Method for welding two pipes to each other, and corresponding pipework

Abstract

The present application relates in particular to a welding method comprising the following successive steps: removing a layer of an inner surface (9) of the end (11, 13) of at least one of the pipes, thereby creating a recessed zone (15) extending in the inner surface (9) to the free edge (14); filling the recessed zone (15) by deposition of a stainless steel (17); welding the pipe ends (11, 13) to each other by end-to-end welding.

Description

[0001] TITLE: Process for welding two pipes together, corresponding piping

[0002] The present invention relates generally to the welding of two conduits to each other.

[0003] The safety injection system circuits of a pressurized water nuclear reactor (circuits designated by the acronyms RIS and RRA) include complex-shaped stainless steel pipes. These pipes consist of several conduits welded together. Cracks resulting from stress corrosion have been observed in some nuclear reactors. These cracks are present on the internal surface of the pipes, near the welds.

[0004] This situation is illustrated in Figure 1. The piping shown in this figure comprises two conduits 1, 3 welded to each other by a peripheral weld 5. The free edges of the ends of the two conduits 1 and 3 delimit between them a chamfer filled by the filler metal. Cracks 7 have been shown on the internal surface 9 of the conduits 1 and 3, near the weld 5.

[0005] One option to repair this pipe and eliminate the cracks would be to completely replace lines 1 and 3 with new lines made of a material resistant to stress corrosion.

[0006] This solution has the disadvantage that it is necessary to qualify the new material, this process being particularly long.

[0007] The cost to a nuclear reactor operator would be particularly high.

[0008] In this context, the invention aims to propose a welding method which can be used to repair cracks in the safety injection system of pressurized water reactors, which does not have the above drawbacks.

[0009] To this end, the invention relates to a method of welding two conduits to each other, the conduits having respective conduit ends having respective free edges, the method comprising the following successive steps:

[0010] - removing a layer from an inner surface of at least one of the conduit ends, thereby creating a recessed area in the inner surface extending to the free edge;

[0011] - filling the hollow area by depositing stainless steel;

[0012] - butt welding of the ends of the conduits to each other.

[0013] In general, the phenomenon of stress corrosion occurs when three main conditions are met:

[0014] - a material subject to the phenomenon of stress corrosion; - a state of tensile stress;

[0015] - an aggressive environment.

[0016] For nuclear reactor safety injection system circuits, the medium is the primary coolant of the nuclear reactor core, the composition of which cannot be changed without compromising the entire operation of the reactor.

[0017] The idea behind the invention is to ensure that the tensile stresses are concentrated only in the resurfacing zone, where the stainless steel has been deposited.

[0018] Figure 2 is a numerical simulation of the axial stress level around the weld in Figure 1.

[0019] Zones a, b and c are zones in which the material exhibits axial tensile stresses, these stresses decreasing from a to c. Zones d, e and f are zones in which the material exhibits axial compressive stresses, these stresses increasing from d to f.

[0020] Zones f to a correspond respectively to the following axial stress ranges: -150 to -100 MPa; -100 to -50 MPa; -50 to 0 MPa; 0 to 50 MPa; 50 to 100 MPa; 100 to 150 MPa.

[0021] The same notation is used in Figures 9, 10 and 13.

[0022] It is observed that the areas closest to the weld present axial tensile stresses (stress levels a, b and c), while the areas further from the weld present axial compressive stresses (stress levels d, e and f).

[0023] Figure 3 is a numerical simulation showing the plastic deformation of the material around the weld in Figure 1.

[0024] The deformation scale is graduated from a to f, the deformations decreasing from a to f. These deformations are expressed in % relative to the work hardening law of the material, here of the material of conduit 1, 3.

[0025] Zones f to a correspond respectively to the following plastic deformation ranges: 0 to 5%; 5 to 10%; 10 to 15%; 15 to 20%; 20 to 25%; 25 to 30%.

[0026] The same notation is used in Figures 11 and 12.

[0027] The highest levels of plastic deformation are observed around the weld (levels a to d). Areas far from the weld show a very moderate level of plastic deformation (levels e and f).

[0028] According to the invention, a layer of stainless steel is deposited on either side of the weld, before welding the two pipe ends to each other. After welding the two pipe ends to each other, the areas of the internal surface exhibiting tensile stresses and / or significant levels of plastic deformation are localized in the deposited stainless steel. The areas of the internal surface further away from the weld, made of the original material of the pipe, exhibit only compressive stresses.

