阅读说明：本技术 增材制造装置的功率辐射源的焦点校准 (Focus calibration of a power radiation source of an additive manufacturing device ) 是由 A·罗布兰 J-P·尼凯斯 I·诺维科夫 于 2018-06-19 设计创作，主要内容包括：本发明涉及一种用于校准增材制造装置的功率辐射源的头部系统的套件,其包括：校准板,其具有多个参考标记；发射支撑件,由至少一种对源辐射敏感的材料制成,当所述支撑件在所述校准板上就位时,所述支撑件使得所述校准板的参考标记可见,其特征在于,所述发射支撑件包括多个窗口,这些窗口分布成使得窗口与所述校准板的各个参考标记重叠,并且当所述发射支撑件在所述校准板上就位时,使得所述校准板的各个参考标记可见。本发明还涉及用于校准这种系统的方法。(The invention relates to a kit for calibrating a head system of a power radiation source of an additive manufacturing device, comprising: a calibration plate having a plurality of reference marks; an emitting support made of at least one material sensitive to source radiation, said support making the reference marks of the calibration plate visible when said support is in place on the calibration plate, characterized in that it comprises a plurality of windows distributed so that they overlap the respective reference marks of the calibration plate and make the respective reference marks of the calibration plate visible when said emitting support is in place on the calibration plate. The invention also relates to a method for calibrating such a system.)

1. A method for calibrating a head system of a power radiation source of an additive manufacturing device, wherein for determining a correction to be applied to a control of the system the following steps are performed:

arranging a calibration plate comprising a plurality of reference marks in an additive manufacturing device,

controlling the source to mark at least one calibration pattern formed by a plurality of impact points on a calibration plate,

acquiring at least one image of the calibration pattern and at least one reference mark,

determining at least one corrective control from the one or more images thus obtained,

characterized in that the step of determining a control of correction determines, in the acquired image, a distribution of diameters of impact points of a calibration pattern appearing in the image, the control of correction depending on the distribution of diameters of the impact points and being focus control.

2. The method of claim 1, wherein in the step of controlling the source, various emissions corresponding to different impact points are generated with different focus controls.

3. The method of claim 2, wherein in the step of controlling the source, the various emissions corresponding to different impact points are generated with focus control that is incremented such that the diameter is gaussian around a midpoint of the pattern.

4. A method according to any one of claims 2 and 3, wherein the step of determining a corrected focus control determines the impact point of minimum diameter or a derived "waist point" from the distribution of diameters of the impact points.

5. The method according to claim 4, wherein the control of the correction associated with the marked mid-point is a focus control for the impact point thus determined or a derived control corresponding to the derived "waist point".

6. The method according to any of claims 1 and 2, wherein the calibration pattern formed by the impact points is a matrix and the focus control values are increased column by column, row by row.

7. A method according to any preceding claim, wherein the source is controlled by a head system to generate on the calibration plate markings of a plurality of calibration patterns theoretically centred on a preset theoretical target point.

8. The method according to claim 7, wherein in the acquisition step an optical measuring device is moved over the calibration plate in order to acquire at least one image of the area in which the reference mark is present on the one hand and the immediate theoretical target position on the other hand for each reference mark of the calibration plate.

9. A method according to claim 7 or 8, wherein for each reference marker, one or more acquired images are processed to derive therefrom the location of the impact point or derived "waist point" of minimum diameter in the orthogonal frame of reference of the reference marker, and the offset between the location of said point and the theoretical target position is determined, and an additional correction is determined in dependence on the offset.

10. The method according to any one of the preceding claims, wherein, for determining a correction, a treatment is performed to convert from a frame of reference of the calibration plate to a frame of reference related to the head.

11. The method of claim 10, wherein the processing comprises: the method comprises identifying a pattern corresponding to two given reference marks on the calibration plate and determining, from said identification, an offset in translation and in angular rotation between a frame of reference of the calibration plate and a frame of reference associated with the head.

Technical Field

The invention relates to selective additive manufacturing.

More particularly, the invention relates to calibration of a focal point of a power radiation source of an additive manufacturing apparatus.

