Additive manufacturing of core-shell structures

Abstract

A method of manufacturing a core shell alloy structure may include scanning one of a core or shell area of the alloy structure to obtain a first microstructure; scanning another of the core or shell area inset area to obtain a second microstructure; and rescanning the entire area to obtain the shell completely surrounding the core. The shell may have a smaller grain structure than the core. The core may be a single crystal.

Description

ADDITIVE MANUFACTURING OF CORE-SHELL STRUCTURESThis application claims priority of provisional application no. 63 / 527,727, filed July 19, 2023, the entire contents of which is hereby incorporated by reference.TECHNICAL FIELD

[0001] Embodiments are generally related to additive manufacturing of core-shell structures by laser scanning of alloys to produce a dual microstructure for improved processability and properties in additive manufactured parts.BACKGROUND

[0002] In recent years there has been a great deal of interest in manufacturing single crystal (SX) parts such as turbine blades utilizing additive manufacturing (AM), a modified form of 3-D printing. A diagram of the additive manufacturing process is shown in FIG. 1 .

[0003] In FIG. 1 a powder feeder 110 deposits powder 112 to produce a powder layer 105 in a powder bed 108. The powder layer that is deposited first can lie directly on the powder bed or on a substrate that may be placed in the powder bed before the powder layers are deposited. A beam scanner 101 can move a beam source 102 or steer an energy beam 1 11 that is produced by the beam source. The energy beam may be a laser beam, an electron beam, etc. The energy beam 111 produces a melt pool 104 where the energy beam 1 1 1 melts some of the powder in the powder bed 108. The melt pool 104 has a melt pool depth 103. The melt pool depth 103 is illustrated as being large enough to also melt some of the patterned layer 106 directly underneath the powder layer 105. The beam scanner moves the melt pool 104 in a path through the topmost powder layer to selectively melt some of the powder and thereby produce a bottom patterned layer. The powder layer that is deposited first becomes the bottom patterned layer 107. A 3D object is printed by iteratively depositinga powder layer and using the energy beam to melt a pattern into that powder layer, thereby producing a patterned layer.

[0004] FIG. 2 is a high level conceptual diagram that illustrates a scan pattern 200 with a melt pool 104 and a hatch distance 203, according to some aspects. The scan pattern 200 can include a large number of scan lines 201 that may be parallel to one another as shown in FIG. 3. The hatch distance 203 is the distance between the scan lines 201. The hatch direction is the direction from a previous scan line to a subsequent scan line in the scan pattern. The hatch direction 204 can be perpendicular to the scan lines 201. The beam scanner moves the melt pool 104 along the scan pattern. For simplicity, the melt pool is shown as a circle where the energy beam 11 1 meets the powder layer and melts the powder at that location. In practice, the melt pool is much longer because the energy beam melts material in a pattern and melted material remains melted for a short period of time.

[0005] That is, laser powder bed fusion AM of metals such as steel and nickel superalloys is conducted in a protective chamber with dynamic inert gas, e.g., argon or nitrogen, flow to avoid oxidation or gas dissolution. To melt specific locations, a laser beam or electron beam is scanned on top of the powder bed with different focal size and power. The absorption of laser beam leads to melting of the powder in the ith layer and also some remelting of the (i-1 )th solid layer. The dynamics of melting and the melt-pool shapes are dictated by the scan strategy in the x-y plane. As the building progresses in the z-direction with pre-determined geometrical cross-sections, one may observe the evolution of porosity, cracks, solidification grain structure, solid-state transformations, and plastic deformations. In turn, all the above physical phenomena lead to spatial variations of microstructure and properties.

[0006] Due to the complex heat history of laser AM and the high Ti and Al content of the y’ phase in nickel superalloys, there are several mechanisms that lead to cracking, pitting and deformation. There are three most likely cracking mechanisms, i.e. solidification cracking, liquation cracking and ductility-dip cracking.