[0029] In other words, in the invention, the areas of the inner surface exhibiting tensile stresses are made of the deposited stainless steel. This material has a structure such that it is not susceptible to developing stress corrosion.

[0030] Areas in the original material of the pipes are subject to compressive stresses. Therefore, they are also not subject to stress corrosion.

[0031] Similarly, plastic deformations are concentrated in the stainless steel deposited in the hollow area. The areas of the internal surface made of the original material of the pipes undergo very moderate plastic deformations, and are therefore not susceptible to developing stress corrosion cracking.

[0032] The method may further have one or more of the following characteristics, considered individually or in all technically possible combinations:

[0033] - stainless steel is an austenitic stainless steel or a nickel-based alloy;

[0034] - the filling step is carried out by depositing the stainless steel by a welding method with filler metal, for example a TIG welding method or a welding method with a coated electrode;

[0035] - the hollow area has a determined length chosen so that, after welding, only the deposited stainless steel has tensile stresses, an intermediate area of ​​the internal surface adjoining the stainless steel having compressive stresses;

[0036] - the method comprises a step of determining the length determined by calculation;

[0037] - the determination step includes the following sub-steps:

[0038] * simulation of the stresses generated on the internal surface of the pipe ends by butt welding said pipe ends to each other in the absence of the prior removal and filling steps;

[0039] * determining on the internal surface a boundary between a tensile zone of the internal surface exhibiting tensile stresses in the simulation, and a compression zone of the internal surface exhibiting compressive stresses in the simulation, the tensile zone extending between a weld of the pipe ends to each other and said boundary, the compression zone extending beyond said boundary;

[0040] * choice of length determined so that the hollow area covers the entire traction zone;

[0041] - the determination step includes the following sub-steps:

[0042] * simulation of plastic deformations generated on the internal surface of the pipe ends by butt welding of said pipe ends to each other in the absence of prior removal and filling steps;

[0043] * determination on the internal surface of a limit between a deformation zone of the internal surface presenting in the simulation a plastic deformation greater than a determined level, and a slightly deformed zone of the internal surface presenting in the simulation a plastic deformation lower than said determined level, the deformation zone extending between a weld of the ends of the conduits to each other and said limit, the slightly deformed zone extending beyond said limit;

[0044] * choice of the length determined so that the hollow zone covers the entire deformation zone;

[0045] - the welding step includes the following sub-steps:

[0046] - machining of the free edges of the ends of conduits, the free edges defining a chamfer between them when the ends of conduits are placed end-to-end;

[0047] - filling the chamfer with said stainless steel;

[0048] - the conduit ends are welded to each other by an original weld, the method comprising, before the removal step, a step of separating the conduit ends and removing the original weld.

[0049] According to a second aspect, the invention relates to a pipe comprising two conduits welded to each other according to the method having the above characteristics.

[0050] Other characteristics and advantages of the invention will emerge from the detailed description given below, for information purposes only and in no way limiting, with reference to the appended figures, among which:

[0051] - Figure 1 is a simplified schematic representation, in section, of a pipe presenting on its internal surface cracks resulting from a phenomenon of stress corrosion;

[0052] - Figure 2 is a numerical simulation of the piping of Figure 1, showing the axial stresses near the weld;

[0053] - Figure 3 is a numerical simulation of the piping of Figure 1, showing the level of plastic deformation near the weld; - Figures 4 to 8 schematically illustrate successive steps of the method of the invention;

[0054] - Figure 9 is a numerical simulation illustrating the level of axial stress in the ends of the conduits at the step of Figure 6;

[0055] - Figure 10 is seen similar to that of Figure 9, showing the axial stress level at the end of the step of Figure 8;

[0056] - Figures 1 1 and 12 are numerical simulations showing the level of plastic deformation at the step of Figure 6 and at the step of Figure 8 respectively;

[0057] - Figure 13 is a numerical simulation showing in the upper part the level of axial stress in the conduit ends for the invention at the step of Figure 8, the lower part showing the level of axial stress in the conduit ends for an alternative process in which the conduit ends are first butt welded, then the internal surface is refaced with stainless steel.

