Misalignment-compensated three-dimensional (3D) object printing system

The 3D printer system addresses alignment errors by using a sensor and controller to measure and correct for misalignment, ensuring precise printing through incremental adjustments, enhancing the accuracy of printed parts.

JP7584320B2Active Publication Date: 2024-11-15XEROX CORP
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
JP2021022075
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Priority Date
2020-03-09
Filing Date
2021-02-15
Publication Date
2024-11-15
Estimated Expiration
2041-02-15

AI Technical Summary

Technical Problem

Existing 3D printers face challenges in maintaining accurate alignment in the cross-process and process directions (X and Y) throughout the height of the build level (Z direction), leading to cumulative errors and misalignment in printed parts, which is particularly acute in low-end printers but can also affect high-end models.

Method used

A 3D printer system with a sensor and controller that measures differences in material deposition at various build levels, calculates shifts in the ejector head to compensate for misalignment, and adjusts the printing process to ensure precise alignment, using diagnostic objects or gauge blocks to determine and correct for errors in the X and Y directions.

Benefits of technology

The system effectively compensates for misalignment errors, resulting in more accurate three-dimensional object printing by adjusting the ejector head position at each build level, improving the precision and quality of printed parts.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007584320000001
    Figure 0007584320000001
  • Figure 0007584320000002
    Figure 0007584320000002
  • Figure 0007584320000003
    Figure 0007584320000003
Patent Text Reader

Abstract

To provide a three-dimension (3D)printer that has an error compensation mechanism, and can operate with high quality and low cost by reducing a setup time and complexity.SOLUTION: A three-dimensional object printer comprises: a platen; a gantry positioned above the platen; an ejector head positioned on the gantry; a sensor; and a controller. The controller is configured to operate the ejector to eject at least one drop of a material toward the platen at a higher build level and to determine a process difference and a cross-process difference between a fiducial point and at least one drop of the material deposited on an upper build surface. The controller is also configured to determine an ejector head shift in a process direction and a cross-process direction associated with each of a plurality of build levels based on the determined process difference and cross-process difference and the number of build levels between the base build level and the higher build level.SELECTED DRAWING: Figure 3
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The apparatus and methods disclosed herein relate to three-dimensional object printing, and more particularly to diagnostic systems in three-dimensional object printers. [Background technology]

[0002] Digital three-dimensional object fabrication, also known as digital additive manufacturing, is the process of creating three-dimensional solid objects of virtually any shape from a digital model. Three-dimensional (3D) object printing is an additive process in which one or more ejector heads of a 3D printer deposit material to build up a part. The material is typically deposited in discrete amounts in a controlled manner to form layers that collectively form the part. An initial layer of material is deposited on a substrate, and subsequent layers are deposited on top of the previous layer. The substrate is supported on a platform that can move relative to the ejection heads so that each layer can be printed, either the substrate is moved via the action of an actuator operably connected to the platform, or the ejector heads are moved via the action of an actuator operably connected to the ejector heads. Three-dimensional object printing is distinguishable from conventional object formation techniques, most of which rely on the removal of material from a workpiece by subtractive processes such as cutting or drilling.

[0003] Care must be taken during the manufacture of printer components and initial setup of a 3D printer to ensure that the rails, platen, and other components of the gantry are properly aligned in the process direction (i.e., Y direction), cross-process direction (i.e., X direction), and build direction (i.e., Z direction). Ideally, the rails in each direction should be perfectly perpendicular to the rails in the other two directions (e.g., the rails in the Z direction should be perfectly perpendicular to the rails in the X and Y directions). In addition, the ejector heads, platens, and associated components configured to dispense and receive build material along the build axis (i.e., Z direction) should ideally be properly aligned with the force of gravity during printer setup so that build material drops fall directly onto the build surface along this axis. If the printer components are not properly manufactured and aligned as intended, each subsequent layer of a 3D printed part will be offset from the previous layer, resulting in cumulative errors with each additional build level (i.e., process and / or cross-process drift occurs at each successive build level and the errors compound) and an insufficiently built part that does not meet design specifications. Such misalignment of any printer component for a variety of reasons may be referred to herein as a printer "misalignment" that results in printer drift in the process and cross-process directions across multiple build levels.

[0004] The above discussion indicates that one challenge with printing 3D objects is maintaining precise alignment in the cross-process and process directions (i.e., X and Y directions) throughout the build level height (Z direction). Maintaining such precision is particularly acute in low-end 3D printers because low-end 3D printers are made with lower quality components that have significant error tolerances. Nevertheless, this same problem can also arise in high-end 3D printers that contain more expensive precision machined parts. Thus, in both high-end and low-end printers, laborious and time-consuming setup procedures must be followed to ensure that all components of the printer are aligned as closely as possible to produce the best possible parts. Unfortunately, these setup procedures are difficult to follow, time-consuming to perform, and add to the cost of the 3D printer.

[0005] In view of the above, it would be beneficial to provide improved 3D printers with error compensation mechanisms and routines that enable low-end printers to produce higher quality parts and that enable high-end printers to be manufactured at low cost and operate with reduced setup time and complexity. It would also be useful in the art to be able to incorporate such error compensation mechanisms into both newly built and existing printers. Summary of the Invention

[0006] In at least one embodiment, a three-dimensional object printer includes a platen, a gantry positioned above the platen, an ejector head positioned on the gantry, a sensor, and a controller. The ejector head is configured to move relative to the platen to a plurality of build levels along a build axis and to a plurality of positions defined by process and cross-process coordinates at each of the build levels. The ejector head has at least one ejector configured to eject a drop of material toward the platen at a plurality of positions. The sensor is configured to determine a difference between a reference and the drop of material at one of the build levels. The controller is operatively connected to the sensor and the ejector head and configured to move the ejector head relative to the platen along the build axis from a base build level to an upper build level. The controller is further configured to operate the ejector to eject at least one drop of material toward the platen at the upper build level such that at least one drop of material is deposited on the build top surface. Additionally, the controller is configured to operate the sensor build top surface to determine process and cross-process differences between the reference and the at least one drop of material deposited thereon. The controller is also configured to determine a shift of an ejector head in a process direction and a cross-process direction associated with each of the plurality of build levels based at least in part on the determined process and cross-process differences and the number of build levels between the base build level and the upper build level. The controller is then configured to operate the ejector head to eject material drops toward the platen at a plurality of predetermined locations associated with each of the plurality of build levels to form the three-dimensional object, where moving the ejector head includes shifting the ejector head relative to the platen by the determined shift of the ejector head associated with the build level relative to each of the plurality of predetermined locations.

[0007] In at least one embodiment, a method for operating a three-dimensional object printer is disclosed. The method includes operating at least one ejector of an ejector head to eject at least one first material drop toward a first build surface provided on a platen at a first build level along a build axis, and then operating a sensor to measure a first distance component of the first material drop from a first fiducial reference on the first build surface, the first distance component including a first process direction component along the process axis and a first cross-process direction component along a cross-process axis. The method further includes moving the ejector head relative to the platen along the build axis, and then disposing a gauge block on the platen, the gauge block providing a second build surface at a second build level along the build axis. Additionally, the method includes operating at least one ejector of the ejector head to eject at least one drop of material toward a second build surface, and then operating a sensor to measure a second distance component of the second drop of material from a second reference on the second build surface, the second distance component including a second process direction component and a second cross-process direction component. The method further includes calculating a difference between the first distance component and the second distance component, the calculated difference including a process direction difference and a cross-process direction difference, and then determining a shift of the ejector head relative to the platen for a plurality of incremental build levels along a build axis between the first build level and the second build level, the determined shift of the ejector head for each incremental build level based on the calculated difference.The method then includes moving the ejector head and operating the multiple ejectors to eject material droplets toward the platen to form the three-dimensional object, and moving the ejector head includes shifting the ejector head relative to the platen by the determined ejector head shift for each of the multiple incremental build levels before ejecting the material droplets at the incremental build levels.

