Friction stir welding tool including an improved pin

Abstract

This description concerns a friction stir welding pin (8), the pin (8) having: - at least one thread (81); - at least one groove (82) transverse to the thread; and - at least one flat (83) cutting into the thread. Illustration: Fig. 3a

Description

Title of the invention: Friction stir welding tool comprising an improved pin. Technical field

[0001] The present description relates to friction stir welding. STATE OF THE ART

[0002] Friction stir welding (FSW) has been used for several years to weld mechanical parts. The weld is created by means of a welding tool comprising a pin and a shoulder, rotating or stationary depending on the welding configurations chosen (in particular according to the materials, thicknesses, and required production rate) and the desired weld quality. The pin rotates and penetrates the parts to be welded. The face of the shoulder is brought into contact with the parts to be welded. Then, the tool moves forward along the joint line to be formed between the parts. This process is particularly advantageous for a low-melting-point alloy, for example, an aluminum alloy, because it does not cause it to melt.This avoids the problems of porosity and hot cracking inherent in fusion welding of alloys, particularly aluminum alloys.

[0003] However, the mixing action performed by the pin can cause defects in the weld, reducing the bond zone between the parts to a few millimeters, or even a few tenths of a millimeter. The mechanical performance of the weld is then reduced. Another consequence of these defects is the appearance of crack initiation points, which weaken the weld and, in some cases, cause its premature failure. GENERAL STATEMENT

[0004] One aim of the present presentation is therefore to improve friction stir welding.

[0005] To this end, according to a first aspect of the present exposition, a friction-mix welding pin is proposed, the pin having: - at least one thread; - at least one groove transverse to the thread; and - at least one flattened section that begins to cut into the fillet.

[0006] Thus, this pin geometry makes it possible to significantly reduce, or even eliminate, the defects caused by mixing. The bonding area is therefore increased and the weld quality improved. It also makes it possible to produce welds for wide ranges of welding parameters. This is referred to as universal pin geometry.

[0007] For example, welding can be performed at a high pin rotation speed, while maintaining a low tool feed rate. Indeed, without this particular geometry, these ranges of values ​​for welding were not considered because they were conducive to the appearance of these defects.

[0008] It can be provided that the groove and the flat form a periodic pattern repeating around the pin.

[0009] It may be provided that the pin has a portion of general frustoconical shape, the flat, the thread and the groove extending over the frustoconical portion.

[0010] It may also be provided that a maximum angle between a principal axis of the pawl and an envelope circumscribed about the pawl is between 4° and 20°, preferably equal to 8°.

[0011] It may be provided that the net extends over at least one turn of the pawn.

[0012] It may be provided that the thread has a pitch less than or equal to 1.2 mm.

[0013] It may be provided that the pawn has several nets.

[0014] It may be provided that the groove is inclined with respect to the main axis of the pin.

[0015] It can be provided that the groove runs along the pin around the main axis of the pin in the opposite direction to the net.

[0016] It can be provided that the groove extends helically around the pin.

[0017] It may be provided that a groove depth is greater than or equal to a net depth.

[0018] It may be provided that the pin has several grooves distributed around the pin.

[0019] It may be provided that the pin has three grooves, in particular helical, regularly arranged around the main axis of the pawn.

[0020] It may be provided that the pawn has several flats.

[0021] It may be provided that the flats are distributed around the pin.

[0022] It can be provided that at least two of the flat surfaces are coplanar.

[0023] It may be provided that a free end of the pin is flat.

[0024] It may be provided that a free end of the pawl is convex.

[0025] It may be provided that a free end of the pin forms a punch.

[0026] According to a second aspect of this presentation, a friction stir welding tool is proposed, the tool comprising: - a pawn conforming to the first aspect; and - a body with a shoulder surrounding the pawn.

[0027] It may be provided that the pin is mounted to rotate freely relative to the shoulder.

[0028] According to a third aspect of this presentation, an electrospindle is proposed comprising: - a frame; and - a tool that conforms to the second aspect, the tool being mounted on a free end of the frame.

[0029] According to a fourth aspect, a process for welding parts by friction stir welding is proposed, the process being implemented by a pin conforming to the first aspect.

[0030] It may be envisaged that the process is applied to parts superimposed one on top of the other.

[0031] It may be provided that the process is applied to two parts, the two parts being at the same time superimposed one on the other and juxtaposed field against field.

[0032] According to a fifth aspect, an assembly comprising welded parts is proposed, the assembly resulting from the implementation of a process according to the fourth aspect.