[0058] The welding process now to be described is particularly suitable for repairing the pipework illustrated in Figure 1.

[0059] The main steps of this process are illustrated in Figures 4 to 7. In these figures, only the upper half of the conduit of Figure 1 is shown, the lower half not being shown.

[0060] As previously described, this piping comprises two conduits 1, 3 having respective conduit ends 11, 13 joined to each other by the weld 5. It has a central axis C.

[0061] The piping is typically a piping of a nuclear reactor. For example, it is a piping of the safety injection system of the nuclear reactor.

[0062] The piping is typically intended to transport the primary heat transfer fluid of the nuclear reactor.

[0063] Alternatively, the piping is intended to transport a fluid other than the primary heat transfer fluid of the nuclear reactor.

[0064] The piping may also belong to an industrial installation other than a nuclear reactor.

[0065] Pipes 1 and 3 are typically made of stainless steel, for example 304L, 316L or type 347 stainless steel.

[0066] The method comprises a step of separating the ends of conduits 11, 13 from each other, and of removing the original weld 5.

[0067] After the separation step, each end of conduit 11, 13 has a free edge 14.

[0068] The situation at the end of this step is shown in Figure 4. The method then comprises a step of removing a layer from the internal surface 9 of each end of the conduit 11, 13, thus creating a hollow zone 15 in the internal surface 9 extending to the free edge 14.

[0069] The hollow zone 15 extends over the entire periphery of the conduit end 11, 13. It has, along the central axis C of the conduit end, a determined length I. It has a determined depth p. The length is taken from the free edge 14, along the central axis C, towards the inside of the conduit.

[0070] The method then comprises a step of filling the hollow zone 15, by depositing a stainless steel 17. The stainless steel 17 is preferably an austenitic stainless steel, in particular a low-carbon austenitic steel, or a nickel-based alloy.

[0071] For example, suitable stainless steel is 316LSi, 316L, 308L or Inconel® 82 or Inconel® 52.

[0072] Preferably stainless steel 17 completely fills the recessed area 15.

[0073] The filling step is typically performed by depositing stainless steel 17 into the recessed area 15 by a filler metal welding method.

[0074] This welding method is, for example, an orbital TIG welding method. Alternatively, the method is a coated electrode welding method.

[0075] The situation at the end of the filling stage is illustrated in Figure 5.

[0076] The method then preferably comprises a step of machining the deposited stainless steel layer 17. This step aims to obtain an equivalent internal diameter on the two ends of conduits 10, 11 to be assembled. The free surface of the stainless steel 17 is machined to a small depth.

[0077] Non-destructive testing of the deposited material, in this case 17 stainless steel, can be carried out.

[0078] The situation at the end of this step is shown in Figure 6.

[0079] The method then comprises a step of butt welding the ends of conduits 11, 13 to each other.

[0080] This welding step includes the following sub-steps:

[0081] - machining of the free edges 14 of the conduit ends 11, 13, the free edges 14 defining between them a chamfer 19 when the conduit ends 11, 13 are placed end to end;

[0082] - filling the chamfer 19 with said stainless steel.

[0083] The situation at the end of the machining sub-step is shown in Figure 7. This figure shows that each free edge 14 is machined so as to have a truncated cone shape, coaxial with the central axis C. The diameter of this truncated cone surface decreases when following the central axis C towards the other end of the conduit.

[0084] The free edges 14 of the two ends of conduits 11, 13 therefore form truncated cone-shaped surfaces of opposite orientation, as visible in figures 7 and 8.

[0085] The chamfer 19 has the shape of a groove with a closed contour, of substantially V-shaped section in a radial plane containing the central axis C. The width of the chamfer 19 decreases radially from the outside towards the inside of the conduit ends.

[0086] The chamfer filling sub-step is performed by depositing the stainless steel by a filler metal welding method, typically by an orbital TIG welding method.

[0087] As indicated above, the stainless steel 21 filling the chamfer 19 is the same as that deposited in the recessed area 15. In other words, it has the same composition.

[0088] Alternatively, the stainless steel 21 filling the chamfer 19 is of a different grade than the stainless steel 17 deposited in the recessed area 15.

[0089] The situation at the end of the sub-step of filling the chamfer 19 is shown in Figure 8.