[0008] In yet another embodiment, a method of operating a three-dimensional object printer includes operating at least one ejector of an ejector head to eject a drop of material toward a platen to form a diagnostic shape of a diagnostic object at a base build level, the diagnostic shape having a process direction dimension and a cross-process direction dimension at the base build level. The method then includes moving the ejector head along a build axis relative to the base build level such that the ejector head moves further away from the base build level, and operating at least one ejector of the ejector head to build a diagnostic shape of the diagnostic object at an upper build level along the build axis. The method further includes sensing the process direction dimension and the cross-process direction dimension of the diagnostic shape at the upper build level, and then determining a difference in at least one dimension at the base build level and the upper build level. Additionally, the method includes determining a shift of the ejector head for each of a number of incremental build levels between the base build level and the upper build level based at least in part on the determined difference. After determining the shift of the ejector head, the method includes moving the ejector head and operating at least one ejector to eject material droplets toward the platen to form the three-dimensional object, wherein moving the ejector head includes shifting the ejector head relative to the platen by the determined ejector head shift for each of a plurality of incremental build levels before ejecting the material droplets at the incremental build levels. [Brief description of the drawings]

[0009] The foregoing aspects and other features of the present method and apparatus are explained in the following description taken in conjunction with the accompanying drawings.

[0010] [Figure 1] A 3D object printer is shown.

[0011] [Diagram 2] 2 illustrates the three-dimensional object printer of FIG. 1 at the start of the build process.

[0012] [Diagram 3] 2 illustrates the three-dimensional object printer of FIG. 1 after multiple layers of the build process have been completed.

[0013] [Figure 4] 2 shows a flowchart of a method of operating the three-dimensional object printer of FIG. 1 .

[0014] [Diagram 5] 2 illustrates the platen of the three-dimensional object printer of FIG. 1 with reference and test drops of build material positioned on the platen.

[0015] [Figure 6] 6 shows the platen of FIG. 5 with a gauge block positioned on the platen and a second reference and a second test drop positioned on the top surface of the gauge block.

[0016] [Figure 7] 2 shows a block diagram of a control system for the three-dimensional object printer of FIG. 1.

[0017] [Figure 8A] 2 is a diagrammatic representation of a series of diagnostic objects formed at a base build level using the printer of FIG. 1;

[0018] [Figure 8B] FIG. 8B is a representation of the diagnostic object of FIG. 8A at the end build level.

[0019] [Figure 9A] 2 shows a perspective view of a first diagnostic object produced by the printer of FIG. 1 , illustrating misalignment of the printer in the process direction.

[0020] [Figure 9B] 10 shows a perspective view of a second diagnostic object produced by the printer of FIG. 1 , illustrating printer misalignment in the cross-process direction.

[0021] [Figure 9C] FIG. 2 shows a perspective view of a third diagnostic object produced by the printer of FIG. 1 illustrating misalignment of the printer in the process direction, resulting in the absence of diagnostic features of the object.

[0022] [Figure 10] 2 illustrates a top view of a diagnostic object produced by the printer of FIG. 1 showing printer misalignment in both the process and cross-process directions.

[0023] [Figure 11A] FIG. 2 illustrates a plan view of a base layer of two diagnostic objects formed by the printer of FIG. 1, including a first diagnostic object elongated in the process direction and a second diagnostic object elongated in the cross-process direction.

[0024] [Figure 11B] 11B illustrates a plan view of an intermediate layer relative to a base layer of the two diagnostic objects of FIG. 11A.

[0025] [Figure 11C] 11B illustrates a top view of the first diagnostic object and the second diagnostic object of FIG. 11A relative to a base layer.

[0026] [Figure 12A]FIG. 2 illustrates a plan view of a base layer of two diagnostic objects formed by the printer of FIG. 1, including a first diagnostic object elongated in the process direction and a second diagnostic object elongated in the cross-process direction.

[0027] [Figure 12B] 12B illustrates a plan view of an intermediate layer relative to a base layer of the two diagnostic objects of FIG. 12A.

[0028] [Figure 12C] 11B illustrates a top view of the top layer of the first diagnostic object and the second diagnostic object of FIG. 11A relative to a base layer.

[0029] [Figure 13] FIG. 2 illustrates a top view of a base layer of four diagnostic objects formed by the printer of FIG. 1, including first and second diagnostic objects elongated in the process direction and third and fourth diagnostic objects elongated in the cross-process direction.

[0030] [Figure 14] 2 shows a flowchart of a method of operating the three-dimensional object printer of FIG. 1 . DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0031] For a general understanding of the environment of the printers and methods disclosed herein, and the details of the printers and methods, reference is made to the drawings, in which like reference numbers represent like elements.

[0032] As disclosed herein and described in further detail below, the three-dimensional object printer includes a sensor configured to determine process and cross-process coordinate differences at a higher build level between the material drop and a reference. The printer further includes a controller configured to determine a shift of the ejector head in a process direction and a cross-process direction associated with each of a plurality of build levels based at least in part on the determined process and cross-process coordinate differences. When building the three-dimensional object, the ejector head is shifted at each successive build level by the determined shift amount. As a result, the printer can manufacture parts with greater precision.

[0033] FIG. 1 illustrates an exemplary three-dimensional object printer 110. The printer 110 includes a platen 120, a gantry 130, one or more ejector heads 140, and a controller 150. The platen 120 has a substantially planar upper surface 122 on which a three-dimensional object, such as a part 112, is formed by the printer 110. The ejector head 140 has a plurality of ejectors 142 configured to eject build material droplets to form the three-dimensional object on the surface 122 of the platen 120. In many embodiments, the plurality of ejectors 142 are arranged in one or more rows in the cross-process direction X, although in other embodiments, the plurality of ejectors 142 may instead include only a single ejector 142. In some embodiments, the plurality of ejectors 142 includes a first plurality of ejectors configured to eject build material droplets and a second plurality of ejectors configured to eject support material droplets, such as wax.

[0034] The gantry 130 provides a framework that allows the ejector head 140 to move above the platen 120. To this end, the gantry includes various frame members 132, such as columns, rails, posts, and other structural members that support and guide the ejector head 140 as it moves above the platen 120. The gantry 130 is generally configured to facilitate movement of the ejector head 140 in the process, cross-process, and build directions (sometimes referred to herein as "Y", "X", and "Z" directions, or along the "Y-axis", "X-axis", and "Z-axis", respectively). In at least some embodiments, the platen 120 may also be configured to move relative to the gantry in the build direction.

[0035] The printer 110 includes a controller 150 operably connected to the ejector head 140. The controller 150 is configured to operate the ejector head 140 to ensure proper alignment of the ejector head 140 in the cross-process and process directions (i.e., X and Y directions) with each successive build of prints in the build direction (i.e., Z direction). To form each layer of the three-dimensional object 112, the controller 150 operates the printer 110 to sweep the ejector head 140 in the process direction (i.e., Y direction) one or more times while ejecting material drops from the ejectors 142 onto the platen 120. For multiple passes, the ejector head 140 shifts in the cross-process direction (i.e., X direction) between each sweep. After each layer is formed, the ejector head 140 moves in the vertical build direction (i.e., Z direction) away from the platen 120 to begin printing the next layer of the part in the next build level. Gantry 130 and various other actuators known in the art (not shown) are configured to facilitate the aforementioned movement of ejector head 140 in the X, Y, and Z directions.

