JP2022533623A - METAL POWDER ANALYZING METHOD AND APPARATUS - Google Patents

Abstract

A method and apparatus for analyzing metal powders used in additive manufacturing processes includes irradiating a region of the powder with electromagnetic radiation, including radiation in the invisible portion of the electromagnetic spectrum, without melting the powder; separately detecting illuminating radiation, including radiation in the invisible portion of the electromagnetic spectrum returned from separate portions of the powder, to produce an output based on the detected radiation; and processing the output to produce one of the powders or and identifying a plurality of characteristics. The powder is illuminated with light containing UV and/or IR and at least the returned UV and/or IR is detected. [Selection drawing] Fig. 1

Description

The present invention relates to methods and apparatus for analyzing metal powders, particularly, but not exclusively, to methods and apparatus for analyzing metal powders used in additive manufacturing (AM) processes.

In known additive manufacturing processes, articles are manufactured from powdered metals or alloys by additive manufacturing machines. In an additive manufacturing machine, layers of powder are deposited on a build platform and the powder is selectively melted by a laser or electron beam to form one or more articles. By repeating this process, the article is formed layer by layer.

Once the build is complete, the unmelted powder can be reused for another build.

The composition and condition of the metal powders used in the molding process can greatly affect the integrity of the article formed by the process.

For example, unmelted powder can degrade during the shaping process. Metal powders can gradually oxidize and change their properties, resulting in changes in the properties of articles made from the powders. Generally, the tendency of powders to oxidize increases with temperature, and exposure to high temperatures can also affect other powder properties. As a result, the closer the unmelted powder is to the article being built or the heat zone, the more likely it is to degrade.

Also, as the powder melts, this process can also cause particles of the heated powder to fly around the article produced from the powder bed, degrading the unmelted powder around the article. .

In order to ensure proper quality of the build, mixed powders can be analyzed by analyzing the used powders and stopping the reuse of powders that have degraded to a certain extent and/or mixing the virgin powders with the recycled powders. are known to have suitable bulk properties for continued use. Another approach places a fixed upper limit on the number of times a batch of powder can be reused.

There are many problems with these approaches.

Powder condition is usually analyzed by making bulk oxygen content measurements. The measurement process involves testing non-reusable powder samples. More importantly, the bulk oxygen content ( (or other bulk) measurements may give erroneous results regarding the suitability of powders for reuse. This is because even though the bulk oxygen content is below the desired threshold, it is not sensitive to the presence of highly oxidized or otherwise degraded particles, which can have a significant negative impact on fabrication. be.

Putting a rough limit on the number of times the powder is reused is a relatively crude approach and does not take into account the amount of powder degradation that can occur with a particular build. The nature and extent of degradation can vary significantly between designs.

Since existing testing methods for powders are destructive, the results are not related to the actual recycled powder and are usually an approximation of the average state of the powder.

Other aspects of powder composition and condition can also affect build quality.

The presence of contaminant particles can have a similar impact as the presence of highly oxidized particles. The size and shape of the particles can also affect the build, as can the compactness with which layers of powder are formed in an additive manufacturing machine.

It is an aim of embodiments of the present invention to address some or all of these issues. In particular, it is an object of embodiments of the present invention to provide improved methods and apparatus for analyzing powder conditions in a non-destructive manner.

SUMMARY OF THE INVENTION According to a first aspect of the present invention, a method is provided for analyzing a metal powder for use in an additive manufacturing process, the method comprising exposing a region of the powder to a non-visible region of the electromagnetic spectrum without melting the powder. irradiating with electromagnetic radiation comprising radiation in portions; separately detecting the irradiating radiation comprising radiation in the invisible portion of the electromagnetic spectrum returned from different portions of the irradiated region of the powder; generating an output; and processing the output to identify one or more properties of the powder.

According to a second aspect of the present invention, there is provided an apparatus for analyzing a metal powder for use in an additive manufacturing process, the apparatus quantifying a region of the powder in the invisible region of the electromagnetic spectrum without melting the powder. and separately detecting the irradiating radiation comprising radiation in the non-visible portion of the electromagnetic spectrum returned from different portions of the irradiated area of the powder and based on the detected radiation. and a processor that processes the output to determine one or more properties of the powder.

This method can be performed in any suitable environment. The powder can be a test powder sample placed in a test container. Alternatively, the analysis can be performed on an additive manufacturing machine or a powder conveying device such as a pipe or conduit. As such, the device may be included in an inspection device, a powder conveying device, or an additive manufacturing device.

Radiation is light, including some invisible light such as ultraviolet and/or infrared light, and may also include visible light. The illumination device may comprise any suitable illumination device, such as lamps containing LEDs, incandescent lamps and/or discharge lamps. It is also possible to irradiate the powder (in low power mode) using a laser such as a laser former that operates to melt the powder in an additive manufacturing machine in the build process. The irradiation means can simultaneously irradiate all areas of the powder or separately irradiate parts of the areas. The irradiation means can scan the beam of radiation over the area of the powder.

Where one or more illuminators are provided, they may be provided within a housing into which the metal powder is introduced or may irradiate the powder sample through a barrier and receive returning illuminating radiation. It is preferably provided within an opening configured to allow for The enclosure can prevent or at least limit penetration of electromagnetic radiation at wavelengths of interest to allow controlled irradiation of the metal powder at wavelengths of interest. The enclosure may be substantially light-tight or configured to be substantially light-tight. The housing may be included in an additive manufacturing machine, or powder conveying, manufacturing, processing, or processing equipment.

A region of the powder can be a surface of the powder, in particular a plane. The method involves placing the powder in a suitable container, build platform, or powder recess of an additive manufacturing machine so that the powder has a substantially planar upper surface and the surface is illuminated. Alternatively, the method may involve illuminating the powder through a suitable transparent barrier forming part of a powder containing structure such as a container for manufacturing, processing or transporting the powder. Such barriers may be planar, but may also be curved, such as the wall or window of a pipe having a generally circular cross-section.