Background

Selective additive manufacturing involves creating a three-dimensional object by consolidating selected regions in successive layers of powder material (metal powder, ceramic powder). The consolidated region corresponds to a continuous cross-section of the three-dimensional object. Consolidation is performed layer-by-layer by selective melting, in whole or in part, using a focused radiation source, such as a light source (e.g., a high power laser) or a particle beam source (e.g., an electron beam in accordance with the technique commonly used in the art by the english term EBM or "electron beam melting").

In the following, reference will be made mainly to light sources (for example, light sources used in SLM or "selective laser melting" according to english terminology).

However, any other type of radiation may be used for implementation.

Generally, as shown in fig. 1, additive manufacturing devices of the type utilizing a light source use a three-axis head system S having three galvanometers (galvano [ tre ]) in order to provide, on the one hand, an increased accuracy with respect to the location of the impact point on the layer of powder material and, on the other hand, an increased accuracy with respect to focusing the light beam on said layer.

The two galvanometers 1, 2 are used to guide the rotation of the two mirrors 3, 4 and make it possible to check the path of the beam leaving the head and thus the position (controlled in X and Y) of the point of impact of the beam on the powder bed (work plane P).

Furthermore, DFMs (called "dynamic focusing modules" according to english terminology) comprise galvanometers, translation modules and lenses, the focus of the beam on the powder bed can be perfectly adjusted (controlled in Z) by adjusting the translation movement of said lenses. Fig. 1 also shows an objective module 6 introducing a fixed focus, and a DFM module 5 makes it possible to vary the focus upstream.

Indeed, it will be appreciated that the better the ability of the laser beam to focus on the powder bed, the more intense and controlled the energy delivered to the melting point.

It will be noted, however, that such a three-axis head system S implies two main deformation types of the powder bed (plane P).

First, the focus of the working plane P, in which the powder bed is located, varies according to the inclination of the light beam. In fact, as shown in fig. 2, the tilt of the beam is a result of variations in the optical path length (which variations are schematically represented by Δ Z in the figure), which themselves must be added to the phenomena caused by the gap caused by the mirrors or by traversing the various glass sheets that may be present in the additive manufacturing device.

Therefore, without any processing of this geometric shift, the focus is not constant over the entire working plane P.

Furthermore, the non-linear deformations in the X and Y directions in the figures are generally seen on the working plane P (according to commonly used technical terminology, such deformations are called "pincushion" -shape PS in fig. 3). This deformation is caused by the geometry of the optical path, in particular by the position of the mirrors and optics and their distance from the plate P.

It is conventionally known practice to correct the control of the head to take account of deformations in the X and Y directions.

For this purpose, a predetermined dedicated correction table is used.

These tables give the control difference by which the control input to the tri-axial head can be corrected. The beam is thus emitted and scanned in the working plane P with the corrected X and Y positions relative to the powder bed.

In order to determine these correction tables, it is known practice to use calibration plates with reference marks. A series of light emissions is performed at predetermined target locations on the plate. A measurement system comprising a camera records the position of these marks relative to the reference marks.

The difference between the position of the mark thus made on the calibration plate and the theoretical target position of the shot is used to calculate the amount of correction that needs to be applied to the control of the beam head system in the X and Y directions.

In this respect, an example is described, for example, in patent EP1048441 or even in patent applications US2015/0100149 and US 2014/0333931.

In EP1048441 in particular, the calibration uses on the one hand a plate with reference marks and on the other hand a sheet intended to receive the marks.

Thus, the plate is divided into two regions: one area receives the sheet sensitive to the light beam and the other area is not covered by the sheet and carries the reference mark.

Patent application CN 101823181 proposes a method of determining the focus of a laser beam by discriminating the form of the optical emission.

However, the proposed processing operation does of course not allow the focusing of the light beam on the powder bed to be optimized.

Furthermore, the corrections made in the X and Y directions are not optimal.

Of course, it will be readily appreciated that similar calibration problems may occur with other radiation sources (e.g., EBM sources).

Disclosure of Invention

It is an object of the invention to propose an automatic calibration solution that makes it possible to optimize the correction applied to the emission.

In particular, it is an object of the invention to propose a solution that allows the focusing of the calibration source (calibration in the Z direction).

Another object of the invention is to propose a calibration solution that allows better calibration in the X and Y directions than the prior art.

