Method of and apparatus for additive layer manufacture

Abstract

A method of additive layer manufacture includes the steps of successively depositing layers of fusible powder material in overlying relationship and introducing energy, such as by an electron beam, into each deposited layer to selectively melt material in the layer so as to fuse the melted material together and to already fused material of a layer thereunder in order to produce a three-dimensional solid article in successive cross-sectional layers. In order to monitor layer quality, structured light defining a fringe pattern is projected onto each deposited layer before and/or after melting of material in that layer and the fringe pattern on each layer is imaged from a perspective different from that of the projection so as to reveal disturbance of the pattern by topographical features of the layer.

Claims

1. Method of additive layer manufacture comprising the steps of successively depositing layers of fusible powder material in overlying relationship, introducing energy into each deposited layer to selectively melt powder material in the layer so as to fuse the melted material together and to already fused material of a layer thereunder for production of a three-dimensional solid article in successive cross-sectional layers, projecting structured light defining a fringe pattern onto each layer in both a time before and a time after the melting of material in that layer, imaging each fringe pattern on each layer from a perspective different from that of the projection so as to reveal disturbance of the pattern by topographical features of the layer, analysing the imaged fringe pattern with respect to any such revealed disturbance to recognise defects in the respective layer of powder material when the structured light was projected onto that layer before the melting of material therein and defects in a cross-sectional layer of the article formed from the respective powder material layer when the structured light was projected onto that layer after the melting of material therein and correctively influencing layer deposition in the case of recognition of defects in the powder material layer and energy introduction in the case of recognition of defects in the article cross-sectional layer.

2. A method according to claim 1, wherein the step of analysing to recognise defects in the powder material layer comprises recognition of local variations in depth of the layer.

3. A method according to claim 2, wherein the recognised local variations in the depth of the powder material layer include depressions in the form of undulations in the layer or voids or areas of missing powder in the layer.

4. A method according to claim 2, wherein the recognised local variations in the depth of the powder material layer include elevations in the form of undulations in the layer or displacements due to protrusions into the layer from below.

5. A method according to claim 1, wherein the step of analysing to recognise defects in the article cross-sectional layer comprises recognition of local variations in depth of the cross-sectional layer.

6. A method according to claim 5, wherein the recognised local variations in the depth of the cross-sectional layer include depressions caused by voids or areas of missing powder in the powder material layer from which the cross-sectional layer was formed or by generation of the cross-sectional layer otherwise than in accordance with specification.

7. A method according to claim 5, wherein the recognised local variations in the depth of the cross-sectional layer include elevations caused by swelling of the article during production in a direction perpendicular to the cross-sectional layer or by generation of the cross-sectional layer otherwise than in accordance with specification.

8. A method according to claim 2, wherein the step of analysing comprises identifying recognised local variations by assignment of Cartesian co-ordinates.

9. A method according to claim 2, wherein the step of analysing comprises subjecting the imaged fringe pattern to a fringe pattern algorithm to obtain a wrapped phase, subjecting the wrapped phase to a phase unwrapping algorithm to obtain an unwrapped phase giving true phase and processing the true phase by a phase-to-height model to give an elevation height or depression depth.

10. A method according to claim 1, wherein the step of analysing to recognise defects in the article cross-sectional layer comprises recognition of departure of the shape of the melt area from a predetermined shape.

11. A method according to claim 1, wherein the step of correctively influencing layer deposition comprises vibrating a powder material layer to resettle the powder material thereof, repetition of smoothing of the top surface of a powder material layer, adding material to or removing material from a powder material layer and replacing a powder material layer.

12. A method according to claim 1, wherein the step of correctively influencing layer deposition is carried out on at least one of the current powder material layer and a subsequent powder material layer.

13. A method according to claim 1, wherein the step of correctively influencing energy introduction comprises at least partial remelting of material in the article cross-sectional layer.

14. A method according to claim 1, wherein the step of projecting comprises forming the fringe pattern to be substantially coincident with the area of the powder bed layer in which selective melting is carried out.

15. A method according to claim 1, wherein the steps of projecting and imaging are carried out with use of dedicated optics dissociated from the energy introduction.

16. A method according to claim 1, wherein the step of introduction of energy into each powder material layer comprises scanning the layer by an electron beam.