The powder can be compact. The degree to which the powder is compacted can be controlled, such as by tapping the powder in a predetermined manner or number of times prior to making the measurement, and/or by passing the powder through a pretreatment funnel prior to the measurement. The powder preferably has sufficient depth from the illuminated surface so that the powder is substantially opaque to the detector.

The apparatus may include a barrier, such as a window, transparent or at least partially transparent to the wavelength of the radiation of interest to separate the powder under investigation from the illuminator and detector. The device is arranged such that the powder is supported by or contained by the barrier such that the powder in compact form is observable through the barrier during use. Thus, the illuminator is positioned to illuminate the window and the detector receives radiation from the window, from a position below the window, or alongside the window if the window is substantially vertical. are arranged as follows.

If the barrier is horizontal or substantially horizontal, the powder will be supported on the barrier in a compact configuration. If the barrier is horizontal or not substantially horizontal, for example a side wall or window of a container or pipe, then the arrangement will be such that the powder is received in a stack against the window. Here, in the case of a pipe or conduit, it may include a valve or flow restrictor that acts to build up powder against the window in the downstream direction away from the window.

A powder is a sample of powder taken from a batch of powder. The powder may be virgin powder or recycled powder that has already been used in the additive manufacturing process. Powders can be analyzed at various stages of the process and can be analyzed at multiple points in the process. Devices for analyzing powders are therefore placed at various points in the additive manufacturing machine and in the powder conveying and handling device. For example, the powder can be analyzed at one or more or any combination of the following process steps. That is, during transport to the additive manufacturing machine, in the powder container of the additive manufacturing machine, on the platform of the additive manufacturing machine, in the powder bed in the additive manufacturing machine, in the molding container in the additive manufacturing machine, and being transported out of the additive manufacturing machine It can be analyzed on-the-fly, after a sieving step, and after mixing with another powder, and so on.

Detectors can take a variety of forms and include any suitable sensor. The detector can be arranged to detect radiation on the same side of the area of the powder as the means for irradiating the powder, in particular on the same side of the irradiated surface. The method includes detecting, the detector positioned to detect radiation reflected and/or scattered by the powder. The detectors can be arranged to detect the illuminating radiation returned from all areas of the powder simultaneously or separately from some of the areas. A detector can be arranged to scan an area of the powder. Relative movement between the detector and the powder can be achieved by moving the detector relative to the stationary powder or by moving the powder past the detector as in the case of a powder conveyer. can be run by

A detector may include a one- or two-dimensional array of sensing elements such as CCD or CMOS sensors. The detector may include one or more or a series of photodiodes, spectrometers or spectrophotometers. The detector may include one or more filters to exclude and/or allow selected wavelengths of electromagnetic radiation.

The detector may include one or more focusing elements, such as lenses, for focusing radiation returned from the powder onto one or more sensors. The one or more focusing elements can focus an image of at least a portion of the area of powder to an image plane in which the one or more sensors are located. The detector can include an image capture device such as a camera or microscope. A detector may include a hyperspectral camera.

These arrangements allow the detector to spatially resolve the source, and/or detect light reflected from part of the area at once, and/or only part of the area. Being irradiated at once, the device can separately detect the irradiation radiation returned from different parts of the irradiated area of the powder.

By processing the output, various different properties of the following powders can be identified. i.e. the state of the individual particles of the powder, the surface properties of the powder and/or individual particles of the powder, the interaction with electromagnetic radiation, in particular the ability of the powder or powder to reflect and/or scatter radiation of specific wavelengths. The properties of individual particles of the powder, the shape of the individual particles of the powder, the surface texture of the powder or individual particles of the powder, the presence or absence of contamination, the compactness of the powder, and the spreadability of the powder can be determined.

The output includes values corresponding to the wavelength of radiation received or detected from each zone of the powder field. The output also includes values corresponding to the intensity of radiation received from each zone of the powder field. The output may include values depending on the intensity of radiation at different wavelengths. The output may include values depending on the intensity of the radiation over a continuous range of wavelengths. and/or the value may depend on the intensity of radiation at one or more discrete wavelengths. The output can be a function of detected wavelength. The output contains values for each of the multiple zones. The areas are substantially the same size. The areas may be contiguous or spatially separated. Together these areas cover all or most of the area. Each zone preferably has an area smaller than the area of the surface of the powder occupied by a single particle of powder of average size, which is 1/4, 1/10, 1/100, 500 of that area. It is preferably very small, such as 1/1000 or 1/1000. As such, each zone can be less than 1000 μm ² or less than 100 μm ² for typical particle sizes in metal powders for additive manufacturing. Thus, at least one, and preferably a plurality of values contained in the output are influenced by a single particle of powder. Of course, the optimum size of the zones will vary with powder type, since the average particle size will vary with powder type. In this way it is possible to identify the state of individual particles.

A data set is formed by the values obtained by analyzing the areas of the powder.

The steps of detecting radiation, generating output, and processing output can be performed at different locations.

The data set substantially includes an image of all or part of the irradiated area of the powder, but at least partially obtained by recording radiation of invisible wavelengths. So in practice it is not possible to make all the data recognizable to the human eye without applying artificial coloration to make it possible to recognize features that are only revealed when examining invisible wavelengths. . For convenience, however, when images are referred to herein, the images include, at least in part, those obtained by detecting invisible electromagnetic radiation, particularly invisible light.

If the dataset substantially contains images, the images may be digital images. A digital image may be formed by or divided into multiple elements such as pixels. Each element corresponds to an area of the powder from which the detector can separately receive radiation and produce an output. The elements are preferably of substantially the same size. Therefore, the ratio of elements in the image to the number of particles in the imaged area or surface of the powder is preferably on the order of at least 1:1, at least 4:1, 10:1, 100:1, 500: 1, or higher such as 1000:1 is even more preferred. The wavelength of the detected radiation represented by the element would then be affected only by the properties of the single particles of the powder.