Thus, according to one aspect, the invention provides a method for calibrating a head system of a power radiation source of an additive manufacturing device, wherein, for determining a correction to be applied to a control of the system, the following steps are performed:

arranging a calibration plate comprising a plurality of reference marks in an additive manufacturing device,

controlling the source to mark at least one calibration pattern formed by a plurality of impact points on a calibration plate,

acquiring at least one image of the calibration pattern and at least one reference mark,

determining at least one corrective control from the one or more images thus obtained,

the step of determining a corrected control that depends on the distribution of the diameters of the impact points and becomes a focus control determines the distribution of the diameters of the impact points of the calibration pattern appearing in the image in the acquired image.

Advantageously, the method is completed by the following various features, which can be implemented alone or in any technically possible combination thereof:

in the step of controlling the source, generating various emissions corresponding to different impact points with different focus controls;

in the step of controlling the source, generating various emissions corresponding to different impact points with focus control that is incremented such that the diameter is gaussian around a midpoint of the pattern;

determining a corrected focus control determines a minimum diameter impact point or a derived "waist point" from the distribution of diameters of the impact points;

the control of the correction associated with the midpoint of the mark is the focus control for the impact point thus determined, or a derivation control corresponding to the derivation of the "waist point";

the calibration pattern formed by the impact points is a matrix, and the focus control values are increased column by column, row by row;

controlling the source with a head system to produce on the calibration plate markings of a plurality of calibration patterns theoretically centered on a preset theoretical target point;

in the acquisition step, the optical measuring device is moved over the calibration plate in order to acquire, for each reference mark of the calibration plate, at least one image of the area in which the reference mark is present on the one hand and in which the immediate theoretical target position is present on the other hand;

for each reference marker, processing one or more of the acquired images to derive therefrom the position of the impact point or of the derived "waist point" with the smallest diameter in the orthogonal reference system of said reference marker, and determining the offset between the position of said point and the theoretical target position, and determining an additional correction according to this offset;

to determine a correction, a process is performed to convert from a frame of reference of the calibration plate to a frame of reference attached to the head;

the process identifies patterns corresponding to two given reference marks on the calibration plate and determines from the identification a translational offset and an angular rotation between the frame of reference of the calibration plate and the frame of reference associated with the head.

Furthermore, according to another aspect, the invention provides a kit for calibrating a head system of a power radiation source of an additive manufacturing device, the kit comprising:

a calibration plate comprising a plurality of reference marks,

an emission support made of at least one material sensitive to the source radiation,

the support makes the reference marks of the calibration plate visible when the support is in place on the calibration plate,

characterized in that said emission support comprises a plurality of windows distributed in such a way that they overlap respective reference marks of the calibration plate and make them visible when the emission support is in position on the calibration plate.

Drawings

Other features and advantages of the present invention will become more apparent from the following description, which is given by way of illustration and not of limitation, and which is to be read in connection with the accompanying drawings, wherein:

FIG. 1 is a schematic view of the optical sequence of the three-axis head of the laser radiation source of the selective manufacturing device;

figures 2 and 3 show the main variants of such a device without correction;

figure 4a schematically shows an example of a calibration kit (calibration assembly) according to the invention, figure 4b itself showing in detail the elements of the optical kit of the calibration kit;

figure 5 is a schematic view showing the superposition of the calibration plate and the emission support;

fig. 6 schematically shows a support with biaxial guides on which the measuring device of the kit of fig. 4a is intended to be mounted.

Figure 7 shows an image in the measurement range of the camera of the set of figure 4 a;

figure 8a shows the impact point on the emitting support within the calibration pattern area;

FIG. 8b shows a Gaussian distribution of impact points and the determination of inferred "waist" points;

figure 9 shows the various steps in a possible embodiment of the method of the invention;

figure 10 shows an emission support showing the effect of various emission patterns;

figures 11 to 14 show a process for transforming from the frame of reference of the plate to the frame of reference of the head;

fig. 15 shows an image in the measurement field of the camera of the calibration kit according to the invention.

Detailed Description

Calibration kit

The calibration kit of fig. 4a, 4b and 5 is a kit comprising a reference calibration plate 7 and a launch support 8.

It also comprises an inspection device 9 involving an optical sensor.