[0007] Generally, solidification cracking occurs in the last period of solidification. The sensitive zone is on the front of as-solidified dendrites, which is named as a mushy region. If the thermal shrinkage strain on the solidified grain boundaries is significant, the residual liquid of mushy region, which always distributes along the solidified boundaries as continuous film, is unable to adapt to the shrinkage strain, therefore cracking occurs.

[0008] Liquation cracking always occurs in the heat affected zone (HAZ), when the nominal solidification finished. It can be attributed to the appearance of liquid film when the bulk material is heated to some extent. Two basic mechanisms, namely, segregation and penetration mechanisms are proposed to explain the appearance of liquid film. In the segregation mechanism, some alloy elements or impurity aggregate along the grain boundaries, thus reduce the actual melting temperature in these areas and facilitate the appearance of continuous or discontinuous liquid film. The segregation mechanism is always suitable for solution-strengthened single-phase alloy. Whereas, for the alloys with precipitated phases such as solute-rich precipitates, carbides and intermetallics, the penetration mechanism is more reasonable.

[0009] Ductility-dip cracking is most likely to happen at a temperature ranging from 0.7 to 0.5 times of the alloy's melting point, at which the alloy always suffers a quick drop of ductility, and it belongs to solid phase cracking. In the additive manufacture of nickel superalloys, it can be observed that these ductility-dip cracks have wedge-shaped crack tips, and locate at the triple junction point of the neighboring grain boundaries or in the bottom of the liquation cracks which have already formed.

[0010] These phenomena are exacerbated when using AM to manufacture core-shell structures. These materials are highly susceptible to the formation of internal defects, for example micro-cracking. Parameters that allow the part to be densified during heat treatment often do not result in the most favorable microstructure.

[0011] Accordingly, there is a need in the art for the AM of core-shell structures with minimized defects such as cracking, pitting, porosity and deformations.BRIEF SUMMARY

[0012] The following summary is provided to facilitate an understanding of some of the innovative features unique to the embodiments disclosed and is not intended to be a full description. A full appreciation of the various aspects of the embodiments can be gained by taking the entire specification, claims, drawings, and abstract as a whole.

[0013] It is, therefore, one aspect of the disclosed embodiments to manufacture a coreshell alloy structure that includes scanning an entire area of the alloy structure to obtain a first microstructure; scanning an insert area within the entire area to obtain a second microstructure; and rescanning the entire area to obtain the shell completely surrounding the core. The scanning may be performed with a laser beam or an electron beam. The shell may have a smaller grain structure than the core. The core may be a single crystal. The core-shell alloy structure may heat treated with hot isostatic pressing, so that surface cracks are reduced. The alloy can be a Co-superalloy, an Fe-superalloy or a Ni-superalloy. The Ni- superalloy may have a y’ or [3 phase. The Ni-superalloy may be selected from IN738LC, IN792, CM247LC, MAR-M247, N4, N5, 125, CMSX-2 and CMSX-4.

[0014] Another aspect of the disclosed embodiments is to manufacture a core-shell alloy structure that includes scanning one of a core or shell area of the alloy structure to obtain a first microstructure; scanning another of the core or shell area inset area to obtain a second microstructure; and rescanning the entire area to obtain the shell completely surrounding the core. The scanning may be performed with a laser beam or an electron beam. The shell may have a smaller grin structure than the core. The core may be a single crystal. The core-shell alloy structure may heat treated with hot isostatic pressing, so that surface cracks are reduced. The alloy can be a Co-superalloy, an Fe-superalloy or a Ni-superalloy. The Ni- superalloy may have a y’ or [3 phase. The Ni-superalloy may be selected from IN738LC, IN792, CM247LC, MAR-M247, N4, N5, 125, CMSX-2 and CMSX-4.

[0015] A further aspect of the disclosed embodiments to manufacture a core-shell alloy structure that includes scanning an insert area of the alloy structure to obtain a first microstructure; scanning an area around the insert to obtain a second microstructure; and rescanning an entire area to obtain the shell completely surrounding the core. The core-shell alloy structure may be heat treated with hot isostatic pressing, whereby surface cracks arereduced.BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying figures, in which like reference numerals refer to identical or functionally similar elements throughout the separate views and which are incorporated in and form a part of the specification, further illustrate the embodiments and, together with the detailed description, serve to explain the embodiments disclosed herein.

