Processing

Please wait...

Settings

Settings

Goto Application

1. WO2020228962 - DETERMINING A SPATIAL DISTRIBUTION OF PHASES IN A PRODUCT TO BE MANUFACTURED BY ADDITIVE MANUFACTURING

Note: Text based on automatic Optical Character Recognition processes. Please use the PDF version for legal matters

[ EN ]

Description

Determining a spatial distribution of phases in a product to be manufactured by additive manufacturing

The invention concerns a method for determining a spatial distribution of phases of a product to be manufactured by ad ditive manufacturing. The invention further concerns a method for manufacturing a product with a target phase composition and a product comprising a spatial distribution of phases ac cording to the method.

Selective Laser Melting (SLM) is an example for an additive manufacturing technique, which uses a laser to locally melt metal powder, creating complex 3D bodies (the product) layer by layer. Such techniques operate by locally melting materi als and include Selective Laser Melting (SLM) , Electron Beam Melting (EBM) , Direct Energy Deposition (DED) or Wire-Arc Ad ditive Manufacturing (WAAM) and the like.

During the manufacturing process, heat is introduced into the product by the laser beam, additional heat can also be ap plied from a heated base plate. Since the process of manufac turing of a typical product is time consuming, the individual temperature history of the material in different areas of the product can be very different. Areas located on top of the product are exposed to higher temperatures, since the heat provided by the energy beam has a longer path to the base plate, as compared to lower positions. The material of the top layers of the product also is younger as compared to the bottom layers, since it is solidified in later stages of the build, so the time of its internal heat treatment is much shorter .

In other words, different positions in the product typically undergo a different heat treatment. It is well-known, that some alloys are strongly depended on heat treatment condi tions, due to build-up of different microstructural phases, which affect the material properties like hardness, yield stress, etc.

The problem addressed by this invention is to improve the prediction of spatial distribution of phases in a product manufactured by additive manufacturing techniques. Further the invention addresses the problem of manufacturing a prod uct with a predetermined target spatial distribution of phas es .

The problem is solved by a computer implemented method for determining a spatial distribution of phases in a product to be manufactured by additive manufacturing. The method com prises the steps of:

- determining one or more areas of the product,

- determining a temperature progression for each of the areas and

- determining the spatial distribution of phases by correlat ing the temperature progression to the phase in each of the areas .

The spatial distribution of phases in the product can be built from values for phase content in the areas. The phase in an area can be a single value, a single averaged value a set of values for a phase content progression for a direction in such an area. The set of phase content in the areas for multiple different areas represents the spatial distribution of phases in the product. The spatial distribution of phases can be evaluated and/or verified and can be used in further engineering and or production steps.

The temperature progressions show temperature data (e.g. in Kelvin) over a determined span of time (e.g. in Seconds) .

For additive manufacturing the product is usually divided in to layers, wherein the product is built layer by layer. The areas can be a single point, e.g. representing a single ele ment in a FEM-mesh. Areas can also be combinations of multi- pie points, polygons, polyhedra or elements, each within a single layer or expanding over adjacent layers. The step of determining areas can comprise selecting multiple areas, preferably in multiple layers.

The spatial distribution of phases in the product can be ob tained with the alloy composition of the product. The alloy composition of each area can be determined separately if nec essary. The determined temperature progressions, e.g. cooling curves, can be fed into a microstructural simulation code yielding the microstructural picture after the end of build for each point of interest.

In a further embodiment determining the areas comprises de termining at least one area for a subset of layers of the product. Multiple layers can be grouped into subsets of lay ers, e.g. with similar properties. It has shown to be advan tageous, e.g. for processing time, to limit the amount of de termined areas by selecting subsets of layers and a limited amount of areas for each subset. In many cases selecting one area for one subset of layers is sufficient.

In a further embodiment determining the areas comprises de termining one area for at least one layer of the product. It also can be beneficial to select one area for one layer, e.g. when the layer has a high complexity compared to the adjacent layers .

In a further embodiment determining the areas comprises de termining multiple areas for one layer of the product. For example, layers can comprise multiple structures that are not connected to each other within this layer or are only con nected via connections with a low material thickness it can be reasonable to select multiple areas in the layer to achieve a meaningful spatial distribution of phases in the layer and eventually in the product.