The reference plate 7 is a plate with a plurality of visible reference marks 10. Preferably, the reference marks 10 are present on both faces of the calibration plate 7, so that the reference plate 7 can be turned over and both sides of the reference plate 7 are used.

These reference marks 10 are distributed over the entire plate at known and checked positions. For example, they may be located near the point of a 65x65 matrix intended to be fully or partially scanned by the power radiation source during calibration. Of course, other distributions are possible.

At each reference point, the visible marking of the marking 10 is defined by two orthogonal axes, the intersection of which corresponds to the reference point and defines an orthogonal frame of reference on the plate.

The emitting support 8 is optional. In the first embodiment, there is no emission support 8. Preferably, the reference plate 7 is a glass plate, the thickness of which may be between 0.1mm and 10mm, and preferably 1 mm.

In a second embodiment, there is an emission support 8 comprising a film of a material sensitive to the radiation beam.

The support 8 has the same dimensions as the plate 7 in the X and Y directions and is placed on the plate 7 during the calibration shot.

The support 8 has a plurality of windows 11 distributed in the same way as the reference 10 on the plate 7. When the emitting support 8 is placed on the plate 7, the centre of the window 11 is centred on the reference mark 10, so that the reference mark 10 is visible.

The device 9 comprises a camera 9a of the CMOS or CCD type, an optical objective 9b and a dedicated illumination system 9 c.

The dedicated illumination system 9c comprises for example a ring of light emitting diodes. The camera 9a and the objective 9b allow the image to have a resolution that allows measurements to be made with tolerances close to the micron scale.

This device 9 is for example incorporated in a selective printing device.

As a variant, it can be installed in the selective printing device only during the calibration phase.

As shown in fig. 6, the selective printing manufacturing apparatus comprises a support 12 for biaxial guides, which accommodates the device 9 and the camera 9 a. This system 12 allows the device 9 and the camera 9a to be guided on the calibration plate 7 and the support 8 (if present) in the X and Y directions, either manually or automatically.

The various images collected are processed by a processor 13 in order to derive therefrom the measurements described hereinafter. The processor 13 is, for example, a processor of a selective printing apparatus. It may also be a processor different from the selective printing device and interchangeable therewith.

Error measurement

During calibration, the camera 9a takes an image for each of the various reference markers 10, which shows the orthogonal marker corresponding to said marker 10 (fig. 7).

As shown in fig. 15, the mark corresponding to the mark 10 of the reference plate 7 may include a plurality of elements having various sizes and shapes to improve the accuracy of positioning, and the numbering of the reference plate 7 may be mentioned.

In the second embodiment, the marking of each reference mark 10 is visible thanks to a window 11 in the emitting support 8, which may be, for example, a rectangular opening, although of course other shapes are possible.

Such reference marks 10 define an orthogonal measurement reference frame on the calibration plate 7, which is used as a reference by the camera 9 b.

Prior to image capture, the source and its head system S have been controlled such that, for each reference marker 10, a given theoretical target position in the vicinity of the respective reference marker 10 is emitted.

In the first embodiment, the emission is directed towards the reference plate 7. In the second embodiment, the emission is directed towards the support 8.

Target position marks may be generated on the reference plate 7 near a given theoretical target position in order to improve the positioning of the emission position and thus the accuracy of the calibration. As shown in fig. 15, the target position marker can be centered at a given theoretical target position. It may constitute a square, which may be surrounded by a cross formed by horizontal and vertical lines, locating a given theoretical target emission position.

The emission is performed successively along several emission points distributed according to a predetermined emission pattern (in the example of fig. 7, a matrix pattern 15 of points of size 7 × 7).

In fig. 7, the theoretical target position (not visible) corresponds to the cross 14.

At the end of the emission of the pattern, an image is taken around each reference mark 10.

For each of these markers 10, the images taken of the area surrounding the marker 10 are processed by a processor 13 in order to derive therefrom the position of the center (midpoint) of the emission pattern 15 in the orthogonal frame of reference of said marker 10 (position measurements Xm and Ym), and hence the offset of this position with respect to the theoretical target position 14.

Furthermore, as shown in fig. 8a, the emission at the various points of the pattern 15 is controlled in such a way that different diameters are produced at the points of impact of the pattern.