[0033] According to a sixth aspect, a battery tray is proposed comprising at least one skin and at least one hollow profile, a weld of the skin to the profile resulting from an implementation of a process according to the fourth aspect. DESCRIPTION OF THE FIGURES

[0034] Other features, purposes and advantages will become apparent from the following description, which is purely illustrative and not limiting, and which should be read in conjunction with the accompanying drawings on which:

[0035] [Fig.1] schematically illustrates a friction stir welding electrospindle mounted on a robot;

[0036] [Fig.2] is an elevation view of a friction-stirring tool mounted on a friction-stirring welding electrospindle;

[0037] [Fig.3a] illustrates in perspective a pawn with an improved geometry;

[0038] [Fig.3b] is a radial cross-sectional view of the pin of [Fig.3a];

[0039] [Fig.3c] is a schematic longitudinal section view of the pawl of [Fig.3a];

[0040] Figures 4a and 4b are schematic perspective views of friction tools- mixing with a static shoulder;

[0041] Figures 5a and 5b schematically illustrate friction-mixing tools;

[0042] [Fig.6] is a flowchart illustrating a method of implementing a process for welding parts by friction stir welding;

[0043] Fig. 7a, Fig. 7b, Fig. 7c, Fig. 7d, Fig. 7e and Fig. 7f schematically illustrate several ways of carrying out friction stir welding;

[0044] [Fig.8] schematically illustrates part of an aircraft;

[0045] Fig. 9a, Fig. 9b and Fig. 9c schematically illustrate several ways of welding aircraft parts by friction stir welding;

[0046] Figure 10 schematically illustrates a welding configuration of an aluminum skin onto a hollow aluminum profile typical of an electric vehicle battery tray; and

[0047] [Fig. 1a] and [Fig. 1b] are photographs of a weld on a battery tray obtained by friction stir welding;

[0048] [Fig. 12] and [Fig. 13] are metallographs of parts welded by friction stir without geometric improvement of the pin;

[0049] [Fig. 14] is a photograph of a weld obtained by a friction-mixing tool with a movable shoulder;

[0050] [Fig.15], [Fig.16], [Fig.17a], [Fig.17b], [Fig.17c] and [Fig.18a] are metallographies of parts welded by friction stir with geometric improvement of the pin;

[0051] [Fig. 18b] and [Fig. 18c] are close-up views of [Fig. 18a];

[0052] [Fig. 19] and [Fig. 20] are respectively a photograph and a metallograph of a weld obtained by a friction-stirring tool with a fixed shoulder and a geometrically improved pin; and

[0053] Figures 21a and 21b are metallographs of a weld obtained by a friction stir tool with geometric improvement of the pin, fixed shoulder, and selection of a maximum cone angle. DETAILED DESCRIPTION

[0054] An industrial robot 100 designed to move within a workspace to weld parts 21, 22 of an assembly 20 is illustrated by way of example in [Fig. 1]. The assembly 20 is further illustrated by way of example in [Fig. 7a]. The robot comprises at least one arm mounted on at least one axis. In the present application, each axis corresponds to one degree of freedom of the robot. In the embodiment illustrated in [Fig. 1], the robot 100 comprises six axes. The robot further comprises a free end 101 configured to receive a spindle, preferably an electrospindle 10, for performing the welding operation. Of course, the electrospindle 10 can be mounted on another device, for example, a machine tool, to perform the welding.

[0055] The parts 21, 22 are mounted in a tool 200 fixed to a table 300. According to one embodiment, notably illustrated in [Fig. 1], the tool 200 is positioned and held in place on the table 300, typically by screws that secure it to the table 300. The tool 200 further includes means for clamping the parts 21, 22. According to one embodiment, the tool 200 includes clamping flanges. configured to hold parts 21, 22 immobile during the friction stir welding operation.

[0056] The welding is carried out by friction stir welding. For this purpose, the electrospindle 10 includes a friction stir tool 1, illustrated in particular in [Fig. 2], mounted on a free end 4 of the electrospindle 10. The electrospindle 10 is known to those skilled in the art and will therefore not be described in further detail. Reference may be made to document FR-3 122 110, which describes an example of an electrospindle 10.

[0057] The friction-mixing tool 1 includes a rotating pin 8, i.e. mounted to rotate about a longitudinal axis Z of the electrospin.

[0058] It is agreed here that an axial plane is a plane parallel to the longitudinal axis Z and passing through it, and a radial plane is a plane orthogonal to the longitudinal axis Z. A circumference is understood as a circle belonging to a radial plane and whose center belongs to the longitudinal axis Z. A circumferential plane is a plane tangent to a circumference.

[0059] An example of pawn 8 has been illustrated in detail in [Fig.3a].

[0060] In this example, the pin 8 is generally frustoconical in shape, meaning that an envelope circumscribed around the pin 8 has a diameter along the longitudinal axis Z, measured perpendicular to the longitudinal axis Z, which is not constant and, in this case, decreases as one moves towards a free end 80 of the pin 8. To facilitate the penetration of the pin 8 into the workpieces, a maximum cone angle measured between the Z axis and a generatrix of the envelope can also be between 4° and 20°. This range of values ​​is also chosen so that the tool 1 does not collide with the workpieces during welding for the configuration illustrated in [Fig. 7e], which will be detailed later in the description. Furthermore, to maximize the bonding area of ​​the workpieces, the cone angle can preferably be 8°. Alternatively, the pin 8 can be generally cylindrical in shape.That is to say, an envelope circumscribed by pin 8 has a constant diameter along the longitudinal axis Z, measured perpendicular to the longitudinal axis Z. Pin 8 may also have a general cylindrical shape on one axial portion and a frustoconical shape on another axial portion. In this case, the free end 80 is flat and extends in a radial plane from pin 8. The free end 80 may have another shape, in particular convex, especially domed, in order to reduce welding forces and facilitate its insertion into the material. To further facilitate its insertion, the convexity of the free end 80 may be accentuated to the point that the free end 80 forms a punch on pin 8.