Description

(1) FIG. 1 is a schematic elevation, partly in block diagram form, of additive layer manufacturing apparatus embodying the present invention;

(2) FIG. 1A is a diagram, to substantially enlarged scale, of a fringe pattern produced by structured light projection in the apparatus;

(3) FIG. 1B is a diagram based on that of FIG. 1A, but showing an image of the pattern from a perspective different from that of the projection; and

(4) FIG. 2 is flow chart showing the steps of a method exemplifying the invention and performable by the apparatus of FIG. 1.

DETAILED DESCRIPTION OF THE ILLUSTRATED EMBODIMENTS

(5) Referring now to the drawings there is shown in FIG. 1 additive layer manufacturing apparatus 10 in an exemplifying form comprising a housing 11 bounding a vacuum chamber 12, the housing having the form of a pressure vessel able to be evacuated to provide a desired sub-atmospheric pressure. Manufacturing is carried out in the chamber 12 on an incrementally lowerable table 13 for supporting an article 14 or artefact to be manufactured by an additive layer or 3D printing process. In this process, a fusible powder material 15, such as a selected metal, usually a metal alloy, is distributed from diagrammatically represented controllable dispensers 16 onto the table 13 and adjoining regions and is spread by a similarly diagrammatically represented controllable spreading and levelling element 17, such as a blade (shown here) or roller, to form a thin powder material layer 15a of uniform depth on the table. The powder material 15 of the layer 15a is then selectively melted and fused by the action of heat in a predefined area. The powder material 15 in that area forms, after solidification, a cross-sectional layer of the article 14 perpendicular to the plane of the drawing. Individual cross-sectional layers of the article 14 formed by fusion of powder material layers prior to and thus below the illustrated layer 15a are denoted by 14a and shown, like the powder material layer 15a, with exaggerated thickness or depth. Residual powder material 15 surrounding the melt area and ultimately the article 14 as a whole plays no further part in the process, but remains as part of the powder bed until recovered and recycled or discarded. After formation of each such cross-sectional layer 14a of the article 14 the table 13 is lowered by the cross-sectional layer depth or thickness, which has a predetermined value, and a further layer 15a of powder material 15 is distributed over the solidified part of the preceding layer to form—by selective melting and fusing together as well as to the fused material thereunder over the same or a different predefined area—as next cross-sectional layer 14a of the article 14. The process is repeated until construction of the article 14 on a layer-by-layer basis is completed.

(6) The heat for melting the powder material 15 to induce fusion can be supplied by, in particular, a laser beam or a charged particle beam, in the case of this apparatus by an electron beam 18 generated and transmitted by an electron beam column 19 mounted on the housing 11. The column 19 comprises a controllable electron beam generating unit 20 containing an electron source, generally a cathode of electron emissive material capable of emitting electrons under applied voltage, and a controllable beam focussing, blanking and deflecting unit 21 comprising, for example, a series of lens for focussing the electrons for propagation as a beam of defined cross-sectional size and shape along the axis of the column 19 and a continuation of that axis, a blanking aperture for effectively switching the beam on and off, and a deflector for deflecting the beam relative to the extended axis. The deflector is operable under programmed computer control in such a way as to cause the beam 18 to scan the instantaneous powder material layer 15a at high speed to melt the powder 15 in an area corresponding with a desired shape of a cross-sectional layer 14a of the article 14 as described in the preceding paragraph. The area to be scanned is generally finely resolved into multiple fields scanned one at a time along a predefined path within each field to melt the powder, either on a single visit or preferably by progressive heat build-up from several visits. Individual control units 16a, 17a, 20a and 21a are provided to control, respectively, the powder material dispensers 16, the spreading and levelling element 17, the beam generating unit 20 and the beam focusing, blanking and deflecting unit 21. The control units are shown as individual items merely by way of convenient example and in practice may be realised by hardware and/or software.

(7) Additive layer manufacture by this process is generally known and has been described only to the extent necessary for general understanding of the construction and operation of the exemplifying form of apparatus 10 used for such a purpose.

