Additive manufacturing systems and methods

Abstract

To deliver an arbitrarily large amount of laser power to the build surface in a controllable profile that can be scanned as a single entity.SOLUTION: An additive manufacturing system may comprise two or more laser energy sources, and associated optical fibers. An optics assembly may be constructed and arranged to form a rectangular laser energy pixel associated with each laser energy source. Each pixel may have a substantially uniform power density, and the pixels may be arranged to form a linear array of laser energy pixels on a build surface with no spacing between the pixels. Exposure of a portion of a layer of material on the build surface to the linear array of laser energy pixels may melt the portion of the layer.SELECTED DRAWING: Figure 26

Description

Technical Field

[0001] Field

[0001] The disclosed embodiments relate to systems and methods for additive manufacturing.

Background Art

[0002] Background

[0002] At present, in the market, many metal additive manufacturing methods are available. These methods can be classified by the source of the material (powder, wire, thin film, etc.) and the form of additional energy to obtain melting / bonding (laser melting, e-beam melting, welding arc, sintering, etc.). The resolution, accuracy, and achievable feature size of the end portion in a given process are based on the form of the initial material and the ability to control the energy distribution for metal fusion. The effective rate of a given process is usually limited by the ability to supply energy into the build surface in a controlled manner.

[0003]

[0003] In a selective laser melting process for metal additive manufacturing, usually, a laser spot is scanned over a thin layer of metal powder. The metal powder scanned by the laser spot is melted and fused into a solid metal structure. When the layer is completed, the structure is indexed, a new layer of metal powder is placed, and the process is repeated. On a new layer located on a previously scanned area on a previous layer, if the area is scanned by the laser spot, the powder is melted and fused onto the individual material of the previous layer. This process can be repeated many times to build almost any form of three-dimensional shape.

Summary of the Invention

Means for Solving the Problems

[0004] Summary

[0004] In one embodiment, an additive manufacturing system includes a build surface, two or more laser energy sources, and two or more optical fibers. Each optical fiber is configured to transmit laser energy from a first end coupled to a respective one of the two or more laser energy sources and out of a second end, and the second ends of the two or more optical fibers are arranged along one line. The additive manufacturing system further includes an optics assembly constructed and configured to shape the laser energy output from each optical fiber to form a respective rectangular laser energy pixel. Each rectangular laser energy pixel has a substantially uniform power density, and the rectangular laser energy pixels are configured to form a linear array of laser energy pixels on the build surface without having a gap between adjacent laser energy pixels, and the exposure of the layer of material on the build surface to the linear array of laser energy pixels melts at least a portion of the layer of material.

[0005]

[0005] In another embodiment, a method for additive manufacturing includes exposing a layer of material on a build surface to a linear array of laser energy pixels. Each laser energy pixel has a rectangular shape and a substantially uniform power density, and in this case, there is no gap between adjacent laser energy pixels. The method further includes melting a portion of the layer of material due to an exposure of a portion to the linear array of laser energy pixels.

[0006]

[0006] It should be understood that the concepts described above and further concepts described hereinafter may be configured in any suitable combination since the present disclosure is not limited in this regard. Further, other advantageous and novel features of the present disclosure will become apparent from the following detailed description of various non-limiting embodiments when considered in connection with the accompanying drawings.

[0007] Brief Description of the Drawings

[0007] The attached drawings are not intended to be drawn to an exact scale. In the drawings, each identical or nearly identical component shown in the various figures may be represented by the same reference numeral. For clarity purposes, not all components are labeled in all of the drawings. The drawings are as follows.

Brief Description of the Drawings

[0008]

Figure 1

[0008] It is a schematic representation of a circular beam having a Gaussian power density.

Figure 2

[0009] It is a schematic representation showing the power density of two adjacent circular beams having a Gaussian power density.

Figure 3

[0010] It is a schematic representation of the power density at different positions of a circular beam having a Gaussian power density.

Figure 4

[0011] It is a schematic representation of a rectangular beam having a uniform power density.

Figure 5

[0012] It is a schematic representation of the power density at different positions of a rectangular beam having a uniform power density.

Figure 6

[0013] Draw a schematic diagram of the incident energy profile of a linear array of rectangular pixels having a uniform power density.

Figure 7

[0013] Draw a schematic diagram of the incident energy profile of a linear array of rectangular pixels having a uniform power density.

Figure 8

[0014] Draw a schematic diagram of the incident energy profile of a linear array of circular pixels having a Gaussian power density.

Figure 9

[0014] Draw a schematic diagram of the incident energy profile of a linear array of circular pixels having a Gaussian power density.

Figure 10

[0015] Depict a schematic representation of the power density of a linear pixel array according to some embodiments.

Figure 11

[0016] Depict a schematic representation of the power density of a linear pixel array and a fixed spot array.

Figure 12

[0017] Depict a schematic representation of the power density output of a linear pixel array according to some embodiments.

Figure 13

[0018] Depict a schematic representation of the power density output of a linear pixel array according to some embodiments.

Figure 14

[0019] It is a schematic representation of a linear pixel array at the focus.

Figure 15

[0020] It is a schematic representation of a linear pixel array away from the focus.

Figure 16

[0021] It is a schematic representation of the pattern of pixels generated by rotationally misaligned square fibers.

Figure 17

[0022] Depict a schematic representation of a rectangular beam profile formed from a circular beam profile according to some embodiments.

Figure 18

[0022] Depict a schematic representation of a rectangular beam profile formed from a circular beam profile according to some embodiments.

Figure 19

[0023] It is a schematic representation of an embodiment of an additive manufacturing system.

Figure 20

[0024] It is a schematic representation of an embodiment of an optics assembly.

Figure 21

[0025] It is a schematic representation of an embodiment of a fiber mount.

Figure 22

[0026] It is a schematic representation of an embodiment of a lens array.

Figure 23

[0027] It is a schematic representation of another embodiment of an optics assembly.

Figure 24

[0028] Schematic representation of an embodiment of an additive manufacturing system including a galvanometer.

Figure 25

[0029] Schematic representation of an embodiment of an additive manufacturing system including a fixed mirror assembly.

Figure 26

[0030] Schematic representation of a linear pixel array formed on a powder bed surface according to some embodiments.

Figure 27

[0031] Schematic representation of an embodiment of an additive manufacturing system.

Figure 28

[0032] Schematic representation of another embodiment of an additive manufacturing system.