For example, several of these emissions are generated using different controls in the Z direction, which means that different focus controls can be used. Typically, the Z control of these various emissions is increased from one point to another in order to obtain a gaussian distribution around the pattern midpoint theoretically.

In the example of fig. 8a, the gaussian distribution is a distribution with an axis of symmetry Y.

As a variant, the distribution may be gaussian in both the X and Y directions (distribution according to a two-dimensional gaussian distribution).

The following table gives an example of the incremental distribution. The emission pattern takes the form of a matrix, the control values in the Z direction increasing column by column (left to right in this case) and row by row (bottom to top).

…

…

-4Δ

+3Δ

…

…

…

…

…

-5Δ

+2Δ

+4Δ

…

…

…

…

-6Δ

+Δ

+3Δ

…

…

…

…

-7Δ

0

+Δ

…

…

…

…

-8Δ

-Δ

-3Δ

…

…

…

…

-9Δ

-2Δ

+5Δ

…

…

…

…

…

-3Δ

+4Δ

…

…

The center point of increment 0 corresponds to the middle point of the pattern, while Δ corresponds to a single increment value controlled in the Z direction. In the case of a 7x7 transmit matrix (given by way of example only), 49Z control values may be tested.

Once the pattern has been generated, the distribution of the diameters of the impact points is analyzed by the processing performed by the processor 13, and the point in the pattern region corresponding to the "waist" is determined based on the distribution.

This "waist" point corresponds to the point of impact of the smallest diameter or to a point of inference derived from the diameter profile of the point of impact.

In the example shown in fig. 8a, various impact points I1 to I4 have been depicted, as well as a gaussian best fit curve G passing through these points.

The inferred point is a point E corresponding to the minimum value of the gaussian G.

Then, the Z control for the impact point thus selected (or the inferred control corresponding to the inferred point thus determined) is used as the control of the theoretical target position of the marker (in the form of correction Δ Zm).

In addition, possible corrections can also be made taking into account the change in the optical path length between this theoretical target point and the impact point of the smallest diameter or determined to correspond to the "waist".

This correction is determined, for example, by reference to a chart which gives this additional focus correction on the basis of a measurement of the distance between the theoretical target point and the impact point of the smallest diameter or determined to correspond to the "waist".

It will be appreciated that this process is particularly facilitated if the impact points as a whole describe a calibration pattern in the form of a matrix. Many other configurations of calibration patterns are of course possible.

The measured values Xm and Ym are then processed in order to derive therefrom the correction values Δ X and Δ Y (in bits) required at the head control system S in order to center the pattern 15 on the target position 14 at the next calibration shot.

In the same manner, a correction Δ Z to be applied to the DFM module to correct the focus is determined from Xm, Ym, and Δ Zm, and loaded to the module in units of bits.

As an order of magnitude concept, the tri-axial galvanometer is typically controlled above 24 bits (this is 2 per axis) ²⁴Leading above the bit) and the step size increments of X, Y on the correction raster are distributed over 4225 values (262144 bit step size). The control Z is also controlled to be more than 24 bits, and the size of the correction table is the same as that of X and Y.

It should also be noted that a pattern in the form of a matrix of the proposed type is particularly advantageous and allows for precise calibration in terms of the position of the emission point and the focusing of the radiation beam on the powder bed. However, other patterns are also possible (patterns in the form of a matrix distributed in a staggered configuration, circular patterns, elliptical patterns, etc.).

Calibration procedure

In a first step (step 21 in fig. 9), the pre-correction tables on the three axes are loaded into the processor 13.

These tables were obtained in advance using theoretical models. This makes it possible to eliminate errors associated with the optical sequence to a large extent. In this way, the calibration emission can be made in a very narrow region compatible with the field of view of the camera, and with sufficient focus to etch.

In a second step (step 22), the plate 7 is mounted in a device for manufacturing objects by selective printing. The calibration plate 7 is dimensioned such that it can be easily and directly mounted on the plate carrying system of the device.

In the second embodiment, the emitting support 8 is located on the calibration plate 7.

In a third step (step 23), the beam source and the head are controlled to emit a series of patterns at different theoretical target points 14. In the first embodiment, a series of emissions is directed at the reference plate 7. In the second embodiment, a series of shots is directed at the support 8.