[0036] The printer 110 further includes a sensor 146 operably connected to the controller 150 and configured to sense dimensions of a layer of material formed by the printer 100. As described in more detail below, the controller 150 is configured to operate the sensor 146 to measure distances of material deposited by the ejector head 140 from various references on the top layer of material. For example, the sensor 146 is configured to measure process and cross-process distances of a deposited drop of material from a reference mark on the platen or an edge of the part 112 being formed. The sensor 146 is further configured to measure dimensions of the top surface of the partially formed portion 112, including X, Y, and Z dimensions at the top surface. Thus, because the sensor 146 can determine dimensions at various locations along the top surface of the part 112, the sensor 146 is also configured to detect the presence or absence of material at any particular location along the top surface. In various embodiments, the specific locations and / or presence of material at one or more locations on the top layer may be used to calibrate the printer 110 to reduce or eliminate process and cross-process drift at each successive build level. In one embodiment, the sensor 146 is an optical profilometer configured to move relative to the platen 120 in the process direction X to scan the entire part 112, one line, or one row at a time. However, other configurations are possible that do not require the sensor 146 to move to scan the part 112. In addition, as shown in FIG. 1, the sensor 146 may be attached to the ejector head 140. However, in other embodiments, the sensor 146 can be configured to move independently of the ejector head 140 and is not attached to the ejector head in such configurations.

[0037] The printer 110 of FIG. 1 is configured to build parts 112 composed of any of a variety of build materials as will be recognized by those skilled in the art, including, by way of example, ABS plastic, PLA, polyamide (nylon), glass-filled polyamide, epoxy resin, silver, titanium, steel, wax, photopolymer, and polycarbonate. As previously mentioned, as the printer moves to each successive build level, drift may occur in the build material deposited in the process and cross-process directions if the printer components are not perfectly manufactured and aligned. In other words, material deposited at one build level that is intended to be directly above material at a previous build level may not be perfectly aligned in the Z direction. As a result of these imperfections in the printer 110, parts produced by the printer may not be as accurate as designed. Thus, as described in more detail below, the printer 110 is configured with error correction components designed to shift the position of the print head at each build level to produce more accurate parts.

[0038] 2 and 3, a printer 110 is shown in an exemplary embodiment where the printer 110 has been tasked with building a rectangular shaped part. However, as illustrated in FIG. 2, the printer 110 includes a vertical frame member 132 that is not perfectly aligned with the other components of the printer. Specifically, the vertical frame member 132 is not perfectly perpendicular to the platen 120, but instead is tilted slightly in the X direction by an angle θ. This tilt causes small errors to accumulate in the part 112 as the printer components move to each new build level.

[0039] To build a part, the controller 150 of the printer 110 is instructed to deposit build material onto the platen 120 at specific coordinates (i.e., cross-process, process, and build level coordinates (x, y, z)). FIG. 2 shows a base layer 114 of material deposited on the surface 122 of the platen 120. For example, material drops forming the base layer 114 include the front left corner A (x1, y1, z1), the right edge B (x2, y2, z1), and the back left corner C (x3, y1, z1). As the part is built, the print head 140 moves through successive build levels (i.e., in the Z direction) ejecting build material drops towards the platen 120. The part in the example of FIGS. 2 and 3 is a rectangular solid shape, and the printer attempts to replicate the same cross-sectional shape at each new build level. As used herein, the term "build level" refers to the location of a component in the Z direction when properly positioned relative to other printer components so that the printer can build a particular layer of a part together. Thus, two components may be at the same "build level" without being at the same location on the Z axis. For example, the print head 140, platen 120, and build portion 112 may all be considered to be at a single "build level," but these components are not at the same location on the z axis.

[0040] FIG. 3 shows the second layer / top layer 116 of the part 112 after the printer 110 has completed many build layers between the base layer 114 and the top layer 116. As previously mentioned, errors in the alignment of the vertical members 132 will cause the part itself to become misaligned as the print head moves to each additional build level. As errors accumulate, the position of the top layer 116 of the build portion 112 will shift by some differential amount dx, as shown in FIG. 3. For example, consider position A (x1, y1, z1) in the base / bottom build layer, and position A' (x1, y1, z1) in the top layer 116 of the part 112. Since the x and y coordinates are the same (i.e., x1 and y1), based on the setup of the printer 110, position A' should be directly above position A, with only the z coordinate changing (i.e., z1 at position A and z2 at position A'). However, as shown in Figure 3, the actual location of x1 on A' has drifted slightly from the original location of x1 by an amount equal to dx. As an example, if dx is 0.1 mm, then the actual print location of A' on the top layer 116 will shift slightly in the cross-process direction, e.g., (0.1, 0, 20). This is even though the printer believes the coordinates of A' are (0, 0, 20). Thus, this setup of the printer 110 will produce parts that are slightly misaligned in the cross-process direction unless the printer misalignment compensation method described below is used.

[0041] How to determine misalignment using gauge blocks

[0042] Referring now to FIG. 4, the controller 150 of the printer 110 is configured to perform a method 400 for compensating for printer misalignment. The method 400 may also be considered to be a method for calibrating the three-dimensional object printer 110 before full-scale use of the printer. In the description of the method, the statement that the method performs some tasks or functions refers to a controller or general-purpose processor executing program instructions stored in a non-transitory computer-readable storage medium operatively connected to the controller or processor to manipulate data or to operate one or more components in the printer to perform the tasks or functions. The controller 150 described above may be such a controller or processor. Alternatively, the controller may be implemented with two or more processors and / or associated circuits and components, each configured to form one or more tasks or functions described herein.

[0043] When method 400 is executed, it begins with the ejector head 140, gantry 130, and platen positioned at a base build level. With the printer components at the base build level, the method begins by ejecting a test drop of build material onto a build surface positioned at the base build level (block 402). For example, the test drop may be printed on the platen 120. The test drop is printed at a particular location on the surface, such as a particular coordinate location, or within a particular distance in the process and cross-process directions from a fiducial. The location of the fiducial is configured to be detected by sensor 146.

[0044] 5 illustrates an example of a configuration that has performed the actions associated with block 402 of FIG. 4. Specifically, FIG. 5 illustrates a first test drop 160 (i.e., point "A") printed on the surface 122 of the platen 120 at a predetermined distance from a fiducial 170. The fiducial 170 is located at a known location on the platen 120. The fiducial 170 may be a small mark, bump, indentation, or other indicia on the platen that can be easily detected by the sensor 146. The test drop 160 is aligned with a designated y z1 and x z1 As shown by y , the printhead is printed at a predetermined distance from the datum in the process and cross-process directions. In at least some embodiments, the predetermined distance may be directly above the datum. In other embodiments, the predetermined distance may be some relatively small or large distance from the datum. In any case, the sensor 146 detects the printhead in the process and cross-process directions (i.e., y z1 and x z1 ) the distance between the first test drop 160 and the reference 170 and verify that this distance is the same as the predetermined distance (or register this distance as the predetermined distance / original distance).

[0045] Returning to FIG. 4 , the method 400 continues by moving the print head 140 to an upper build level relative to the platen 120 (block 404). At this upper build level, the print head 140 is further separated from the platen 120 along the Z axis. This separation between the print head 140 and the platen allows a gauge block to be inserted onto the platen 120. The gauge block is provided in a simple shape (e.g., a rectangular parallelepiped) with a known height (i.e., Z-direction component) and a flat top surface (i.e., the plane defined by the X-direction and Y-direction components). The gauge block is a precision machined component and may be manufactured from any of a variety of metals that can be grounded to high tolerances, thus providing very accurate dimensions in the Z direction. Similar to the surface of the platen, the top surface of the gauge block also includes fiducials that can be detected by a sensor to identify the location of fiducials on the top surface. Because the gauge block has a known height, the gauge block provides a target build surface at a second build level that is a known offset height above the platen surface 122 in the Z direction. The gauge block and platen 120 preferably include alignment features for precisely locating the gauge block on the platen. For example, the platen may include a number of alignment pins and the gauge block may include a number of complementary cavities, each configured to receive one of the pins. This allows the gauge block to be positioned on the platen in a very precise position in the process and cross-process directions. Of course, any of a variety of other means for alignment may also be used, as will be appreciated by those skilled in the art.