DETAILED DESCRIPTION OF THE INVENTION

[0009] Detailed Description

[0033] The inventors have recognized and understood that the thickness of the powder layer, the size of the laser spot, and the accuracy of the laser spot motion can all be combined to affect the dimensional accuracy and precision of the final part produced by a selective laser melting process. Also, the size and power of the laser spot can affect the rate limit of a given mechanical process. For example, for a given laser spot size, there is a minimum spot power to enable melting of the powder layer. Laser power below this point cannot supply sufficient energy to melt and fuse the powder below the laser spot, even when the laser spot is stationary on the powder surface. This minimum energy depends on the material (e.g., type of metal powder), the thermal properties of the powder bed, and the absorption properties of the powder surface associated with a given laser wavelength. As a result, when the laser spot is scanned over the powder surface, this minimum laser power must increase such that the laser beam can supply sufficient energy to the powder for a given period of time to melt and fuse the powder while a given point is below the laser beam spot profile. The greater the scanning speed, the greater the minimum power required to maintain a continuous melt pool below the scanning spot.

[0010]

[0034] In addition, the inventors have understood that there is a limit to the amount of energy that can be supplied under a given laser spot size. The incident laser energy is typically absorbed in a very narrow layer on the surface of the powder. This energy is converted to thermal energy within this thin layer, and then conduction and convection allow this absorbed energy to spread further downward within the powder layer. Conduction is limited by the fine contact points between individual powder particles, and convection occurs through the gaps between the particles. If excessive energy is incident on the layer surface, the energy cannot spread throughout the powder layer fast enough, and the powder surface temperature will reach a temperature high enough to evaporate each part of the metal powder surface. This rapid evaporation at the layer surface will result in the release of powder particles from the powder surface.

[0011]

[0035] High incident energy can cause surface evaporation on the powder surface, while also causing evaporation of the metal within the melt pool. The melt pool is a predetermined amount of molten metal generated by laser heating of the metal powder while the laser spot is being scanned on the powder surface. The melt pool can also be represented as a certain amount of metal powder after the metal powder has melted and before the molten metal has cooled sufficiently to solidify again. If a portion of the metal from the melt pool evaporates due to excessive energy from the laser spot, the rapid expansion of the metal during the evaporation process can result in the release of the molten metal from the melt pool. These released particles can still deform the powder surface within the area to be scanned, thereby causing further melt pool instability due to the non-uniform surface.

[0012]

[0036] In addition, the released particles of the molten metal may tend to form oxides and / or other compounds due to their large surface area, elevated temperature, and exposure to the surrounding gas during the release process. These particles may ultimately remain in the area being scanned and melted in the continuous scanning process and may affect the microstructure and mechanical properties of the final part.

[0013]

[0037] If the incident laser spot energy is excessively large and the scan speed is increased to compensate to reduce the above-mentioned problems of resulting metal evaporation, the thermal energy may not be able to propagate within the powder layer at a speed fast enough to fully melt and fuse all of the metal powder. As a result, the final part will contain voids of unmelted metal powder, which can impair the properties of the final part.

[0014]

[0038] In some instances, the above-mentioned problems can be addressed by reducing the thickness of the powder layer so as to ensure sufficient melting of all of the metal powder at a given laser spot energy and scan speed. However, since a large number of relatively thin layers are required to build a given part, such an approach reduces the effective rate of the process. Also, there are limitations on the layer thickness since it cannot be less than the average powder particle size (usually in the range of 15 - 45 μm).

[0015]

[0039] As the laser spot power and speed increase, another problem occurs. While ensuring sufficient melting of all the metal powder below the scan track, when the incident power and scanning speed are increased sufficiently, the melt pool can still become unstable. Instead of generating a continuous melt pool that is then cooled into a continuous solid metal track, the melt pool begins to destabilize and break up into individual molten droplets, which then cool into jagged balls on the surface of the treated area. This phenomenon, often referred to as "balling," is a mechanism for forming individual droplets due to Rayleigh instability. This instability is observed when the length-to-width ratio of the melt pool reaches a certain critical value. As a result, a high surface tension gradient can lead to the formation of voids within the tail of the melt pool. As the laser spot scanning speed increases, the length of the melt pool grows significantly more rapidly relative to the width of the melt pool, causing the length-to-width ratio to increase. Further factors contributing to this melt pool instability phenomenon are the local composition, wetting, Marangoni flow, and gravity of specific powder particles.

[0016]

[0040] In addition to the above, another possible way to increase the rate of a laser-based powder bed fusion / melting process is to increase the spot size. As the spot size increases, the total spot power increases while the average power density of the spot remains constant, thereby enabling an increase in the effective machine rate. However, the inventors recognized that this ability to increase the spot size is limited because at certain sizes, the additional power results in metal evaporation rather than an increase in the net effective powder processing rate. Also, as the spot size increases, the spatial resolution of the final finished part decreases. If the resolution of the feature size decreases sufficiently, the resulting part will require intensive post-machining and processing to obtain the final desired and useful state, thus eliminating the benefits of the selective laser melting process.

[0017]

[0041] In view of the above, the selective laser melting process in current machines is constrained by the physical limitations of the process. It is not possible to achieve a relatively large net processing rate while maintaining a high-quality process state. One approach attempted to address these limitations was to add multiple laser beams to a single machine in a state where the beams are scanned completely independently using rotating mirrors. For example, machines with two, or in some cases, four, beams are available with a total power in the range of 2 - 4 kW. This approach can allow for a relatively large effective machine rate by increasing the number of beam spots, but each spot is still constrained by the energy limits outlined above. Furthermore, the positional accuracy between each spot location on the powder layer can make multi-spot processing of a single point difficult without a reduction in part accuracy and resolution.

[0018]

[0042] A multi-spot laser system may have sufficiently independent laser spot control that can be useful for generating multiple independent parts within a single powder bed. However, the effective rate for each part / spot combination is still limited by the process power combination. Also, it is possible to use multiple independent spots on a single part. In this case, the positional accuracy between the spots becomes very important and the control is also very difficult. The part accuracy and resolution may be reduced and care must be taken to avoid interference between the beams and the beam positioning mechanics. As the machine size increases to accommodate a relatively large build volume, it becomes even more difficult to maintain the positional accuracy between multiple independent spots.

[0019]

[0043] Also, the multi-spot laser system can be constructed to have final delivery optics that are fixed relative to each other. This approach can assist in solving the problem of positional accuracy between spot locations, but due to the size of the delivery optics, the layout of the spot locations may not be optimal. For example, in order to keep the output spots close to each other, the delivery optics are necessarily placed at a sufficient distance from the spot size due to its size. In some examples, the optics are arranged at a predetermined angle to each other so that the beams converge at the incident area on the powder surface or at another point in front of the final reflecting mirror or galvo scanner assembly, thereby focusing the beams into one. However, the problem associated with this approach is that the incidence with the angles of multiple spots results in different melting behaviors below each laser spot. Also, the different incident angles limit the proximity with which each spot can be placed and maintained adjacent to each adjacent spot. To obtain a uniform melt front from two adjacent circular spots, the energy density within each spot must be uniform, which can be very difficult to achieve with different incident angles and separate delivery optics. For this reason, a fixed relative position multi-spot laser system is typically configured to operate with individual separate spots each generating a separate melt pool. As a result, the scanning pattern is designed to be incrementally stepped in the gaps between spots on a continuous scan path. Furthermore, the problem associated with placing multiple fixed optical heads adjacent to each other places a practical limit on the number of laser units that can be utilized at once with this approach, and thus, a fixed independent laser optics head system is typically limited to 4 to 5 individual laser spots.