[0090] The two ends of conduits 11, 13 are welded to each other. The root 25 of the weld is flush between the layers of stainless steel 17 deposited in the hollow areas 15 of the two ends of conduits 11, 13.

[0091] These layers of stainless steel 17 extend over the internal surface 9, on either side of the root 25, substantially over an axial length I. They extend over the entire periphery of the ends of conduits 11, 13.

[0092] The axial stresses in the two ends of conduits 11, 13 are shown in Figure 9 in the situation of Figure 6, i.e. after machining the stainless steel layer 17 deposited in the hollow area 15.

[0093] It appears in Figure 9 that the stainless steel 17 deposited in each recessed area 15 has low compressive stresses towards the free end 14, the stainless steel 17 deposited opposite the free end 14 having moderate tensile stresses. The area of ​​the inner surface 9 adjoining the deposited stainless steel 17 has low tensile stresses, and beyond the recessed area 15 has compressive stresses.

[0094] Figure 10 shows the axial stresses in the ends of conduits 11, 13 at the end of the welding step, i.e. in the situation of Figure 8. It can be seen that the tensile stresses are concentrated at the root 25 of the weld, and in the layer of stainless steel 17 deposited on either side of it. The internal surface areas 9 of the ends of conduits 11, 13 which are not formed by the deposited stainless steel 17 only exhibit compressive stresses.

[0095] Figure 11 shows the plastic deformation in the ends of conduits 11, 13, after machining of the stainless steel layer 17 deposited in the hollow zone, i.e. in the situation of Figure 6.

[0096] Figure 11 shows that the material constituting the conduit ends 11, 13 is only very little plastically deformed at this stage of the process.

[0097] Figure 12 shows the plastic deformation in the ends of conduits 11,13 at the end of the welding step, i.e. in the situation of Figure 8.

[0098] It appears that the areas of the internal surface 9 of the ends of the conduits exhibiting a plastic deformation greater than 10% relative to the work hardening law of the material of the conduit 1, 3 are all located in the stainless steel layer 17 deposited in the hollow zone 15. The areas of the internal surface 9 which do not belong to the stainless steel layer 17 exhibit plastic deformations less than 10% relative to the work hardening law of the material of the conduit 1, 3. These plastic deformations are of level e or f.

[0099] The stainless steel 17 deposited in the hollow zone 15 is not susceptible to stress corrosion, due to its nature.

[0100] What distinguishes the base metal of the pipe 1, 3 and the deposited stainless steel 17 or even the stainless steel 21 of the weld, from a stress corrosion point of view, is mainly:

[0101] - the microstructure of the materials considered, namely the grain size and the crystalline structure;

[0102] - the chemical nature of the materials considered;

[0103] - possibly, the presence of ferrite can be considered as favorable to resistance to stress corrosion.

[0104] The conditions for resistance to stress corrosion are obtained by choosing appropriate welding parameters for the deposition operation of stainless steel 17.

[0105] These parameters are determined experimentally. This determination is a routine operation for welders. It should be noted that the austenitic steels used for 17 stainless steel are low carbon, generally less than 0.03%, to avoid intergranular corrosion by dechromization of the grain boundaries following the precipitation of chromium carbides.

[0106] There is therefore on the internal surface 9 of the ends of conduits 11, 13 an area having tensile stresses, entirely constituted by the layer of stainless steel 17 filling the hollow area 15. The areas of the internal surface 9 which are not constituted by the stainless steel 17 deposited in the filling step are only subjected to compressive stresses. As a result, they are also not likely to develop stress corrosion.

[0107] Plastic deformations of a level greater than 10% compared to the work hardening law of the material of the conduit 1, 3 are concentrated in a zone of the internal surface 9 belonging to the stainless steel layer 17 filling the hollow zone 15. They cannot therefore contribute to the appearance of stress corrosion on the internal surface 9 of the conduits 1, 3.

[0108] Figure 13 compares the results obtained with the method of the invention and with an alternative method not in accordance with the invention. The upper part of Figure 13 is identical to Figure 10. It shows the level of axial stresses in the pipe ends 11, 13 at the end of the welding step, i.e. in the situation of Figure 8.

[0109] The lower part of Figure 13 shows the level of axial stresses in the conduit ends 11, 13, for the alternative welding method. In this method, the removal, filling and welding steps are not carried out in the same order as in the method of the invention.