The data set for processing typically contains values for each of the 2 to 6 million areas, representing about 5000 particles, typically in the size range of 10 to 110 μm or 40 to 50 μm. However, the size of the data set varies greatly depending on the application.

A data set if the image is formed in visible colors, or if a standard capable of representing invisible electromagnetic radiation of interest is used, or if the detected invisible wavelengths are represented in the data set by artificial coloring contains data representing each area or image element of the powder according to an established color standard, eg RGB or CIELAB.

Multiple regions of the same volume or powder sample can be analyzed. The regions may be contiguous or separate. In embodiments, at least 2, 3, 4, 5, 10, 50, 100 or more regions of the same powder are analyzed to form multiple data sets of the powder. The number of datasets depends on the amount of data required for statistical significance. Analyzing multiple areas helps assess how well the powder is mixed.

The data in the set are processed by comparing the data values of one area of the powder region with the data values of one or more other areas and/or reference data.

Deleting data in the dataset allows the remaining data to represent selected sub-regions of the powder, similar to cropping an image. This is useful when the detector contains a focusing element that produces an image on the detector by allowing distortion to be eliminated. It also allows easier and more reliable comparisons between different datasets, especially those relating to the same sample. As an example, by reducing the size of the data set, 2000×2000 contiguous areas of powder can be defined, defining an image composed of 2000×2000 image elements.

Identifying whether the processing of a dataset is of sufficient quality for further processing and/or whether all independently acquired data in a particular set are sufficiently similar can be specified. Data that do not meet predetermined quality criteria are filtered out and not processed further. This may involve, for example, obtaining statistics from the data and determining whether those statistics are within a given range, or whether the data sets differ by more than a given threshold. include. In one example, the average intensity of a particular detected wavelength, wavelength range, or color channel used to define the detected radiation is determined for all areas of the powder where the radiation is detected, or the threshold brightness For all areas greater than , the deviation from the mean is calculated as well.

Data in the data set can be identified that relate to spaces between particles of the powder. This allows the data to be excluded from subsequent processing. Such data can be identified by identifying data defining areas of the powder region where the light intensity is below a predetermined threshold. The spaces between the particles of the powder tend to appear darker than the particles and therefore are less luminous. Such areas of powder can be considered background areas and the remaining lighter areas are foreground areas. If the dataset contains images, this step includes identifying darker image elements. Darker image elements are considered background elements and lighter elements are considered foreground elements.

By processing the data set, it is also possible to identify, or at least estimate, the number of particles of the powder it represents. If the dataset contains images, this can be done by watershed segmentation. This step is preferably performed after removing background elements from the image, if any.

Alternatively, or in addition, the number of particles represented by the data set may be determined by another suitable method independent of the data set, e.g., the size of the area represented by the data set, the average size of the powder particles, and/or the can be estimated by information such as the density of

The data representing the foreground area of the powder represent properties of the particles of the powder that affect how the particles interact with the applied radiation, including surface properties.

The number of particles of the powder represented by the data set that have surface properties that fall outside a predetermined range can be identified. The percentage of powder with measured surface properties that fall outside the selected range can be identified.

This can be done by identifying areas, preferably foreground areas, defined by the data where the wavelength of the detected radiation is outside a selected range.

Zone selection is based on statistical analysis of the detected wavelengths. The selected area is the area that defines the detected wavelengths as outliers of the wavelength distribution over all foreground areas. For example, elements can be selected by specifying how much their wavelength deviates from the average wavelength of all areas. The selected area preferably represents no more than 5%, no more than 1%, or no more than 0.1% outside the wavelength distribution of the foreground area.

Identifying data reflecting a property within the range of interest allows identification of individual particles with that property. To that end, groups of connected areas exceeding a predetermined number are identified. The number is chosen to be a number representing the total area expected to be occupied by a single particle of powder. A group of identified areas can therefore be considered to represent at least one, but usually one, particle having properties that meet the selected criteria.

If the area to particle ratio is close to 1:1, this step can be skipped and a single image element is considered to represent a single particle.

Data can then be stored for each identified particle, which is data relating to one or more areas of the powder that the particle occupies. The number of zones within an identified group indicates the size of the particle that group represents. The average wavelength of radiation detected from areas within a group is indicative of the color or other characteristic of the particles represented by that group. This allows the number of particles identified, along with the properties of the particles, to be determined, and by analyzing this data, various information related to the powder can be determined or inferred. Particles identified in this manner are further sorted by surface characteristics, such as color, to identify particles having surface characteristics falling within a particular range. In doing so, other techniques may be used to select particles of interest from the identified particles.

The ratio of background area to foreground area provides an indication of the ratio between the particles of the powder and the space between them, and thus the density of the particles. Changes in compactness observed in images of batches of powder taken over time can reveal changes that affect the flow properties of the powder.

Since oxidation of metal powders usually affects their interaction with radiation, it is possible to determine the chemical properties of the powder, especially the The degree of oxidation of is indicated. Information about how oxidation affects the color of a particular type of powder can be used to determine the bulk oxygen content of the powder.

The number of particles identified as returning wavelengths of radiation outside the selected range can reveal whether the powder contains highly oxidized particles or is contaminated so that the powder is It may be desirable not to be used or reused. That number, along with the calculated or estimated total number of particles imaged, can be used to calculate the percentage of particles with measured properties that fall outside the selected range.

This information can be used to inform or control subsequent processing of the analyzed powder or the powder from which the analyzed powder was extracted.