[0046] FIG. 6 shows an example of a configuration for performing the actions associated with block 404 of FIG. 4. FIG. 6 illustrates a gauge block 180 positioned on the surface 122 of the platen 120. In this embodiment, a front left vertical edge 188 of the gauge block 180 extends directly upward from a first datum 170 on the platen 120 until it terminates at a top surface 182 of the gauge block. Because the gauge block 180 has a precise rectangular parallelepiped shape, the upper left corner of the gauge block serves as a second datum 172 having the same xy location (i.e., cross-process, process location) as the first datum 170 on the platen. In addition, because the gauge block 180 is precisely machined in a rectangular parallelepiped shape and precisely positioned on the platen 120 (e.g., via alignment pins or other alignment features), the horizontal edges 184 and 186 of the gauge block are also adapted to provide fiducial reference markers (i.e., known locations relative to the platen surface) at the upper build level.

[0047] Referring again to FIG. 4, after the gauge block 180 is positioned on the platen 120, a second test drop of build material is ejected onto the gauge block (block 406). The printer deposits this second test drop onto the gauge block at a location that the printer determines to be directly above the first test drop on the platen. In other words, the second test drop is ejected when the print head is at a location having the same process and cross-process coordinates as the first test drop (but with a different Z-direction coordinate). The second test drop is ejected without regard to a second fiducial at the second build level. The sensor 146 then measures the distance of the second test drop from the second fiducial at the second build level. Alternatively, the printer deposits the second test drop at another predetermined location on the gauge block (i.e., a location having different process and cross-process coordinates than the first test drop), and the controller considers different positions of the test drop relative to the fiducial mark for different build levels.

[0048] FIG. 6 illustrates an example of a configuration that performs the actions associated with block 406 of FIG. 4. FIG. 6 illustrates a second test drop 162 (i.e., point "B") deposited on a gauge block 180. As described above, the printer deposits this test drop at a location that has the same process and cross-process coordinates relative to the printer as the first test drop, regardless of the location of the second fiducial 172. As an example, if the printer coordinates of the first test drop 160 are (x1, y1, z1), then the printer coordinates of the second test drop 162 are (x1, y1, z2) (i.e., only the z-direction coordinates are changed). The sensor 146 then measures the actual distance in the process and cross-process directions between the second test drop 162 and the second fiducial 172 at the second build level. In this case, the fiducial 172 at the second build level may be considered to include the upper front horizontal edge 184 and the upper left horizontal edge 186 of the gauge block 180. As shown in FIG. 6, the measured process and cross-process dimensions between the second test drop 162 and the reference 172 are y z2 and x z2 It is.

[0049] Returning again to FIG. 4, after the second test drop 162 is printed on the gauge block and the distance dimension is measured to the second reference 172, the controller 150 calculates (block 408) the measured difference / shift in the process and cross-process directions between the first and second test drop. In other words, the controller 150 calculates the process and cross-process drift in the printer between the base build level (e.g., platen surface 122) and the upper build level (e.g., top surface 182 of the gauge block). In the example of FIGS. 5 and 6, the controller 150 calculates y z2 andy z1 (i.e., dy = y z2 -y z1 ) and x z2 and x z1 (i.e., dx=x z2 -x z1 ) to calculate the total difference value. When there is no difference (i.e., no drift), y z2 =y z1and x z2 =x z1 (i.e., there is no actual change in process and cross-process coordinate positions between the base layer and the top layer). However, y z2 ≠y z1 and / or x z2 ≠x z1 As an example of the above-described method for determining cross-process and process drift, in FIG. z1 = 10mm and y z1 Consider the situation where x = 10 mm. Also, in Figure 6, z1 = 11 mm and y z1 = 12 mm. Next, the controller will set dx to 1.0 mm (i.e., dx = x z2 -x z1 =11mm-10mm=1.0mm), dy is set to 2.0mm (i.e., dy=y z2 -y z1 = 12mm - 10mm = 2.0mm).

[0050] With the differences known, the controller 150 then calculates the incremental delta for each incremental build level (block 408). The incremental delta is calculated by dividing each difference (i.e., dx, dy) by the total number of build levels between the base build level and the higher build level. In other words, Δx=dx / z and Δy=dy / z. Continuing with the previous example of FIG. 5 and FIG. 6 where dx=1.0 mm and dy=2.0 mm, further consider the total number of build levels between the base build level and the higher build level to be 20 levels (i.e., z=20 levels). As a result, the controller sets Δx to 0.05 mm / level and Δy to 0.10 mm / level (i.e., Δx=dx / z=1.0 mm / 20 levels=0.05 mm / level and Δy=dy / z=2.0 mm / 20 levels=0.1 mm / level). With this information calculated, the controller determines that the average difference / drift for each level between z=0 and z=20 is 0.05 mm / level in the cross-process direction and 0.1 mm / level in the process direction.

[0051] Returning again to FIG. 4, after Δx and Δy are calculated across the height of the first gauge block (e.g., z=20 in the example of FIGS. 5 and 6), a decision is made as to whether to use additional gauge blocks to determine any additional alignment of the printer or drift across additional build levels (e.g., build levels higher than the build level associated with the top surface 182 of the gauge block 180) or whether the already calculated Δx and Δy are used to determine drift for all build levels of the printer (block 410). If additional gauge blocks are used in block 410, the method returns to block 404 and the ejector head moves again to make room for another gauge block. The next gauge block may then be placed on the first gauge block, and the actions of blocks 404-408 are repeated for an additional set of build levels. Any additional Δx and Δy values ​​may then be associated with a particular build level of the printer (e.g., z=21-z=40, etc.). Additional gauge blocks are particularly useful when the misalignment of the printer 110 is not linear from top to bottom. Specifically, when the errors are non-linear (e.g., considering the possibility of curved rails on the gantry), the additional gauge blocks are useful in determining which particular build levels are associated with more or less drift than other build levels. When no additional gauge blocks are used to determine printer misalignment, the method continues to block 412.

[0052] In block 412, the controller 150 takes all the calculated Δx and Δy information and determines the ejector head shift (i.e., Δx and Δy) for each build level of the printer. In various embodiments, the incremental Δx and Δy values ​​may be the same for each build level. This is the case when only one gauge block is used to determine the average shift across the height of the gauge block and the average shift is applied to all of the build levels of the printer. For example, if a printer has 50 build levels and a gauge block is used to determine the shift across the first 20 build levels, the average shift across those 20 build levels may also be applied to the remaining 30 build levels. In other embodiments where multiple gauge blocks are used, the Δx and Δy values ​​may be different across different build levels. For example, if three gauge blocks are used to determine the shift in the same printer with 50 levels, a first Δx and Δy value may be used across the first 20 build levels, a second Δx and Δy value may be used across the next 20 build levels, and a third Δx and Δy value may be used across the last 10 build levels. In any event, the controller 150 associates specific Δx and Δy values ​​with each build level of the printer prior to printing a part.

[0053] In at least one embodiment, the calculated Δx and Δy values ​​are used to fine-tune the ejector head on the gantry to provide the appropriate shift for the ejector. Thus, in the above example where the controller specifies Δx as 0.05 mm / level and Δy as 0.10 mm / level, adjustment of the ejector can be achieved by simply fine-tuning the print head by the determined shift of 0.05 mm / level in the cross-process direction and 0.1 mm per level in the process direction.