In a further embodiment the method comprises a step of providing a geometric model of the product, wherein determin ing the areas is at least partly based on the geometric mod el. A geometric model can be provided as a 3D-CAD model of the product. Additionally, or as an alternative to a layer-based selection of the areas, a geometric model-based selec tion of the areas can be conducted. If the geometric model has anticipated weaknesses, e.g. due to constructive re strictions, the phase composition of those weaknesses can be essential. Therefore, determining the areas based on the geo metric model can further improve the significance of the spa tial distribution of the phases.

In a further embodiment a strain and/or stress profile of the product is provided, wherein determining the areas is at least partly based on the strain and/or stress profile of the product. The strain and/or stress profile can be combined with a geometric model and can be combined with a material (alloy) distribution within the product. Considering the strain and/or stress profile of the product can indicate are as with the need for a certain phase composition in the prod uct and therefore further improves the significance of the spatial distribution of the phases.

In a further embodiment the temperature progressions are de termined by macroscale thermal simulation of the areas and/or by macroscale thermal simulation of certain layers of the product. Determining the temperature progressions can be based on cooling curves based on a series of quasi static thermal simulations conducted layer by layer for each layer during buildup of the product. To reduce the overall necessi ty for large numbers of simulations, only the selected areas can be determined. It is also possible that the determination of temperature profiles is prepared in a CAD/CAE stage for the product to be manufactured by additive manufacturing in a previous step. Quasi static is to be understood as static relatively to the short period of time required for solidifi- cation compared long time for the further cooldown and tem pering processes.

To determine a (macro scale) thermal distribution for every step passed during the process of manufacturing the product, thermal calculation can be performed for a pre-set number of heights. The height directly corresponds to the number of layers/slices that already have been applied during the manu facturing process. For example, thermal distributions for every layer, for subsets of layers, for layers with areas can be conducted. For each height, which represents a time point in the overall product build, a thermal distribution can be calculated. The time point of each height corresponds to the time of its solidification, and can be calculated using the part geometry, the process parameters (scanning speed and hatch distance) as well as the machine-depended idle times (skywriting and recoating times) . It has proven to be very efficient to use a boundary condition that the layer itself (or the sublayers respectively) have a quasi-static state. With the layer by layer build, the temperature-progression of the product itself can be adequately reproduced.

In a further embodiment effects with a duration of under one second are averaged out of the temperature progression data. Therefore, effects with duration of milliseconds, such as the solidification process of the melting pool are not considered in detail. Effects with a duration over one second, especial ly over 10 seconds can be considered. A time scale of the temperature progression is selected assuming a temporarily averaged quasi-stationary value for the temperature. The tem perature progression data can be provided until the prod uct/the current layer and its adjacent layers reach a target temperature, e.g. an ambient temperature, such as room tem perature .

In a further embodiment the temperature progressions comprise at least temperature data until a martensite finish tempera ture (also called "martensite finish" MF) of the area. To en- sure a complete view of the phase content of the area the martensite finish temperature of the alloy used in the area can be used as end point or as trigger point for reaching an end in a determined span of time, such as 10s, minutes or even hours after falling below the martensite finish tempera ture. The temperature progressions therefore comprise data from solidification, martensite start temperature down to a martensite finish temperature.

In a further embodiment the temperature progressions comprise temperature data of a post build heat treatment of the area and/or of the product. Post build is the time span after the product has been built by the additive manufacturing process and cools down. The post build heat treatment can be an ac tive temperature treatment with a controlled temperature pro file or a passive cooling down to a controlled ambient tem perature. Phase content of the product and/or the areas is significantly influenced by the heat treatment conditions af ter the actual solidification. Therefore, it has proven to be advantageous to include the period of time where a heat treatment of the whole product or single areas is conducted. It is also possible to decide whether a heat treatment is necessary based on a spatial distribution of phases (SP) in the product before a heat treatment has been performed.

In a further embodiment correlating the temperature progres sion to the phase is at least partly based on a CCT- and/or TTT-diagram. Continuous cooling transformation diagrams (CCT) , isothermal transformation diagrams, also known as time-temperature-transformation diagrams (TTT) indicate which phase content has formed due to the temperature conditions present in the areas. It has proven, that using CCT- and/or TTT-diagrams can be an efficient way to determine the phase content. Combinations with further models of phase content are possible.

