (intermediate) top surface 31 of a cylindrically shaped calibration object 21 is shown. An image of the top surface 31 can be acquired with a calibration camera 23. With reference to the following discussion of the calibration concepts disclosed herein, a laser-generated calibration pattern 33 is schematically indicated on the top surface 31 shown.

[0111] Specifically, if Figure 2 As shown in more detail in , each laser beam 15A, 15B can be used to create a cross-shaped structure (two pairs of mutually intersecting linear lines or linear lines associated with the identified structure) on the surface of the calibration object 21 by a remelting process (ie after the intermediate layers have been processed).

[0112] The calibration camera 23 takes a picture of the top surface 31 of the cross-shaped structure with the markings. The controller 20 is data-connected to the calibration camera 23 (and optionally the monitoring camera 19) for receiving image data. Similarly, the controller 20 is data-connected to the two illuminators 25, 27 for controlling the lighting conditions. (See Figure 1 20C in the exemplary data line.)

[0113] The calibration procedure uses, for example, a calibration algorithm that may be executed in the microprocessor 20A of the controller 20. The calibration algorithm may include an image analysis subroutine and a calibration setup subroutine.

[0114] The image analysis subroutine is programmed to identify a reference point (e.g., a center or intersection) for each cross and calculate the offset between the two crosses. The calibration setup subroutine is programmed to derive an offset compensation value as a control parameter for one or both of the scanning optics 17A, 17B. By applying the control parameter, the regulation of the laser beams 15A, 15B is improved.

[0115] For calibration, the beam spot size of each of the laser beams 15A, 15B (as given in the area of ​​the calibration object 21 at the working surface 9B) can be, for example, in the range of 50 μm up to 100 μm. The calibration camera 23 can have, for example, a resolution of 25 μm per pixel and can be dedicated exclusively to the calibration procedure. Although Figure 1 The configuration of apparatus 1 shown has monitoring camera 19 positioned over the center of building platform 11 (also referred to as substrate) and calibration camera 23 located over a border area of ​​the substrate, but in some embodiments monitoring camera 19 may also be used for the calibration procedure.

[0116] Figure 2 An overlay image 35 of the top surface 31 is shown. The overlay image 35 is based on a combination of Figure 3 The results of the two images acquired by the explained procedure. Specifically, two images imaged with different lighting settings are superimposed on each other.

[0117] Superimposed image 35 shows two crosses 37, 39 (calibration patterns) formed by corresponding pairs of orthogonal straight line segments (short straight lines) 37A, 37B and 39A, 39B (these lines are line-shaped surface areas created by remelting top surface 31). Each of the illumination settings increases the contrast of straight lines that extend substantially along the corresponding illumination direction, such as lines 37A, 39A and lines 37B, 39B.

[0118] For each of the crosses 37, 39, the intersection points 37C, 39C of the corresponding straight lines are schematically indicated. The intersection points 37C, 39C can be considered as reference points in the manufacturing coordinate system, which should be at specific preset positions on the work surface 9B. For example, the intersection points 37C, 39C should coincide to achieve precise adjustment. At least the offset (spatial distance) between the intersection points 37C, 39C should be less than a preset value, such as less than 50% of the beam spot size. When determining the offset associated with the irradiation system of the device 1, in particular the scanning optical devices 17A, 17B used to generate the crosses 37, 39, the image analysis subroutine derives the image points corresponding to the intersection points, and the calibration setting subroutine derives the distance between the image points in the image plane from the distance between the image points, which is related to the offset of the scanning optical devices 17A, 17B.

[0119] As mentioned above, a remelting process may be used to increase the visibility of the crosses 37 and 39 in images acquired with the calibration camera 23 .

[0120] Figure 3 A flow chart for an exemplary calibration procedure is shown. As a first step 101, the remelting process requires irradiating the sample with a first irradiation beam ( Figure 1 The laser beam 15A in the second irradiation beam ( Figure 1 At least one of the laser beams 15B in the irradiation irradiates the calibration area 21A of the layer of powdered material 5 and forms the intermediate top surface 31 of the calibration object 21; the calibration area 21A is associated with the cross section of the calibration object 21 in the XY plane (see Figure 1 ).

[0121] As mentioned, the formation of the intermediate layer 29A of the calibration object 21 occurs during the layer-by-layer manufacturing of the object 3, for example, during the irradiation with at least one of the first and second irradiation beams ( Figure 2 Step 103 in the above example) The object region 3A of the powdered material 5 layers (see Figure 1 ); object region 3A is associated with a cross section of the (target) object 3 in the XY layer plane. In other words, steps 101 and 103 occur during the manufacture of the layer of powdered material 5.

[0122] It should be noted that, in principle, an intermediate layer 29A of the calibration object 21 processed in step 101 can comprise several layers of one of the objects 3, for example if a thinner layer is required. That is, step 103 can be performed multiple times before a powder layer of sufficient thickness is accumulated for the calibration procedure. In principle, this can also be the case in reverse.

[0123] Once intermediate top surface 31 is formed, in step 105A, a first irradiation beam is directed onto calibration region 21A using first scanning optics 17A. Consequently, the first irradiation beam melts a first calibration pattern having first geometric features (e.g., a cross 37 with line intersections) into top surface 31. Similarly, in step 105B, a second irradiation beam is directed onto calibration region 21A using second scanning optics 17B. Consequently, the second irradiation beam melts a second calibration pattern having second geometric features (e.g., a cross 39 with line intersections) into top surface 31. Steps 105A and 105B can be performed sequentially or simultaneously.

[0124] Now, with both calibration patterns remelted on the intermediate calibration layer 29A, the calibration camera 23 captures one or more images from the top surface 31 (step 107). The one or more images are processed by the image analysis subroutine. The image analysis subroutine identifies image points associated with the first feature of the first calibration pattern (e.g., Figure 2 The intersection point 37C in the second calibration pattern) and the image point of the second feature of the second calibration pattern (e.g. Figure 2 39C) (step 109).

[0125] Based on the image points (corresponding to intersection points 37C, 39C), the calibration setup subroutine derives a spatial offset defined in the layer of powdered material 5 (i.e., at the level of work surface 9B) in step 111, typically based on an offset based on image data. Thus, an offset is defined between the XY positions of work surface 9B corresponding to the first and second features identified in the acquired image. Using information regarding the derived spatial offset O, at least one of first scanning optics 17A and second scanning optics 17B is set to reduce the spatial offset, i.e., to improve the regulation of laser beams 15A, 15B (step 113).

