WO2018007770A2

WO2018007770A2 - Additive manufacturing method comprising removal of material from between two layers

Info

Publication number: WO2018007770A2
Application number: PCT/FR2017/051861
Authority: WO
Inventors: Jean-Philippe ROZENBAUM; Olivier Martin
Original assignee: Mecachrome France
Priority date: 2016-07-08
Filing date: 2017-07-07
Publication date: 2018-01-11
Also published as: FR3053632A1; FR3053632B1; WO2018007770A3; EP3481570A2

Abstract

The invention relates to a method for the additive manufacturing of a preform of a metal object comprising a stack of layers, based on a digital object, said method comprising successive steps of producing a layer (21 to 24), in which metal material is added locally to a metal support (1), combined with the local supply of energy. The method comprises a step of defining and saving a marker surface (10) for the stack and at least one step of removing material, carried out under dry conditions, from between two successive layers n and n+1 of the stack, said step consisting in removing the material on the upper face of layer n in order to create a reference surface for the deposition of layer n+1, the distance between the reference surface and the marker surface being substantially constant.

Description

ADDITIVE MANUFACTURING METHOD WITH REMOVAL OF MATERIAL BETWEEN TWO LAYERS

The present invention relates to the technical field of additive manufacturing.

Classically, this term denotes the set of methods for making by adding material, layer by layer, a physical object from a previously defined digital object.

These additive manufacturing processes are the opposite of material removal or subtractive processes, which make it possible to obtain metal objects from thick plates or hollow cylinders.

Metal objects can also be obtained from forged blanks or foundry crudes near the ribs.

However, all these methods require the use of a large amount of material relative to that present in the final part. The corresponding ratio can be as high as 30: 1, particularly in the field of aeronautics, where it is called "Buy to Fly ratio" in English terminology.

In addition, they require the implementation of heavy industrial means and lead to very high realization costs that increase with the dimensions of the object to be manufactured.

It has been proposed (CN 104 959 603) to produce a metal object by the so-called powder bed technique, which consists in successively depositing layers of powder in a cylinder and then providing locally energy, for example by a laser beam that follows a specific path, so as to locally melt the powder and form a layer of the final object.

This technique has many limitations, in particular because it does not allow to produce parts on a non-planar substrate, for example cylindrical.

This is why additive manufacturing processes are developed to obtain metal objects which proceed by local supply of metal for example in the form of powder or wire, and energy for example in the form of a laser beam, a electron beam, arc or plasma. These methods are commonly referred to as DMD (Direct Metal Deposition) in English terminology.

It is thus possible to provide a flow of gas driving metal powder, moving above a metal support with a laser beam that causes the powder to melt, which is deposited in the molten state on the support itself. slightly melted and solidifies on this one. Each movement of the beams creates a layer that is superimposed on the previous ones. In general, the thickness of each layer after melting is less than the width of the deposited bead.

These methods are in particular proposed in the field of aeronautics to reduce the amount of total material used relative to that of the final part and generally reduce manufacturing costs. These methods, however, have disadvantages.

Indeed, the deposition of molten metal leads to high stresses during its solidification and thus the deformation of the support.

In addition, the control of the melting is difficult and the deposited material may, after solidification, have defects. That is why it is necessary to practice a nondestructive control of any volume built on the support.

Finally, it is found that the macroscopic and microscopic surface states are not very good due to the presence of droplets, projections or hollows and that the surface is often polluted.

To remedy this, it has been envisaged to better regulate the various implementation parameters of these processes, including the power of energy supply, the construction strategy or the temperature by the introduction of external cooling or heating. preheating the substrate. It has also been experimented with different types of filler materials whose characteristics have been modified with regard for example their particle size or morphology for powder materials, or their size for materials in the form of wire. The processes were finally implemented under a protective atmosphere to avoid surface pollution.

These changes, however, are of limited effect and result in significant additional costs, has relatively long feedstock losses and cycle times.

The object of the invention is to overcome these drawbacks by proposing a radically different and highly efficient solution by introducing one or more intermediate steps of catching of the dimensions during the manufacture of the blank.

Thus, the invention relates to a method of additive manufacturing of a blank of a metal object comprising a stack of superposed layers, from a digital object, said method comprising successive steps of producing a layer by local input of metal material on a support itself metal, combined with a local supply of energy, characterized in that it comprises a step of defining and storing a reference surface for the stack and at least one step of removing the material made dry between two successive layers n and n + 1 of the stack, this step of removing material on the upper face of the layer n to create a reference surface for the deposition of the layer n +1, the distance between the reference surface and the reference surface being substantially constant.