[0061] The pin 8 further has a thread 81 extending along the longitudinal axis around the pin 8, in this case over several turns, either completely or partially. Thus, the thread 81 allows the pin to mix the kneaded material all the way to the free end 80. In this application, the net 81 is defined as a pattern having an edge 801, more commonly referred to as a top of the net 81, and a trough 802, more commonly referred to as a bottom of the net 81, extending helically around the pin. Thus, in any axial plane passing through the pin 8, the net 81 exhibits a periodic variation in the portions of the top and portions of the trough. Of course, the net is not limited to any particular typology and can have any known shape, for example, metric, trapezoidal (the edge and trough are generally trapezoidal), round (the edge and trough are generally rounded), gaseous, or even sawtooth.

[0062] The pin 8 may also have several threads arranged successively along its length. In other words, the edges of each thread can follow one another to form a continuous or discontinuous helix by offsetting each thread along the Z-axis. The number of turns of the thread can vary from one pin to another. In other words, the thread 81 can have one or more turns. A pitch value of the thread 81, that is, a maximum distance between two opposite sides 803a, 803b of the edge 801, is preferably less than or equal to 1.2 mm. Alternatively, the pin 8 may have several threads arranged successively around it, each thread extending over less than one turn of the pin 8.

[0063] The pin 8 also has three grooves 82 regularly arranged around the Z-axis. In other words, the grooves 82 are distributed around the pin 8 isometrically. That is to say, the grooves 82 are mutually spaced by a constant angle around the pin 8. In this case, the angle is 120° because there are three grooves. But it can take other values, for example, 180° for two grooves or 90° for four grooves. This is also referred to as the circular repetition of the grooves 82 around the pin 8. It is also said that the grooves 82 are distributed around the pin 8 isometrically in a radial plane passing through the pin 8. Each groove 81 runs around the pin 8 around the Z-axis in a direction opposite to the thread 81 in this example. Thus, the grooves allow for additional mixing which opposes that of the net during mixing by pin 8. The mixing is then further improved.The pin 8 may have fewer than three grooves 82, more than three grooves 82, or even only one groove 82. Each groove 82 is transverse to the thread and is inclined. The grooves are all straight but may alternatively all be helical, or each groove may be selectively straight or helical. They may also have other shapes. The grooves 82 also have a depth Pr greater than or equal to a depth of the thread Pf (see [Fig. 3b]).

[0064] The pin 8 also has flats 83. Each flat 83 extends partially onto the thread 81, here in a circumferential plane passing through the thread 81. The flats 83 cut into the thread 81 because a height Hm, taken in an axial plane between a The base of the hollow 802 and the flat 83 are less than a nominal height Hn measured in the same axial plane between the base of the hollow 802 and a crest of the edge 801 ([Fig. 3c]). The flats 83 accentuate the mixing effect of the grooves 82 and the thread 81. The flats 83 are regularly arranged around the pin 8 between the grooves 82. That is to say, the flats 83 are mutually spaced in the radial plane by a constant angle. In this case, the angle is 120° because there are three flats. But it can take other values, for example, 180° for two flats or 90° for four flats. We also refer to the circular repetition of the flats 83 around the pin 8. Furthermore, since they are between the grooves 82 which are themselves regularly arranged around the pin 8, the flats 83 are angularly offset with respect to the grooves 82. The grooves 82 and the flats 83 therefore repeat periodically around the pin 8.

[0065] In the present example, the flats 83 are arranged along three lines 820, 821, 822 lying in non-coplanar circumferential planes. Of course, the flats 83 can be arranged over more than three lines, fewer than three lines, or even just one line. On each line, the pawn 8 has four flats 83. It can also have more than four flats, or fewer than four flats. The flats 83 are evenly spaced on each line. In other words, they are homothetic images of each other with the same ratio. The homothety ratio can also vary from one line to another, for example, to arrange the flats 83 in a staggered pattern around the pawn 8. Alternatively, the flats 83 can be evenly arranged around the pawn 8 on a circumference. In other words, the flats 83 are distributed around the pin 8 isometrically in a radial plane passing through the pin 8.

[0066] The friction-mixing tool 1 further comprises a body 5 separate from the pin 8. The body 5 can be fixed relative to the pin 8. The body 5 comprises a base 6 and a shoulder 3. The base 6 is mounted on the electrospindle 10, and the shoulder 3 is configured to be in contact with the parts 21, 22 during welding. The body 5 is axially delimited from the base 6 to the shoulder 3 and radially delimited by a lateral face 7 extending around the longitudinal axis Z. The shoulder 3 surrounds the pin 8 and has a diameter D2 greater than a diameter DI of the pin 8. Thus, the shoulder forms a variation in cross-section between the body 5 and the pin 8.