[0126] like Figure 3 As further shown, a "calibrated" scanning optics setup can be used to process multiple layers of the object 3 (additional step 103). This generation also results in the generation of the calibration object 21 (only the additional layer formation step 101' is performed). The calibration can then be started again (step 101), followed by steps 105A and 105B, etc., as described above.

[0127] In order to further increase the visibility of the crosses 37 and 39 in the images acquired with the calibration camera 23, the remelting process can be adapted to affect the reflective properties of the intermediate top surface 31 (here the area that was not remelted in steps 105A and 105B). For example, when forming the intermediate layer 29A (top surface 31), it is recommended to apply a specific hatching strategy to the powder bed in the calibration area 21A, which increases the image contrast for specific lighting directions. Figures 4A to 6B The basic concept is described.

[0128] Figure 4A A schematic top view of the top surface 31 (and the surrounding powder bed / powdered material 5) is shown. The illumination beams 25A, 27A are indicated to fall onto the top surface 31 from an orthogonal direction (or at least a nearly orthogonal direction, e.g., at 60° to 90°). The projection of the central ray of the incident illumination beam 25A, 27A onto the work surface 9B is shown in FIG. Figures 4A to 4E This is shown by lines 25B, 27B (extending in the Y-direction and the X-direction, respectively; ie, lines 25B and 27B extend orthogonally to each other in the plane of the working surface 9B).

[0129] The hatching strategy divides top surface 31 into four sections I, II, III, and IV. Sections I, II, III, and IV are segments of, for example, 90° angular widths (i.e., quadrants) of the circular cross-section of calibration object 21. Sections I, II, III, and IV are adjusted relative to illumination beams 25A and 27A so that the respective bisectors of sections I and IV extend along line 25B, while the respective bisectors of sections II and IIII extend along line 27B.

[0130] In each of the four sections I, II, III, IV, the hatching strategy may be specifically chosen to increase the reflection of light from the illumination beams 25A, 27A towards the calibration camera 23. For example, the hatching strategy may be chosen by scanning along a (substantially) parallel scan vector 41 or along a scan vector 41' that deviates from parallel by less than 10° (e.g. Figure 4A 15B) directs the irradiation beam to irradiate portions I, II, III, IV. For example, this hatching strategy within the calibration region 21A may be applied to either or both of the laser beams 15A, 15B.

[0131] like Figure 5A As shown, when using a system with parallel or largely parallel scan vectors (see e.g. Figure 4A When a laser beam (e.g., a scanning vector 41, 41' in the X direction in parts I and IV or a scanning vector 42 in the Y direction in parts II and III) irradiates a portion of the calibration area 21A, this portion may form a wave-shaped surface 43 (e.g., as shown in parts II and III) with ridges 43A and valleys 43B extending in the Y direction. As a result, the light of the illumination beam 27B is reflected in multiple directions (arrows 45), in particular towards the calibration camera 23, which may even be reflected, for example, as shown in FIG. Figure 6A This is particularly true if a particular azimuth direction of incidence is chosen, for example if the projection line 27B of the light beam extends in the X direction. Typically, the azimuth component θ may be in the range of 45° to 90° relative to the scan vectors 41, 42; 53A, 53B of one of the portions of the calibration area 21A (see Figure 4A , wherein θ is equal to approximately 90°). Furthermore, the tilt component φ is in the range of 25° to 80°, preferably in the range of 40° to 60°, relative to the normal direction n of the layer of powdery material 5 .

[0132] Under such lighting conditions, the top surface 31 appears bright in the image 47 of the calibration camera 23 (see Figure 5C , parts II and III are white, except for the dark line 49 explained below).

[0133] Return to reference Figure 4A , in the image captured by calibration camera 23 when illuminator 27 is activated, parts II and III appear bright, while parts I and IV appear dark. In contrast, when illuminator 25 is activated, parts I and IV appear bright, while parts II and III appear dark.

[0134] Exemplary “real” images 51A, 51B taken under these two lighting conditions are shown in Figure 5. Figure 6A and Figure 6B In each lighting condition, you will see opposing quadrants that appear bright or dim.

[0135] In other words, the black or white quadrant of the top surface 31 is associated with a particular orientation of the hatching and the direction of illumination. Figure 4A The hatching strategy shown can be used to capture two different images under different lighting conditions (ie, different lighting directions) to improve the contrast of the calibration pattern in the images.

[0136] Figure 4B Shown Figure 2 A first calibration pattern 37 is shown, with two intersecting lines 37A, 37B shown as hatched. Each line is an example of a straight line segment. Intersecting lines 37A, 37B are marked / created by a remelting process immediately after processing of the intermediate layer 29 is completed and without applying an additional layer of powdered material to the calibration area 21A. Each of lines 37A, 37B is created by moving (the same) laser beam 15A linearly, i.e., with a linear scan vector 44, across the calibration area 21A.

[0137] Reference Figure 4C , further illustrating the use of scan vectors. Each of the lines 37A, 37B can be irradiated by a plurality (e.g., eight or ten, etc.) of parallel scan vectors 44A, 44B. In addition, in some embodiments, the central region 58 of the top surface 31 may not be used for the lines 37A, 37B, because overheating of the powder / calibration object may occur during irradiation in the intersection region of the lines 37A, 37B. Therefore, the scan trajectory 44A ends before the central region 58 (e.g., by blocking the laser beam) and continues scanning on the other side of the central region, thereby forming the scan trajectory 44B respectively. It should be noted that in some embodiments, the image analysis subroutine may not require the lines 37A, 37B to extend in the central region 58, because the intersection point 37C of the lines 37A, 37B can also be determined based solely on the remaining (imaged) portion of the lines 37A, 37B outside the central region 58.

[0138] In any case, linear movement of the laser beam along the linear scan vector 44, 44A, 44B results in a flat surface region extending in a linear shape across the contoured surface 43 (at least in the outer portion as discussed above).