(8) As outlined in the introduction, a particular challenge to successful performance of additive layer manufacturing is represented by various faults which may appear at the stage of deposition of each powder layer 15a and at the stage of selective melting of the deposited layer. In order to address these issues, the apparatus 10 comprises a system capable of optically monitoring each powder material layer 15a deposited on the table 13 and the article cross-sectional layer 14a produced therefrom, analysing the results of the optical monitoring to recognise aberrations in the manufacturing process and/or to assure continuing adherence to manufacturing specifications, and to carry out—when and where necessary—corrective action via components involved in the manufacturing process, in particular a process with the monitoring and feedback steps described below with reference to FIG. 2.

(9) For this purpose, the apparatus shown in FIG. 1 includes a non-contact measurement system comprising an optical projector 22 for projecting structured light on an optical path A through the vacuum chamber 12 via a window (not illustrated) to define a fringe pattern on each deposited powder material layer 15a. Projection is carried out before beam action on the layer for melting constituent powder material, after such beam action or, as in the case of the described embodiment of the apparatus 10, both before and after beam action. The fringe pattern, which is generated by computer software, can take various forms, but expediently consists of an array of regularly spaced stripes 23 of contrasting tone (fringes), an example of such a projected fringe pattern being shown in FIG. 1A in the form of a small segment, to greatly enlarged scale, of the pattern. As the path A of the projection beam in FIG. 1 evidences, the incident fringe pattern is substantially coterminous with the area of the powder material layer 15a in which melting is carried out to form a cross-sectional layer 14a of the article 14, thus the whole of the general zone of scanning action of the electron beam 18 on the powder material 15 in each layer 15a.

(10) The measurement system further includes an optical imaging camera 24 for imaging, the projected fringe pattern on an optical path B through the vacuum chamber 12 via a further window (also not illustrated), the imaged pattern being shown in FIG. 1B with, by comparison with FIG. 1A, an obliquity signifying the angle of the optical imaging path B relative to the optical projection path A. Since the imaging undertaken from a perspective different from that of the projection, as represented by the relatively angled optical paths A and B, topographical features, in particular elevations and depressions, in the area of the layer 14a or 15a onto which the fringe pattern was projected manifest themselves as disturbances in or distortions of the individual stripes 23 of the imaged fringe pattern. Such disturbances are reproduced, as shown in FIG. 1B, in the form of local elevations 23a and depressions 23b in specific stripes. The elevations 23a in the case of the unmelted powder material layer 15a may be formed by the crests of undulations in the layer, agglomerations of particles, displacements due to protrusions from an underlying layer and other such faults, and the depressions 23b may be formed by the valleys of such undulations, voids, areas of missing powder, scores due to particle drag by or damage of the spreading and levelling element 17 and various other issues. Similarly, elevations 23a in the case of the article cross-sectional layer 14a produced by selective melting of material of the powder material layer 15a may be caused by swelling of the article 14 during production, faults in scanning or in other aspects of beam action in departure from the prescribed article design parameters, and other problems. Depressions 23b in the article cross-sectional layer 14a may arise from pre-existing faults with the powder material layer 15a, such as voids and missing areas of powder, and again from departures from design parameters, to mention just the principal causes.

(11) Recognition of defects of these kinds is achieved by analysis of each imaged fringe pattern by an analysing unit 25 employing analysis techniques as described below in connection with FIG. 2. The analysing unit 25 produces first data which characterise, faults in the powder material layer 15a prior to beam action and second data which characterise faults in the article cross-sectional layer 14a evident after beam action, the first data being supplied to and processed by a first processor 26 for determining corrective action in relation to the powder disposition in the powder material layer 15a and the second data being supplied to and processed by a second processor 27 for determining corrective action in relation to the operation of the beam 18. Accordingly, the first processor 26 generates and issues encoded instructions, insofar as they are needed, for corrective action to the control units 16a, 17a for the powder material dispensers 16 and the spreading and levelling element 17 and the second processor 27 generates and issues encoded instructions, insofar as they are needed, for corrective action to the control units 20a, 21a for the beam generating unit 20 and the beam focusing, blanking and deflecting unit 21. These corrective actions may be additional to the usual control actions performed by the control units 16a, 17a, 20a and 21a, as represented by the illustrated further inputs to the units for usual control variables, or may, appropriate, be provided by suitable modification of those variables. It is to be understood that, like the control units, the illustrated processors 26, 27 are merely graphical representations of functional capabilities of the apparatus and in practice the processors could be realised by, for example, part of the software of a general control system.