The theoretical target points are selected to correspond to points similar to all or some of the reference marks on the calibration plate.

By using a matrix of theoretical points corresponding to the matrix of points used for the correction table, the effect of the correction can be limited.

However, the number of reference points used may be lower than the number of points to be determined in the correction table. In this case, the missing point can be deduced from the result relating to the target point.

In a second embodiment, the emission of the pattern is aimed at 7.5 mm on the Y-plate to emit onto the material of the support 8 instead of the window 11, so that the impact can be seen on the support 8 and can be measured with the camera 9b (fig. 10).

In a fourth step (step 24 in fig. 9), the measuring device 9 is introduced into the selective printing device in order to perform an optical measurement allowing correction.

The operator moves the device 9, using the two-axis support 12, by hand or with an electric device, to take images of the set of various emission points in succession. In the first embodiment, the emission point is located on the reference plate 7. In the second embodiment, the emission point is located on the support 8. It should be noted that the calibration plate 7 may be provided with a numbering system near each reference mark 10, so that scanning may be performed without imposing a measurement sequence on the reference marks.

The images thus captured may be processed (step 25) by the processor 13 in real time or stored for later processing.

The measurement and processing were performed as follows.

Analysis of the image on a given reference mark can determine the coordinates Xm and Ym of the center (middle) point of the emission pattern.

This also allows the correction value Δ Zm to be determined.

The head cannot be mechanically positioned within one micron relative to the calibration plate 7 and so the processor 13 performs a processing operation that changes the frame of reference to convert the measurements to the frame of reference of the head.

Fig. 11 shows in an exaggerated manner the possible translational and rotational offsets between the two frames of reference.

To evaluate the translational shift amount, the center (midpoint) of one of the patterns is used as a reference point (point a in fig. 12).

This point is preferably selected to be located below the source head (coordinates (0,0) in bits of the galvanometer controlling the head in the X and Y directions).

Image processing corresponding to reference marker 10 allows the processor to determine the positions Xm0 and Ym0 of the corresponding points in the frame of reference of the plate and derive therefrom a translation correction to be applied to the measurement in order to convert it into the frame of reference of the head.

To evaluate the amount of rotational offset, the processor 13 uses a pattern corresponding to point A and a second emission point (point B in FIG. 12) that is theoretically located on the same line in the X direction as point A.

Since the mirror control Y is zero in both positions, the vector connecting the two points is transverse to the head frame of reference.

The location of point B in the frame of reference of the plate 7 allows the processor to determine the angle θ 0 (fig. 13) between the frame of reference of the plate and the frame of reference of the head system S.

Using these parameters, the processor 13 performs the desired frame of reference conversion on all recorded images.

Thus, for each measurement point with coordinates Xm and Ym, it determines the corresponding coordinates Xmt and Ymt in the frame of reference of the head system.

These measurements are then processed to obtain correction values in bits to control the movement of the two mirrors (X and Y directions).

This correction of X and Y itself causes a shift of the impact point on the plate 7 and therefore a change of focus (this change corresponds to Zd in fig. 14). The processor 13 estimates this change Zd using the laws of optics and determines the corresponding value of the galvanometer bit (Δ Zd).

For each pattern 15 (i.e. each calibration point), for the relevant pattern, the correction is added to the correction control value for Z determined by the processor 13 to derive therefrom the correct value for the calibration in the Z direction.

When the correction has been established, the processor 13 stores the new correction table in the memory.

The device is then ready for a check transmission (step 26).

In the first embodiment, once the correction table is obtained, the calibration plate 7 is turned over and the operator starts a new inspection emission sequence. The position of the impact (shot pattern) on the calibration plate 7 is recorded and a new correction is determined by the processor 13.

In the second embodiment, once the correction table is thus obtained, the launch support 8 is turned over and the operator starts to check a new sequence of launches. The position of the impact (shot pattern) on the firing support 8 is recorded and a further correction is determined by the processor 13.

If appropriate, test reports may be issued.

If the inspection test indicates that the position and focus accuracy is insufficient relative to the expected accuracy, a new correction table may be calculated and then a new inspection transmission performed.

Thus, the process is iteratively repeated until the calibration is deemed sufficient to meet the expected positioning tolerances.