[0139] like Figure 5B As shown, when irradiating a substantially flat surface area 46 generated by the calibration pattern, the incident light of the illumination beam 27A is specularly reflected. As long as the calibration camera 23 is not located in the region of the specularly reflected beam 27A', the calibration camera 23 does not receive the specularly reflected beam 27A'. This situation can occur for segments of the intersection lines 37A, 37B that are oriented along a specific beam incidence direction. For example, in Figure 4BIn the example shown in FIG. 2 , the projection line 27B and the line 37B of the light beam extend in the X direction. Under such lighting conditions, a dark line 49 appears in the image 47 acquired by the calibration camera 23, as shown in FIG. Figure 5C This is shown as an example for a straight line segment extending through parts II and III.

[0140] like Figure 4B As shown, line 37A extends primarily in sections I and IV and is oriented at an intersection angle α in the range of 45° to 90°, preferably in the range of 60° to 80°, and exemplarily at 70°, relative to the scan vectors 41, 41′ for these sections. Similarly, line 37B extends primarily in sections II and III and is oriented at an intersection angle β in the range of 45° to 90°, preferably in the range of 60° to 80°, and exemplarily at 70°, relative to the scan vector 42 for these sections.

[0141] Figure 4B It is further shown that intersection 37C of lines 37A, 37B is located in portion II and is displaced ΔX and ΔY relative to the center of top surface 31. Intersection 37C is a feature of the first calibration pattern (here cross 37) to be identified based on the image(s) taken with calibration camera 23.

[0142] It is noted that the surface of the intermediate layer 29A may differ from the general surface characteristics described above, for example, in the area at the center of the top surface 31 (where all the segments meet) and in areas where straight line segments intersect. Therefore, the image analysis subroutine can focus on the segments within the less affected segments and show the expected surface characteristics, typically the interior areas of these segments.

[0143] Regarding the identification of the location of intersection 37C, the image analysis subroutine can identify two linear segments of line 37A in sections I and IV in the image data acquired under illumination by illuminator 25. Similarly, the image analysis subroutine can identify linear segments of line 37B in sections II and III in the image data acquired under illumination by illuminator 27. The image analysis subroutine can then calculate the extent of the linear segments and first identify the location in the image plane (i.e., in the image data) and then translate that location onto work surface 9B.

[0144] exist Figure 4E In FIG, the top surface 31 is additionally shown with a second calibration pattern, here a cross formed by lines 39A, 39B. Figure 4B As discussed, the intersection point 39C can be identified and its location in the working surface 9B derived.

[0145] In addition, Figure 4E3. An enlarged portion 59 of the central region of the top surface 31 is shown in FIG. 1. The spatial offset O between the intersection point 37C and the intersection point 39C can be seen. Based on this, the scanning optics 17A, 17B can be adjusted to correctly direct the laser beam onto the work surface 9B.

[0146] Refer again Figure 6A and Figure 6B Images 51A and 51B of calibration region 21A are shown, showing two dark lines extending through illuminated sections II and III and I and IV, respectively. These dark lines are associated with lines 37B and 39B and lines 37A and 39A, respectively. From the images, it is clear that the intersecting lines should be rotated relative to one another, preferably extending on opposite sides relative to the projection line of the light beam, since the optical conditions for these lines are comparable.

[0147] It is to be noted that in corresponding cases also lighted lines can be seen in the dark quadrants, due to the presence of curvature in the otherwise flat surface causing the light beam to be partially redirected towards the calibration camera 23 .

[0148] For example, the images 51A, 51B may be processed by an image analysis subroutine to derive image points corresponding to respective intersection points as described above.

[0149] The above-described exemplary implementation of the calibration is based on a cylindrically shaped calibration object and a division of the top surface into equal quadrants.

[0150] However, these concepts can be applied to a variety of geometries.

[0151] Reference Figure 7A Another implementation is based on a rectangular cross-section of the calibration object. The cross-section is divided into two rectangular parts I', II', and each of these parts is subject to a specific hatching strategy as indicated by the scan vectors 53A and 53B (e.g., a configuration of parallel scan vectors or substantially parallel scan vectors). The scan vectors 53A, 53B are oriented relative to each other at an angle γ' in the range of 45° to 90°, preferably in the range of 80° to 90°.

[0152] The calibration patterns each comprise linear line segments 37A', 37B'; 39A', 39B' generated by moving the respective irradiation beam along a linear scan vector 44'. Figure 7A, the calibration pattern is, for example, V-shaped and comprises two oblique line segments respectively. The V-shaped pattern does not need to cross within the calibration area 21A, but rather defines reference points 37C', 39C' which are generally located next to the calibration object 21 by extending the detected linear segments (extensions 55, 57). The linear extensions 55, 57 preferably cross at an angle δ' in the range of 80° to 90°, respectively. For the cross, the orientation of the linear scan vector 44' of the first V-shaped calibration pattern is inclined relative to the orientation of the linear scan vector 44' of the second V-shaped calibration pattern. The linear scan vectors 44' of the first calibration pattern and the second calibration pattern extend in at least two parts I', II', respectively.

[0153] When first scanning optical device 17A and second scanning optical device 17B are adjusted, first reference point 37C' and second reference point 39C' differ in position by less than a threshold value (i.e., have an offset less than a threshold value). The threshold value may be equal to or less than the laser beam spot size (e.g., approximately 50 μm to 100 μm).

[0154] To enhance contrast, again by applying a specific lighting direction with respect to the hatching strategy / scan vector orientation, one or the other of the rectangular portions I', II' will appear bright / dark in the image acquired by the calibration camera 23. Similarly, the straight line segments 37A', 37B'; 39A', 39B' may appear dark in the corresponding illuminated rectangular portions I', II', allowing the image analysis subroutine to derive corresponding image points for reference points of the calibration pattern and initiate adjustments of the scanning optics 17A, 17B.

[0155] Reference Figure 7B In addition to determining the offset between corresponding scanning optics by comparing intersection lines associated with different laser beams, the concepts described herein also allow calibration of one scanning optic when comparing patterns generated with the same laser beam. Figure 7B In the example, straight line segments 61A, 61B and 63A, 63B are generated with one of the laser beams 15A, 15B. This allows adjusting the scanning optics and / or setting the zero point of the scanning optics to correctly shift / move the laser beam spot across the powder bed according to the machine coordinate system.

[0156] In general, the concepts disclosed herein can achieve accuracy in the range of 20 μm and below for a 50% spot size of approximately 30 μm.