The method may include indicating that the powder is not suitable for reuse if the number or percentage of particles identified as having properties outside a predetermined range exceeds a predetermined range. The given ranges and percentages are determined according to the particular powder being analyzed and its intended use. However, it is usually intended to identify the presence of significantly oxidized, degraded or contaminated particles, so the range is caused by exposure to oxygen or cold temperatures from being used in the build process. It is preferably set to include values for properties that indicate powders that are considered to be in normal deterioration. Particles whose measured surface properties fall outside this range are therefore outliers. Their measured surface properties reflect the fact that they have been abnormally degraded, usually as a result of exposure to high temperatures, but without melting to form part of the construct. is. If the population of such outliers exceeds a certain percentage of the total population of particles, it can be considered that recycling the powder will increase the risk.

The method may also include identifying an average property of the percentage of powders for which the measured property is within a predetermined range. This measurement gives an overall average degradation of the powder, excluding outlier particles that are significantly degraded. This measurement is therefore an indication of the level of degradation due to normal degradation.

The method also detects if the measured average property of a percentage of the powder, i.e. the approximate percentage of particles whose measured surface property is within a given range, is greater than (or less than) a given threshold. indicating that the removed powder is not suitable for reuse. It can therefore be shown that the powder is no longer suitable for reuse, possibly as a result of cumulative normal degradation effects.

The method may also include identifying average properties of all powders, and thus all measured particles, by processing data for all foreground areas. Such measurements show the overall average degradation of all particles, while bulk oxygen measurements affect the surface properties of the particles by building up not only the oxygen in the outer oxide layer, but also the " It is possible to obtain similar indices as bulk oxygen measurements, except that "internal" oxygen, ie oxygen present inside, can also be measured.

The method may also include indicating that the powder is not suitable for reuse if the average property of all measured particles is greater than or less than a predetermined threshold. The method may include controlling equipment based on the measured powder properties, such as controlling an additive manufacturing machine or powder processing, handling or conveying equipment.

If a measured average property of a powder, i.e. a particle of interest, is desired, this is the average and is obtained by specifying the average measured surface property represented by the area of the powder region in which the particles of interest reside.

Information about powders analyzed by this method can be used to automatically control equipment including additive manufacturing equipment and/or powder handling, conveying and processing equipment. The device can, for example, cause the powder to be discarded and/or processed depending on the detected state of the powder.

The processor is a programmed computer and may be configured to cause the apparatus to perform some or all of the method steps described above.

In all aspects of the invention, the particles of the powder that return radiation containing wavelengths within a particular range can be identified and used to determine the particle size in view of the experimental data associated with the powder type concerned. can infer the degree of oxidation of

Embodiments of aspects of the present invention provide non-destructive methods and apparatus for characterizing powder conditions and determining whether a powder sample is suitable for reuse in a particular fabrication process. If identified by looking at the proportion of outlier particles, it provides a novel and useful measurement of powder condition that allows for improved decision-making and thus use of powders over current measurements of bulk powder properties. be done.

The method and apparatus are also useful for identifying the presence of contaminant particles that have surface properties that differ from similar properties of the particles of interest.

According to a third aspect of the invention, there is provided a method for analyzing metal powder used in an additive manufacturing process, the method comprising the steps of: providing a dense powder proximate to a barrier; irradiating a region of the powder; separately detecting illuminating radiation returned from different portions of the irradiated region of the powder through the barrier to generate an output in accordance with the detected radiation; and determining one or more properties of the powder.

According to a fourth aspect of the invention, there is provided an apparatus for analyzing metal powders used in an additive manufacturing process, the apparatus having a barrier for densely containing the powder to be analyzed in use. A container or conduit for the metal powder, an irradiation device for irradiating an area of the powder with electromagnetic radiation through a barrier, and separately detecting the irradiating radiation returned through the barrier from different parts of the irradiated area of the powder. , a detector for generating an output in accordance with the detected radiation, and a processor configured to identify one or more properties of the powder by processing the output.

The barrier can be the wall of a container or conduit. A barrier is a window in the wall of the container or conduit, the window being at least partially transparent to the electromagnetic radiation of interest. Electromagnetic radiation can be light, including infrared, visible, and ultraviolet. The third and fourth aspects of the present invention desirably or optionally follow the first aspect of the present invention, except that it is not necessary that the powder be irradiated with invisible electromagnetic radiation or that invisible radiation be detected. and any or all features of the second aspect.

For a clearer understanding of the invention, embodiments thereof will be described, by way of example only, with reference to the accompanying drawings.
1 is a schematic diagram showing an embodiment of an apparatus for analyzing the state of powder; FIG. 1 is a schematic diagram showing an embodiment of an apparatus for analyzing the state of powder; FIG. 1 is a schematic diagram showing an embodiment of an apparatus for analyzing the state of powder; FIG. 1 is a schematic diagram illustrating an embodiment of a powder processing and conveying device including a device for analyzing powder condition; FIG. 1 is a schematic diagram illustrating an embodiment of a powder processing and conveying device including a device for analyzing powder condition; FIG. 1 is a schematic diagram illustrating an embodiment of a powder processing and conveying device including a device for analyzing powder condition; FIG. 1 is a schematic side view showing an additive manufacturing machine including various devices for analyzing powder conditions; FIG. Figure 8 is a schematic plan view of the apparatus of Figure 7; 4 is a flow chart showing the steps associated with processing an image of powder; Fig. 3 is a graph showing the number of particles versus wavelength;

Terms such as top, bottom, top, bottom, etc., are hereinafter used to describe the device illustrated in the orientation shown in the figures, which orientation is intended for use. However, it is not limited to that direction. Similar reference numbers are used to denote similar components in all figures. Each drawing is not to scale.