This method thus makes it possible to overcome the deformations of the metallic support and of the successive layers and to reduce the defects of the surface states during the formation of the stack of layers, to ensure the quality of the blank obtained while optimizing productivity through the use of simplified parameters. In addition, the removal of dry material prevents contamination from a lubricant and it is efficient.

In advantageous embodiments, one and / or the other of the following provisions is also used:

the removal of material is of the cryogenic type;

the removal of material is carried out with an addition of liquid nitrogen or CO ₂ ;

after the deposition of a layer, a step is taken to measure the distance between the free face of this layer and the reference surface and at a step of comparison between the measured distance and a threshold value, said removal step of material being performed if the measured distance is greater than said threshold value;

at least one layer is produced by adding metal powder;

at least one layer is made by adding metal wire;

at least one layer is produced by welding discrete surface elements;

said layer having a predetermined thickness, said discrete surface elements have at least one dimension greater than said thickness;

said layer is obtained with successively welded elements, a complementary step of dry material removal being carried out after the welding of an element, for at least a part of said elements; said layer is obtained by welding a first series of elements, at least two of which are spaced apart, and a second series of elements which are interposed in the space or spaces formed between the elements of the first series, a step complementary dry material removal being provided in said one or more spaces, between the welding of the first series and that of the second series;

welding is performed without fusion;

said elements and said support come from the same piece;

the reference surface corresponds to the surface of the support before any step of material supply.

The invention also relates to a device for the additive manufacturing of a blank of a metallic object comprising a stack of superposed layers, from a digital object, said device comprising means for successively producing a layer by local supply of metallic material. on a support itself metal, combined with a local supply of energy, means for storing a reference surface and means for the removal of material on the upper face of a layer of the stack.

The invention also relates to a method of manufacturing a metal object from a blank obtained by the method according to the invention, this manufacturing method comprising a step of machining this blank to obtain this metal object. It also relates to a blank of a metal object and a metal object obtained by the methods according to the invention.

In particular, the invention relates to a metallic component of an aircraft obtained by the methods according to the invention, that is to say a component of the structure of the aircraft or its propulsion means.

The invention will be better understood and other objects, advantages and characteristics thereof will appear more clearly on reading the following description of non-limiting examples of implementation of the invention, with reference to the accompanying drawings. on which ones :

FIG. 1 comprises FIGS. 1A to 1E which illustrate the successive deposition of layers by a conventional method of additive manufacturing, with an addition of material in the form of powder or wire.

FIG. 2 comprises FIGS. 2A to 2E which illustrate the implementation of the method according to the invention on a flat support with a contribution of material in the form of powder or metal wire.

Figures 3 and 4 are perspective views illustrating the implementation of the method according to the invention on a plane support with a contribution of material in the form of discrete surface elements.

The elements common to the different figures will be designated by the same references.

Reference is first made to FIG. 1, which schematically illustrates various steps of a conventional method of additive manufacturing comprising depositing four successive layers by adding metallic material to a support 1, to obtain a stack.

The dotted line 10 materializes the plane in which extends the upper face 11 of the support on which a first layer will be deposited. This plane defines a marker surface that is stored.

FIG. 1B illustrates support 1 and this first layer 12 which is obtained by a local supply of metal, for example in the form of powder or wire, and energy, for example in the form of a laser beam, a beam of electrons or plasma. It is the same of the different layers that will be described.

In all cases, the metal is deposited on the support 1 while it is melt, which causes a deformation of the support.

FIG. 1B thus shows that the upper face 11 of the support 1 has a concave shape, the concavity of which is turned towards the first layer 12, the maximum height between the upper face 11 and the plane 10 being identified by the height hi.

The first layer 12 matches the geometry of the support 1 and thus also has a slightly concave shape whose concavity is turned away from the support 1.

FIG. 1C illustrates the following step in which a second layer 13 is formed on the upper face 120 of the first layer 12, that is to say the face opposite to the support. FIG. 1C shows that, during the deposition of the second layer, the deformation of the support 1 and of the first layer 11 is amplified because of the heat provided, the maximum height h ₂ between the upper face 11 of the support and the plane 10 being superior to hi.

The second layer 13 marries, again, the geometry of the first layer 12 and has a greater concavity than the first layer shown in Figure 1B.

FIG. 1D illustrates the following step of deposition of the third layer 14 on the upper face 130 of the second layer 13, opposite the support 1.