[0157] As will be explained, the above-described calibration of the individual scanning optics, as well as the relative calibration between the scanning optics, may degrade as the response time varies.

[0158] like Figure 8AAs shown, four types of calibration measurements are performed in a repetitive manner using two laser beams L and R guided by corresponding scanning optics. For example, the sequence is set so that a specific type of measurement is repeated every twelve layers. Thus, in this example, crosses can be generated for the following situations:

[0159] (1) Relative alignment of laser beams L and R at the initial layer 1,

[0160] (2) calibration of the laser beam L at layer 4,

[0161] (3) Relative alignment of the laser beams R and L at layer 7,

[0162] (4) Calibration of the laser beam R at layer 10, and so on...

[0163] It is noted that in measurements (1) and (3) the laser beams L and R are back-referenced, according to which crosses are associated with the laser beams L and R in an alternating manner in order to reduce systematic errors.

[0164] Note that measurements (2) and (4) can be used to zero the measurement system, specifically each scanning optic. This is analogous to a scale: before weighing something, you must press the "zero" button. Thus, you "press the zero button" for each measurement (12 layers).

[0165] Based on such a calibration measurement sequence, the offset O can be kept below a preset threshold. Figure 8B The "Adjustment Accuracy" graph shows an exemplary development of the offset O measurement over a number of layers N. Prior to layer S, no calibration procedure was performed, resulting in a large offset of more than 50 μm being measured at layer S. Once the correction was activated, the offset O decreased and stabilized below an acceptable threshold 71 (e.g., a threshold of 20 μm).

[0166] While the foregoing discussion exemplarily relates to the calibration of two laser systems, large powder beds may even require more than two laser beams and, therefore, more than two scanning optics to be calibrated. Similarly, the use of several laser beams to accelerate manufacturing may also be the case.

[0167] Those skilled in the art will recognize that it is possible to perform a paired calibration of the scanning optics. This can be performed by alternating the irradiation of the laser beams to be calibrated during the generation of the calibration object. Alternatively, several calibration objects can be generated, one for each laser beam pair to be calibrated, or a large calibration object can be used with different segments assigned to the respective laser beam pairs (e.g., a rectangular shape with aligned segments or a circular shape with evenly distributed segments). In this case, one of the laser beams is typically considered the primary laser beam, and the "other" scanning optics are calibrated separately relative to the scanning optics of the primary laser beam. For example, for four laser beams 1, 2, 3, and 4, laser beams 1:2, 1:3, and 1:4 can be calibrated with laser beam 1 being the primary laser beam.

[0168] In addition, refer to Figure 4D It will be appreciated that by increasing the size (e.g. diameter) of the calibration object, it is possible to distinguish between multiple different lines marked into only one calibration object's top surface 31. This can be achieved even if the different lines differ only slightly in orientation (e.g., by an angle of only a few degrees). Figure 4D Four pairs of cross lines are shown, each consisting of two segments: lines 81A / 81B-81C / 81D, lines 83A / 83B-83C / 83D, lines 85A / 85B-85C / 85D, and lines 87A / 87B-87C / 87D, each forming a cross and defining a corresponding intersection point for calibration. These lines are used to mark calibration objects of a size selected based on the resolution of the calibration camera and the quality of the image analysis. Figure 4D As shown, it is still possible to mark all crossing pairs with an angle close to 45 degrees. In this case in particular, the laser irradiation / marking in the central area can be reduced to avoid increasing the energy density beyond a certain threshold.

[0169] Even in this case, a selected one of the laser beams is considered the main laser beam.All other laser beams are calibrated, ie the scanning optics are shifted relative to the zero point so that they themselves are over-positioned relative to the main scanning optics.

[0170] In summary, during an additive manufacturing process, for example, a cylindrical calibration object can be generated along with one or more target objects. During the additive manufacturing process, laser markers are repeatedly (e.g., periodically) written onto a specifically selected intermediate calibration layer of the calibration object. The laser markers can be, for example, calibration patterns, such as a cross, that is melted into the top surface of the cylindrical calibration object at a specific moment in time. A calibration camera is then used to acquire an image of the laser markers.

[0171] Contrast enhancement can be achieved by specifically adapting the additive manufacturing process of the calibration layer to the available lighting conditions. For example, vectors that scan essentially in parallel can be used for specific parts, thereby increasing the reflection of light towards the calibration camera. In contrast, if the laser markers include linear segments in the direction of the incident light, the calibration camera can receive less scattered light from these laser markers. Based on the acquired one or more images, an algorithm can process and analyze the image data to calculate the offset between laser markers produced with different laser beams (relative calibration), i.e. different laser sources guided by different scanning optics. Corresponding feedback can then be sent to the scanning optics to maintain the correct adjustment of the laser beam / scanning optics. Alternatively, a single scanning optic can be calibrated when the laser markers are produced with the same laser beam.

[0172] Another aspect disclosed herein relates to image acquisition, and in particular to image acquisition based on multiple cameras. For image acquisition and metrology applications in mechanical engineering, images often consist of several separate images taken of portions of the area to be imaged. In some cases, the lack of high-contrast structures can make automatic stitching difficult or lead to a loss of accuracy in subsequent metrological measurements. If images are stitched together by detecting and specifying edges and / or corners, and typically marked points (as examples of characteristic features), the stitching may not be efficient and applicable.

[0173] Particularly for additive manufacturing devices with large build platforms, a single monitoring camera / viewpoint may not be sufficient to provide a complete view of the build platform. For example, if the build platform exceeds a certain size (e.g., a diameter in the range of approximately 500 mm and above), a single monitoring camera will not be able to provide the required detail resolution for the entire area of ​​the build platform. Therefore, the required monitoring image is stitched together from several images taken by different monitoring cameras or by a camera that moves above the build platform.

[0174] exist Figure 9A In FIG, a plurality of monitoring cameras 19, 19' are schematically shown. In addition to the number and position of the monitoring cameras 19, 19', Figure 9A Corresponding to Figure 1 For the corresponding description of the remaining features of the additive manufacturing device, reference is made to the above Figure 1 Description.