[0017] FIG. 1 depicts the formation of a single crystal workpiece by additive manufacturing.

[0018] FIG. 2 depicts the scan geometry of additive manufacturing

[0019] FIG. 3 depicts two methods for scanning a core-shell part according to an embodiment of the disclosure.

[0020] FIG. 4 depicts an exploded view of scanning a core-shell part according to an embodiment of the disclosure.

[0021] FIG. 5 depicts a finished part according to an embodiment of the disclosure.

[0022] FIG. 6 depicts the core-shell structure of a part having an opening according to an embodiment of the disclosure.

[0023] FIG. 7 is a schematic showing crack structures in the surface and interior of an alloy according to an embodiment of the disclosure.

[0024] FIG. 8 shows the grain structure of the core and the shell of the alloy according to an embodiment of the disclosure.

[0025] FIG. 9 depicts A) grain size effect on creep rate (lower rate is better), B) effect of crystal orientation on modulus (lower modulus is better), C) microstructure used in polycrystalline turbine blades.

[0026] FIG. 10 is a photograph of a part with a core-shell structure according to an embodiment of the disclosure.DETAILED DESCRIPTION

[0027] The particular values and configurations discussed in the following non-limiting examples can be varied and are cited merely to illustrate one or more embodiments and are not intended to limit the scope thereof.

[0028] Parts manufactured using conventional casting techniques typically have large grain sizes that are beneficial for time-dependent properties such as creep, while having lower strength and fatigue resistance. This can be addressed by adding coatings to mold walls that encourage the nucleation of fine grains at the surface of a part during casting, while allowing for coarser grains within the core.

[0029] While techniques must be used in conventional casting to encourage fine grain formation in a cast part, additive manufactured parts suffer from the opposite problem. Additive manufactured parts often have smaller grains compared to cast parts and display good low temperature ductility, low temperature tensile strength, and dynamic strength. However, the small grains result in poorer time-dependent properties such as creep strength. Similar to cast parts, there is a desire to have fine grains near the surface and large grains in the core to achieve the desired combination of properties. However, in additive manufacturing, it is the large grains in the core that are more difficult to achieve.

[0030] When additive manufacturing (AM) materials that are highly susceptible to the formation of internal defects (for example: microcracking), parameters that allow the part to be densified during heat treatment (for example: Hot Isostatic Pressing or HIP) create a microstructure with small grains. HIP compresses materials by applying high temperature of several hundreds to 2000 °C and isostatic pressure of several tens to 200 MPa at the same time. Argon is the most commonly used pressure medium. Alternately, nitrogen and in few cases air or hydrogen can be used. With HIP, the sample will contract while keeping its initial shape until pores inside disappear, and bond together due to diffusion effects. With nickel- based alloys, HIP is performed at about 1100 to 1400 °C at a pressure of about 100 to 150 MPa. HIP treatment to castings can improve creep fracture lifetime by 1 .3 to 3.5 times.

[0031] One approach to forming large grains entail forming a melt pool in a first powder layer along a first layer scan pattern that includes a number of scan lines, where the scan lines are parallel and spaced by a hatch distance along a hatch direction that is perpendicularto the scan lines If a dendrite grows as the melt pool cools, the scan pattern promotes the dendrite growth direction orthogonal to the scan lines, and the hatch distance discourages the dendrite growth direction that has a directional component that is opposite the hatch direction.