In a further embodiment the temperature progression for each of the areas comprise at least temperature data beginning from a solidification of a melt pool in the area. It has proven to be efficient when the data starts after solidifica tion of the melt pool of the alloy present in the area. This has the advantage that simplifications, such as averaging on a larger scale, can be used. It is possible to assume solidi fication when the beam has moved past the area. In this case the time scale can be chosen accordingly.

In a further embodiment the temperature progressions comprise at least 100 seconds, 1000 seconds, lh or longer of tempera ture data. Wherein the temperature data comprises post build heat treatment. The data preferably begins after solidifica tion of a melt pool. When a predefined time scale is chosen for all areas the comparability increases. Areas that are built in a later stage of the additive manufacturing process of the product due to their location in higher layers com prise less data and therefore the temperature of points which have not been yet created is undefined. In other words, high er points will have a shorter thermal history, as compared to lower points in the part. This can be considered in determin ing the spatial distribution of phases. For each stage of the layer by layer build, in other words for each height of the product, the temperature of the determined areas of interest can be recorded.

In a further embodiment the temperature the temperature pro gressions are derived from a thermal build calculation of the product. The temperature of the areas can therefore be ex tracted from existing simulations, further enhancing the ef ficiency of the current method.

At the end of the build, the whole product will cool down, so in a further embodiment a temperature fall-off to ambient (e.g. room) temperature can be determined for the temperature progressions. For each point of interest, a complete cooling curve can be recorded.

The problem of manufacturing a product with a target spatial distribution of phases by additive manufacturing comprising is solved by a method comprising the steps of:

- defining the target spatial distribution of phases,

- determining, with a first set of parameters, the spatial distribution of phases of the product by a method according to the invention,

- optionally determining the step of simulating with a second set of parameters until the target spatial distribution of phases is achieved and

- manufacturing the product with the determined parameter settings .

In other words, a target phase composition can be provided by an e.g. engineering tool, considering e.g. stress, strain and further parameters. The method according to the invention can be carried out multiple times until a spatial distribution of phases is found, that comes closest to the specification of the target spatial distribution of phases. This process can be automatically performed. The obtained parameters can be reused for future products that have similar structures.

The problem further is solved by a product comprising a spa tial distribution of phases determined by a method according to the invention.

The invention is now described and explained in more detail referring to the exemplary embodiments shown in the figures. Shown are in:

FIG 1 a schematic overview of an embodiment of the inven tion,

FIG 2 determining areas of a product and a macro scale

thermal simulation of a geometric model of the prod uct,

FIG 3 determining temperature progressions for each of the areas ,

FIG 4 a possible flowchart of the method, and

FIG 5 a flow chart of a method for manufacturing a product .

The illustration in the drawings is in schematic form. It is noted that in different figures, similar or identical elements may be provided with the same reference signs.

FIG 1 shows a schematic overview of an embodiment of the invention. In this view a 2D model serves as a geometric model GM of a product 100. A step of determining S2 one or more areas A1,...,A4 of the product 100 is indicated. A first area A1 is determined in the narrow column 102 which is on top of a base 101 of the product 100. A second area A2 is determined in the area where the narrow column 102 changes thickness into a triangular shape, a pyramid element 103. The change of thickness could indicate an interesting location for an area A1,...,A4. A third area A3 is determined in the lower third of the upper pyramid element 103 of the product 100. A fourth area A4 is determined on the very top of the product 100, possibly indicating a last layer of the product and consequently the end of the manufacturing process. For each of the determined areas A1,...,A4 a temperature progression T1,...,T4 is determined in a further step S3. The temperature progressions can be stored in an intermediate format, which is suitable to store temperature data with different time lengths. Schematically determining S4 a spatial distribution of phases SP by correlating the temperature progressions T1,...,T4 to the phase P1,...,P4 in each of the areas A1,...,A4. Therefore the phase content of each area A1,...,A4 is known. The combination of the phases P1,...,P4 in each of the areas A1,...,A4 with their different locations within the product 100 allows the spatial distribution of phases SP to indicate the distribution of phases that is reached with a certain manufacturing process with certain parameters. This allows the engineering to be more concise and a first quality assessment can be conducted before producing the product 100 for the first time. Parts of the product 100, which are non-critical can be ignored by not determining S2 an area A1,...,A4 in the non-critical parts of the product 100 and therefore saving processing time.