[0175] Monitoring cameras 19, 19' can be used to monitor the manufacturing process. Monitoring cameras 19, 19' are positioned above build platform 11, for example symmetrically relative to the center of build platform 11. Build platform 11, and in particular the powder bed, can be considered the area to be imaged. Due to their purpose, monitoring cameras 19, 19' can also be referred to as powder bed cameras. Monitoring cameras 19, 19' can have a resolution of approximately 90 μm per pixel and can be configured to capture images of work surface 9B with low distortion.

[0176] If combined Figure 1 As explained, the controller 20 is data-connected to the monitoring cameras 19 , 19 ′ for receiving image data.

[0177] Each of the monitoring cameras 19 , 19 ′ is associated with a portion of the building platform 11 and provides a respective image of the associated portion.

[0178] Figure 9B Image 201 of build platform 11 and portions 219, 219' of build platform 11, each associated with monitoring cameras 19, 19', are shown. To obtain a complete image 201 of build platform 11, the images from monitoring cameras 19, 19' need to be combined, also known as image stitching or image recording. However, if combined image 201 is to be used, for example, to analyze the surface of a powder bed disposed on build platform 11, the stitching needs to be as accurate as possible.

[0179] In additive manufacturing, for example, Figure 9B As can be seen in FIG, the surface of the powder bed segment does not have specific characteristic features that can be used for image analysis due to the diffuse scattering of the powder 5. Therefore, the simple conventional procedure mentioned above for image combination based on specific features cannot be applied.

[0180] Even if no recorded high-contrast features are available or located Figure 9B Within the overlapping region 221 shown, the stitching procedure described below can also form a combined image.

[0181] Specifically, we recommend using a correlation algorithm that compares the textures of the images to be recorded. For the recording, we identify the point of maximum cross-correlation. Based on this point of maximum cross-correlation, we can derive a shift vector that indicates the positional change between the images. Therefore, the presence of specific features and their corresponding identification are not required.

[0182] Furthermore, in other areas of the metal or steel processing industry, similar optical situations may arise, so stitching together a combined image from several individual images becomes a metrological task. An example is seam tracking in the monitoring of laser cutting or laser welding processes, especially since the corresponding laser cutting or laser welding devices typically have (one or more) on-axis viewing cameras. For example, correctly stitched seam images can be used to control the path of a robotic arm that guides the laser head across the material to be processed. Furthermore, in these types of applications, the individual images to be recorded have no or only few high-contrast features, since the welded metal surface can have a very smooth and uniform appearance.

[0183] In the following, the stitching procedure is described exemplarily for an image of a metal surface with laser markers, so that the observer has at least some sense of relative position. However, it is noted that an image of a powder bed (e.g. Figure 9B ) can in particular be spliced ​​in a corresponding manner as will be recognized by a person skilled in the art.

[0184] In this context, another object of the invention disclosed herein is to provide an image stitching program that allows combining images, in particular images with a relatively smooth appearance, such as images of a powder bed. For example, if, for example, multiple calibration objects are created at different areas of the powder bed (as the positioning of multiple calibration objects may typically depend on the specific manufacturing operation and the area available for calibration objects), the above-mentioned calibration analysis can be performed on the stitched image.

[0185] The inventors have recognized that although images of powders or metal surfaces may appear uniformly gray, they contain a texture typical of the respective material. Whereas in the case of a powder bed, texture can be created due to the different orientations of the powder particles, which results in, for example, diffuse scattering, the texture of a metal surface is given by traces of processing of the metal surface during the manufacture of the metal workpiece.

[0186] The concept disclosed herein is based on the fact that the resulting texture is physically connected to the workpiece. The mentioned registration does not require individual features, since a control procedure based on maximizing the cross-correlation function of the individual images to be stitched is applied. The procedure is as follows.

[0187] (i) In the search window ( Figure 9B In the portion 223 in the search window, the image data of the two images to be stitched are shifted relative to each other stepwise, for example within an assumed unadjusted range. For the (remaining) overlap within the search window, a common cross-correlation value Corr is calculated. An exemplary cross-correlation calculation can be based on the following definition of the cross-correlation value Corr:

[0188] Corr=Sum_i,j(IMG1_ij*IMG2_ij) / N(IMG1,IMG2),

[0189] Wherein, i and j represent pixel coordinates, IMG1 and IMG2 represent the images to be spliced. Therefore, IMG1_ij and IMG2_ij represent the pixel values ​​of the corresponding images at pixel positions i and j. The correlation value is calculated for a given relative position change between the images. Obviously, the position change can be given in one or two dimensions. N(IMG1, IMG2) represents an image-specific scaling. N(IMG1, IMG2) can be, for example, derived from an image-specific scaling, such as norm(IMG1)*norm(IMG2), for example, the average pixel value of the image derived within a search window or a complete window. In other words, the correlation value can be derived by pixel-by-pixel multiplication of the pixel values ​​of the image and summing the multiplication results. According to the above, the correlation value Corr will be derived for a predefined set of potential position changes to be considered.

[0190] (ii) The calculated correlation value Corr may be stored in a correlation data map that relates the correlation value to its underlying position change. In other words, the correlation data map stores the correlation value Corr as a function of the image position change.

[0191] (iii) The maximum value among the calculated correlation values ​​Corr may be derived, for example, based on an entry of a correlation data map.The maximum value among the calculated correlation values ​​Corr indicates an image offset having a larger (highest measured) overlap of image information.

[0192] (iv) Registration of the images is then performed based on the position of the maximum value among the calculated correlation values ​​Corr in the correlation data map, which indicates the positional change between the image data.

[0193] It is important to note that all pixels of the image contribute to the correlation peak, since all image information from within the search window is encoded in the underlying correlation function. Therefore, even very small offsets (position changes) can be detected immediately.

[0194] Furthermore, it is noted that position changes can use translations and rotations to account for linear position changes as well as rotations within the camera facility.

[0195] Reference Figure 10A and Figure 10BAs well as step (i) mentioned above, two slightly displaced images are shown. These images include image data 225A and 225B, which have data points / image points / pixels associated with pixel values ​​indicated on the side in a grayscale. These images show a metal surface with a laser engraved outline 227. These images were taken with the help of a robotic arm that moves the camera substantially along the engraving. Thus, the images represent first and second image data related to image acquisition performed by one or more cameras at different positions relative to the area to be imaged. When stitching the images, the stitching is based on the relative positions of the images. The relative position indicates the displacement of the pixels of one image relative to the pixels of the other image within a coordinate system generated by the pixels.