[0032] However, when using this previous technology on a crack susceptible material such as CM247LC (a Nil-based superalloy), long surface-connected microcracks are present in the microstructure that do not allow for densification of the part with HIP. Therefore, there is a desire to print with parameters that form larger grains in the core of a part due to the beneficial properties they confer, while also printing a shell region around the part that allows the internal defects in the core to close with HIP. Materials with this type of crack susceptibility include Ni-superalloys (IN738LC, IN792, CM247LC, MAR-M247, N4, N5, CMSX-2, CMSX- 4), Co-superalloys, and Fe-superalloys. Ni-superalloys have an austenitic matrix structure. Many of these contain molybdenum and / or tungsten, which are soluble in the matrix and act as solid-solution strengtheners. In nickel-base superalloys, y' and y" lead to precipitation strengthening; The y' phase, is identified as Nia(AI,Ti) with its crystal structure as ordered f.c.c of L12-type ordering and is responsible for antiphase boundary (APB) strengthening. This implies that these superalloys are heat-treated for precipitation hardening by way of solutionizing followed by aging. The strength of y' phase increases with increasing temperature and so the control of aging temperature is important. A typical composition of these type of Ni-superalloys (Rene N5 or Rene 125) can be found in Table 1.Table 1. Typical Alloy Compositions.

[0033] Crack closure during heat treatment can occur (as one example) during HIP, in which high temperature and high pressure is applied for a duration of time to soften and compress the material. However, if a crack or defect is exposed to the surface of the part, the high-pressure atmosphere can penetrate into the defect and prevent it from closing. Therefore, surface connected cracks are not closable during common HIP treatments.

[0034] The scanning method of the disclosure addresses these concerns.

[0035] FIG. 3 shows two methods for scanning a core-shell part 300. One method scans the entire part area 310 first, and then rescans an inset area (core) 320 within the first area to obtain a different microstructure. A second approach to scanning a shell part 330 scans the core 340 and shell 350 separately. Either one can be scanned first. There may be any amount of overlap (or no overlap) between the shell and core. Finally, the entire part may be scanned to have the shell completely surrounding the core.

[0036] In either case, any upskin or downskin regions may be scanned with the parameters used for the shell, so that the core is fully enclosed in 3 dimensions. For example, the top and bottom few layers of a part (could be one layer, but more likely more than one) will not have a core-shell structure. It will be purely shell so that the part has a shell surrounding it in 3D. The thickness of the shell, either at the top, bottom, or sides, is not defined and can be as thin or as thick as needed.

[0037] Referring to the image in FIG. 4, the part is built layer by layer from the bottom. The first few layers (m layers) 410 and last few layers (p layers) 420 are scanned with the “shell” process parameters to enclose the core of the part. Layers m and p can be any number of layers equal to or greater than 1 . Typically m=p=25 in some parts, or m=p=750 in other parts, or m=500 and p=75 in others. So, m does not have to equal p, but it can, and they technically have no maximum limit.

[0038] The region in the middle (the n layers) 430 are scanned in two steps using twopotential approaches. This corresponds to the innage in FIG. 3 and its description.

[0039] Step 1 ) The entire layer is scanned with shell parameters, and then the core is scanned with core parameters. This essentially overwrites the part of the layer in the core that was first scanned with the shell parameters. This does not work backwards. One cannot scan the core first, and then scan the entire layer with the shell parameters, since this would overwrite the part of the layer in the core that was first scanned with core parameters.

[0040] Step 2) The core is scanned first and then the shell, or the shell is scanned first and then the core. Since they are scanned separately, either one can be done first and nothing is overwritten.

[0041] After either of these two approaches are used to scan a layer, powder is recoated onto the build chamber and the next layer is scanned. A diagram of a finished part 500 showing the core 510 completely surrounded by the shell 520 is in FIG. 5.

[0042] A more complicated situation is with the geometry shown in FIG. 6. FIG. 6 shows a part 600 with a through-hole or opening 610 exposed to the external atmosphere. The cross sectional view shows the core 620 surrounded by an outer shell 630 and an inner shell 640. Despite this geometry, the process is similar to that depicted in FIG 4, where there is no central opening.