[0020]

[0044] Yet another approach used to address the above power density limitations in single or multi-spot selective melting systems is to scan the spot very rapidly in a reciprocating direction along a linear path to produce a heating area having an average line shape. However, even with a very high scan speed, the resulting average line shape is still limited in power input due to the heat transfer characteristics of the powdered metal layer as described above. For example, excessive power still causes metal evaporation and metal pool instability.

[0021]

[0045] In view of the above, the inventors have recognized and appreciated a number of benefits associated with a selective laser melting system that utilizes a line-shaped incident energy source. For example, such a system can provide an increase in the effective material processing rate in comparison to the approaches described above. In some embodiments, the instantaneous shape of the incident energy is a line having a first dimension (e.g., the length of the line) that is greater than a second dimension (e.g., the width of the line). The power density profile across and along the line can be capable of being controlled to be substantially uniform, and the line can be scanned in at least two directions with the primary scanning direction perpendicular to the long dimension of the line. In some embodiments, the power density profile of the line can be modulated over time.

[0022]

[0046] The embodiments described herein address many of the above - mentioned problems by increasing the described effective laser power supply limits while maintaining sub - resolution and accuracy over a relatively large build volume in powder - bed metal additive manufacturing. The systems and methods described herein can supply any large amount of laser power to the build surface in a controllable profile that can be scanned as a single entity. In some embodiments, the ideal shape of this incident laser power is a line having a long dimension and a short dimension, in which case the primary scanning direction is perpendicular to the long dimension of the line. The power density in the long direction of the line can be made substantially uniform.

[0023]

[0047] In some aspects, the line of incident laser energy is composed of a plurality of individual laser energy pixels arranged adjacent to each other, where the individual power levels of the pixels can be individually controlled. Each laser energy pixel can be independently turned on or off, and the power of each pixel can be independently controlled. The power density across any single pixel can be substantially uniform such that the pixel has a top - hat energy profile when the pixel is turned on. In some embodiments, the power density of two adjacent pixels can result in a uniform power density over the length of the resulting line when both pixels are turned on and set to the same power density.

[0024]

[0048] As a result, the pixel-based lines obtained are mainly scanned perpendicular to the long axis of the lines, so the forward speed and pixel power density are still constrained by power and speed limits that are approximately the same as those of conventional single-spot laser selective melting processes. However, since there are multiple spots directly adjacent to each other, the effective process rate can be approximately N times that of a single pixel rate, where N is the number of available pixels in this case. Also, since each pixel can be individually turned on and off, the effective part resolution and accuracy remain comparable to those of a single-spot system. The system can operate as a single-spot system by turning on only a single pixel, but in this case, the effective system rate is substantially the same as that of a single-spot system.

[0025]

[0049] In some embodiments, the additive manufacturing system includes an optical path configured within an optics assembly (e.g., an optics box) for generating lines, the optics assembly including a set of lenses arranged in series. Alternatively, or in addition to this, one or more mirrors can be added to the beam path for beam rotation or bending, and / or a galvo scanner can be added to the beam path for one-axis powder bed scanning.

[0026]

[0050] According to some aspects, the laser energy of each laser energy pixel is generated by an independently controllable laser energy source and supplied to the optics assembly through individual optical fibers associated with each laser energy source. In some examples, each individual optical fiber may include fiber segments joined together into a single optical fiber. Alternatively, or in addition to this, a single optical fiber path can also be generated by using an optical connector to join the ends of two fibers together. All of the optical fibers from multiple (e.g., two or more) independent laser energy sources are routed to the optics assembly. Within the optics assembly, the ends of the optical fibers are received within a mounting fixture (e.g., a fiber holder) that ensures that the ends of the optical fibers are parallel and aligned. The ends of the multiple optical fiber cables can be cut and polished to ensure a clear and uniform optical path at the exit from each optical fiber. Also, the ends of the multiple optical fiber cables can be coated with an anti-reflection coating. From these fiber terminations, all individual laser beams pass through a single set of optical lenses and mirrors within the optics assembly. This configuration maintains the optical paths of all laser energy sources in the same state and ensures that the individual pixels from the individual laser energy sources remain adjacent while minimizing the size of the optical system.

[0027]

[0051] With all optics following the same optical path through the same lens array, the resulting laser energy lines maintain their shape and can be reflected by a single mirror assembly and scanned using a mirror attached to a single-axis galvo scanner. As a result, any number of laser systems can be combined into pixel-based lines to generate a high-rate selective laser melting process.

[0028]

[0052] As will be described in more detail below, in some embodiments, the output from the optics assembly can be directed towards the powder layer using a galvanometer scanner and then passed through a lens or lens assembly, such as an f-θ or telecentric lens, to minimize beam shape distortion at non-vertical incidence on the powder layer.

[0029]

[0053] In some embodiments, the output of the optics assembly can be scanned in the primary direction using a galvanometer scanner, while the entire optics assembly is scanned in a secondary direction perpendicular to the primary direction using a motorized stage actuator. Alternatively, instead of this, the output of the optics assembly can be scanned with high-speed motion using a galvanometer scanner in the primary direction, while the optics box is scanned with relatively slow motion in both the primary direction and the secondary direction perpendicular to the primary direction using a motorized stage mounted orthogonally. In other embodiments, the output from the optics assembly can be scanned using only motorized stage motion without any accompanying galvanometer scanner stage. In a further embodiment, the optics assembly can be mounted such that the pixel array line output from the optics assembly is oriented at a fixed angle in relation to the motion stage to move a line perpendicular to the long axis of the pixel line. Alternatively, instead of this, this can also be achieved by scanning the output from the optics assembly using a galvanometer scanner. In other embodiments, the output from the optics assembly can be rotated dynamically in relation to the motion stage during motion. Alternatively, instead of this, the dynamic rotation of the optics box can also be coupled to a fixed galvanometer scanner in relation to the optics box.

[0030]

[0054] In certain embodiments, the optics assembly can also be mounted on a motorized stage that allows for upward and downward motion perpendicular to the powder layer on the build surface, in combination with motorized motion along other axes. This can allow for improved focusing of the output beam as the beam is scanned back and forth by a galvo scanner. Alternatively, instead of this, the optical path within the optics assembly can include an autofocus array to enable fast dynamic focus adjustment. This can allow for a relatively large focus adjustment to accommodate a relatively wide galvo scanning range.