[0110] Butt welding of the pipe ends 11, 13 to each other is first performed. Then, the removal and filling steps are performed, in that order.

[0111] It appears in the lower part of Figure 13 that the deposited stainless steel layer 17 is subjected to a very high level of tensile stresses. The area of ​​the internal surface 9 located around the stainless steel layer 17 is also exposed to a very high level of tensile stresses. This latter area is therefore likely to develop stress corrosion, due to its level of tensile stresses.

[0112] According to an advantageous aspect of the invention, the determined length I of the hollow zone 15, as indicated above, is chosen so that, after welding, only the deposited stainless steel 17 exhibits tensile stresses.

[0113] An intermediate zone of the internal surface 9, adjoining the deposited stainless steel 17, has only compressive stresses. Advantageously, the method comprises a step of determining the determined length I, by calculation.

[0114] This determination step includes at least the following sub-steps:

[0115] - simulation of the stresses generated on the internal surface 9 of the conduit ends 11, 13 by butt welding of said conduit ends to each other, in the absence of the prior removal and filling steps;

[0116] - determination on the internal surface 9 of a limit L between a traction zone 27 of the internal surface 9 presenting traction stresses in the simulation, and a compression zone 29 of the internal surface 9 presenting compression stresses in the simulation;

[0117] - choice of the determined length I such that the hollow zone 15 covers the entire traction zone 27.

[0118] The result of the simulation sub-step is shown in Figure 2. The tensile zone 27 extends between the weld 30 connecting the conduit ends 11, 13 to each other and said boundary L.

[0119] The compression zone 29 extends beyond said limit.

[0120] The traction zone 27 only comprises portions of the internal surface 9 undergoing tensile stresses, or on the contrary may comprise both portions of the internal surface 9 undergoing tensile stresses and others undergoing compressive stresses.

[0121] On the other hand, the compression zone 29 only comprises portions of the internal surface 9 exhibiting compressive stresses.

[0122] Preferably, the boundary L is placed as close as possible to the weld 30.

[0123] Advantageously, the determination step includes, in addition to or instead of the sub-steps defined above, the following other sub-steps:

[0124] - simulation of the plastic deformations generated on the internal surface 9 of the ends of conduits 11, 13 by the butt welding of said ends of conduits 11, 13 to each other in the absence of the prior steps of elimination and filling;

[0125] - determination on the internal surface 9 of a limit L' between a deformation zone 31 of the internal surface 9 presenting in the simulation a plastic deformation greater than a determined level, and a slightly deformed zone 33 of the internal surface 9 presenting in the simulation a plastic deformation lower than said determined level;

[0126] - choice of the determined length I such that the hollow zone 15 covers the entire deformation zone 31.

[0127] The determined level is typically 10% relative to the work hardening law of the pipe material 1, 3. The result of the simulation step is shown in Figure 3. The deformation zone 31 extends between the weld 30 and the limit L'. The slightly deformed zone 33 extends beyond the limit L'.

[0128] The deformation zone 31 only comprises portions of the internal surface 9 in which the level of plastic deformation is greater than the determined level, or on the contrary comprises both portions in which the plastic deformation is greater than the determined level and other portions for which the plastic deformation is less than the determined level. On the other hand, the slightly deformed zone 33 only comprises portions of the internal surface 9 in which the plastic deformation is less than the determined level.

[0129] This method is particularly suitable, because the applicant has found that the level of axial stresses in the pipe ends is little impacted by the presence of the stainless steel layer deposited in the hollow area. The tensile stresses are essentially generated by the butt welding of the pipe ends 11, 13 to each other.

[0130] This is also true for plastic deformations.

[0131] Alternatively, the determined length I of the hollow zone 15 is fixed arbitrarily, i.e. on the basis of the operator's experience, without prior simulation.

[0132] The depth p of the hollow zone 15 is chosen so as to guarantee, at the end of the various stages of the process, the expected metallurgical and chemical properties of resistance to stress corrosion. This depth must also allow for the absorption of the plastic deformations which will be generated during the butt welding stage.

[0133] Typically, the determined length I is between 20 mm and 40 mm. The outside diameters of the pipes are generally between 200 mm and 350 mm and have a thickness between 20 mm and 37 mm.