[0067] According to an exemplary embodiment of the body 5, notably illustrated in Figures 4a and 4b, the base 6 of the body 5 has a first cylindrical section 60 and a second cylindrical section 61 having a diameter greater than the diameter of the first cylindrical section 60. The first cylindrical section 60 includes flats 63 configured to prevent the body 5 from rotating within the electrospindle 10. Thus, when mounting the body 5 on the electrospindle 10, the body 5 is positioned with the electrospindle 10 by centering it on the first section 60 and providing a flat support on the second section 61. The body 5 is held in position by a clamping means allowing the body 5 to be secured with the electrospindle 10.

[0068] According to an exemplary embodiment of the lateral face 7, notably illustrated in [Fig. 5a], the lateral face 7 is cylindrical. It has a constant diameter D2 along the longitudinal axis Z, measured perpendicular to the longitudinal axis Z. Alternatively, the lateral face 7 may be non-cylindrical. For example, it may be frustoconical, as illustrated by way of example in [Fig. 5b]. It has a diameter D2 measured perpendicular to the axis that is not constant and, in this case, decreases as one moves towards the pin. The lateral face 7 may also be non-cylindrical on a first portion 70 and cylindrical on a second portion 71. For example, it may be frustoconical on the first portion 70, as illustrated by way of example in [Fig. 4a] and [Fig. 4b].

[0069] The pin 8 can be mounted to move relative to the shoulder 3, which is then said to be fixed, for example by assembling the shoulder 3 to a base of the electrospindle 10. Thus, the tool 1 is more robust thanks to the fixed shoulder 3. In addition, the weld 23 has a better surface finish. It is therefore more resistant. Furthermore, the tool 1 welds faster. The cutting speed of the tool 1 is therefore improved. Alternatively, the pin 8 can be mounted to be rigidly fixed to the shoulder 3, for example by assembling the pin 8 and the shoulder 3 on the same axis of rotation of the electrospindle 10.

[0070] Thus, the pin 8 extends outward from the shoulder 3. The thread 81 and the grooves 82 then extend to the maximum up to the shoulder 3. As for the flats 83, they can be arranged between the shoulder 3 and the free end 80.

[0071] The shoulder 3 has at least one axial end face 30 of generally flat shape and configured to be in contact with the surface of the parts 21, 22 to be welded.

[0072] The axial end face 30 may have another shape. For example, a junction between the axial end face 30 and the lateral face may be rounded or chamfered. The face 30 may have a concave shape.

[0073] Thanks to the fixed shoulder, the shape of the shoulder can be particularly well adapted to the shape and configuration of the parts to be welded. This shoulder then constitutes a "counterform" of the parting line and contains the material being welded (see, for example, the embodiment of Figures 7e and 7f described below). For each particular welding configuration, it is possible to adapt the shape of the shoulder to that of the joint to be produced.

[0074] The device carrying the tool 1, for example the aforementioned robot, is configured to move the tool vertically, in order to bring the axial end face 30, typically flat, into contact with a portion of the parts 21, 22, and horizontally, in order to advance the tool along the parts to be welded. Thus, during welding, particularly when the shoulder is fixed, the face 30 slides on the portion of the parts 21, 22 and at the same time, the pin 8 animated by a rotational movement around the longitudinal axis heats and kneads the joint plane, to create a weld 23 of parts 21, 22. The face 30 then allows to contain the material during the welding operation.

[0075] The shoulder 3 may further comprise several axial end faces, in particular planar, non-coplanar 30, 31, 32. Thus, it is possible to carry out the weld 23 between parts extending angularly, for example perpendicularly or presenting an angle less than 90°, such as 70°, to each other.

[0076] According to one embodiment, notably illustrated in [Fig.4b], the face 3 is decomposed into a first axial end face 30 which extends in a radial end plane and two other planar faces 31, 32 which extend symmetrically from each other with respect to the axial plane giving rise to two distinct oblique planes from the face 30. Alternatively, the two other planar faces 31, 32 extend asymmetrically from each other with respect to the axial plane for welding parts not extending perpendicularly but having an angle of less than 90°, for example 70°.

[0077] A method of implementing a welding process of a first part 21 and a second part 22 of an assembly 20 is described with reference to [Fig.6].

[0078] The welding process is implemented by tool 1 of this presentation.

[0079] In the case of a removable tool 200, the tool 200 is fixed to a table 300 during step E20. According to an example of step E20 implementation, tooling 200 is placed and held in position on table 300, typically by screws. Alternatively, the tooling is part of table 300. In this case, tooling 200 is already mounted on table 300.

[0080] During step E21, the parts 21, 22 to be welded are positioned in the tooling 200 and then held in position by fastening means, for example clamping flanges placed in the tooling 200. The parts 21, 22 can be positioned according to different configurations: - in juxtaposition field against field ([Fig.7a]), the pieces 21, 22 then having coplanar principal faces 210, 220; - superimposed one on top of the other parallel ([Fig. 7b]), one of the parts 21 then extending over the other part 22. The lower principal face 211 of the upper part 21 is in surface contact with the upper principal face 220 of the lower part 22; or - superimposed one on the other angularly, for example perpendicularly or at an angle less than 90° such as 70°, or face to face (figures 7c, 7e and 7f), one of the parts 21 extending above the other part 22, the lower principal face 211 of the upper part 21 being in surface contact with a field 222 of the lower part 22 ([Fig.7c]) or a field 212 of the upper part 21 being in contact with the upper principal face 220 of the lower part 22 ([Fig.7e] and 7f). The principal faces 220, 221 of the latter are perpendicular to those of the upper part 21.