[0031]

[0055] Depending on the embodiment, a linear array of laser energy pixels (i.e., a line array) can have a uniform power density along the length of the line array in the length direction. In some examples, the line output pixel array can have a non-uniform power density along the length of the line array by setting different power output levels for each associated laser energy source of each pixel. Further, individual pixels on the ends of the linear array can be selectively turned off or on to create a line array having a relatively short length or a relatively long length in the length direction. In some embodiments, all pixels except a single pixel can be selectively turned off to obtain a single point pixel for microfeature profiling. In certain embodiments, the linear array can be divided into a plurality of relatively small linear arrays that have the same power density along each respective relatively small linear array. Alternatively, instead of or in addition to this, the plurality of relatively small linear arrays can have different power densities that are uniform within each respective relatively small array but different from array to array. In some examples, the plurality of relatively small linear arrays can have different power densities between the small arrays and different power densities across each respective relatively small array.

[0032]

[0056] Further, in some embodiments, the power levels of the various pixels within the linear array of laser energy can be independently controlled throughout the additive manufacturing process. For example, the various pixels can be selectively turned on, turned off, or operated at an intermediate power level to provide a desired power density along the length of the linear array.

[0033]

[0057] According to some aspects of the present disclosure, the optical path of the incident laser beam after exiting the optical fiber may be important for obtaining a uniform line shape on the powder surface. In some embodiments, the optical path of the additive manufacturing system includes a lens array that is followed by one or more micro-lenses (e.g., one or more micro-lens arrays) by one or more objective lenses. All beams from an independent laser energy source pass through the same set of lenses within the lens array and the same objective lens within the optics assembly. As will be described in more detail below, the lens array (including one or more micro-lenses) can be configured to collimate the laser energy output from each optical fiber source and to convert the beam shape from a circular beam profile having a Gaussian power distribution to a rectangular beam profile having a uniform power distribution in both axes (e.g., a top-hat power distribution). In this way, the lens array can convert the laser energy output into an array of rectangular laser energy pixels. The objective lens can be configured to define the focal length of the combined line array and to function to reduce the output from the lens array. This reduction changes the pixel spacing from an initial spacing set by the distance between adjacent fibers in the fiber holder to a desired pixel spacing on the powder surface, reducing the pixel spacing. For example, the objective lens can be configured to reduce the array such that there is no spacing between adjacent pixels.

[0034]

[0058] According to some aspects, the laser line output from the optics assembly may be a linear array having two or more rectangular pixels, in which case the line is longer than its width when all of the pixels are in the on state. For example, in a system having two laser energy sources, when both lasers are controlled to output the same laser power, the power density across the length of the line is substantially uniform. Alternatively, instead, the power levels of these two pixels can be controlled to have different levels, which results in a line output having a power density that varies across the length of the line. In other embodiments, more than two pixels can be utilized to achieve a relatively long laser line. In some such embodiments, by controlling each pixel to have the same power output, the power density across the length of the line can be maintained in a uniform state. Alternatively, instead, by independently controlling the pixels to have different power levels, any combination of power density along the line can also be obtained.

[0035]

[0059] The inventors have understood that the conversion of the beam output from each laser energy source from a circular Gaussian profile to a square top-hat profile is important to enable multi-track single-pass melt pool stability on the powder surface. For example, FIG. 1 shows a circular beam having a Gaussian power density around its center line. As shown in FIG. 2, when two circular Gaussian beams are arranged adjacent to each other, the power densities are combined according to the exact beam power density. The sum of the combined power densities on the line passing through the centers of both circular profiles is as shown in the graph on the right in FIG. 2. The combined power density defines peaks and valleys and is not uniform. As shown in FIG. 3, when these two spots move in a direction orthogonal to their stacked direction, different tracks on the powder surface will be exposed to different power profiles depending on the position of the track in relation to the center line of the incident beam. The track directly on the center of the individual circular beam (Track A in FIG. 3) will correspond to a relatively wide beam shape with a relatively large peak intensity in comparison to the track offset from the center line of the individual circular beam (Track B in FIG. 3). The largest difference between different tracks is between the track on the center line of the individual beam and the track at the point where two adjacent circular beams intersect. The differences in the effective incident beam width and peak power density in different tracks under this moving line will consequently result in different melting rates and melt pool instabilities such as those described above.

[0036]

[0060] In contrast to the above circular Gaussian beam shape, when the output from each independent laser energy source is shaped into a rectangular beam shape (i.e., a top-hat profile) having a uniform power density (e.g., using one or more microlenses), the resulting beam shape and power density are those shown in FIG. 4. In both the X and Y directions of the resulting rectangular beam shape, the power profile is substantially uniform. As a result, when two rectangular profiles are arranged adjacent to each other and moved so as to be orthogonal to the length of their lengthwise stack, the power profiles of the adjacent beam profiles are combined to provide a substantially uniform profile not only over the length of the line projection but also over the short width of the line. This means that each track line (tracks C, D, and E in FIG. 5) corresponds to the same effective beam width and peak power as all the other tracks as the line is scanned over the powder surface.

[0037]

[0061] Furthermore, the inventors have understood that in order to produce a high-quality fused material track, the uniformity of the net power exposure of the incident laser energy line on the powder surface is important. As shown in FIGS. 6 to 9, a line (FIGS. 6 to 7) composed of a plurality of adjacent rectangular top-hat profile pixels produces a relatively uniform incident energy profile at all points within the scanned beam in comparison with a similar line (FIGS. 8 to 9) composed of adjacent circular Gaussian pixels. According to the rectangular top-hat profile, not only is the centerline power density relatively uniform across the width of the line, but the power density deviating from the centerline across the width of the line is also relatively uniform. Even when the circular Gaussian profile is converted into a circular top-hat profile (i.e., a circular beam shape having a uniform power density), while the centerline power density profile between the rectangular and circular beam shapes becomes the same, the off-centerline power density of the rectangular beam shape is significantly more relatively uniform than in the case of the off-centerline power density of the top-hat circular beam shape. Therefore, in all cases, the rectangular top-hat beam shape for each pixel produces a relatively uniform power density at all points within the multi-pixel line compared to a similar multi-pixel line composed of circular beam shapes of Gaussian or top-hat profiles. Furthermore, a relatively uniform power density can produce a relatively small melt pool instability and can allow for a relatively high processing speed, a relatively high power density, and a relatively wide process window.