[0054] In an alternative approach, instead of fine-tuning the position of the ejector head 140, the calculated Δx and Δy values ​​may be provided by shifting the image to be printed in one pixel increments across the range of travel, depending on the resolution of the ejector head. For example, if the ejector head has a resolution of 300 dpi, one pixel adjustment may be made approximately every 84 μm. Thus, a shift of 0.05 mm (i.e., 50 μm) per build level may result in one pixel shift for one build level, but no additional pixel shifts at adjacent levels. Following this example, consider the planned pixel shift for a 300 dpi ejector head across three levels. If the desired shift is 0.05 mm (50 μm) per build level, the total desired shift across the three levels is 0.15 mm (150 μm). However, if one pixel shift is performed for each build level, the total shift will be 252 μm (i.e., 84×3=252 μm), which is a larger shift than is desired across these three levels. Therefore, to arrive at a total shift closer to the desired 150 μm shift, one pixel shift may be performed on only two of the three build levels (e.g., on the first and third build levels, with no additional pixel shift on the second build level). As a result, the total shift across the three build levels will be 168 μm (i.e., 84×2=168 μm), which is much closer to the desired 0.15 mm shift. Thus, the controller 150 is configured to even better align the pixels at multiple build levels across the entire build height of the printer. In other words, with multiple build levels across the total build, the controller can easily determine the appropriate ratio of build level to pixel adjustment (e.g., 5 pixel shift across 7 build levels, 11 pixel shift across 12 build levels, etc.) to average the desired Δx and Δy per build level across the entire build.

[0055] After determining the shift of the ejector head for each build level of the printer (block 412), the method then continues as the actual three-dimensional object is formed based on the determined shift of the ejector head (block 414). Specifically, the printer 110 retrieves a set of instructions for building the part, including predetermined coordinates where the ejectors should fire to build the part, and adjusts the instructions based on the determined Δx and Δy values. As a result, the printer 110 is configured to compensate for misregistration errors and form a more accurate three-dimensional object than would be possible without such compensation.

[0056] 7 shows a control system diagram of one embodiment of the printer 110. The control system is essentially a closed loop feedback system that uses sensors 146 to determine errors that the system compensates for by adjusting process and cross-process components at each build level. In the illustrated embodiment, the controller 150 includes a position control component 704, an ejector control component 708, and a sensor control component 712. The reader should note that the particular configuration shown and described with respect to FIG. 5 is merely exemplary. Those skilled in the art will appreciate that many alternative and equivalent configurations can be employed to achieve similar functionality.

[0057] The position control component 704 is configured to provide control signals for operating actuators responsible for providing relative motion of the ejector 142 and the platen 120, and, if necessary, for providing relative motion of the sensor 146 and the platen 120. Additionally, in one embodiment, the position control component 704 provides position information related to the ejector control component 708 and the sensor control component 712. Specifically, the position control component 704 provides position information (X E , Z E) to the ejector control component 708. The ejector control component 708 also includes instructions for building a particular part (i.e., a set of predetermined coordinates indicating where build material should be deposited to build a desired part). When the ejector control component 708 receives an indication that build material is in position to be ejected, based on instructions for the three-dimensional object being generated or based on instructions associated with a diagnostic process, the ejector control component 708 sends a fire signal to the ejector 142 to eject the build material therefrom. The position control component 704 also provides position information (X S , Z S ) to the sensor control component 712.

[0058] The sensor control component 712 is configured to operate the sensor 146 to measure the positions of the various build material drops in the process and cross-process directions at various build levels. Specifically, the sensor control component 712 is configured to measure the positions of the drops relative to a fiducial (e.g., a mark, edge, or other reference point) on a particular build surface. The sensor control component 712 is further configured to determine the existence (i.e., presence) or non-existence (i.e., absence) of the various build materials and fiducials at any particular cross-process, process location (i.e., (x, y) location) at each build height. The sensor control component 712 is configured to receive signals from the sensor 146 corresponding to the presence or absence of the materials and fiducials at the various build locations. The sensor control component 712 also receives position information (X, y) from the position control component 704 regarding the position of the sensor 132. S , Y S ) from the sensor 146 and position information (X S , Y S), the sensor control component 712 is configured to generate a measured distance between the deposited build material and the detected fiducial marker at a given build location, shown as MP(x,y) in Figure 7. The sensor control component 712 is also configured to indicate a null state associated with the absence of build material at a particular build location, if desired.

[0059] The controller 150 is configured to compare the measured distance of the build material from the reference MP(x,y) to the target position for the build material in such layer TP(x,y) to arrive at a difference coordinate for the build material in that layer (dx,dy). These difference coordinates are then sent to a coordinate shift engine 718 that determines the amount of ejector head shift associated with each build level. The coordinate shift engine 718 may be configured to divide each difference coordinate by the total number of build levels for which the difference coordinate was taken (i.e., based on the known target positions) and then multiply those coordinates by the current build level to arrive at a coordinate shift (Δx,Δy) for each build level. For example, if the last known target position TP(x,y) was at the base build level (e.g., z=0) and the measured build position MP(x,y) was at build level 20 (i.e., z=20), the coordinate shift engine 718 divides the difference coordinate (dx,dy) by 20 to determine the incremental shift coordinate. It is then multiplied by the current build level (e.g., z=5) to arrive at a coordinate shift (Δx, Δy) for such build level. The coordinate shift is then delivered to the ejector control 708, which sends a signal to the ejector 142 to fire when a given cross-process and process coordinate (x, y) at the current build level is reached. As the ejector fires, a build material drop is ejected therefrom and deposited on a build surface (e.g., platen, part, support material, etc.). As will be appreciated, the coordinate shift (Δx, Δy) is incremental with each successive build level, such that the total shift in coordinates is incrementally increased with each successive build level.

[0060] As an example of the above, the controller operation may be considered in relation to the examples of Figures 5 and 6 discussed above. In this example, the controller determined Δx to be 0.05 mm / level and Δy to be 0.10 mm / level. The controller then determined that the average difference / drift for each level is 0.05 mm / layer in the cross-process direction and 0.1 mm / layer in the process direction. Thus, at build level 5 (i.e., z=5), the shift of the ejector head in the cross-process direction will be 0.25 mm (i.e., 0.05 mm / level x 5 levels = 0.25 mm) and the shift of the ejector head in the process direction will be 0.5 mm (i.e., 0.1 mm / level x 5 levels = 0.5 mm).

[0061] Considering the above example, it is noted that the coordinate shifting engine 718 can determine pixel adjustments for each incremental build level based on the calculated differences. As will be appreciated by those skilled in the art, the specific mathematics and operations of the controller 150 may be expressed in many alternative but equivalent forms. For example, the coordinate adjustment for each build level may be expressed as the sum of all previous coordinate adjustments for previously formed layers.

[0062] The ejector control 708 is configured to receive coordinate adjustments for each build layer (i.e., Δx, Δy) from the coordinate shift engine 718. The ejector control 708 takes predefined coordinates associated with a planned build and shifts those coordinates based on input from the coordinate shift engine 718. For example, consider the above example where at z=5, the cross-process shift is 0.25 mm and the process shift is 0.5 mm. If the ejector control 708 has instructions to build a part where the ejector head should fire at position (10, 20, 5), the ejector control 708 will adjust those coordinates to (10.25, 20.5, 5) and wait until the ejector reaches this position before firing the ejector. Thus, the ejector control component 708 receives position information (X,Y) from the position control component 704 about the current position of the ejector 142. E , Y E ) is configured to receive each layer (Δx, Δy) and the current position information (X E , Y E ), the ejector control component 708 is configured to provide appropriate firing signals to the ejector 142. In particular, the ejector control component 708 is configured to determine a new ejector firing position at each build level based on the coordinate adjustment (Δx, Δy) for that level. In this manner, the ejector control component 708 eliminates build orientation errors that result from poor manufacturing tolerances or poor settings of the 3D object printer 110.

[0063] The three-dimensional object printer described above in connection with Figures 1-7 is one exemplary embodiment of a printing system and method of operation that is configured to compensate for misregistration. Various alternative embodiments of such printers and methods are possible and are contemplated by the disclosure provided herein.