Here again, the deformation of the support 1 is accentuated, the maximum height h ₃ between its upper face 11 and the plane 10 being greater than the height h ₂ . Likewise, the concavity of the third layer 14 which matches the geometry of the second layer 13 is accentuated.

Finally, FIG. 1E illustrates the deposition of a fourth layer 15 on the upper face 140 of the third layer 14.

The contribution of this fourth layer amplifies the deformation of the support and layers already deposited, because of the heat input during the formation of this fourth layer.

Figure 1E shows that the concavity of the support 1 is accentuated, the height h ₄ between the upper face 11 of the support and the plane 10 being greater than .3.

The fourth layer 15 matches the geometry of the third layer 14.

In practice, the deformation of the assembly obtained can reach several millimeters and requires then a mechanical recovery. In addition, to obtain the desired geometry, it is necessary to deposit excess material.

Reference is now made to FIGS. 2A to 2E which illustrate an exemplary implementation of the method according to the invention.

Figure 2A is identical to Figure 1A.

This FIG. 2A illustrates a first step of the method in which a reference surface for the stack is defined and stored.

This step is performed before any input of matter and energy.

This reference surface here consists of the upper face 11 of the support 1, before any deposition of metal layer and is represented by the dotted line 10.

This reference surface therefore remains identical in positioning and in shape throughout the implementation of the method, while the upper face of the support may itself be deformed.

Figure 2B is similar to Figure 1B and illustrates the deposition of a first layer 21 on the upper face 11 of the support.

This first layer 21 is made by local supply of metal, for example in the form of powder or wire, and energy, used to heat the material, for example in the form of a laser beam, an electron beam an electric arc or plasma. These two contributions are simultaneous and made close to each other. Thus the material supply zone is located at a distance from the energy supply zone of between 0 and 50 mm. A local contribution is classically understood as a contribution that is made only at the level of the draft to be produced. This is not particularly the case when powder bed techniques are used.

In addition, it is noted that the material is moving relative to the support during the supply of energy for melting, since the inputs are made at the same time.

It can also be noted that, in the case of powder bed techniques, the support on which the blank is made is independent of the latter.

On the contrary, with the method according to the invention, the support on which the material is made will be part of the blank so that the mass of the blank is greater than that of the mass of metallic material provided.

It is the same for the other layers which will be described with reference to Figures 2C to 2E.

Many devices are known capable of producing such metal layers. It is particularly noteworthy the machine marketed by the company BeAM, under the name Magic 2.0; by the company Oerlikon Metco under the name Metco-Clad; by the company Trumpf under the name Tru- Laser LMD or by the company Sciaky under the name EBAM.

Before the deposition of the second layer 22, the upper face 210 of the layer 21, free face opposite to the support, undergoes a step of removing material, such as machining, so as to create a reference surface 211 for the deposition of the layer 22.

This removal of material is adjusted so that the distance between this reference surface 211 and the reference surface 10 is constant or substantially constant. This is verified throughout the layer 21, or at any point of the reference surface 211.

In other words, when the reference surface is plane, as in the example illustrated in FIG. 2, the reference surface 211 is parallel to the reference surface 10.

Moreover, following this material removal step, the thickness of the layer 21 is not necessarily constant. In particular, FIG. 2C shows that the thickness is smaller at the periphery of the stack than at its center.

In general, the method according to the invention produces a blank comprising locally a stack of layers, at least one of which does not have a constant thickness.

The presence of this layer of variable thickness can also be detected on the object obtained after machining the blank.

This material removal step is performed dry, so as not to pollute the support.

Preferably, the removal of material is of the type assisted by a cryogenic fluid, for example liquid nitrogen or CO ₂ .

Due to the cooling induced by this material removal step, the deposition of the second layer 22 on the reference surface 211 causes limited deformation of the support.

Moreover, this cryogenic fluid makes it possible to cool the first layer 21 to a temperature that will be substantially constant over the entire layer.

This makes it possible to carry out the material removal step as soon as the first layer is produced, without waiting for it to be completely cooled.

Another advantage is to achieve the removal of material at the same reference temperature over the entire surface of the first layer and therefore under the same conditions.

It is therefore understood that the step of removing material made with a cryogenic fluid makes it possible to create, for a given layer, a geometrical reference with the reference surface and a thermal reference, conferring substantially the same temperature on the whole of the layer, that is to say that the temperature difference between two points of the layer is less than 20%.

This reference temperature also makes it possible to freeze the microstructure of the formed metal layer and this, over the entire surface of the layer.