[0038]

[0062] According to some aspects, it may be important to operate an additive manufacturing process in accordance with a powder surface positioned within the focus. For example, this can ensure the maximum possible power density as the incident beam area is minimized on the powder surface. Also, this can minimize the size of each pixel and allow for the highest possible resolution of the resulting melt pool and fused part features. Operation outside the focus can result in a low effective power density of the incident beam, which can result in a relatively low scanning speed being required and can slow down the net processing rate. Alternatively, in order to operate at the same speed and obtain a similar processing rate, a relatively high-power laser energy source may be required. These large net powers can, as described above, result in melt pool instability, which will also result in a relatively small processing speed being required to operate within a stable set of process parameters. Also, operating within the focus can help maintain the power density profile across the width of the multi-pixel line in a substantially uniform state. According to operation outside the focus, pixels can overlap and the power density profile across the length of the line can have undesirable peaks and valleys. Depending on the particular embodiment, the type, number, and spacing of the objective lenses depend on the desired degree of reduction, the desired focal length, and the desired size of the focus.

[0039]

[0063] Referring to FIGS. 10-13, another example of the influence of the power distribution within a linear array of laser energy pixels will be described. Specifically, FIG. 10 depicts a linear array 200 that can be projected onto a surface (e.g., onto a powder bed surface), and the linear array includes a series of adjacent rectangular laser energy pixels 201-205. As described above, the power density across the width of an individual pixel can have different distributions, and the interaction of the power densities of the different distributions of adjacent pixels will determine the uniformity of the power density of the linear array 200. When the power density distribution for an individual pixel takes on a Gaussian form 210, when a plurality of adjacent pixels are set to the same power level, the individual pixels 220, 221, 222 are combined to form a single output line 230 having an output power density distribution that is the spatial sum of the individual pixels. The variations in density on this single output line 230 are due to variations not only within each pixel but also to the sum of the overlap areas between the pixels. When the power density distribution for an individual pixel takes on a uniform distribution (i.e., a top-hat form) 215, when adjacent pixels within the linear array 200 are set to the same power level, the individual pixels 225, 226, 227 are combined to form a single output line 235. In comparison to the Gaussian pixels, since there is relatively little variation across the width of each individual pixel and relatively little overlap area between adjacent pixels, the resulting output line 235 has a relatively uniform power density distribution in comparison to the line 230 from the Gaussian pixels. The same content is applicable to the case of the power distribution across the width of a single output line. The top-hat profile 215 across the width of the pixel results in a relatively uniform power density distribution compared to the case of the Gaussian distribution 210 across the width.

[0040]

[0064] As shown in FIG. 11, the use of a single output linear array 250, which includes independently controllable laser energy pixels 251-255, has advantages compared to the use of individual fixed spot arrays 270, where the individual laser energy inputs must be tilted relative to each other to correspond to the size of the optical array. In the single output linear array 250, the individual pixels are all incident on the surface (e.g., the powder bed surface) at the same angle. This angle can be perpendicular to the surface (280), or can have a sharp angle with respect to the surface, away from the primary scan direction (281), or towards the primary scan direction (282). All pixels always maintain the same angle. In the case of a top hat power density profile across the width of this line, the power density for the sharp angle (283, 284) has a low peak value depending on the actual angle of incidence, but has a relatively wide width. In contrast, in the case of a Gaussian distribution, the power density across the width of the line shows an elongated lead-in 286 or an elongated tail 288 with a sharp angle of incidence, in comparison to the perpendicular incident pixel 287.

[0041]

[0065] Furthermore, according to each individual spot array 270, due to the size limitation of the optical array, the incident beam will always be composed of a combination of sharp incident angles in both the direction (301) towards and the direction (302) away from the primary scan direction, and may also include a vertically incident spot (300). In the case of an individual spot having a top-hat power density distribution, spots having the same incident angles 303, 305 from different directions will result in similar power distributions, but for the vertical spot 304, different power densities will result. In the case of a Gaussian power distribution, different incident angles will result in several spots being scanned simultaneously by elongated lead-ins 306 and elongated tails 308, and a potential vertically incident spot 307 without a lead-in or a tail. This hinders the optimization of the scan speed in different incident beams. As the incident angle increases, the difficulty of managing different scan power density profiles increases.

[0042]

[0066] Figure 12 shows a line array with different possible outputs. The upper plot shows a potential array of 8 pixels forming a line. Each pixel within this line can be set to the same power density to generate a full-width line with a uniform power density along the length of the line, as shown in the uppermost plot of Figure 12. In some examples, individual pixels can be turned off to generate relatively short individual lines, as shown in the central plot of Figure 12. Alternatively, instead of this, the power level of each pixel can be set individually to generate a line of almost any power density along the length of the line, or along a plurality of relatively small lines. As the total line width increases and the pixel count increases, the ability to set a non-uniform power density along the length of the line can be useful for adapting the process to account for edge effects. Pixels on the powder bed surface that are adjacent on both sides to other pixels may require relatively less total incident power to obtain a stable melt pool state than pixels on the edge of the line. This is due to convective and conductive heat losses at the edge of the line that must be countered by relatively more power. Also, the presence or absence of fused material below an individual pixel from a previous scan layer can affect the local thermal characteristics of the melt pool of a given pixel. One solution to these edge effects and previous scan effects is to set a custom power density across the length of the line depending on the boundary conditions of that pixel at that point during the scan. The lower plot of Figure 12 shows a line power density with relatively high power at the two edges and a stepped profile across the remaining pixels of the line.

[0043]

[0067] Figure 13 depicts another example of the ability to turn off individual pixels within a linear array. Similar to the configuration described above in relation to Figure 12, this also allows for the use of not only line segments of different lengths but also multiple relatively short segments. As shown in the upper plot of Figure 13, the linear array can be divided downward to the resolution of individual pixels. The ultimate resolution of 3D features formed from a powder bed fusion process by turning individual pixels on or off at various points during the scanning motion of the optics assembly is that of a single pixel. Also, the line length can be formed from multiple pixels shown in the two central plots of Figure 13 to a single plot shown in the lower plot of Figure 13. The single pixel resolution can be generated from any of the pixels within the linear array. As a result, by activating any single pixel at any point, it is possible to generate fine resolution features at any point in the scanning motion of the linear array. Also, the single pixel resolution can be obtained from a single scan of the line array, which has the characteristic of having a predetermined angle with respect to the direction of the scanning motion by turning on and off consecutive adjacent pixels in the optics unit moving on the powder bed surface.

[0044]

[0068] As will be described in more detail below, one or more objective lenses after a lens array including one or more microlenses can define the focal length and focus of a combined beam (e.g., a linear array of laser energy pixels). The focal length of the beam is the distance from the last objective lens of the optical assembly to the focus. At the focus, each beam or pixel has its minimum dimension, and each beam is directly adjacent to an adjacent beam. For example, there may be no gap between adjacent beams such that the edges of adjacent laser energy pixels are in contact with each other. FIG. 14 schematically shows the state at this focus. As the beam moves away from the focus (in any direction along the beam path), the size of the beam increases, and the beams can begin to overlap with each other as shown in FIG. 15. The focus is typically defined as a region along the beam path within which the beam size is defined. In this example, this can be defined as a region within the beam path where the rectangular profile dimension is less than a predefined value. In some examples, the focus can be defined as a length along the beam path rather than an actual single point, and is defined to start when the size of an individual beam (e.g., pixel size) decreases below a predefined focus size and end when the beam expands back beyond the predefined focus size.