[0196] exist Figure 10A and Figure 10B In the specific example of images 225A, 225B, the task of image analysis is to derive information about the camera movement, i.e. the direction and length of the displacement of the imaging areas relative to each other, or generally the relative position changes of a set of essentially identical image data within the two images 225A, 225B.

[0197] For image analysis, a search window of, for example, 200 x 200 pixels is selected in the area falling within each of the two images 225A, 225B. Within the search window, the cross-correlation value is calculated based on the above formula.

[0198] Figure 11A Exemplary graphs of the correlation values ​​Corr of the relative position changes in the X direction (graph 229A) and the Y direction (graph 229B) are indicated in FIG. Exemplarily, the graphs are selected so as to pass through the highest correlation value Corr_max (see Figure 11B ).

[0199] Figure 11B The associated data map 331 is shown for all possible position changes based on the selected size and possible combinations of position changes in the X and Y directions. Figure 11B The peaks of the highest correlation values ​​Corr_max are shown as white dots in the correlation data map 331 .

[0200] Reference Figure 12A , a superimposed image 333 can be created by superimposing the images 225A, 225B subjected to the shift based on the correlation value Corr_max. In addition to the large central overlapping area, a border-like stripe originating from only one of the images is also seen.

[0201] Reference Figure 12B, a combined image 335 as superimposed image 333 is based on the correlation value Corr_max. Combined image 335 comprises at each pixel the contribution from only one of the images. In combined image 335 no transition lines or cracks indicating stitching can be seen.

[0202] Reference Figure 13 , shows another example of an overlay image 337 , indicating that the algorithm also provides good stitching results for the angular laser mark 337A.

[0203] The following summarizes aspects related to image recording:

[0204] Aspect 1. A method for stitching together two images of an area to be imaged, wherein the images include first image data (225A) and second image data (225B), each of the first image data (225A) and the second image data (225B) including image data of a respective portion (219, 219') of the area to be imaged, the portions (219, 219') overlapping in a common overlap region (221) of the area to be imaged, the method comprising:

[0205] defining a search window portion (223) within each of the first image data (225A) and the second image data (225B);

[0206] For a plurality of relative positions of the first image data (225A) and the second image data (225B), a correlation value (Corr) between the image data within the search window portion (223) of the first image data (225A) and the second image data (225B) is obtained;

[0207] identifying a maximum correlation value (Corr_max) among the derived correlation values; and

[0208] The two images are stitched for the relative position of the first image data (225A) and the second image data (225B) associated with the maximum correlation value (Corr_max).

[0209] Aspect 2. A method as described in Aspect 1, wherein obtaining the correlation value (Corr) includes pixel-by-pixel multiplication of the pixel values ​​of the image data within the corresponding search window portion (223), in particular according to the corresponding relative position, and summing the results of the pixel-by-pixel multiplication.

[0210] Aspect 3. The method of aspect 1 or 2, wherein deriving the correlation value (Corr) comprises normalizing based on the image data or the corresponding image data within the search window portion (223).

[0211] Aspect 4. The method of any one of Aspects 1 to 3, wherein the plurality of relative positions are generated by relative translation and / or rotation of the first image data (225A) and the second image data (225B).

[0212] Aspect 5. The method according to any one of aspects 1 to 4, further comprising:

[0213] The correlation values ​​(Corr) obtained are stored in a correlation data map (331) or a function of relative position, which relates the correlation values ​​(Corr) to a base relative position, in particular a position change between the images.

[0214] Aspect 6. A method as described in any of aspects 1 to 5, wherein the search window portion (223) is selected to extend over subsets of pixels in the first image data and the second image data, respectively, which subsets correspond to sub-areas of the area to be imaged having the same size, and wherein these subsets in particular include the same number of pixels; and / or

[0215] Therein, the search window portion (223) is preferably selected to include image data associated with the common overlapping region.

[0216] Aspect 7. A method as described in any one of Aspects 1 to 6, wherein the pixels (also referred to as data points or image points) in each image data substantially correspond to pixel areas of the same size in the area to be imaged.

[0217] Aspect 8. The method according to any one of aspects 1 to 7, further comprising:

[0218] acquiring an image of a first portion (219, 219') of the area to be imaged, the image comprising first image data (225A),

[0219] acquiring an image of a second portion (219, 219') of the area to be imaged, the image comprising second image data (225B),

[0220] The images are acquired such that each of the first image data (225A) and the second image data (225B) includes image data of a common overlapping region (221) of the region to be imaged.

[0221] Aspect 9. A method as described in any of Aspects 1 to 8, wherein the first image data (225A) and the second image data (225B) relate to images acquired with two monitoring cameras (19, 19') located at different positions relative to the building platform (11) or with one monitoring camera (19, 19') moved to two positions relative to the area to be imaged.

[0222] Aspect 10. A method as described in any of Aspects 1 to 9, wherein the area to be imaged is in particular a substantially flat surface area, and / or wherein the area to be imaged is a building platform (119) having a powder bed or a metal surface, which powder bed or metal surface is optionally obtained during additive manufacturing of the object.

[0223] Aspect 11. An apparatus (1) for additively manufacturing a three-dimensional object (3) from a powdered material (5), the apparatus (1) comprising:

[0224] An object forming chamber (7) having a working surface (9B) and a building platform (11), wherein the building platform (11) is adjustable in a vertical position relative to a building platform opening (11A) for manufacturing a three-dimensional object (3) and a calibration object (21) layer by layer on the building platform (11);

[0225] two monitoring cameras (19, 19'), each of which is configured to capture an image of at least a portion of the building platform (11); and

[0226] an irradiation system for directing an irradiation beam (15A, 15B) onto a layer of powdered material (5) distributed on top of a building platform (11); and

[0227] A controller (20) having at least one microprocessor (20A) and at least one storage device (20B) storing instructions which, when executed by the at least one microprocessor (20A), are operable to perform the method according to any one of aspects 1 to 9.

[0228] Finally, examples of additive manufacturing machines to which the concepts disclosed herein can be applied include selective laser sintering machines or selective laser melting machines, such as "mysint100," "mysint200," "mysint300," "TruPrint1000," "TruPrint2000," "TruPrint3000," and "TruPrint5000" manufactured by TRUMPF SISMA and / or TRUMPF. Manufacturing metal- or ceramic-based components according to the concepts disclosed herein can find applications in a variety of technical fields, such as medical, dental, aerospace, and automotive applications.