Referring to the drawings, Figure 1 shows a first apparatus for analyzing metal powders. The device comprises a substantially light-tight enclosure 1 that can be opened and closed. The housing contains a container 2 for powder 3 which can take the form of a dish or slide or any other suitable form. The container is open at the top and has a substantially square opening with sides of about 10 mm and a cross-sectional area of about 100 mm ² . Also, its depth is at least 2 mm. Although the vessel shown is shallow, it could be deeper so that the vessel is longer. The container is removable from the housing. The enclosure also contains a lens (or microscope) 4 attached to (or included in) a digital camera 5 that is sensitive to UV and IR in addition to visible light, and a lamp 6 that emits UV, IR and visible light. do. Camera 4 includes a substantially square sensor, such as a CCD sensor having approximately 6 megapixels (12 megapixels in another embodiment), includes a keyboard and mouse or other user interface, display 8 and/or other is connected to a computer 7 which is connected to an output device of the The lamps are arranged to provide diffused light. These lamps are indicated as dome or flat dome lamps. In another configuration (and other embodiments), a ring light can be used.

In use, a sample of powder 3 taken from a batch of powder to be analyzed is introduced into container 2 either inside or outside housing 1 . The powder is introduced in an amount sufficient to obtain a dense powder depth that completely covers the bottom of the container 2 when viewed from above. Thus, the powder depth usually comprises at least two, preferably two or more layers of particles. The powder is flattened within the container, such as by tapping the container, so that it has a substantially flat top surface. If the powder was introduced into the container outside the housing, the container is placed under the microscope in the housing and the housing is closed.

The lamp 6 is then activated. The lamp can be controlled by computer 7 . The lamp is arranged to illuminate the upper surface of powder 3 in container 2 . By illuminating the powder with a lamp in a substantially lighttight enclosure, the powder can be analyzed under controllable and reproducible light conditions. It should be noted that in this and other described embodiments, the lamps may be provided in different arrangements and placed in front of the camera and/or lens.

Camera 5 then takes a digital image of the illuminated powder surface in the container and sends it to computer 7 . A digital image contains information relating to all wavelengths of light detected by the camera. The camera and microscope are positioned to take an image of approximately 12 mm ² of the surface of the powder within the container. Multiple images of the surface of the powder are taken until substantially all of the surface of the powder is imaged. In another embodiment, the camera and lens have different fields of view. Therefore, the surface of the powder can be captured in a single image.

Metal powders used in additive manufacturing processes are typically on the order of tens of microns in average diameter. And the number of particles visible on the surface of the powder imaged by the camera is on the order of thousands, thus about three orders of magnitude less than the number of pixels of the sensor. Thus, the camera can produce a digital image of the powder surface in which there are approximately 1000 times as many pixels as there are particles of the powder shown in the image.

The images taken by the camera are then sent to the computer 7 and processed.

FIG. 2 shows a second device for analyzing metal powders. In this case, a light-tight housing 1 is provided below a container 2 for containing powder 3 . The housing upwardly houses a camera 5 and a lens 4 and a lamp 6 similar to those of the embodiment of FIG. The base of the container is transparent or at least partially transparent to the wavelengths of light produced by the lamp and sensitive to the camera. The camera can thus take an image of the surface of the powder 3 in the container 2 through the base of the container. Since the top surface of the base of the container in which the powder is supported is substantially planar, the camera is directed at the flat surface of the powder to be imaged. The bottom surface of the base of the container is substantially parallel to its top surface.

FIG. 3 shows another device for analyzing powder 3 . The device comprises an elongated tube 2 closed at one end for containing the powder. Tube 2 has one flat side wall 2a. For convenience, the tube has a square or rectangular cross-section, but other cross-sections are possible, such as a D-shaped cross-section. A camera 5 and lens 4 assembly is provided in a substantially lighttight enclosure 1 with a lamp 6, similar to the apparatus of FIGS. The camera is aimed at the flat sidewall of the container. The sidewalls, which are at least partially transparent to the wavelengths of light produced by the lamp and which are detectable by the camera, are substantially parallel and opposed to each other, similar to the base of container 2 of FIG. It has flat sides. A camera is positioned to image the compacted powder contained in the container through the side wall. The container is mounted for movement relative to the housing and the camera so that the camera can image different portions of the surface of the powder sample contained in the container.

FIG. 4 shows a powder conveying device comprising a pipe 10 leading into and out of the powder screen and further into the powder mixing device. Each pipe 10 includes a flat or substantially flat transparent window 11 on which is mounted a suitable lens 4 for taking an image of the powder in the pipe 10 through the window 11. A light-shielding housing 1 for housing a digital camera 5 is attached. A lamp 6 is provided around the lens 4 within the housing to irradiate the powder through the window 11 . The camera outputs its image containing information relating to all wavelengths of light detected by the camera to a computer 7 with an output device 8 . Alternatively, or in addition, the output may be sent to a computer or processor that controls the powder conveying device or its associated equipment. A pipe 10 downstream of the mixing device is provided with a valve 12 (such as a butterfly valve) to allow and stop the flow of powder through the pipe. This allows the pipe to be filled with powder and directs the densely packed powder to the window 11 of the pipe which is transparent to the wavelengths of light produced by the lamp that can be sensed by the camera.

Similar to the apparatus shown in FIG. 1, the digital camera 5 estimates the expected number of powder particles visible in the area of the window 11 imaged by the camera when the pipe is filled with compact powder to be analyzed, approximately 1000. It has a sensor with twice the number of pixels. Similarly, lamp 6 emits light and camera 4 senses broad spectrum light, including ultraviolet, visible and infrared light. Window 11 is substantially transparent to this broad spectrum light.

Both the pipes 10 leading into and out of the sieve are provided with windows 11 and associated housings 1 with cameras 5 to allow analysis of the state of the powder before and after the sieve. The camera can also analyze the condition of the powder entering both inlets to the powder mixing device.

The windows can be provided in powder transport conduits or powder storage containers of other types of equipment, such as, for example, additive manufacturing machines.