[0045]

[0069] In some examples, a rectangular top-hat beam shape can be obtained without the use of a lens array including microlenses. For example, when a square optical fiber is used to supply power from an independent laser energy source, the output from the fiber in the fiber holder already has a square top-hat shape with a uniform power density in both directions of the square. The independent output from a square optical fiber mounted within a fiber mount can be directed straight into an objective lens stack for reduction and focusing on the fiber surface.

[0046]

[0070] However, the inventors understood that the use of square fibers can impose additional requirements regarding the alignment and attachment of the fibers within the fiber mount. Specifically, for square fibers where the square top-hat shape is established by the fiber itself, the fiber mount must establish not only the spacing and axial alignment, but also the rotational alignment between the square shapes. If the square fibers are not rotationally aligned, the resulting pixel pattern will also be misaligned, as shown in FIG. 16. In contrast, according to a lens array including a circular fiber and one or more microlenses to shape a circular beam into a rectangular beam, as shown in FIGS. 17-18, the orientation of the fiber around its axis is not a critical parameter. Since the ultimate rectangular shape and orientation of the pixels are established and maintained by the lens array, a circular fiber can be in any orientation and still be suitable. Since each microlens of the lens array can be manufactured as a single monolithic part, maintaining the alignment between the rectangular beam shapes is inherent in the manufacture of each lens and can be controlled. Thus, after the rectangular pixels pass through the objective lens, the output from the fiber optic assembly will produce a uniform pixel array on the powder surface without the need to rotationally align the fibers.

[0047]

[0071] Next, referring to FIGS. 19-28, specific non-limiting embodiments of an additive manufacturing system according to the present disclosure will be described in further detail. It should be understood that since the present disclosure is not limited only to the specific embodiments described herein, the various systems, components, features, and methods described in relation to these embodiments can be used individually and / or in any desired combination.

[0048]

[0072] FIG. 19 depicts one embodiment of an additive manufacturing system 10. The system includes two or more independent laser energy sources 1 coupled to associated optical fibers 2. For example, the independent laser energy source 1 may be a Nd:Yag fiber laser having a maximum output power of 10 to 2000 W. In some examples, the maximum output power of each laser energy source may be 200 to 1000 W. In other embodiments, the laser source may include a fiber-coupled diode laser. Each independent optical fiber 2 for each independent laser energy source 1 is routed to an optics assembly 3 such that a first end of each optical fiber is coupled to the associated energy source and a second end of each optical fiber is received within the optics assembly. As will be described in more detail below, the optics assembly 3 generates a combined optical output 6 that is directed onto a powder bed surface 7 disposed on a build surface.

[0049]

[0073] Each independent laser energy source 1 is connected to a central control unit 4 by using a control cable 5. The central control unit 4 is configured to independently control each laser energy source 1. For example, the central control unit can provide an on / off signal and a power output signal to each independent laser energy source. The power output signal to each independent laser energy source 1 can control the output power of the laser energy source from a minimum power level to a maximum output power level. For example, the output power range may be 10% to 100% of the maximum power output of each independent laser energy source 1.

[0050]

[0074] Figure 20 depicts an embodiment of the optics assembly 3. Within the optics assembly 3, the ends of the respective optical fibers 2 are received within the fiber mount 20 (see Figure 21). The fiber mount 20 fixes the spacing between adjacent optical fibers 2 and ensures that the axes of the optical fibers 2 are parallel and aligned. At the output side of the fiber mount 20, the ends of the optical fibers 2 are cut, cleaned, and polished to ensure that a uniform and consistent beam exits each individual optical fiber 2. The combined laser output 34 from the individual optical fibers 2 is directed towards a series of optical lenses 21 - 29. The combined laser output 34 forms a single line from the projection of adjacent independent laser sources 1 through the individual optical fibers 2 linearly arranged within the fiber mount 20.

[0051]

[0075] As best shown in Figure 21, when all independent laser sources 1 are on, the combined laser output 34 forms a continuous line 55 with individual pixels 50 - 54. At points along the optical path within this optics assembly, adjacent pixels 50 - 54 may overlap, or the pixels may be separated from each other. The degree of overlap of the pixels 50 - 54 in the combined laser energy output 34 may depend on the optical path from when the laser energy output exits the fiber 2 in the fiber mount 20 until it passes through the pixels 50 - 54. As a result, any suitable number of independent laser energy sources 1 can be combined into a single combined laser output 34. For example, in some embodiments, the number of independent laser energy sources 1 ranges from 2 - 20, 5 - 50, or 10 - 100.

[0052]

[0076] The combined laser output 34 within the optics assembly 3 passes through a series of lenses 21-29. The number and type of lenses 21-29 depend on the desired output shape and focal length of the combined optical output 6 from the optics assembly. As shown in FIG. 22, the first lens 21 is a fast-axis collimator followed by a slow-axis collimator 22, which are used to form the combined laser output 34 into a line shape having a controlled overlap between adjacent pixels 60, 61, 62, where in this case each pixel is generated by the output of an independent laser energy source 1. Also, the fast-axis collimator 21 and the slow-axis collimator 22 are used to modify the beam shape of each independent laser source 1, for example, to convert a circular beam profile from an optical fiber into rectangular laser energy pixels. The resulting pixel shapes 60, 61, 62 form a continuous output line 63 when all the independent laser sources 1 are turned on. This continuous output line 63 then passes through a further series of lenses 23-29 (see FIG. 20), which shape and focus the output line within the optics assembly 3 to form the controlled and combined optical output 6.

[0053]

[0077] Optical assembly 3 may include certain mirrors 30, 31, 32 configured to bend the continuous output line 63 of laser energy. In some cases, this bending helps to maintain the dimensions of the optical assembly 3 in a limited state. In the depicted embodiment, the continuous output line 63 is bent twice within the optical assembly 3. However, since the present disclosure is not limited in this regard, it should be understood that the output line 63 may not be bent, or may be bent once, three times, or any other appropriate number of times. The optical assembly 3 further includes a frame 36 that includes mounting and alignment features for the fiber mounts 20, lenses 20-29, and mirrors 30-32. Additionally, the optical assembly 3 can include an adjustable focus array 37 having a set of adjustable lenses 27, 28, 29 that can adjust the focal length of the combined optical output 6.