(12) The corrective action able to be undertaken by the spreading and levelling element 17 under the control of its control unit 17a includes repeat smoothing of the surface of the powder material layer 15a or repeat spreading of the layer, whereas the dispensers 16 can be controlled by the respective control unit 16a to discharge material for addition to the layer 15a or even complete replacement of the layer. If the apparatus 10 includes a facility for imparting vibration to the layer to resettle the constituent powder material 15 this can also be carried out under the control of one of the control units 16a, 17a or a further such control unit. Similarly, the corrective action able to be undertaken by way of the beam generating unit 20 under the control of its control unit 20a includes adjustment of beam power and hence the energy introducible into the layer 15a to melt the powder material 15, whereas the corrective action able to be exercised by the beam focussing, blanking and deflecting unit 21 under the control of the respective control unit 21a includes change in beam spot size, change in dwell or exposure time during scanning, variation of scanning rate, repeat scanning of selected areas and various other possibilities connected with beam operation.

(13) It should be noted that the projector 22 is arranged as closely as possible to the column axis, the axis being the ideal axis for the optical path A. Compensation for any influence on the projected fringe pattern by the offset of the axis of the path A from the column axis can be included in the analysis undertaken by the analysing unit 25.

(14) FIG. 2 shows, in the form of a flow chart, the steps of an exemplifying method performable by the apparatus 10 of FIG. 1 or other such apparatus with equivalent or otherwise appropriate features; in the following, the method is described specifically on the basis of performance by the apparatus 10 of FIG. 1. In a first step S1 a layer of fusible metallic powder material is deposited on the table, in an uppermost position thereof, by co-ordinated operation of the dispensers and the spreading and levelling element, the former to discharge the powder from reserves into a region adjacent to the table and the latter to spread the powder over the table in a layer of uniform depth and with a level or smooth surface. As already mentioned, in practice the formed layer may exhibit various faults or departures from the ideal, such as voids, areas of missing powder, scores, lines, hollows, troughs, pits, bumps, mounds, ridges and generally uneven or irregular depth, as well as possibly an overall depth differing, in positive or negative sense from a prescribed depth. Accordingly, prior to action—in this embodiment by an electron beam—on the powder material layer for selective melting of a predefined area in correspondence with an intended cross sectional layer of the article to be produced, structured light defining a fringe pattern with stripes of alternating contrast as shown in exemplifying form in FIG. 1A is projected onto the layer in a step S2. This is followed by a step S3 in which the pattern is imaged from a perspective different from that of the projection to reveal disturbances in the linearity of the pattern stripes due to local variations in the layer depth, specifically elevations and depressions (cf. FIG. 1B) in the layer surface.

(15) The image is then analysed in a step S4 to assign X and Y co-ordinates to the revealed disturbances and recognise the specific faults responsible for the disturbances. Analysis can be carried out by known procedures, taking into account that in the case of the disturbances such as shown in FIG. 1B the height of a particular elevation may be encoded as a function of the spatial phase of the fringe pattern (φ), i.e. spacing of the fringe pattern stripes, projected onto the layer surface. Consequently, the geometry of the elevation will modulate the intensity distribution of the fringe pattern, i.e. contrast of the pattern stripes, as expressed by the equation:
g(x,y)=a(x,y)+b(x,y)cos(2πf.sub.0x+φ(x,y))
in which a(x,y) is the background illumination, b(x,y) is the amplitude modulation of the fringe pattern stripes, f.sub.0 is the spatial frequency carrier, φ(x,y) is the phase modulation of the fringe pattern stripes (the required phase distribution) and x and y are the sample indices for the X and Y axes, respectively.