FIG. 5 shows a powder storage device with a hopper having a lower frusto-conical wall and a cylindrical side wall. An upright pipe 10 is connected to an inlet at the top of the hopper via a valve 12 for turning on and off powder feeding through the pipe 10 to the hopper. The bottom of the hopper is provided with an outlet and is also fitted with a valve 12 to allow and stop powder from flowing out of the hopper. The hopper is for storing the metal powder and can be used with or form part of the atomizer used to produce the metal powder. The frusto-conical base of the hopper is provided with two windows 11, each with a lamp 6 for imaging the powder in the hopper packed against the window (see FIG. 4). A housing containing a camera 5 with a lens 4 (of the type shown) is attached. As in the apparatus of FIG. 4, the windows have substantially parallel, opposed planes that are transparent to the wavelengths of light produced by the lamps that can be sensed by the camera.

The pipe 10 leading to the hopper 12a has five pairs of windows equally spaced along the portion of the pipe above the valve at the entrance to the hopper. Attached to each window is a respective enclosure 1 containing a camera 5 with a lens 4 and a lamp 6, similar to the enclosure 1 at the bottom of the hopper 12a. These cameras can image the densely packed metal powder in the pipe at various locations along the length of the pipe. This is useful, for example, when it is possible to identify whether the powder fed into the pipe 10 has changed over time, for example during the manufacturing process.

Each camera provides output to a computer or processor 7, which provides output to a user via a suitable user interface 8 or to equipment controlled by this output. Instead of having multiple housings 1 and cameras 4 along the length of the pipe 10, a single camera (or fewer cameras) is movably mounted relative to the pipe, along the length of the pipe. It can also be arranged to image the powder at different locations.

By closing the valve 12 at the entrance to the hopper, the powder to be imaged can be packed densely into the pipe.

FIG. 6 shows part of a pipe for conveying powder. The pipe is fitted with a bypass conduit 10a which is provided with a valve 12 which allows control of powder flow into and out of the bypass. A further valve 12 is provided in the pipe between the connections with the bypass conduit. The bypass conduit comprises a substantially planar window 11 fitted with a housing 1 containing a camera with a lens and a lamp, similar to the housing of FIGS. The output of the camera is connected to a computer or processor 7, a display or other output device 8, or other device controlled in response to the output of the camera. The bypass conduit is formed by a pipe with a smaller cross-section than pipe 10 . The valve is controlled to allow the powder flowing in the pipe 10 to flow into and fill the conduit, allowing the powder to be imaged by the camera 4 in the housing 1 in a compacted state. The bypass allows the powder flowing in the conduit to be efficiently sampled for analysis, allowing powder to be imaged in dense conditions even when the powder is not moving through the pipe in a dense state.

7 and 8 show the laminate molding machine 13. FIG. The additive manufacturing machine includes a housing 1 that is or is configured to be substantially light-shielding. The housing comprises a powder delivery container 14 with a powder delivery piston 15, a building container 14a with a building platform 16, and a wiper mounted on a movable support for transporting powder from the powder delivery container to the building container. The blade 17 is accommodated. The housing also houses laser output optics 18 for selectively fusing the powder on the build platform 16 . All of these features are common to known selective layer fused additive manufacturing machines.

The housing 1 further comprises two cameras 5 with suitable lenses 4 for taking images of the area of the upper surface of the powder in the powder delivery container and the build container, and a lamp 6 arranged around each camera. accommodates An image sensor 19 and a lamp 20 are also mounted on the moveable support of the wiper blade 17 and arranged to scan an image of the surface of the powder in the powder delivery or build container as the wiper blade traverses the container. It is As with the other embodiments, the lamps 6, 20 emit light and the camera 5 and sensor 19 sense broad spectrum light, including ultraviolet, visible and infrared.

A camera 5 with a lens 4 and an associated lamp 6 are also mounted in the housing 1 located behind the windows through the sidewalls of the powder and build containers, through which they are closely spaced within the container. powder can be imaged.

A camera 5 with a lens 4 and an associated lamp 6 are also mounted on the housing 1 located below a window formed in a platform extending between the powder container and the build container. , the powder on its platform can be imaged.

All three windows are transparent to the wavelengths of light produced by the lamps that can be sensed by the camera. The platform windows have substantially parallel, opposed planes. The window into the powder container and the shaped container may also have such a surface, or if it is not flat it may be molded to match the shape of the inner wall of the container.

The camera 5 and sensor 19 are arranged to output images to a connected computer or processor 7 with an output device 8 or arranged to control an additive manufacturing machine in response to the output. Similar to the apparatus shown in FIGS. 1 and 2, the digital camera 5 and sensor 19 are arranged to produce an image of the powder having approximately 1000 times the number of pixels shown in the imaged area of the powder. be done.

Of course, the additive manufacturing machine shown in FIGS. 7 and 8 need not be equipped with all of the imaging devices shown. However, multiple devices can be used to image the powder at various stages of its use.

In use, each embodiment of the device produces a digital image of the surface of the powder within the device. The digital image defines the characteristics of the image elements containing information relating to all wavelengths of light detected by the camera, i.e. wavelengths of light received by the camera from each area of the surface of the powder corresponding to an element of the image. contains a set of data to The ratio of image elements to the number of particles of the powder shown in the image is about 1000. The image data is sent to a computer where the detected wavelength is represented by the wavelength or combination of wavelengths of light, or the luminosity of each element of the image, such as in the CIELAB color space with variables L, a and b. Stored in a computer in a defined way.

The computer is configured to process the image data to identify information about the state of the powder shown in the image by performing at least some of the steps shown in FIG.

As an optional first step 21, the image is cropped to a predetermined size, removing elements outside the borders (or other selected area) of the original image. This optional step allows you to filter out distorted areas of the image, or to combine images taken with different cameras or sensors to represent the same area or have the same number of pixels.