[0081] Of course, it is also possible to mix the configurations. For example, parts 21 and 22 can be simultaneously superimposed on one another and juxtaposed face to face ([Fig. 7d]). In this case, the upper part 21 is supported against the lower part 22, with their faces and the lower principal face 211 of the upper part 21 coinciding with one of the upper faces 220 of the lower part 22.

[0082] During a step E22, the tool 1 is moved in vertical translation until the end face 30 is in contact with at least one of the first and second parts 21, 22.

[0083] In juxtaposition, the end face 30 is in contact with both parts 21, 22 simultaneously. In superposition, the tool 1 can be in contact with one or both parts 21, 22. According to one embodiment, notably illustrated in [Fig. 7b], the parts are superimposed one on top of the other parallel to each other, and the end face 30 is in contact with the upper principal face 210 of the upper part 21.

[0084] According to another embodiment, notably illustrated in [Fig.7c], one of the parts 21 extends perpendicularly above the other part 22, the lower principal face 211 of the upper part 21 being in surface contact with a field 222 of the lower part 22. Thus, the end face 30 is in contact with the upper principal face 210 of the upper part 21.

[0085] Alternatively, as illustrated in particular in Figures 7e and 7f, the upper part 21 has a main flat portion 213, having principal faces extending parallel to each other, and a flared edge 214 in which the principal faces move away as one approaches the field. This field 212 of the upper part 21 is in contact with the upper principal face 220 of the lower part 22. Thus: - the axial end face 30 and the first oblique face 31 of the shoulder 3 are respectively in contact with the edge 214 and a flat face of the main part 213 of the upper piece 21; and - the second oblique face 32 of the shoulder 3 is in contact with the upper main face of the lower part 22.

[0086] Tool 1 is ready to weld. During a step E23, the pin 8 is rotated and tool 1 is moved in a direction, typically horizontal, to follow a joint line so as to make a pass and form the weld 23 of the first and second parts 21, 22. Several passes can be made to form the weld 23, typically two passes.

[0087] In juxtaposition, an example of which is illustrated in particular in [Fig.7a], and when the configurations are mixed, for example when the parts 21, 22 are at the same time superimposed on each other and juxtaposed field to field ([Fig.7d]), the end face 30 passes simultaneously over the coplanar principal faces 210, 220 of the parts 21, 22.

[0088] In the case where the parts 21, 22 are superimposed one on the other parallel, a configuration of which an example is illustrated in particular in [Fig.7b], the end face 30 passes over the main external face 210 of the upper part 21 and welds the two parts together.

[0089] In the case where one of the parts 21 extends angularly above the other part 22, for example perpendicularly in an example illustrated in [Fig.7c] or presenting an angle less than 90° such as 70°, the lower principal face 211 of the upper part 21 being in surface contact with a field 222 of the lower part 22, the end face 30 passes over the upper principal face 210 of the upper part. Alternatively, illustrated by way of example in [Fig.7e] and 7f, the field 212 of the upper part 21 can be in contact with the upper main face 220 of the lower part 22, the end face 30 passes over the main face 214 of the upper part 21 and the oblique faces 31, 32 pass respectively over the main face 213 of the upper part 21 and the upper main face 220 of the lower part 22.

[0090] Parts 21 and 22 are then welded. Steps E20 to E23 can be repeated until assembly 20 is completely welded. Consequently, assembly 20 then comprises at least two parts.

[0091] In what follows, we detail two concrete applications of the welding process.

[0092] An aircraft fuselage is illustrated in [Fig. 8]. It includes stringers 51 extending following a longitudinal direction of the fuselage, frames 55 extending in a transverse direction of the fuselage, and a skin 52 covering the stringers 51 and frames 55. The frames 55 and stringers 51 act here as stiffeners of the skin 52 of the fuselage.

[0093] For the sake of brevity, frames 55 and rails 51 will be referred to as stiffeners in the remainder of this description.

[0094] The stiffeners therefore extend over an inner face 520 of the skin 52, while an outer face 521 of the skin 52 is free and opposite the stiffeners. The skin and each stiffener are welded to each other by the welding process implemented by the tool 1. It is therefore unnecessary to fasten them to each other using additional fasteners. Thus, the mass of the aircraft 50 is reduced and the efficiency of the aircraft 50 is improved. The aircraft 50 is also assembled more quickly.

[0095] According to a first embodiment, notably illustrated in [Fig.9a], the stiffener has a wedge-shaped profile and includes a flange 53, the flange 53 being welded in overlap onto the fuselage skin 52. The welding takes place with the axial end face 30 in contact with the flange 53, on the side of the stiffener 51, 55 opposite to the skin 52.