[0064] Method for determining misalignment by forming a diagnostic object - Patents.com

[0065] 8A-13, in at least one alternative embodiment, a three-dimensional printer is configured to determine process and cross-process drift associated with various build levels along the Z-axis by printing one or more diagnostic objects. The one or more diagnostic objects are printed instead of using one or more gauge blocks as described in FIGS. 4-6. Specifically, instead of using the gauge block 180 described above to detect process and cross-process differences between a base build level and a higher build level, the diagnostic object may be used to determine the differences. According to this embodiment, the sensor 146 of the printer 110 is used to measure various dimensional components and / or features of the diagnostic object. The diagnostic object is printed with appropriate features to communicate the appropriate settings to an operator via graduated part angles and / or dimensions. By way of example, and as described in more detail below, visual feature failures such as the absence of build material at a particular build level may be used to quantify adjustments for feedback to a control system.

[0066] 8A and 8B show a series of diagnostic objects 200 printed by the printer 110 on the platen 120. FIG. 8A shows the diagnostic object 200 at the base build level (i.e., Z=0). FIG. 8B shows the diagnostic object 200 at the end build level (i.e., Z=Z2). The diagnostic object 200 is provided in the form of a number of column structures of different sizes and shapes. The diagnostic object includes a series of first cylindrical columns 210 positioned along the cross-process direction and a series of second cylindrical columns 220 positioned along the process direction. As best shown in FIG. 8A, the first cylindrical columns 210 all have a circular cross-section with the cross-sectional diameter of each succeeding column being larger along the x-axis than the cross-sectional diameter of the previous column. The second cylindrical columns 220 all have a rectangular cross-section with the cross-sectional diameter of each succeeding column being larger along the y-axis than the cross-sectional diameter of the previous column. Depending on the diagnostic features incorporated into column 200, the columns may be constructed from different materials, with different drop sizes / spacing, etc.

[0067] While each of the columns in Figures 8A and 8B are shown in an idealized state with no misalignment in the printer that would result in drift (i.e., no drift in the x or y directions in successive build levels), each of the columns 200 in Figures 8A and 8B includes a diagnostic feature provided by the cross-sectional shape and size of the column. Specifically, the columns 200 are configured such that as errors in the x and y directions accumulate during the build, any jetted material that is beyond the pillar feature boundary built in the previous layer will be unsupported and will jet over the edge of the pillar and into the gap. Thus, the gap designed between the columns 200 takes this into account so as not to jet onto adjacent pillars. For example, assuming the columns are configured to detect a maximum error of 5 mm in either the x or y direction, the gap between the columns 200 is a 6-7 mm gap, providing multiple spacings that allow build material to drop to the bottom platen 120 between the columns 200.

[0068] 9A-9C, three exemplary diagnostic objects are shown with diagnostic features illustrating printer misalignment. In the example of FIG. 9A, the diagnostic object is column 222 and the diagnostic feature is the cross-sectional shape of the column. In the base build layer 224, the cross-process dimension of column 222 is x1 and the process dimension of the column is y1. In the upper build layer 226, the cross-process dimension of column 222 is x2 and the process dimension of the column is y2. As shown in FIG. 9A, x2 is very similar in dimension to x1, but y2 is much smaller than y1. This illustrates a first order error / misalignment for the printer in the process direction (i.e., y direction). In other words, with each successive layer built by the printer, the process direction error / drift accumulates. As the process coordinates drift with each successive build layer (i.e., to the right in FIG. 9A), build material falls off the right boundary / edge 228 of the object and onto the platen (or other substrate on which the base layer 224 rests). Thus, by observing the cross-sectional shape of the columns 222 as a diagnostic feature, misregistration / errors in the printer are indicated in the process direction. To determine the exact amount of error between the base build level and the upper build levels, the difference in the process and cross-process dimensions may be taken (i.e., dy=y2-y1, dx=x2-x1).

[0069] Referring now to the example of FIG. 9B, the diagnostic object is again the column 212 and the diagnostic feature is the cross-sectional shape of the column. In the base build layer 214, the cross-process dimension of the column 212 is x1 and the process dimension of the column is y1. In the upper build layer 216, the cross-process dimension of the column 212 is x2 and the process dimension of the column is y2. As shown in FIG. 9B, x2 is much smaller than x1, but y2 is similar to y1. This illustrates a first order error / misalignment for the printer in the cross-process direction (i.e., x-direction). In other words, as each successive layer is built by the printer, an error / drift in the cross-process direction accumulates. As the cross-process coordinates drift with each successive build layer (i.e., toward the rear of FIG. 9B), build material falls off the rear boundary / edge 218 of the object and onto the platen (or other substrate on which the base layer 214 rests). By observing the cross-sectional shape of the columns 212 as a diagnostic feature, misregistration / errors in the printer are indicated in the cross-process direction. To determine the exact amount of error between the base build level and the upper build level, the difference in the process and cross-process dimensions may be taken (i.e., dy=y2-y1, dx=x2-x1).

[0070] Referring now to the example of FIG. 9C, another diagnostic column 232 is shown similar to that of FIGS. 9A and 9B, but is a more extreme example of a printer with errors in the process direction. In this example, drift in the process direction has accumulated to the extent that the upper cross-sectional shape of column 232 is absent / non-existent. In this example, in the base build layer 234, the cross-process dimension of column 232 is x1, and the process dimension of the column is y1. However, all build material in the upper build levels (e.g., similar to the height of columns 222 and 212 in FIGS. 9A and 9B) is missing / absent once all off the right boundary / edge 238. The shape of this edge 238 indicates that it was primarily in the process direction (y) since the edge was aligned in the cross-process direction (x). In other words, at the right edge 238, x2 is approximately the same as x1, but 、y2 is null. By observing the build level where y2 is null, a process direction error / drift equal to y2 may be recorded at that build level. If the difference between x2 and x1 is also measured at this build level, the difference in the cross-process direction may also be determined.

[0071] 10 illustrates a specific example of how the cross-sectional shape of a diagnostic column 200 can be used as a diagnostic feature indicative of process and cross-process drift in the printer 110 between the base build level and higher build levels. FIG. 10 is a top / plan view of a diagnostic column 200. The rectangular shape shown by the dotted lines indicates the dimensions of the diagnostic column 200 at the base build layer 202 (i.e., the base build level relative to the printer 110). The dimensions at the base build level are x z1 , y z1 The solid rectangular shape indicates the dimension of the diagnostic column 200 in the upper build layer 204 (i.e., the higher build level relative to the printer 110). The dimension at the higher build level is x z2 , y z2 To determine the printer 110 error at the higher build levels, the printer subtracts the dimension at the higher build level from the dimension at the base build level (i.e., dx=x z2 -x z1 , dy=y z2 -y z1 ). For example, x z1 = 20 mm, and y z1 = 20 mm 、 x z2 = 15 mm, and y z2Consider a situation where dx, dy = 8 mm. In this case, dx = 5 mm (i.e., dx = 20-15 = 5 mm) and dy = 12 mm (i.e., dy = 20-12 = 8 mm). Based on these measured differences, the diagnostic column 200 is used to determine the drift associated with each build level of the printer between the upper build level and the base build level. For example, in the above example, consider that the base build level is z = 1 and the upper build level is z = 50. In this example, there are 50 build levels. To determine the drift associated with each build level, each difference dx, dy is simply divided by 50 to arrive at the drift for each build level. In other words, Δx = dx / 50 = 5 mm / 50 levels = 0.1 mm / level and Δy = dy / 50 = 8 mm / 50 levels = 0.16 mm / level. 10, the printer drift / misalignment for each level of the printer may be determined by printing a diagnostic column, measuring the dimensions of the diagnostic column at the upper build levels and the base build level, calculating the associated difference in the process and cross-process directions, and then dividing the difference by the number of build levels. In various embodiments, this process may be repeated across multiple subsets of build levels (e.g., every 10 build levels), such that different amounts of drift are calculated in relation to different sets of build levels.