The remaining image data is then inspected at 22 to ensure that it is of sufficient quality for further processing. If not, it is disqualified at 23 and a new image is acquired.

If the image data indicates sufficient quality, the computer then identifies elements whose brightness is below a predetermined threshold and replaces the spaces between particles of powder (or other background material) in the image. ) are removed from the image data at 24 . The actual threshold will depend on the characteristics of the particular equipment being used and the type of powder being inspected. Once the image elements outside the threshold are removed, the remaining image elements represent the powder particles in the foreground of the image.

The data for the remaining image elements can then be processed at 25 to estimate the number of particles they represent using a suitable technique such as watershed segmentation. The total number of particles represented can also be estimated in other ways. For a given powder and device, the number of particles expected to be visible in the area of the surface of the powder corresponding to the number represented by the image data is a measure of the expected powder particle size and expected powder compaction. Information can be used for calculations.

The data for the remaining image elements are then statistically analyzed at 26 to determine how the wavelengths represented or detected by each image element are distributed around the corresponding mean value of all remaining elements. and detect outlier elements representing wavelengths or detected wavelengths outside a threshold percentage of the total population of elements. This can be done using the chi-square test for outlier detection. Other techniques can also be used, including using machine learning. An appropriate proportion of the total population can be selected depending on the type of powder to be analyzed, but a typical proportion is 0.1% of the element of interest, i.e. the outlier population accounts for 0.1% of the total population of elements.

A visualization of this step is shown in FIG. 10, where the number of image elements is represented on the vertical axis against the detected wavelength (or color) measurements on the horizontal axis. Shown here is a generally bell-shaped distribution curve centered on the mean value indicated at 27, and a low threshold 28 and a high threshold 29, outside both thresholds and below the curve. The area corresponding to the side indicates 0.1% of the total outlier population.

The outlier elements are then subjected to a connected component filter at 30 to determine if they are spatially connected in the image they define. Any group of connected image elements exceeding a predetermined number of elements is considered to represent a single particle. The data representing each such identified group is associated with a unique particle identifier, the first identifier identifying the largest group of connected elements and the second identifier identifying the next largest connected element group. Identify the group, and so on.

At this stage, the computer creates a series of definitions for the size and surface properties of individual particles that affect their interaction with various wavelengths of light that impinge on them, revealing statistical outliers within the powder. Generating image data. Its surface properties include color as far as visible light is concerned, but are broader properties in that they include properties that alter the interaction with ultraviolet and infrared light. The inclusion of these non-visible wavelengths broadens the range of particle properties that can be detected by the instrument, particularly in relation to particle composition.

This data is then analyzed at 31 to extract useful data relating to the state of the analyzed powder.

That data includes
the number and proportion of particles with surface properties that fall outside of a given range; and
an average wavelength of light received from particles within a predetermined range;
and the average wavelength received from all image elements.

It has been found that how metal particles interact with light changes as the particles age. In particular, it changes when the particles oxidize or are exposed to heat. The more the particles are oxidized, or the higher the temperature to which they are exposed, the greater the change in effect. Therefore, the amount of change is related to the degree of deterioration that the particles have undergone and thus also to their suitability for reuse. For some types of particles, such property changes can be seen by an observer as a change in the color of the particles, but further examination of the particles using non-visible wavelengths reveals a broader range of types. is more likely to be able to detect changes in the particles of

Moreover, it has been found that the presence of highly degraded particles in spite of the normal condition of a batch of powder can identify the batch as unsuitable for reuse. This is because if even one highly degraded particle is contained in a modeled object, the characteristics of the modeled object may be significantly affected. If highly degraded particles are incorporated into the article, the article can become unsafe, especially if the particles are incorporated into the article where stress concentrations occur during use.

The first measurement above is the total sample size, assuming that the batch of powder from which the sample is taken is well mixed, or that the imaged area of the powder represents a typical composition of the powder as a whole. and the percentage of significantly degraded particles in the entire batch of powder tested. Accuracy can be improved by taking multiple samples from a particular batch and analyzing them separately or performing multiple tests on a batch of powder, such as taking multiple images of the surface of the powder. and/or a particular sample may be analyzed, mixed and then re-analyzed. For a particular powder and build, an appropriate wavelength range and a minimum percentage outside its threshold can be determined, and if the percentage of particles outside the threshold exceeds the selected limit, the batch of powder is , at least for the object modeled object, can be regarded as unsuitable for reuse.

Therefore, this measurement allows determination of the state of the powder regardless of bulk quantity.

The second measurement gives an indication of the average deterioration of the remaining powder when particles outside the threshold are not taken into account. Such measurements are similar to the results of conventional bulk oxygen measurements, but are obtained in a more convenient and non-destructive manner, except that they eliminate the effects of significantly degraded particles (or internal oxygen). be done. A powder is considered unsuitable for reuse when the average wavelength received from the remaining particles is outside another predetermined range, not taking into account the particles that are outside the predetermined range.

A third measurement is similar to the second, but takes into account significantly degraded particles. If the average wavelength of light received from the particles is outside another predetermined range, the powder is considered unsuitable for reuse.

The decision to recycle the powder can be made based on one or more of the three measurements described above. Typically, if it is determined that the powder should not be reused based on any measurements, the powder is not reused. In one embodiment, the first and second measurements are calculated and the powder is considered unsuitable for reuse if either measurement indicates that it is unsuitable for reuse.

It is useful to analyze the new powder before it is used and then after it has been used in the building process and subsequent building processes. Analysis of fresh powders can provide useful control data for comparison with data from subsequent analyses.