FIG 2 shows an embodiment of the step of determining S2 one or more areas A1,...,A4 of a product 100 in greater detail.

This embodiment is based on a macro scale thermal simulation of a geometric model GM, in this case a 3D model of the 2D model known from FIG 1, during the build of the product 100. The shown macro scale thermal simulation depicts the progress of the layer by layer build of the product 100 in three phas es P1,P2,P3, wherein in each of the phases P1,P2,P3 multiple layers are added. The areas with higher temperature are de picted with a higher density of dots than the areas with a lower temperature. The areas that are not dotted yet have not been built/simulated yet.

The first phase PI shows the product 100, already comprising the base 101 and the narrow column 102, the build just begin ning to manufacture the pyramid element 103 on top of the product 100. The temperature distribution within the part of the product 100 just finished with completion of the first phase PI shows an already cooler base 101 with temperature rising to the top at the upper end of the narrow column 102. Temperature differences within single layers Ll,...,Ln are not part of this simulation but can be taken into consideration by dividing layers into sublayers. It is possible to divide only specific layers into sublayers, wherein other layers can be simulated as a whole. A first area A1 has been determined in the lower part of the narrow column 102. A second area A2 has been determined at the connection between the pyramid el ement 103 and the narrow column 102. It is possible to simu late temperature distributions within layers of a product 100, especially when simulating a product that is not as sym metrical as the one currently shown.

In the second phase P2 the pyramid element 103 on top of the narrow column 102 of the product 100 is in process of being built. A third area A3 has been determined about halfway into the pyramid element 103. As can be seen, the areas A1,A2,A3 do not necessarily have to be determined in already simulated layers. All areas A1,...,A4 can be determined before any thermal simulation. Areas A1,...,A4 can also be determined during or after thermal simulation. After thermal simulation areas A1,...,A4 can be determined for parts of the product 100 which comprise time series of thermal data.

In the third phase P3 the product 100 has been built up to the third area A3, which at this point of time has the highest temperature. It is shown, that a fourth area A4 has been determined at the outer edge of the pyramid element 103. This is an example for an area A4 not covering a whole layer of the product 100.

To sum up the thermal calculations, a quasi-static (macroscale) thermal calculation is performed for a pre-set number of heights, corresponding to the number of layers. For each height, which represents a time point in the overall part build, a thermal distribution within the part of the product 100 is calculated. The first time point of each height corresponds to the time of its solidification, and can be calculated using the geometric model GM, the process parameters (e.g. scanning speed and hatch distance) as well as the machine-depended idle times (e.g. skywriting and recoating times) .

FIG 3 shows determining S3 temperature progressions T1,...,T4 for each of the areas A1,...,A4. A visual representation of the temperature progressions T1,...,T4 is shown on the right side, wherein each area A1,...,A4 is allocated to a corresponding temperature progressions T1,...,T4. A simulation SIM can be conducted as shown and described in FIG 2. Further a thermal macro scale analysis TM and a build time calculator BT can be used to create the temperature progressions T1,...,T4. If known products 100 are to be produced and e.g. an alloy composition has to be changed preexisting thermal data can be used or slightly amended.

The first phase PI and the third phase P3 of the build as known from FIG 2 are shown. The determined temperature data of the temperature progressions T1,...,T4 starts at the point of time the layers that contain the respective area A1,...,A4 have actually been simulated. During manufacturing, areas A1,...,A4 which have not been manufactured yet therefore do not have an own temperature yet. This leads to the temperature progressions T1,...,T4 having different lengths, e.g. as shown the data of the second temperature progression T2 starting at phase PI and the data of the third temperature progression T3 starting at phase P3. Over time the temperature progressions converge to a constant temperature value, since the product 100 will eventually cool down to ambient temperature. As an additional, optional step, the temperature progression

T1,...,T4 of the build can be complemented by additional thermal curves, representing a subsequent heat treatment. When parts of the product 100 are tempered after manufacturing, the temperature progressions T1,...,T4 can diverge again but will converge when ambient temperature is reached. Since the phase content might be different after the part build, the subsequent heat treatment might or might not lead to different consequences in the heat-treated state, depending on the alloy used in the areas A1,...,A4 of the product 100.