[0072] In at least one embodiment, instead of measuring the cross-sectional dimensions of the diagnostic columns at the base build level and at higher build levels, diagnostic features may be incorporated into the diagnostic columns such that a sensor recognizes a nominal error when the feature disappears. This is similar to tread depth marks on a tire, where the feature disappears when the error is comparable to the feature dimension. As an example of this, consider a configuration such as in Figures 11A-11C, where two flat / top diagnostic targets 250, 260 are shown, each with a diagnostic feature provided by a cross-sectional shape. In this example, the disappearance of one of the shapes indicates a predetermined amount of error. In Figures 11A-11C, the two diagnostic columns 250 and 260 each have a rectangular cross-sectional shape with the first diagnostic column 250 elongated in the cross-process direction and the second diagnostic column 260 elongated in the process direction. FIG. 11A shows a base layer of diagnostic columns, FIG. 11B shows a middle layer of diagnostic columns, and FIG. 11C shows a top layer of diagnostic columns (the base layer is shown in dotted lines in FIG. 11B and FIG. 11C for illustrative purposes). In the example of FIG. 11A-FIG. 11C, the printer has a first order error in the process direction and little or no error in the cross-process direction. Thus, as shown in FIG. 11B, in the middle layer, diagnostic columns 250 and 260 are both y s 11C , the first diagnostic column 250 is shifted to the right (i.e., in the process direction) by an amount equal to y . In this intermediate layer, the right edges 251, 261 of the diagnostic columns 250, 260 remain in a known position because the absence of support material to the right of these edges causes build material to fall to the right of the diagnostic columns and onto the platen at these locations. (As in FIG. 10 , the base layer boundaries are shown by dotted lines in FIGS. 11B and 11C .) At some upper build level, as shown in FIG. 11C , the error in the process direction is significant and prevents additional build material from being deposited on the first diagnostic column 250. Specifically, if the error in the process direction is greater than y , the first diagnostic column 250 is shifted to the right (i.e., in the process direction) by an amount equal to y . c(i.e., the original process direction dimension of the first diagnostic column 250), no additional build material is deposited on the first diagnostic column 250. Thus, as shown in FIG. 11C, by detecting the absence of additional build material at the upper build level, the system can determine that the process direction error at the upper build level is equal to y c Recognize that it is equal to.

[0073] 12A-12C, a similar configuration to that of FIGS. 11A-11C is shown, but in this embodiment, the printer has drift errors in both the process and cross-process directions. As shown in FIG. 12A, two diagnostic columns 250 and 260 extend along the cross-process and process directions, respectively. As such, they are configured to recognize nominal errors in both the cross-process and process directions based on loss of build material in the respective columns. Diagnostic column 250 is y c The diagnostic column 260 has a smaller dimension in the process direction equal to x c As shown in FIG. 12B, as the diagnostic columns are constructed, the cross-sectional dimension is reduced in the intermediate build layers in both the cross-process and process dimensions. As shown in FIG. 12C, in the first upper build layer, the material associated with diagnostic column 250 is absent, and the difference in the process direction is y c (i.e., dy = y c ). If the system then continues to build the second diagnostic column 260, the build material associated with this column 260 will eventually be absent (although this is not shown in the figure). At this build level where the build material for diagnostic column 260 is absent, the system will determine that the cross-process direction difference is x c (i.e., dx=x c12A-12C, the printer configuration is therefore configured such that the sensor forms a diagnostic column having a cross-sectional shape that recognizes the nominal error / difference in either the process or cross-process direction when a feature is missing.

[0074] In the example of Figures 12A-12C, the error in one direction may be so small that the disappearance of one of the diagnostic columns may be erroneously associated with a false error. For example, consider the configuration of Figures 11A-11C, where the error is primarily in the process direction. In this case, if the diagnostic column 260 continues to be built after the layer associated with Figure 11C until the build material of the column 260 is absent / missing at the upper level, the reason for such disappearance is due to a process direction error. If the system erroneously recognizes this disappearance as a cross-process direction error, the system also erroneously identifies a difference in the cross-process direction. Thus, in at least some embodiments, the system is configured to measure the rate of change of process and cross-process dimensions in different layers to determine whether the absence of a particular feature can reasonably be attributed to a particular type of error. For example, in the build level associated with Figure 12C, the system may measure the rate of change of the X-axis of the diagnostic column 260 relative to the remaining length of the diagnostic column 260. s2 The rate of change in the diagnostic column 260 at a particular future build level is X s2 It may be appreciated that it is sufficient to indicate a difference in the cross-process direction equal to

[0075] Another example of a printer configuration is shown in FIG. 13, which is configured to form a diagnostic column having a cross-sectional shape such that the sensor recognizes nominal errors / differences in both the process and cross-process directions. In this embodiment, four diagnostic columns are constructed, including two in the cross-process direction (identified as columns I and III) and two in the cross-process direction (identified as columns II and IV). Both columns I and II have dimensions "a" and "b", but in opposite directions (i.e., dimension "a" is in the process direction for column I but in the cross-process direction for column II, and "c" is in the cross-process direction for column I but in the process direction for column II). Similarly, both columns III and IV have dimensions "b" and "d", but in opposite directions. Further, the magnitudes of these dimensions are such that a < b < c < d. Similar to FIGS. 12A - 12C, the printer constructs columns I - IV until one of columns I or II disappears (i.e., the build material is absent at that build level). At that point, the system recognizes the following. - If column I disappears first, the error is a process direction error having a process direction difference corresponding to "a" at the build level of disappearance, - If column II disappears first, the error is a cross-process direction error having a difference corresponding to "a" at the build level of disappearance.

[0076] After column I or II has disappeared, the printer continues to build diagnostic columns until another column disappears. Depending on which column disappears next, the system can recognize the following. - If column I disappears first and column II disappears second, the error associated with the disappearance of column II is due to a cross-process error having a difference corresponding to "a" at that build level associated with the disappearance of column II, - if column I is lost first and column III is lost second, the error associated with the loss of column III continues to be a process direction error and a determination of the cross-process direction error cannot be made based on this loss of column III; - if column II is lost first and column I is lost second, the error associated with the loss of column I is due to a process error having a difference equivalent to "a" at that build level associated with the loss of column I; If column II is lost first and column IV is lost second, the error associated with the loss of column IV remains a cross-process error and no determination of the process direction error can be made based on this loss of column IV.

[0077] The order of column disappearance described above may provide useful insight for the printer to provide the best possible accuracy for the printer. For example, when columns I and II are the first two columns to disappear, the printer's incremental shifts (Δx and Δy) for each build level may be determined using the method described below. Alternatively, when either columns III and IV disappear before one of columns I or II, it may be determined that (i) the error in one direction is minimal enough that an incremental shift is not required, or (ii) the error in the opposite direction is large enough that instructions should be provided to the user to perform the printer's mechanical setup again in an attempt to minimize the error by mechanical means.

[0078] The above examples of Figures 11A-13 are just a few configurations of diagnostic objects having features that can be used to determine the appropriate coordinate shift for each build level of the printer. As previously mentioned, various shapes and sizes of diagnostic objects may be used in numerous configurations to provide the user with different types of information regarding drift / error between build levels. Furthermore, in at least one alternative embodiment, both build and support material may be used when printing the diagnostic object. In such an embodiment, material jetted off the column boundaries can be supported by the support material, which allows the object to maintain a dimensionally accurate cross-section by simply being shifted in the x and / or y directions.