It should be noted that the cryogenic fluid also fulfills the role of a machining lubricant, which does not pollute the support.

It is recalled that a lubricant machining is intended to reduce friction on cutting tools and evacuate a portion of the heat generated by deformation, fracture and friction during cutting. However, a cryogenic fluid necessarily has the effect of evacuating heat efficiently and quickly and without contaminating the support.

FIG. 2D illustrates a next step of the method, corresponding to the deposition of the third layer 23.

Before the deposition of this third layer, the upper face 220 of the second layer 22, free face opposite the support 1, is subjected to a material removal step to create a reference surface 221 on which the third layer 23 will be deposited. This reference surface 221 is defined such that the distance between this surface 221 and the reference surface represented by the line 10 is constant. This is verified throughout the second layer 22 or at any point of the reference surface 221.

Figure 2D illustrates the third layer 23, after its deposition on the reference surface 221.

As indicated above, the material removal step leading to the obtaining of the reference surface 221 is preferably assisted by a cryogenic type fluid. Thus, the deposition of the third layer 23 generates little deformation of the first and second layers 21 and 22, because of the induced cooling.

FIG. 2E illustrates the deposition of a fourth layer 24 on the third layer 23.

As before, this deposit does not intervene on the upper face 230 of the layer 23, that is to say the free face of the layer 23 opposite the support 1. Again, a material removal step is provided to create a reference surface 231, on which the fourth layer 24 will be deposited.

This reference surface is obtained by a step of removing material, preferably assisted by a cryogenic fluid.

As a result, the deposition of the fourth layer 24 causes little deformation of the lower layers, thanks to the cooling induced during the production of the reference surface 231.

Figure 2E illustrates the stack obtained. It appears that the free face 240 of the fourth layer 24, opposite the support 1, is substantially parallel to the reference surface materialized by the line 10.

It is therefore understood that a stack of layers, obtained by the method according to the invention does not require a significant mechanical recovery on the side of the upper face of the support, unlike the stack of layers illustrated in Figure 1E and obtained with a classical process. Moreover, it is not useful to deposit excess material to obtain a stack having the desired geometry.

The invention is however not limited to the embodiment which has been described.

In particular, a material removal step does not occur systematically between two successive layers of the stack.

In practice, the method may consist, after each deposition of a layer, in measuring with a sensor the distance between the free face, opposite to the support, of the last layer of the stack and the surface reference 10. When this distance is greater than a threshold value, previously identified, the material removal step is then implemented to achieve a reference surface.

Another solution is to model the behavior of the stack during its production, depending for example on the thickness of the substrate, the energy supply or the materials used. This modeling makes it possible to know in advance the deformations induced during the deposition of the successive layers and thus to predict between which layers formation steps a material removal step will be performed to form a reference surface.

This material removal step may consist of machining or rectification and be performed by conventional devices.

Mention may in particular be made of machines marketed by Mazak under the trade name HV800, by the company Fives under the name Flexiax. Mention may also be made of a turning machine marketed by Mazak under the name A 16, or a turning and milling machine also marketed by Mazak under the name Integrex 1550 RAM.

In addition, it should be noted that the reference surface is not necessarily flat, as in the exemplary implementation of the method illustrated in Figure 2.

Indeed, before any step of supply of material, the support can be plane on all its surface or only on a part of its surface. It may also have a non-planar shape, for example a form of revolution in particular cylindrical or may have a left shape.

The blank that will be obtained at the end of the process will have a shape similar to that of the support, as the final object obtained after machining the blank, for example a truncated cone shape.

On the other hand, the material removal steps are preferably provided such that at the given point of the layer of the stack on which the material is removed the loss of height is less than the height of the layer. in question.

Another method of producing a layer of a stack will now be described with reference to FIGS. 3 and 4.

In this method, the first layer of the stack produced on the support 3 is formed based on discrete surface elements 4 and 4 '.

In the example illustrated in FIGS. 3 and 4, the discrete surface elements 4 and 4 'have an isosceles trapezoidal section.

The element 4 is defined by two faces 46 forming an isosceles trapezium, with two parallel bases 40 and 41, the length L of the base 40 being greater than the length 1_ of the base 41.

The two bases 40 and 41 are connected by two sides 42 and 43 having the same inclination with respect to the base 40 and relative to the base 41.

Each element 4 has a height h corresponding to the distance between the bases 40 and 41. Each element 4 also has a thickness e which is counted between the two faces 46 or in a plane substantially perpendicular to that defined by the height h and the length L or _1.