[0096] According to a second embodiment, notably illustrated in [Fig.9b], the stiffener 51, 55 is welded in overlap onto the fuselage skin 52 and the tool is this time in contact with the fuselage skin 52 during welding, extending over the external face 521 of the skin, opposite the stiffener 51, 55. A field of the stiffener 513 is in contact with the internal face 520 of the skin 52.

[0097] According to a third embodiment, notably illustrated in [Fig. 9c], the stiffener 51, 55 is welded in overlap onto the fuselage skin 52, and the tool is in contact with both the fuselage skin 52 and the stiffener 51, 55 during welding. The welding takes place from the inner, non-visible face 520 of the skin, representing a considerable advantage over the configuration of [Fig. 9b]. The welding takes place in the configuration shown in Figures 7e and 7f. The oblique faces 31, 32 of the shoulder 3 come into contact with both the main face of the stiffener 511 and the inner main face 520 of the skin 52, and the end face 30 of face 3 is in contact with the portion 512 of the stiffener 51, 55. The tool 1 therefore performs the weld by simultaneously passing over the flat portions 511, 512 of the stiffener 51, 55 as well as the inner face 520 of the skin 52, for example on the left in [Fig. 9c]. Another pass allows the same to be done on the right.Since the weld is performed on the internal side of the aircraft, a finishing step, for example machining or grinding of weld 23, is eliminated. There is also no longer a need for a base 53, which reduces the overall weight.

[0098] The welding process implemented by tool 1 of this presentation also makes it possible to obtain an assembly 90 having a hollow welded structure and illustrated in [Fig. 10]. This architecture is typical of the aluminum battery trays used in electric vehicles and consists of welding a skin element 91 onto a hollow aluminum profile 92. The hollow profile 92 makes it possible to save weight in order to improve the power-to-weight ratio of the electric vehicle and to consume less energy.

[0099] According to an example of an embodiment, in particular illustrated in figures 11a and 11b, the skin element 91 and the hollow aluminium profile 92 are placed one on top of the other parallel, the skin element 91 extending then above the profile 92. The weld 23 of the skin 91 with the profile 92 is then obtained by a passage of the tool on a main upper face 910 of the skin 91 opposite the hollow aluminium profile 92.

[0100] The welding process implemented by tool 1 of this presentation offers numerous advantages. Indeed, the welded assembly 20 differs macroscopically from that obtained by known friction-stir tools. The weld 23 no longer exhibits defects at the interface between parts 21, 22. The mixed material of parts 21, 22 at the weld 23 is then homogeneous. Consequently, the bond zone between parts 21, 22 is increased. The mechanical performance of the weld is also improved. Crack initiation points are reduced, or even eliminated. The weld is therefore more robust and no longer at risk of premature failure.

[0101] In the case of a fixed shoulder, the weld 23 of the assembly 20 also exhibits an improved surface quality. This has the effect of reducing, or even eliminating, surface defects of the weld that may be encountered in the case of a rotating shoulder. Consequently, the mechanical performance of the weld 23 is further improved.

[0102] Thus, the improved geometry pin 8 and the fixed shoulder 3 together make it possible to obtain a weld 23 which has no defects either on the surface or within it. EXAMPLES

[0103] Concrete, but by no means limiting, examples illustrating the advantages of the geometric improvement of pawn 8 are detailed in the following examples. Examples without geometric improvement of pawn 8

[0104] Fig. 12 is a metallographic cross-section of a weld of a first 1 mm thick 7075 aluminum plate 21 (7000 series) onto a second 2024 aluminum plate (2000 series). The plates are superimposed one on top of the other and the cross-section is taken in a plane orthogonal to the feed direction of the tool 1.

[0105] This figure shows defects on either side of the weld 31 in the form of a wave 25 (better known by the Anglo-Saxon term "cold lap") and a hook 26 (better known by the Anglo-Saxon term "hook"). These defects reduce the bond zone 27. It is small (on the order of a few millimeters, or even a few tenths of a millimeter) and heterogeneous.

[0106] Wave 25 is caused by insufficient shearing of the plate material to completely break the initial joint plane 24. This insufficient shearing is due to a lack of mixing of the pin 8. It takes the form of a wave of varying height depending on the degree of mixing.

[0107] The hook 26 takes the form of a hook, or a comma, caused by the shearing of the material. The hook causes a sudden upward movement of the joint plane 24. In extreme cases, the hook 26 can be completely vertical and rise up to the top plate.

[0108] These defects are the cause of poor quality welding which is the cause of all the aforementioned disadvantages in this presentation.

[0109] To reduce these defects, it is possible to adjust the welding parameters as shown in [Fig. 13]. Indeed, the wave 25 is less pronounced and the bond zone 27 is more extensive and homogeneous. Nevertheless, the hook 26 remains pronounced.

[0110] In particular, due to a particularly low feed rate per revolution (defined as the ratio between the tool feed rate and the pin rotation speed), the hook 26 tends to rise excessively. Conversely, the wave 25 is generally reduced under these conditions (better mixing). This is referred to as "hot parameterization." Therefore, adjusting the welding parameters to simultaneously limit both the hook 26 and the wave 25 is very complex.

[0111] However, this method is tedious because it requires a large number of welding tests and has low robustness. Furthermore, it is necessary to redefine the parameters, and therefore repeat the tests, for each new configuration (thicknesses, materials, etc.).