[0054]

[0078] Depending on a particular embodiment, the combined optical output 6 of the optical assembly 3 can exit the assembly in any suitable orientation, such as parallel to the individual optical fibers 2, perpendicular to the individual optical fibers 2 in relation to the input of the individual optical fibers 2, or at any angle in relation to the optical fibers 2.

[0055]

[0079] Figure 23 depicts another embodiment of the optics assembly 500. In this embodiment, an independent optical fiber 600 enters on one side of the optics assembly 500, and the end 601 of the optical fiber is received by a fiber mount 602. The fiber mount ensures that the ends of the individual fibers 601 are firmly held, aligned within the optical assembly, and parallel. Depending on the particular embodiment, the spacing between the individual fibers within the fiber mount may be from about 500 microns to about 10 mm. For example, in some embodiments, the spacing between adjacent fibers may be about 1 mm or about 2 mm. The ends of the individual fibers 601 are cleaved, abraded, and polished at the exit of the fiber mount 602. The beam output from the ends of the individual fibers 601 is then directed to a cylindrical lens 603 that collimates the beam in a first direction perpendicular to the plane passing through the center of the end of the fiber 601. The beam then passes through a cylindrical microlens array 604 that collimates the beam in a second direction orthogonal to the first direction. The beam then passes through another cylindrical lens 605 that shapes the beam from a circular Gaussian to a rectangular top-hat shape in the first direction. The beam then passes through a cylindrical microlens array 606 that shapes the beam from a circular Gaussian to a rectangular top-hat shape in the second direction. After this, the beam then passes through cylindrical microlenses 607 and 608 that generate Fourier transforms in the first and second directions, respectively. At this point, the beam can be bent using a mirror 610, or it can be allowed to continue in a straight path. The beam passes through a set of spherical objective lenses 611 - 616 that reduce the beam and set the focal length and focus. The beam path can be bent by a mirror 613 before, during, or after passing through the objective lenses. The reduction can be by a factor of 10:1, or 5:1, or 20:1, or any range of these factors.The focal length can be set at 100 mm, 200 mm, 300 mm, or any other suitable value by adjusting the type and spacing of the objective lens. The output beam 620 from the last objective lens may be directed straight towards the powder surface, or may be directed towards the powder surface by using a mirror or a mirror controlled by a scanning galvanometer.

[0056]

[0080] The cylindrical lens and the microlens have been described above as collimating and shaping the beam output, but it should be understood that other types of lenses may also be suitable. For example, in other embodiments, any suitable combination of cylindrical lenses, spherical lenses, and conical lenses can be utilized to collimate and / or shape the beam output. Similarly, although a spherical objective lens has been described above, other lens shapes (e.g., cylindrical and / or conical lenses) can be used for the objective lens. Thus, it should be understood that the present disclosure is not limited to any particular lens shape or combination of lens shapes.

[0057]

[0081] Next, referring to FIG. 24, the combined optical output 6 can be directed towards a mirror 80 mounted on the galvanometer. In the depicted embodiment, the combined optical output 6 is redirected towards the powder bed surface 7 by a mirror 80 mounted on the galvanometer. Since the mirror 80 mounted on the galvanometer rotates over an angle 84, the combined optical outputs 81, 82 (after reflection at the mirror 80) are scanned over a distance on the powder bed surface 7 defined by the angle 83. The position of the mirror 80 mounted on the galvanometer is driven by a galvanometer controller 85, and the galvanometer controller is connected to the central control unit 4 via an electrical control cable 86. The angle 83 over which the combined optical outputs 81, 82 can be scanned depends on the focal length of the combined optical output 6 from the optics assembly 3. In some embodiments, the use of an adjustable focus array 37 within the optics assembly 3 (see FIG. 16) allows for a relatively wide angle 83 for scanning the powder bed surface 7. Also, in some examples, the optics box 3 having a mirror assembly 80 mounted on the galvanometer can be moved relative to the powder bed surface 7, thereby enabling multiple scanning modes.

[0058]

[0082] As depicted in FIG. 25, in some embodiments, the combined optical output 6 from the optics assembly 3 can be directed towards the powder bed surface 7 using a fixed mirror assembly 90 such that the optical output 91 is fixed in relation to the position of the optics assembly 3. The optics assembly 3 can then be moved in relation to the powder bed surface 7 to obtain a desired scanning pattern.

[0059]

[0083] As described above, the combined optical output from the optics assembly that is directed towards the powder bed surface is in the form of a linear array of two or more adjacent rectangular laser energy pixels formed from the output of independent laser energy sources, and the linear array can be scanned over the powder bed surface to selectively melt a portion of the material on the powder bed to form the desired part.

[0060]

[0084] FIG. 26 depicts a schematic representation of a linear array 110 on a powder bed surface 100 that includes a plurality of adjacent rectangular laser energy pixels 101-105. Depending on the particular embodiment, the laser energy pixels can be shaped (e.g., via a lens array that includes one or more microlenses) to have any suitable size. For example, in some embodiments, each rectangular pixel can have a width of from about 50 microns to about 200 microns. In one embodiment, each pixel can have a width of about 100 microns. Further, the power level of each individual pixel 101-105 can be independently controlled by adjusting the output of the individual independent laser energy sources. Also, each pixel 101-105 can be independently turned on and off by controlling the associated laser energy source.

[0061]

[0085] The resulting linear array 110 can be scanned on the powder bed surface 100 not only in at least the primary direction 115 that is perpendicular to the longer dimension of the linear array 110, but also in the secondary direction 116 that is parallel to the longer direction of the linear array 110. For example, scanning in the primary direction 115 can be achieved by moving the optics assembly or by scanning the combined optical output from the optics assembly using a mirror assembly attached to a galvanometer or other suitable configuration. Additionally, in some examples, a combination of moving the optics assembly while scanning with a mirror assembly attached to a galvanometer can also be used to scan the combined optical output on the powder bed surface in the primary direction. Scanning in the secondary direction 116 can be achieved by moving the optics assembly in relation to the powder bed surface 100. It should be understood that simultaneous scanning in both the primary and secondary directions can be used to scan any desired pattern on the powder bed surface 100.

[0062]

[0086] Figure 27 depicts another embodiment in which the optics assembly 3 is mounted on a rotary stage that allows rotation of the optics assembly 3 about an axis. This configuration allows the linear array 110 to rotate to a new position 111 on the powder bed surface 100. In some examples, this can allow subsequent passes on the powder bed surface to be performed in different directions, which can result in sufficient utilization of the line array in the scanning of both layers, while causing the melt tracks on successive layers 112, 113 to be rotated relative to each other at a predetermined angle. For example, in some embodiments, the rotation angle of the optics assembly 3 between successive layers is from about 30° to about 90° (e.g., about 30°, about 45°, or about 90°). The axis of rotation of the optics assembly 3 can be aligned with an axis 130 that passes through the center of the linear array 110. Alternatively, instead, the axis of rotation of the optics assembly can be positioned away from the center of the linear array 110, such as along axis 131. In such an embodiment, the position shift of the linear array 110 can be calculated based on the displacement of the axis of rotation from the center of the linear array 110. This offset can then be accommodated by applying an offset to the primary and secondary scan positions. In some embodiments, the optics assembly 3 can be rotated while simultaneously moving in the primary and secondary scan directions 115, 116 to obtain a linear array 110 in any desired orientation on the powder bed surface 100.