FIG 4 shows a possible flowchart of the method according to the present invention. The order shown in the flowchart is to be seen as an example and can be changed. Some steps may be skipped or can be performed by different entities such as cloud computing entities or local entities on a PC with a simulation hardware and simulation software. All steps can be performed on a single local PC or can be performed remote by a single server or distributed cloud hardware.

In an example the method can be performed in the following order. The step of providing S10 a geometric model GM of the product 100 provides the basis for the engineer to analyze the product 100. Additionally or optionally the step of providing S12 a strain and/or stress profile of the product 100 can be performed before, after or in parallel to step providing S10 a geometric model GM. The strain and/or stress profile and the geometric model GM can be obtained from a single data source, such as an CAE and/or CAD system, as one model or as multiple models. The step of determining S2 one or more areas A1,...,A4 of the product 100 can then be performed on basis of the strain and/or stress profile and the geometric model GM or on basis of empirical values, older similar projects or simply experience. Following the step of determining S3 a temperature progression T1,...,T4 for each of the areas A1,...,A4 can be performed by simulating each of the layers built in a quasi-static manner. Next the determining S4 the spatial distribution of phases SP by correlating the temperature progression T1,...,T4 to the phase P1,...,P4 in each of the areas A1,...,A4 is conducted and the spatial distribution of phases SP is available for further processing, such as quality management/assurance, production itself or further design and/or engineering.

The advantage of the method as proposed is the ability to predict phase composition in different points of the production process. This can be either utilized to ensure, that the material quality and properties will be adequate (or equal) within the whole part or use the fact of locally different thermal histories to design parts with locally different properties (e.g. hard/ductile regions). The construction of temperature progressions, e.g. cooling curves, based on a series of quasi-static thermal simulations and the combination of these curves with thermo-dynamical microstructural codes has proven to be of great advantage.

FIG 5 shows a flow chart of a method for manufacturing a product 100 with a target spatial distribution of phases TSP by additive manufacturing comprising the following steps in an exemplary order.

A first step of defining MSI the target spatial distribution of phases TSP is conducted. For example, a desired spatial distribution of phases can be selected as a target - the tar get spatial distribution of phases TSP. The desired spatial distribution of phases could be dependent on the intended use of the product 100. In a second step of determining MS2 a spatial distribution of phases of the product 100 is deter mined with a first set of parameters P101 by a method accord ing to the invention, e.g. according to the embodiment of FIG 4. When the spatial distribution of phases with the first set of parameters P101 is not sufficiently close to the target, the step of determining MS2 can optionally be repeated one or more times with a second set of parameters P102 in a step of repeating MS25 until the target spatial distribution of phas es TSP is achieved.

A step of comparing MS20 can be conducted to compare the spa tial distribution of phases with the first set of parameters P101 and/or the spatial distribution of phases with the sec ond set of parameters P102 with the target spatial distribu tion of phases TSP.

If comparing MS20 shows that the spatial distribution of phases currently compared with the target spatial distribu tion of phases TSP is sufficiently close to the target, the currently compared set of parameters P101, P102 can be stored and used in a further step. Otherwise the method can be re peated until the target is reached or reconsiderations of the overall product design must be made when the deviations re main too large.

The last step is manufacturing MS4 the product 100 with the determined parameters P101, P102. The parameters P101, P102 can be process parameters (e.g. scanning speed and hatch dis tance) as well as the machine-depended idle times (e.g. sky writing and recoating times) as well as material dependent parameters (e.g. alloy composition/powder metal composition) as well as tempering parameters during manufacturing and post build heat treatment. The geometric design of the product 100 also can influence the process parameters P101, P102.

In summary the invention concerns a method for determining a spatial distribution of phases SP of a product 100 to be manufactured by additive manufacturing. To improve the prediction of spatial distribution of phases SP in the product 100 manufactured by additive manufacturing techniques the proposed method comprises the following steps:

- determining S2 one or more areas A1,...,A4 of the product

100,

- determining S3 a temperature progression T1,...,T4 for each of the areas A1,...,A4 and

- determining S4 the spatial distribution of phases SP by correlating the temperature progression T1,...,T4 to the phase P1,...,P4 in each of the areas A1,...,A4. Furthermore, the invention addresses the problem of manufacturing a product 100 with a predetermined target spatial distribution TSP of phases. The invention further concerns a method for manufacturing a product 100 with a target phase composition and a product 100 comprising a spatial distribution of phases SP according to the method.