In the example illustrated in Figures 3 and 4, the height h and the length L or 1. are greater than the thickness e.

In general, the height h of the elements 4 is greater than their thickness e. The discrete surface elements are then made, for example by cutting in a metal sheet which may be constituted by the support 3 itself.

FIG. 3 illustrates a first step of the method in which metallic elements 4 are welded to the support 3, also metallic, in a first arc of a circle. This is a T (or L) assembly.

The elements 4 are welded to the support 3 by their face 47 defined by the bases 40, that is to say along their great length L.

Moreover, they are spaced from each other by a distance corresponding to their short length 1.

This step corresponds to a first series of elements 4.

The welding of the elements 4 can be carried out by any suitable method and in particular a linear friction welding (LFW for Linear Friction Welding in the English terminology) or a process using high temperature plasmas (for example SPS type processes). for Spark Plasma Sintering in terminology English or sputter welding or flash butt welding in English terminology).

In a known manner, a linear friction welding process is a solid phase welding process, the materials not being melted, and which does not require filler materials.

The assembly is performed by rubbing against one another the surfaces to be assembled, under a controlled pressure. The advantages of solid state welding are the preservation of the properties of the materials as well as the possibility of assembling between heterogeneous materials.

However, the method of linear friction welding leads to the formation of burrs at the interface of the welded parts, here between the elements 4 and the support 3.

Also known, a method of the SPS type is based on the use of high temperature plasmas, momentarily generated between powder particles, by an electric discharge.

The heating and cooling speeds are high and the temperature maintenance is generally short. The densification of the material can therefore be done at a relatively low temperature, which qualifies this method of fusion welding process.

It therefore has the same advantages as non-fusion welding processes. Compared to a method of the LFW type, it has the additional advantage of avoiding the backflow of material at the interface between two welded parts and thus the formation of burrs. In all these processes, a discrete element provided is in motion relative to the support during the implementation of the welding process and in particular during the supply of energy for melting.

Once the elements 4 have been welded thanks to a local supply of energy on the support 3, as illustrated in FIG. 1, the next step of the process consists of a removal of material, such as machining, on the elements 4. The removal of material can also be achieved by grinding.

This material removal step is performed dry. It may in particular be assisted by a cryogenic fluid, that is to say with a supply of liquid nitrogen or CO2.

This step of removing material is intended to remove the burrs that may result from the previous welding step and, in general, to prepare the side faces 44 and 45 of the elements 1 already welded which will be in contact and welded together with other elements that will subsequently be welded to the support.

In the method illustrated in Figures 3 and 4, at least the faces 44 and 45 opposite two successive elements 4 of the first series of elements welded to the support will be machined. Similarly, the support 3 will be subjected to a material removal step, such as machining, between two successive elements 1 of this first series and in a direction substantially perpendicular to that in which these elements 4 extend. stage will allow the passage of the material that will escape from the interface between the elements 4 'subsequently welded and the support 3. FIG. 4 illustrates the next step of the method, in which is welded a second series of elements 4 'which are interposed in the spaces formed between the elements 4 of the first series, illustrated in FIG. possible in that all the elements 4, 4 'have the same shape and the elements 4 are spaced a length _1.

These elements 4 'are identical to the elements 4 and the references concerning the elements 4' will be the same as those concerning the elements 4, assigned the index '' '.

FIG. 4 thus illustrates the elements 4 of the first series and the elements 4 'of the second series which are all welded on the support 3 with the exception of a single element 4'.

FIG. 4 shows that the elements 4 'are welded on the support 3 by their short length _1, or else by their surface 48' defined between their bases 41 ', and on the faces 45, 44 of the adjacent elements 1 by their faces 45 'and 44'. The elements 4 and 4 'are therefore arranged head to tail.

The elements 4 'are welded to both the support and the elements 4 by any suitable welding process and, preferably, by a non-fusion welding process, as previously described.

With this step of the method, a full arc 30 is obtained which forms a layer of one stack.

Depending on the geometry of the blank that is desired, the layers that will be formed on the layer 30 obtained will be made from discrete surface elements or by adding metal powder or wire.

Before the deposition of another layer on the layer 30, the upper face 300 of the layer 30, free face opposite to the support, undergoes a step of removal of material, such as machining, so as to create a surface of reference for the deposition of the next layer. As previously explained, this removal of material is adjusted so that the distance between the reference surface and the reference surface consisting of the upper face 31 of the support, before any discrete surface element deposition, is constant. This is verified, after creation of the reference surface, throughout the layer 30 or at any point of the reference surface.