[0043] In the scanning method of the disclosure, the shell is scanned with parameters that allow for defects to be easily closed with heat treatment. These parameters may, for example, result in a microstructure that has small grains and shallow melt pools, which results in small cracks that are disconnected from the surface. When used to scan the shell, this disconnects the defects in the core from the surface as well and allows them to heal during HIP or other heat treatments. A schematic of this is shown in FIG. 7. With this method, process parameters for the core can be chosen such that a more desirable microstructure (grain size, texture, etc.) is obtained after heat treatment.

[0044] As can be seen in FIG. 7, a large surface crack 710 is not closable with heat treatment. In the core-shell configuration, the crack 710 is no longer connected or in the vicinity of the surface. Small cracks 720 in the shell region 730 are closable using a heat treatment such as HIP. The large cracks 710 are isolated in the core region 740. The small cracks have a length of less than about 500 pm and can usually range from about 10 pm to 100 pm. The large cracks can be greater than 0.5 mm and can be several mm in length.

[0045] The crack sizes are material dependent - some materials crack more than others. In the core, crack lengths have no limit. The cracks can technically span the entire width of the core. In practice, cracks can have a length of up to about 4 mm. In the shell, the crack length should be limited to less than the thickness of the shell and ideally less than half the thickness of the shell. This is because one wants to isolate the core cracks from the outside of the part, and this gives a greater assurance that the cracks in the shell don't connect the external atmosphere to the cracks in the core. In practice, crack lengths are up to about 0.5 mm in the shell, but shells can be printed that are about 1 mm in thickness. But it should be noted that some materials crack more than others. HIP may be used to reduce crack and grain size in the shell.

[0046] The grain size depends on how the material is heat treated after printing. Without being limited, the grain size in the core should - at the very least - be two times larger than in the shell. The maximum grain size in the center is essentially the entire size of the core, which is true in the case of a core that is a single crystal.

[0047] The crystalline structure of the core and shell can be either polycrystalline (PX) or single crystal (SX). The structure can also be directional, i.e., uniaxial, by promoting dendrite formation from the melt pool as the part is scanned. In one embodiment, the core can be SX while the shell is PX. In another embodiment, both core and shell can be PX with the PX grains of the shell being larger than the PX grains of the core.

[0048] There are a number of phases that can be found in superalloys, including the desirable y’ or p fraction. These phases are summarized in Table 2.Table 2. Superalloy Phases.

[0049] An example of some preferred microstructures allowed by this disclosure are shown in FIG. 8. The shell has small grains 810, while the core has either A) several large grains with random orientations (achieved before or after a recrystallization heat treatment) or B) a highly textured microstructure composed of one or several grains. The main purpose of the disclosure was shown FIGs. 3 to 6, but the results are shown in FIG. 8, which shows microstructure in the shell and (small grains 810) and potential microstructures in the core (single crystal, highly textured, large grains 820). Specifically, the fine-grained shell is a fortunate and unexpected outcome. The size of the photographs in FIG. 8 is 1 mm x 1 mm.

[0050] While the small grains sizes that additive manufactured parts tend to form give an advantage in fatigue life, in some applications it is preferable to trade some of that fatigue performance for improved time-dependent properties (creep, dwell-fatigue, etc.). One of the possible results of this is that the disclosure creates components with process parameters that create larger grain sizes using crack-susceptible materials that otherwise would not be possible.

[0051] These larger grain sizes are required for improved creep properties, as shown inFIG. 9a. At small grain sizes the mechanisms that result in creep (i.e., grain boundary slippage) are more prevalent throughout the material and creep occurs at a faster rate. The specific grain diameters where rapid vs. slow creep occurs are material dependent, and the example in this image is for an austenitic steel.

[0052] The technology of the disclosure yields beneficial properties if the process parameters used in the core result in a highly textured or single crystal material. As FIG. 9b shows, some crystallographic orientations exhibit lower modulus. The result is that when a strain is applied to a part, it experiences a lower stress than it would if it had a random texture. This microstructure is useful for the full range of properties (tensile strength, fatigue, creep), and the process parameters that allow it to form can suffer from cracking in crack-susceptible materials. Printing an outer shell allows one to heat treat the part and close those cracks.