[0112] It is possible to reduce the size of the hook 26 by using a greater forging force (force along the Z-axis when the pin 8 is driven into the parts 21, 22). This reduces the upward movement of the interface. However, using a higher forging force will generate burrs 28 (visible in particular in [Fig. 14]), also called "weld flash" due to the relief formed by the projected pieces of material which agglomerate at the periphery of the weld during the rotation of the shoulder 3, due to an increased penetration of the tool 1 into the parts 21, 22. In addition, a higher forging force generates an increase in stresses in the parts 21, 22, and therefore greater deformations reducing the quality of the welded assembly 20. Finally, this range of force may prove incompatible with the use of the tool 1 on the industrial robot 100 which is generally limited in load capacity.

[0113] It is also possible to reduce the wave 25 by increasing the rotation speed of the pin 8. However, this leads to significant weld flash 28, or even weld collapse, as well as cavities 29 (visible in particular in [Fig. 14]) due to excessive energy at the surface. Increasing the rotation speed of the pin 8 also locally raises the joint plane 24 and therefore generates a larger hook 26.

[0114] Reaching a compromise to reduce these defects can prove tedious. Indeed, the surface defects and defects within the weld 30 reduce its mechanical properties. The flashes 28 also form corrosion initiation points on the parts 21, 22. Furthermore, the weld 23 is irregular. Indeed, as illustrated in the example in [Fig. 14], weld 23 has many reliefs in its center, because of the cavities 29 and in its periphery because of the weld flashes 28. Examples with geometric improvement of the pawn

[0115] Various tests were carried out to highlight the advantages of the geometric improvement of the pin. In these tests, different thicknesses and material types of the parts were tested. The parts were also welded for several ranges of pin 8 rotational speed and tool 1 feed rates. Thus, a wide range of feed per revolution values, defined as the ratio between the tool 1 feed rate and the pin 8 rotational speed, was tested. Other parameters, such as the forging force, varied between the tests.

[0116] Fig. 15 is a metallographic cross-section of a weld of a 2 mm thick 6061 T6 (6000 series) aluminum cover 21 to a 6061 T6 (6000 series) aluminum support 22 approximately 10 mm thick. The cover 21 and the support 22 are welded in the configuration of Fig. 7d, and the cross-section is taken in a plane orthogonal to the feed direction of tool 1. The rotational speed of pin 8 is 3500 rpm, and the feed rate of tool 1 is 100 cm / min.

[0117] Fig. 16 is a metallographic cross-section of a weld of a 1 mm thick 6061 T6 (6000 series) aluminum sheet 21 to a 2 mm thick 6061 T6 (6000 series) aluminum plate 22. The sheet 21 and the plate 22 are welded in the configuration of Fig. 7b, and the cross-section is taken in a plane orthogonal to the feed direction of tool 1. The rotational speed of pin 8 is 3000 rpm, and the feed rate of tool 1 is 48 cm / min.

[0118] Figures 17a, 17b, and 17c are metallographic cross-sections of several welds of a 3 mm thick 6061 T6 (6000 series) aluminum plate 21 to a 15 mm thick 6061 T6 (6000 series) aluminum support 22. These welds require significant forging forces. The plate 21 and the support 22 are welded in the configuration of [Fig. 7b], and the cross-section is taken in a plane orthogonal to the feed direction of tool 1. The rotational speed of the pin 8 is 3000 rpm, and the feed rate of tool 1 is 78 cm / min. [Fig. 17a] illustrates a weld 23 obtained with a forging force of 350 kg, [Fig. 17b] 400 Kg and [Fig. 17c] 450 kg (25% variation).

[0119] To differentiate itself from previous application tests, which were performed on 6000 series aluminum alloys, Figure 18a is a metallographic cross-section of a weld between two 2000 series aluminum parts 21, 22. Parts 21, 22 are welded in the configuration of [Fig. 7b], and the cross-section is taken in a plane orthogonal to the feed direction of tool 1. Weld 23 is obtained by a double welding pass (a back-and-forth pass). The material is therefore heated considerably, which promotes the formation of a significant weld bead. Figures 18b and 18c are close-up views. of the same metallographic section on both sides of weld 23 at the joint plane 24.

[0120] In all these configurations, there is an absence of hook and wave-type defects, and a contrast ranging from light to dark is observed in the weld 23 along the parting line 24, indicating a homogeneous mixture of the material of the two parts 21, 22. Thus, the bond zone 27 is extensive and homogeneous. The improved geometry of the pin 8 makes it possible to eliminate defects within the weld 23 for a wide range of welding parameters (rotation speed of the pin 8, feed rate of the tool 1, forging force), for different welding configurations, and for different part thicknesses.

[0121] Examples with geometric improvement of the pawn and fixed shoulder

[0122] We also carried out a weld 23 with a geometric improvement of the pin, so as not to have hook and wave type defects, as well as with a fixed shoulder 3.

[0123] Figure 19 illustrates weld 23 obtained under conditions similar to those illustrated in Figure 17b. Here, in particular, the forging force is 400 kg. Weld 23 shows no flash or cavities.