[0229] It is expressly stated that all features disclosed in the description and / or claims are intended to be disclosed individually and independently of each other for the purpose of original disclosure and for the purpose of limiting the claimed invention independently of the combination of features in the embodiments and / or claims. It is expressly stated that all value ranges or indications of groups of entities disclose every possible intermediate value or intermediate entity for the purpose of original disclosure and for the purpose of limiting the claimed invention, in particular as limitations of value ranges.

[0230] Although preferred embodiments of the present invention have been described herein, improvements and modifications may be incorporated without departing from the scope of the appended claims.

Claims

1. A method of automatically adjusting a first scanning optic (17A) associated with a first irradiation beam (15A) and a second scanning optic (17B) associated with a second irradiation beam (15B) during additive manufacturing of a target object (3), the method comprising: irradiating (step 101) an object region (3A) of a layer of powdered material (5) arranged on a building platform (11) with at least one of the first irradiation beam (15A) and the second irradiation beam (15B), wherein the object region (3A) is associated with a cross section of the target object (3); irradiating (step 103) a calibration area (21A) of the layer of powdered material (5) with at least one of the first irradiation beam (15A) and the second irradiation beam (15B), wherein the calibration area (21A) is associated with a cross section of a calibration object (21) thereby forming an intermediate top surface (31) of the calibration object (21); directing (step 105A) the first irradiation beam (15A) onto the intermediate top surface (31) with the first scanning optics (17A), thereby melting a first calibration pattern (37) having a first geometric feature in the intermediate top surface (31); directing (step 105B) the second irradiation beam (15B) onto the intermediate top surface (31) with the second scanning optics (17B), thereby melting a second calibration pattern (39) having a second geometric feature in the intermediate top surface (31); Acquiring (step 107) at least one image (35, 51A, 51B) of the intermediate top surface (31), the at least one image (35, 51A, 51B) having the first calibration pattern (37) and the second calibration pattern (39); identifying (step 109) image points associated with the first geometric feature and the second geometric feature using the at least one image (35, 51A, 51B); Determining (step 111 ) the spatial offset (O) between the positions of the first geometrical feature and the second geometrical feature within the layer of powdered material (5) from these image points; and adjusting (step 113) at least one of the first scanning optics (17A) and the second scanning optics (17B) taking into account the spatial offset (O), wherein, when forming the intermediate top surface (31) of the calibration object (21), the calibration area (21A) is irradiated in a plurality of parts (I, II, III, IV), wherein: - at least one of the plurality of parts (I, II, III, IV) is manufactured to have a corrugated surface (43); and / or - irradiating a first portion (I) of the plurality of portions (I, II, III, IV) by directing at least one of the first irradiation beam (15A) and the second irradiation beam (15B) along scan vectors (41, 41') extending parallel to each other or extending less than 10° out of parallel relation, The method further comprises: Directing (step 115) an illumination beam (25A, 27A) onto an intermediate top surface (31) having the first calibration pattern (37) and the second calibration pattern (39), wherein the incident direction of the illumination beam (25A, 27A) has an azimuth component (θ) in the range of 45° to 90° relative to a scan vector (41, 42; 53A, 53B) of one of the portions of the calibration area (21A), and Tilt component The tilt component is in the range of 25° to 80° relative to the normal direction (n) of the layer of powdered material (5), and, When the illumination light beam (25A, 27A) is directed onto the intermediate top surface (31), at least one image of the intermediate top surface (31) is acquired with the first calibration pattern (37) and the second calibration pattern (39).

2. The method according to claim 1, wherein For each orientation of parallel scan vectors (41, 42, 41'; 53A, 53B), directing an illumination beam (25A, 27A) onto the intermediate top surface (31) and acquiring an image (51A, 51B), the method further comprising: generating a superposition of images (51A, 51B) acquired for each configuration of scan vectors (41, 42, 41'; 53A, 53B), and Based on the superposition of the images (51A, 51B), image points of the first feature and the second feature are identified.

3. The method according to claim 1 or 2, wherein The method further comprises: For each of the first calibration pattern (37) and the second calibration pattern (39), a plurality of straight line segments (37A, 37B; 39A, 39B) having a flat surface area (46) are melted, wherein For each of the first calibration pattern (37) and the second calibration pattern (39), at least one of the plurality of straight line segments (37A, 37B; 39A, 39B) extends in the first portion (I) and is oriented at a first intersection angle (α) in the range of 45° to 90° relative to a scan vector (41, 41') for the first portion (I).

4. The method according to claim 3, wherein: irradiating a second portion (II) of the plurality of portions (I, II, III, IV) by directing at least one of the first irradiation beam (15A) and the second irradiation beam (15B) along a scan vector (42) extending parallel to each other or extending less than 10° out of parallel relation, and The scanning vector (42) of the second portion (II) differs in orientation from the scanning vector (41, 41') of the first portion (I) by an angle (γ) in the range of 45° to 90°.

5. The method according to claim 4, wherein For each of the first calibration pattern (37) and the second calibration pattern (39), at least one of the plurality of straight line segments (37A, 37B; 39A, 39B) extends in the second portion (II) and is oriented at an intersection angle (β) in the range of 45° to 90° relative to a scan vector (42) for the second portion (II).

6. The method according to any one of claims 1, 2, 4 and 5, wherein: The calibration area (21A) comprises two opposing portion pairs, each portion pair being irradiated by directing at least one of the first irradiation beam (15A) and the second irradiation beam (15B) along a scan vector (41, 42), wherein the scan vectors (41, 42) of the two portion pairs are oriented at an angle (γ) in the range of 45° to 90° relative to each other, The first calibration pattern (37) comprises a pair of straight line segments (37A, 37B) intersecting at a first reference point (37C) defining the first geometric feature, each of the straight line segments (37A, 37B) extending within one of the pairs of opposing portions of the calibration area (21A) for a length sufficient to enable assessment of the orientation of the straight line segments, The second calibration pattern (39) includes a pair of straight line segments (39A, 39B) intersecting at a second reference point (39C) defining the second geometric feature, each of the straight line segments (39A, 39B) extending within one of the pairs of opposing portions of the calibration area (21A) for a length sufficient to enable assessment of the orientation of the straight line segments, The spatial offset (O) is the difference between the positions of the first reference point (37C) and the second reference point (39C), These straight line segments (37A, 37B, 39A, 39B) are respectively generated by moving the corresponding irradiation beam along a linear scan vector (44), wherein the orientation of the linear scan vector (44) of the first calibration pattern (37) is inclined relative to the orientation of the linear scan vector (44) of the second calibration pattern (39).