[0053] FIG. 9c shows the microstructure used in polycrystalline turbine blades. The fine grain zone near the treated surface is intended to improve the thermal fatigue behavior. The fine grain zone is located on the untreated art surface.

[0054] A photograph of a finished part according to an embodiment of the disclosure can be seen in FIG. 10. Because of the way a part is scanned with a shell encapsulating the core in 3D, a fully, properly printed part will have no visible indication that it has a core scanned with different parameters. The technology of the disclosure can scan some parts and then stop halfway through the build, such that an “n layer” was exposed on the top surface, as can be seen in FIG. 10. In this example, the shell is about 1 mm thick, and is uniform around the surfaces of the visible layer. The shell does not necessarily have to be the same thickness everywhere. One can print shells on the side of the part as thin as about 0.1 mm and as thick as about 5 mm. The minimum shell thickness would be the width of a melt pool and the maximum is half the width of the part, just based on geometry limitations. Much thicker shells can be formed on the bottom and top of a part, ranging from about 1 mm to 30 mm. There is no reason one couldn’t go thinner or thicker, if desired. The minimum is one layer thickness, which is often between about 0.02 mm and 0.1 mm. The maximum is half the height of the part, again for geometry reasons.

[0055] The end result is a microstructure that resembles that commonly used in aerospace components, where a part with a large grained core is surrounded by a fine-grained region on the outside. This provides good creep properties while improving the fatigue properties incomparison to a part that was solely composed of large grains.

Claims

CLAIMSWhat is claimed is:1 . A method of manufacturing a core-shell alloy structure, comprising: scanning an entire area of the alloy structure to obtain a first microstructure; scanning an insert area within the entire area to obtain a second microstructure; and rescanning the entire area to obtain the shell completely surrounding the core.

2. The method of claim 1 , wherein the scanning is performed with a laser beam or an electron beam.

3. The method of claim 1 , wherein the shell has a smaller grain structure than the core.

4. The method of claim 1 , wherein the core comprises a single crystal.

5. The method of claim 1 , wherein the core-shell alloy structure is heat treated with hot isostatic pressing, whereby surface cracks are reduced.

6. The method of claim 1 , wherein the alloy is a Co-superalloy or an Fe-superalloy.

7. The method of claim 1 , wherein the alloy is a Ni-superalloy.

8. The method of claim 7, wherein the nickel superalloy has a y’ phase or a p phase.

9. The method of claim 7, wherein the nickel superalloy is selected from the group consisting of IN738LC, IN792, CM247LC, MAR-M247, N4, N5, 125, CMSX-2 and CMSX-4.

10. A method of manufacturing a core-shell alloy structure, comprising: scanning one of a core or shell area of the alloy structure to obtain a first microstructure;scanning another of the core or shell area inset area to obtain a second microstructure; and rescanning the entire area to obtain the shell completely surrounding the core.1 1 . The method of claim 10, wherein the scanning is performed with a laser beam or an electron beam.

12. The method of claim 10, wherein the shell has a smaller grain structure than the core.

13. The method of claim 10, wherein the core comprises a single crystal.

14. The method of claim 10, wherein the core-shell alloy structure is heat treated with hot isostatic pressing, whereby surface cracks are reduced.

15. The method of claim 10, wherein the alloy is a Co-superalloy or an Fe-superalloy.

16. The method of claim 10, wherein the alloy is a nickel superalloy.

17. The method of claim 16, wherein the nickel superalloy has a y’ phase.

18. The method of claim 16, wherein the nickel superalloy is selected from the group consisting of IN738LC, IN792, CM247LC, MAR-M247, N4, N5, 125, CMSX-2 and CMSX-4.

19. A method of manufacturing a core-shell alloy structure, comprising: scanning an insert area of the alloy structure to obtain a first microstructure; scanning an area around the insert to obtain a second microstructure; and rescanning an entire area to obtain the shell completely surrounding the core.

20. The method of claim 19, wherein the core-shell alloy structure is heat treated with hot isostatic pressing, whereby surface cracks are reduced.

Images