Other information about the powder can be determined by detecting light returned by the powder. Artifacts other than the powder may also be detected in the powder. Unusual powder particles, such as particles made of a different material than intended, may also be detected, and the abnormal particles may be identified by observable characteristics. An estimate of the total incident energy experienced by the powder can also be made. If multiple images of a sample of powder or a batch of powder are made and analyzed, comparing the results between the images can determine how well the powder is mixed. Images of powders processed by different machines, such as additive manufacturing machines, can be used to compare machine performance and determine machine condition. Images taken at different time periods and/or different locations within the device are useful for tracking powder through the device.

If the device is incorporated into a powder processing machine or an additive manufacturing machine, the device can automatically perform certain functions based on the output of the inspection results. For example, a powder can be disqualified for continued use, or it can be refreshed by combining it with other powders for continued use. It is also possible to stop the additive manufacturing machine or use a wiper blade to remove the layer of powder and replace it before the powder melts.

Data associated with powder analysis can be stored to validate builds using the powder. In particular, storing an analysis of at least a portion of the surface of a layer of powder deposited during the build process can provide consistency or other facts about the powder used throughout the build process. Time-stamped data can also be compared from multiple images of powder taken at various points throughout the powder conveying system to determine, for example, how efficiently oxidized or contaminated powder moves through the system. It is possible to examine the performance of the system, such as whether

In each example, the computer is equipped with appropriate software that causes the camera to take an image, processes the image to identify the color distribution between image elements, and allows the user to specify ranges, percentages or other values. It can be entered, calculates one or more of three measurements, takes into account user-specified ranges and percentages to determine if a particular sample is reusable, and displays this result on a display such as 8. can be output to the user via

The above embodiments are described as examples only. Many variations are possible without departing from the scope of the invention as defined in the appended claims.

Claims (25)

A method of analyzing a metal powder used in an additive manufacturing process, comprising:
irradiating a region of the powder with electromagnetic radiation, including radiation in the non-visible portion of the electromagnetic spectrum, without melting the powder;
separately detecting irradiating radiation, including radiation in the non-visible portion of the electromagnetic spectrum returned from different portions of the irradiated area of the powder, to produce an output based on the detected radiation;
processing the output to identify one or more properties of the powder;
How to prepare. characterized in that a region of said powder is illuminated with light containing ultraviolet and/or infrared radiation and at least the returned ultraviolet and/or infrared radiation is detected.
The method of claim 1. The powder is in a dense state, and the surface of the powder is irradiated,
3. A method according to any of claims 1 or 2. further comprising detecting irradiating radiation reflected and/or scattered by said powder;
4. The method of claim 3. wherein the output is processed to identify properties of individual particles of the powder;
5. A method according to any one of claims 1-4. characterized in that said property is the ability of the particles to reflect and/or scatter radiation of a particular wavelength and/or the shape of the individual particles of the powder,
6. The method of claim 5. The output is processed to identify the density of the powder,
7. A method according to any one of claims 1-6. wherein the output includes values based on one or more wavelengths of radiation received or detected from respective zones of the powder field;
8. A method according to any one of claims 1-7. characterized in that the zones are of substantially the same size,
9. The method of claim 8. characterized in that each zone has an area smaller than the area of the surface of the powder occupied by a single particle of the powder of average size,
10. A method according to any of claims 8 or 9. characterized in that the area of each zone is no more than 10 times, 500 times, or 1000 times less than the area of the surface of the powder occupied by a single particle of average size powder,
11. The method of claim 10. further comprising storing the values as a dataset;
12. A method according to any one of claims 8-11. processing the data set to identify values associated with spaces between particles by identifying values associated with areas of regions of the powder having luminous intensities below a predetermined threshold; 13. The method of claim 12, wherein processing values other than the identified values to identify values representing particles having properties that are outside of a predetermined range;
14. The method of claim 13. wherein the processing identifies a value representing an area that defines the detected wavelength as an outlier in the wavelength distribution of all processed values;
15. The method of claim 14. wherein a particle or group of particles of interest is identified by identifying groups in which more than a predetermined number of identified areas are connected;
16. The method of claim 15. An apparatus for analyzing metal powders used in an additive manufacturing process, comprising:
means for irradiating regions of said powder with electromagnetic radiation, including radiation in the non-visible portion of the electromagnetic spectrum, without melting said powder;
a detector for separately detecting irradiating radiation, including radiation in the non-visible portion of the electromagnetic spectrum returned from different portions of the irradiated region of the powder, and producing an output based on the detected radiation;
a processor that processes the output to identify one or more properties of the powder;
A device comprising 18. Apparatus according to claim 17, provided in a powder conveying apparatus or an additive manufacturing apparatus. further comprising an irradiation device configured to irradiate the surface of the region of the powder with light containing ultraviolet and/or infrared light;
The detector is arranged so as to detect radiation to the same side as the irradiation surface of the irradiation device,
19. Apparatus according to any of claims 17 or 18. The irradiator and detector are housed in a housing into which the metal powder can be introduced, and the housing allows controlled irradiation of the metal powder at a wavelength of interest to enable controlled irradiation of the metal powder at that wavelength. characterized in that the penetration of electromagnetic radiation can be prevented or at least limited,
20. Apparatus according to claim 19. characterized in that said detector comprises a focusing element for focusing an image of at least part of the area of the powder onto an image plane in which the one or more sensors are arranged;
21. Apparatus according to any one of claims 17-20. wherein said detector is capable of spatially resolving the position of radiation returned from said powder region,
21. Apparatus according to any one of claims 17-20. wherein the detector detects radiation returned from a portion of the powder area at a time;
21. Apparatus according to any one of claims 17-20. configured to store the output as a data set containing values corresponding to wavelengths of radiation received or detected from respective zones of the powder field,
21. Apparatus according to any one of claims 17-20. characterized in that the processor is configured to cause a device to perform the method according to any one of claims 1 to 16,
25. Apparatus according to any one of claims 17-24.

Images