As explained above, this material removal step is performed dry, so as not to pollute the support.

Preferably, this removal of material is of the type assisted by a cryogenic fluid, for example liquid nitrogen or CO2.

The fact of using a cryogenic fluid has the same advantages as those described for the process involving a supply of metal in powder or wire form.

The use of such a fluid is all the more advantageous as the discrete elements are welded by a fusion welding process.

The realization of a layer, from discrete elements is not necessarily obtained by the welding of two series of elements but with elements welded successively one after the other on the support 3 or with series having a lower number of elements.

Moreover, a layer can be obtained from surface elements having a shape different from that illustrated in FIGS. 3 and 4.

It should be noted that the upper face of the support on which the material is provided is preferably a finished surface whose dimensions correspond to the metal object obtained by machining the blank made with the method according to the invention.

In other words, the upper face may be devoid of any extra thickness and it will not be necessary to machine it.

On the other hand, the lower face of the support, opposite to the blank, may be machined, in particular to obtain the desired dimension for the object, taking into account the deformations of the support.

As is obvious and as also follows from the above, the present invention is not limited to the embodiments more particularly described. On the contrary, it embraces all variants.

Claims

1. Method of additive manufacturing of a blank of a metal object comprising a stack of superposed layers, from a digital object, said method comprising successive steps of producing a layer (21 to 24) by local input of metallic material on a support (1) itself metal, combined with a local supply of energy, characterized in that it comprises a step of defining and memorizing a reference surface (10) for the stacking and at least one step of removing material made dry between two successive layers n and n + 1 of the stack, this step of removing material on the upper face of the layer n to create a reference surface for the deposition of the n + 1 layer, the distance between the reference surface and the reference surface being substantially constant.

2. Method according to claim 1, characterized in that the removal of material is of the type assisted by a cryogenic fluid.

3. Method according to claim 2, characterized in that the removal of material is carried out with a supply of liquid nitrogen or CO ₂ .

4. Method according to any one of claims 1 to 3, characterized in that it comprises, after the deposition of a layer, a step of measuring the distance between the free face of this layer and the reference surface (10). ) and a comparison step between the measured distance and a value threshold, said material removal step being performed on said layer if the measured distance is greater than said threshold value.

5. Method according to any one of claims 1 to 4, characterized in that at least one layer (21 to 24) is formed by supplying metal powder.

6. Method according to any one of claims 1 to 5, characterized in that at least one layer (21 to 24) is formed by providing wire.

7. Method according to any one of claims 1 to 6, characterized in that at least one layer is produced by welding discrete surface elements (4, 4 ').

8. The method of claim 7, characterized in that said layer (30) having a predetermined thickness, said discrete surface elements have at least one dimension greater than said thickness.

9. A method according to claim 7 or 8, characterized in that said layer is obtained with welded elements (4, 4 ') successively, a complementary dry material removal step being carried out after the welding of an element. , for at least a part of said elements.

10. The method of claim 7 or 8, characterized in that said layer is obtained by welding a first series of elements (4) of which at least two are spaced apart and a second series of elements (4 '). interspersed in the space or spaces formed between the elements of the first series, a complementary dry material removal step being provided in said one or more spaces, between the welding of the first series and that of the second series.

11. A method according to any one of claims 9 or 10, characterized in that the dry material removal complementary step is of the type assisted by a cryogenic fluid.

12. The method of claim 11, characterized in that the complementary step of removal of material produced is carried out with a liquid nitrogen supply or CO ₂ .

13. Method according to any one of claims 7 to 12, characterized in that the welding is performed without fusion.

14. Method according to any one of claims 7 to 13, characterized in that said elements (4, 4 ') and said support (3) come from the same room.

15. Method according to any one of claims 1 to 14, characterized in that the reference surface corresponds to the surface of the support before any material supply step.

16. A device for the additive manufacturing of a blank of a metallic object comprising a stack of superimposed layers, from a digital object, said device comprising means for successively producing a layer by local supply of metallic material on a support - even metal, combined with a local supply of energy, means for memorizing a surface landmark and means for the removal of material on the upper face of a layer of the stack.

17. A method of manufacturing a metal object from a blank obtained by the method according to any one of claims 1 to 15, comprising a step of machining this blank to obtain said metal object.

18. A blank of a metal object obtained by the method according to any one of claims 1 to 15.

Metal object obtained by the process according to claim 17.