[0079] FIG. 14 is a block diagram illustrating a method 1400 of operating a three-dimensional object printer in view of the disclosure of FIGS. 8A-13. As shown in block 1402, the method includes operating at least one ejector of an ejector head to eject a material drop toward a platen to form a diagnostic shape of a diagnostic object at a base build level, the diagnostic shape having a process direction dimension and a cross-process direction dimension at the base build level (see, e.g., FIGS. 10 and 12A). As shown in block 1404, the method further includes moving the ejector head relative to the base build level along the build axis such that the ejector head moves further away from the base build level. Next, as also shown in block 1404, the method includes operating at least one ejector of the ejector head to build a diagnostic shape of the diagnostic object at a higher build level along the build axis (see, e.g., FIGS. 10 and 12B). Moving to block 1406, the method continues by sensing the process direction dimension and the cross-process direction dimension of the diagnostic feature at the upper build level. As shown in block 1408, the method then proceeds by determining a difference (dx, dy) between at least one dimension at the base build level and the upper build level (see, e.g., FIG. 10 and FIG. 12C). This difference is similar to that determined in connection with the method of FIG. 4 and the acts of block 408 described above. In this embodiment, the edge of the diagnostic object at the upper build level may actually be considered the reference edge because the position of this edge is known (i.e., because the dimension at the base build level is known, and because the build material along one edge of the diagnostic object falls off the edge with error at each successive build level, and the true position of the edge at the upper level is known and serves as a reference). The method then continues at block 410 by determining an ejector head shift (Δx, Δy) for each incremental number of accumulation levels between the base build level and the upper build level based at least in part on the determined difference.After determining the ejector head shift, the method continues in block 412 by moving the ejector head and operating at least one ejector to eject material droplets toward the platen to form the three-dimensional object, where moving the ejector head includes shifting the ejector head relative to the platen by the determined ejector head shift for each of a plurality of incremental build levels before ejecting material droplets at the incremental build level. Blocks 410 and 412 involve the performance of similar acts as those described above in relation to FIG. 4 in blocks 412 and 414.

[0080] The above-described embodiments are just a few exemplary embodiments of a 3D object printing system with compensation for misalignment. It will be understood that variations of the above-disclosed and other features and functions, or alternatives thereof, may be desirably combined into many other different systems, applications, or methods. Moreover, various presently unanticipated or unforeseen alternatives, modifications, variations, or improvements may be subsequently made by those skilled in the art, which are also intended to be encompassed by the following claims.

Claims

1. 1. A method of operating a three-dimensional object printer, comprising: operating at least one ejector of the ejector head to eject at least one first droplet of material toward a first build surface provided on a platen at a first build level along a build axis; operating a sensor to measure a first distance component of the first drop of material from a first fiducial reference on the first build surface, the first distance component including a first process direction component along a process axis and a first cross-process direction component along a cross-process axis; moving the ejector head relative to the platen along the build axis; placing a gauge block on the platen, the gauge block providing a second build surface at a second build level along the build axis; operating the at least one ejector of the ejector head to eject at least one second drop of material toward the second build surface; operating the sensor to measure a second distance component of the second drop of material from a second fiducial reference on the second build surface, the second distance component including a second process direction component and a second cross-process direction component; calculating a difference between the first distance component and the second distance component, the calculated difference including a process direction difference and a cross-process direction difference; determining an ejector head shift relative to the platen for a plurality of incremental build levels along the build axis between the first build level and the second build level, the determined ejector head shift for each of the incremental build levels being based on the calculated difference; and operating the at least one ejector to move the ejector head and eject a droplet of material toward the platen to form a three-dimensional object, wherein moving the ejector head includes shifting the ejector head relative to the platen by the determined ejector head shift for each of the plurality of incremental build levels before ejecting a droplet of material at the incremental build level.

2. The method of claim 1 , wherein placing the gauge block on the platen includes aligning registration features on the platen with complementary features on the gauge block.

3. The method of claim 1 , wherein moving the ejector head relative to the platen along the build axis comprises moving the platen along the build axis.

4. The method of claim 1 , wherein moving the ejector head relative to the platen along the build axis comprises moving the gantry along the build axis with the ejector head positioned on a gantry.

5. 2. The method of claim 1, wherein determining a shift of an ejector head relative to the platen for multiple incremental build levels along the build axis comprises: determining a shift of an ejector head along the process axis based on the process orientation difference; and determining a shift of an ejector head along the cross-process axis based on the cross-process orientation difference.

6. 2. The method of claim 1, wherein determining a shift of the ejector head includes inputting the calculated difference into a printer control system, the printer control system being configured to (i) determine a pixel adjustment for each of the incremental build levels based on the calculated difference, and (ii) shift the ejector head by the determined pixel adjustment for each of the incremental build levels.

7. 2. The method of claim 1, wherein determining a shift of the ejector head includes inputting the calculated difference into a printer control system, the printer control system being configured to (i) determine a shift amount for each of the incremental build levels based on the calculated difference, and (ii) shift a position of the ejector head on a gantry or shift a position of the platen by the determined shift amount for each of the incremental build levels.

8. placing a second gauge block on the platen, the second gauge block providing a third build surface at a third build level along the build axis; operating the at least one ejector of the ejector head to eject at least one third drop of material toward the third build surface; operating the sensor to measure a third distance component of the third drop of material from a third fiducial reference on the third build surface, the third distance component including a third process direction component and a third cross-process direction component; calculating a difference between the second distance component and the third distance component; 2. The method of claim 1, further comprising: determining a shift of an ejector head relative to the platen for a second plurality of incremental build levels along the build axis between the second build level and the third build level based on the calculated difference between the second distance component and the third distance component.

9. 1. A method of operating a three-dimensional object printer, comprising: operating at least one ejector of an ejector head to eject a drop of material toward a platen to form a diagnostic shape of a diagnostic object at a base build level, the diagnostic shape having a process direction dimension and a cross-process direction dimension at the base build level; moving the ejector head relative to the base build level along a build axis such that the ejector head moves further away from the base build level; operating the at least one ejector of the ejector head to build the diagnostic shape of the diagnostic object at an upper build level along the build axis; sensing process direction and cross-process direction dimensions of the diagnostic feature at the upper build level; determining a difference in at least one dimension between the base build level and the upper build level; determining an ejector head shift for each of a number of incremental build levels between the base build level and the upper build level based at least in part on the determined difference; after determining the shift of the ejector head, moving the ejector head and operating the at least one ejector to eject a drop of material toward the platen to form a three-dimensional object, wherein moving the ejector head includes shifting the ejector head relative to the platen by the determined shift of the ejector head for each of the incremental build levels before ejecting a drop of material at the incremental build level; determining the difference in at least one dimension at the base build level and the upper build level includes sensing an absence of the diagnostic object at the upper build level such that the determined difference is the same as the dimension at the base build level; the diagnostic object is a first diagnostic object having a first diagnostic shape, the process direction dimension of the first diagnostic shape being greater than the cross-process direction dimension at the base build level; The method further comprising: operating the at least one ejector to build a second diagnostic object, the second diagnostic object including a second diagnostic shape having a process direction dimension at the base build level that is less than a cross-process direction dimension; sensing the absence of one of the first and second diagnostic objects at one of the incremental build levels; determining a difference equivalent to the process direction dimension or the cross-process direction dimension of the first or second diagnostic object at the base build level based on an absence of one of the first and second diagnostic objects at one of the incremental build levels. method.

10. 10. The method of claim 9, wherein the diagnostic object is a column and the diagnostic shape is a rectangular cross-sectional shape of the column.

11. The method of claim 9 , wherein determining the shift of the ejector head for each of the incremental build levels comprises dividing the determined difference by a number of the incremental build levels.

12. 2. The method of claim 1, wherein the process axis is perpendicular to the cross-process axis, the build axis is perpendicular to both the process axis and the cross-process axis, and the platen is planar, the planar surface being perpendicular to the build axis.

Citation Information

Patent Citations

  • System and apparatus for evaluating inkjet performance and alignment in direct-to-object printer

    JP2019104239A

  • Device, system and method for calibration in three-dimensional model printing

    US20060111807A1