[0124] Thus, thanks to the fixed shoulder, the surface quality of the weld 23 is greatly improved. The resulting weld 23 therefore no longer exhibits any surface or internal defects.

[0125] Figure 20 is a metallographic cross-section of a weld 23 produced in two passes and in a configuration similar to that illustrated in Figure 7e. The cross-section is taken in a plane orthogonal to the feed direction of the tool 1. An overlap zone 40 of the two weld passes can be seen in this figure. This overlap zone notably improves the robustness of the weld.

[0126] Examples with geometric improvement of the pawl, fixed shoulder and selection of the maximum cone angle

[0127] A weld 23 was also produced with, as in the previous configuration, a geometric improvement of the pin and a fixed shoulder 3. In addition, the maximum cone angle is chosen so that the weld 23 extends over a fraction of the thickness of the lower part 22.

[0128] Figures 21a and 21b are metallographic sections under conditions analogous to those of [Fig. 20]. In [Fig. 21a], the maximum angle is arbitrary and the lower part is 5 mm thick. In [Fig. 21b], the maximum angle is selected and the lower part 22 is 2 mm thick. It can be seen in [Fig. 21a] that the weld 23 extends further into the thickness of the part 22 compared to the weld 23 in [Fig. 21b]. Furthermore, if the lower part shown in [Fig. 21a] were 2 mm thick, the pin 8 would penetrate almost all of the thickness of the lower piece. In extreme cases, the 8-piece can even pass through the lower piece when the maximum angle is arbitrary.

[0129] Thus, choosing the maximum cone angle reduces the penetration of the pin 8 into the lower part 22 in order to maintain a sufficient margin to avoid penetrating the lower part 22. Here, the margin is defined as a minimum difference between the thickness of the part 22, measured perpendicularly between the upper principal face 220 and the lower principal face 221 or as specified by the manufacturer, and the penetration of the weld 23 measured perpendicularly between a point P1 on the upper principal face 220 and a point P2 on the weld 23. Typically, the margin is between four and five tenths of a millimeter. The margin can, for example, be measured on the metallograph using known geometric constructions.

Claims

Demands

1. Friction stir welding tool (1), the tool (1) extending along a longitudinal axis (Z), the tool (1) comprising: - a pin (8) having: - at least one thread (81); - at least one groove (82) transverse to the thread; and - at least one flat (83) cutting into the thread; and - a body (5) having a shoulder (5) surrounding the pin, the pin (8) being mounted to rotate freely relative to the shoulder (3) around the longitudinal axis (Z).

2. Tool (1) according to claim 1, wherein the groove (82) and the flat (83) form a periodic pattern repeating around the pin.

3. Tool (1) according to any one of claims 1 or 2, wherein the pin is generally frustoconical in shape, the flat (83), the thread (81) and the groove (82) extending over the frustoconical portion.

4. Tool (1) according to any one of claims 1 to 3, wherein the net (81) extends over at least one turn of the pawl.

5. Tool (1) according to any one of claims 1 to 4, wherein the thread (81) has a pitch less than or equal to 1.2 mm.

6. Tool (1) according to any one of claims 1 to 5, wherein the pin has several threads (81).

7. Tool (1) according to any one of claims 1 to 6, wherein the groove (82) is inclined with respect to a principal axis (Z) of the pin.

8. Tool (1) according to any one of claims 1 to 7, wherein the groove (82) runs along the pin around the main axis (Z) of the pin in a direction opposite to the thread (81).

9. Tool (1) according to any one of claims 1 to 8, wherein the groove (82) extends helically around the pin.

10. Tool (1) according to any one of claims 1 to 9, wherein a groove depth (Pr) is greater than or equal to a thread depth (Pf).

11. Tool (1) according to any one of claims 1 to 10, wherein the pin has several grooves (82) distributed around the pin.

12. Tool (1) according to any one of claims 1 to 11, wherein the pin has several flats (83).

13.

14.

15.

16.

17.

18.

19.

20.

21.

22. Tool (1) according to claim 12, in which the flats (83) are distributed around the pin. Tool (1) according to any one of claims 12 or 13, wherein at least two of the flats (83) are coplanar. Tool (1) according to any one of claims 1 to 14, wherein a free end (80) of the pin is flat. Tool (1) according to any one of claims 1 to 14, wherein a free end (80) of the pin is convex. Electrospindle (10) comprising: - a building; and - a tool (1) conforming to one of claims 1 to 16, the tool being mounted on a free end (4) of the frame. A process for welding parts (21, 22) by friction stir welding, the process being carried out by a tool (1) according to one of claims 1 to 16. Method according to claim 18, applied to parts (21, 22) superimposed one on top of the other. Method according to claim 18, applied to two parts (21, 22), the two parts being at the same time superimposed one on top of the other and juxtaposed field to field. Assembly (20) comprising welded parts (21, 22), the assembly resulting from the implementation of a process according to one of claims 18 to 20. Battery tray (90) comprising at least one skin (91) and at least one hollow profile (92), a weld of the skin (91) to the profile (92) resulting from an implementation of a method according to any one of claims 18 to 20.

Images