7. The method according to any one of claims 1, 2, 4 and 5, wherein: The calibration area (21A) comprises at least two portions (I', II'), the at least two portions (I', II') being irradiated by directing at least one of the first irradiation beam (15A) and the second irradiation beam (15B) along a scan vector (53A, 53B), wherein the scan vector (53A, 53B) in one of the at least two portions (I', II') is oriented at an angle (γ') in the range of 45° to 90° with respect to the scan vector (53A, 53B) in the other of the at least two portions (I', II'), and The first and second calibration patterns each comprise straight line segments (37A', 37B'; 39A', 39B') generated by directing a respective irradiation beam along a linear scan vector (44'), wherein the orientation of the linear scan vector (44') of the first calibration pattern is tilted relative to the orientation of the linear scan vector (44') of the second calibration pattern, and These straight line segments (37A', 37B'; 39A', 39B') are associated with linear extensions (55, 57) that intersect at a first reference point (37C') defining the first geometric feature or a second reference point (39C') defining the second geometric feature, respectively.

8. The method according to any one of claims 1, 2, 4 and 5, wherein: The first calibration pattern (37) and the second calibration pattern (39) each comprise straight line segments, the method further comprising: In the at least one image, deriving straight lines extending through the image of the straight line segments and identifying image points at intersections of the straight lines, and The spatial offset (O) in the layer of powdery material (5) is determined from the distance between these image points in the image.

9. The method according to any one of claims 1, 2, 4 and 5, wherein: Adjusting (step 113) the first scanning optics (17A) or the second scanning optics (17B) comprises setting a beam path through the respective scanning optics or a zero point associated with the respective scanning optics for directing the laser beam across the powder bed according to a machine coordinate system; and / or attaching a first layer (29_1) associated with the calibration object (21) to a build platform (11); and / or For multiple layers of powdered material (5) applied for manufacturing a target object (3), the first calibration pattern (37) and the second calibration pattern (39) are melted into the corresponding intermediate top surface (31) of the calibration object (21) to repeatedly perform the adjustment of the first scanning optical device (17A) and the second scanning optical device (17B) throughout the additive manufacturing.

10. The method according to claim 1, wherein The tilt component is in the range of 40° to 60° relative to the normal direction (n) of the layer of powdered material (5).

11. The method according to claim 3, wherein: For each of the first calibration pattern (37) and the second calibration pattern (39), at least one of the plurality of straight line segments (37A, 37B; 39A, 39B) is oriented at a first intersection angle (α) in the range of 60° to 90° or 60° to 80° relative to a scan vector (41, 41′) for the first portion (I).

12. The method according to claim 4, wherein: The scanning vector (42) of the second portion (II) differs in orientation relative to the scanning vector (41, 41') of the first portion (I) by an angle (γ) in the range of 80° to 90°.

13. The method according to claim 5, wherein: For each of the first calibration pattern (37) and the second calibration pattern (39), at least one of the plurality of straight line segments (37A, 37B; 39A, 39B) is oriented at an intersection angle (β) in the range of 60° to 90° or 60° to 80° relative to a scan vector (42) for the second portion (II).

14. The method according to claim 6, wherein The scanning vectors (41, 42) of the two partial pairs are oriented relative to one another at an angle (γ) in the range of 80° to 90°.

15. The method according to claim 6, wherein The linear scan vector (44) of the first calibration pattern (37) and the linear scan vector (44) of the second calibration pattern (39) intersect at an angle (δ) in the range of 80° to 90°.

16. The method according to claim 7, wherein The scanning vector (53A, 53B) in one of the at least two parts (I', II') is oriented at an angle (γ') in the range of 80° to 90° relative to the scanning vector (53A, 53B) in the other of the at least two parts (I', II').

17. The method according to claim 7, wherein: These linear extensions (55, 57) intersect at an angle (δ') in the range of 80° to 90° for the first calibration pattern and the second calibration pattern, respectively.

18. The method according to claim 9, wherein The calibration object (21) is positioned at a border area of ​​the building platform (11).

19. A device (1) for additively manufacturing a target object (3) from a powdered material (5), the device (1) comprising: An object forming chamber (7) having a working surface (9B) and a building platform (11), wherein the building platform (11) is adjustable in a vertical position relative to a building platform opening (11A) for manufacturing a target object (3) and a calibration object (21) layer by layer on the building platform (11); a calibration camera (23) configured to acquire an image of at least one calibration area (21A) of the work surface (9B) associated with the calibration object (21); an illuminator (25, 27) positioned to direct an illumination beam (25A, 27A) onto the calibration area (21A); an irradiation system comprising at least two scanning optics (17A, 17B) configured to direct respective irradiation beams (15A, 15B) to a layer of powdered material (5) distributed on top of the building platform (11); and A controller (20) having at least one microprocessor (20A) and at least one storage device (20B) for storing instructions, the instructions being operable when executed by the at least one microprocessor (20A) to cause the apparatus (1) to control the vertical position of the building platform (11), to activate the illuminators (25, 27) to illuminate the calibration area (21A), to control the calibration camera (23) to acquire an image when the illuminators (25, 27) are activated, to receive image data from the calibration camera (23), to control the irradiation system, and to perform the method of any one of claims 1 to 18.

20. The device (1) according to claim 19, wherein Having at least one of the following characteristics: The calibration camera (23) is positioned vertically above the calibration area (21A), The calibration camera (23) is configured to image the entire working surface (9B), The calibration area (21A) is located at the boundary area of ​​the building platform opening (11A), and The apparatus (1) further comprises a powder system configured to distribute powdered material (5) on the working surface (9B) using a powder deposition tool (13) to provide the powdered material (5) layer by layer onto the building platform (11) during additive manufacturing.

Images