(16) The important aspect of the analysis to recognise faults represented by the disturbances in the fringe pattern image is extraction of the phase from the intensity distribution, as given by the equation. Various fringe pattern techniques have been proposed for demodulation of fringe patterns, such as phase stepping procedures, Fourier fringe analysis (ETA), direct phase detection (DPD), wavelet transform fringe analysis and other algorithms. These techniques produce a wrapped phase rather than the required phase; consequently, phase unwrapping algorithms are also required to recover the true phase from the wrapped phase. Finally, the calculated phase difference, which gives relative height information, is converted into absolute height by use of a reliable height calibration model. The steps of fringe pattern analysis can accordingly be summarised as subjecting the imaged fringe pattern g(x,y) to an appropriate fringe pattern algorithm of one of the mentioned kinds to obtain the wrapped phase ψ(x,y), which is then subjected to a phase unwrapping algorithm to obtain the unwrapped phase φ(x,y). Finally, processing of this by a phase-to-height model gives the absolute height h(x,y) of an individual elevation, on the basis of which the presence and location of a fault is recognised and from that the need for layer correction.

(17) Depending on the nature of a recognised fault a decision can be made in step S5 whether correction of the powder dispersion in the current layer is needed, which will be the case particularly if a fault located within the area of powder material to be melted is liable to impact on the integrity of the article cross-sectional layer to be produced by the selective melting. If correction of the current layer is required, this is undertaken in a succeeding step S6, in which, for example, the spreading and levelling element 17 is operated to redistribute, relevel or even partially remove powder making up the layer, or to reconstitute the layer with the addition of supplementary powder from the dispensers 16. If the apparatus includes equipment for vibrating the powder bed containing the layer the equipment can be operated to gently shake the layer and in that way resettle the powder to eliminate small elevations.

(18) If a particular fault is recognised, but is determined—such as by reference to fault thresholds—to be of a kind not requiring, immediate rectification, a decision can be made in a step S7 to carry out correction in the next layer and this is scheduled, in a step S8, for implementation at the appropriate time. Accordingly, in step S8 suitable adjustments can be made, for example to future pass speed or number of passes of the spreading and levelling element or to future aspects of the discharge of powder from the dispensers, the powder composition and the powder flowability, so as to avoid repetition of a fault of a kind which may be non-critical in the current layer, but potentially critical in subsequent layers, especially if there is a cumulative effect.

(19) When the appropriate actions have been carried out in the step S6 and/or step S8 or if no such actions are needed the method can proceed to a step S9 in which selective melting of powder in the originally laid or improved layer is carried out by electron beam scanning under control of a program determining the powder area to be melted so as to form the respective shape of the article cross-sectional layer, in a step S10, structured light projection to superimpose the fringe pattern on the layer, now containing an area of molten material, is carried out analogously to step S2 and the pattern is then imaged in a step S11 analogously to step S3 to reveal disturbances in the pattern, but now disturbances representing faults connected with melting, for example an area erroneously missed from melting and having a height signifying unmelted powder rather than a melt pool. In similar manner to step S4, the obtained image is now subjected to analysis in a step S12 to identify whether layer correction is needed and, if so, decisions are made in steps S13 and S15, analogously to steps S5 and S7, as to whether correction is required to the current layer and/or to the next layer. If correction is required to the current layer then in a step S14, for example, the material is remelted to at least some extent to remove the fault or faults in the current layer and if correction is required to the next layer then, for example, the beam generation and/or transmission control is scheduled in step S16 for adjustment to change such factors as beam power, beam scanning (deflection) rate, beam spot size or even shape, beam dwell time, number of visits to each site (melt point) and so forth so that the fault does not or should not recur.

(20) After the action or actions in step S14 and/or step S16 has or have been undertaken or if no action is required the method proceeds to a step S17 in which, after lowering of the table by the depth of the article cross-sectional layer, the next powder layer is deposited on top of the first layer as the first step in creating the next cross-sectional layer in layer-by-layer manufacture. In a step S18, the steps S2 to S16 are repeated in relation to that next layer and subsequently after each deposition, through repetition of step S17, of a fresh powder layer.

(21) Although the example of the method described with reference to FIG. 2 includes structured light projection and pattern imaging to enable layer analysis both before and after selective melting of layer, it is entirely possible, depending on requirements, to undertake projection, imaging and analysis only before or only after the melting phase.

(22) The described apparatus embodying and method exemplifying the present invention enable continuous optical monitoring of critical aspects of the additive layer manufacturing process and determination of corrective actions in real time, which may allow significant improvement in the quality of articles produced by the process or otherwise ensure that a desired standard of quality is maintained.