[0063]

[0087] Figure 28 shows one possible layout for a multi-source powder bed laser fusion system. The build surface 510 can be indexed up and down using a vertical slide system 560. The recoater head 570 is configured to add a layer of material (e.g., powdered metal) onto the build surface after the vertical slide has been indexed downward. The optics assembly 500, which receives optical fibers from a plurality of laser sources, is mounted on another vertical slide 550 that is mounted on a cross slide 540 that can generate lateral motion. Each end of the cross slide is mounted in a gantry style on linear slides 520, 530 that allow the cross slide to move in the other lateral dimension. As a result, the output 501 from the optics assembly can be scanned across the entire build surface as appropriate.

[0064]

[0088] While the present teachings have been described in connection with various embodiments and examples, this is not intended to limit the present teachings to such embodiments or examples. On the contrary, the present teachings include various alternatives, modifications, and equivalents, as will be understood by those skilled in the art. Accordingly, the foregoing description and drawings are for illustrative purposes only.

Claims

1. 1. An additive manufacturing system, comprising: A build surface; two or more laser energy sources; two or more optical fibers, each optical fiber configured to transmit laser energy from a first end coupled to an associated one of the two or more laser energy sources and out a second end, the second ends of the two or more optical fibers being arranged along a line; an optics assembly constructed and configured to shape the laser energy output from each optical fiber to form a rectangular laser energy pixel associated with a respective laser energy source, each rectangular laser energy pixel having a substantially uniform power density, the rectangular laser energy pixels configured to form a linear array of laser energy pixels on the build surface without spacing between adjacent laser energy pixels, and exposure of a layer of material on the build surface to the linear array of laser energy pixels melts at least a portion of the layer of material; Including, the system.

2. 2. The additive manufacturing system of claim 1, wherein the optics assembly includes a lens array including one or more microlenses and one or more objective lenses positioned after the lens array along an optical path of the optics assembly.

3. 3. The additive manufacturing system of claim 2, wherein the laser energy output from each optical fiber has a circular beam shape, and the lens array is configured to convert the laser energy output from each optical fiber into a rectangular beam shape having a substantially uniform power density.

4. 4. The additive manufacturing system of claim 3, wherein the one or more objective lenses are configured to demagnify the laser energy output from the lens array to form a linear array of the laser energy pixels on the build surface.

5. 10. The additive manufacturing system of claim 1, wherein each laser energy source of the two or more energy sources is independently controllable to adjust a power level of an associated laser energy pixel.

6. 6. The additive manufacturing system of claim 5, wherein a power density is adjusted along the linear array of laser energy pixels by adjusting a power level of the one or more laser energy pixels.

7. 2. The additive manufacturing system of claim 1, wherein the optics assembly is configured to move the linear array of laser energy pixels along a direction perpendicular to the linear array of laser energy pixels.

8. 2. The additive manufacturing system of claim 1, further comprising a fiber mount coupled to the optics assembly, the fiber mount configured to receive the second ends of the optical fibers and define a spacing between adjacent second ends of the optical fibers.

9. 10. The additive manufacturing system of claim 1 , wherein each rectangular laser energy pixel has a width between about 50 micrometers and about 200 micrometers.

10. 10. The additive manufacturing system of claim 9, wherein each rectangular laser energy pixel has a width of about 100 micrometers.

11. 2. The additive manufacturing system of claim 1, wherein the optics assembly is one of translatable and rotatable relative to the build surface.

12. 2. The additive manufacturing system of claim 1 , wherein the rectangular laser energy pixel is a square energy pixel.

13. 10. The additive manufacturing system of claim 1 , wherein each laser energy source comprises a fiber laser.

14. 10. The additive manufacturing system of claim 1 , wherein each laser energy source comprises a diode laser.

15. 10. The additive manufacturing system of claim 1, wherein a maximum power output from each laser source is between about 200 W and about 1000 W.

16. 2. The additive manufacturing system of claim 1 , wherein the linear array of laser energy pixels forms a homogenous line of laser energy on the build surface.

17. 10. The additive manufacturing system of claim 1, further comprising a material deposition system configured to deposit the layer of material onto the build surface.

18. 20. The additive manufacturing system of claim 17, wherein the layer of material comprises a metal powder, and the material deposition system is configured to distribute the metal powder on the build surface.

19. 1. A method for additive manufacturing, comprising: exposing a layer of material on a build surface to a linear array of laser energy pixels, each energy pixel having a rectangular shape and a substantially uniform power density, and no spacing between adjacent laser energy pixels; melting a portion of the layer of material due to exposure of the portion to the linear array of laser energy pixels; The method includes:

20. 20. The method of claim 19, further comprising controlling a power level of each laser energy pixel to adjust a power density along the linear array of laser energy pixels.

21. 20. The method of claim 19, further comprising moving the linear array of laser energy pixels relative to the build surface.

22. 22. The method of claim 21, wherein moving the linear array of laser energy pixels comprises at least one of translating and rotating the linear array of laser energy pixels.

23. 20. The method of claim 19, wherein the linear array of laser energy pixels forms a homogenous line of laser energy on the build surface.

24. 20. The method of claim 19, wherein each laser energy pixel has a width of about 50 micrometers to about 200 micrometers.

25. 25. The method of claim 24, wherein each laser energy pixel has a width of about 100 micrometers.

26. The method of claim 19 , wherein each laser energy pixel has a square shape.

27. The method of claim 19 , further comprising depositing a layer of material on the build surface with a material deposition system.

28. 30. The method of claim 27, wherein the layer of material comprises a metal powder, and depositing the layer of material comprises sprinkling the metal powder on the build surface.

29. 20. The method of claim 19, further comprising forming a linear array of the laser energy pixels.

30. forming the linear array of laser energy pixels transmitting laser energy from two or more laser energy sources through two or more optical fibers, each optical fiber being coupled to one laser energy source at a first end of the optical fiber; transmitting the laser energy output from the second end of each optical fiber through a lens array including one or more microlenses to convert the laser energy output of each optical fiber from a circular beam shape to a rectangular beam shape having a uniform power density; transmitting the rectangular beam through one or more objective lenses to reduce the rectangular beam; 30. The method of claim 29, comprising:

Images