[0134] The depth p in this case is typically between 2 mm and 6 mm.

[0135] The welding process can have multiple variations.

[0136] In the preceding description, it was specified that a hollow area was provided in the internal surface of the two conduit ends, and that these two hollow areas were filled by depositing a stainless steel.

[0137] Alternatively, the recessed area is created only on the inner surface of one of the two pipe ends and then filled with stainless steel. The other pipe end does not undergo the stainless steel resurfacing operation.

[0138] In the foregoing description, the chamfer has been described as having a general groove shape, of V-shaped cross-section. Alternatively, the chamfer may have any other suitable shape. The cross-section may be U-shaped, or rectangular, etc. As described above, the method is typically applied to repair a pipe consisting of two conduits whose ends are welded to each other.

[0139] It also applies to the production of new piping, for welding two pipes of this piping to each other. The process in this case does not include the step of separating the ends of the pipes and removing the original weld.

[0140] The process described above has multiple advantages.

[0141] Using austenitic stainless steel or a nickel-based alloy to fill the hollow area ensures excellent resistance to stress corrosion on the internal surface of the pipes.

[0142] Carrying out the filling step by depositing the molten stainless steel using a filler metal welding method is particularly convenient, as this method is well mastered. Molten stainless steel, after solidification, has a high resistance to stress corrosion cracking.

[0143] Choosing the determined length of the recessed area so that, after welding, only the deposited stainless steel has tensile stresses, with the intermediate area of ​​the internal surface adjacent to the recessed area having compressive stresses, ensures high resistance to stress corrosion at the internal surface of the pipes.

[0144] Including a calculation length determination step in the process ensures that tensile stresses are fully localized in the deposited stainless steel.

[0145] Determining this length by simulating the stresses generated on the internal surface of the conduit ends by butt welding is particularly convenient.

[0146] Determining this length by simulating plastic deformations is also particularly convenient.

[0147] When the determined length is chosen taking into consideration both stresses and plastic deformations, the choice of the determined length is particularly robust.

[0148] The fact that the welding step comprises both the machining of the free edges of the pipe ends and the filling of the chamfer thus formed with said stainless steel makes it possible to obtain a weld of the pipe ends of particularly good quality. In particular, the connection between the weld and the layer of stainless steel deposited in the hollow area is particularly good. The machining of the free edges makes it possible to adapt the chamfer to the welding process by promoting the connection between the ends of the pipes. The filling area of ​​the chamfer carried out by depositing the molten stainless steel by the welding method with filler metal can be easily controlled by ultrasound in order to ensure the good quality of the weld.

Claims

DEMANDS 1. A method for welding two conduits (1, 3) to each other, the conduits (1, 3) having respective conduit ends (11, 13) having respective free edges (14), the method comprising the following successive steps: - removal of a layer from an internal surface (9) of at least one of the ends of conduits (11, 13), thus creating a hollow zone (15) in the internal surface (9) extending to the free edge (14); - filling the hollow area (15) by depositing a stainless steel (17); - butt welding of the ends of conduits (11, 13) to each other, in which the hollow area (15) has a determined length (I) chosen so that, after welding, only the deposited stainless steel (17) has tensile stresses, an intermediate area of ​​the internal surface (9) adjoining the stainless steel (17) having compressive stresses.

2. A method according to claim 1, wherein the stainless steel (17) is an austenitic stainless steel or a nickel-based alloy.

3. A method according to claim 1 or 2, wherein the filling step is carried out by depositing the stainless steel (17) by a filler metal welding method, for example a TIG welding method or a coated electrode welding method.

4. A method according to any one of the preceding claims, wherein the method includes a step of determining the length determined (I) by calculation.

5. A method according to claim 4, wherein the determination step comprises the following substeps: - simulation of the stresses generated on the internal surface (9) of the ends of conduits (11, 13) by butt welding of said ends of conduits (11, 13) to each other in the absence of the prior steps of removal and filling; - determination on the internal surface (9) of a limit (L) between a tension zone (27) of the internal surface (9) exhibiting tensile stresses in the simulation, and a compression zone (29) of the internal surface (9) exhibiting compression stresses in the simulation, the tension zone (27) extending between a weld (30) of the ends of conduits (11, 13) to each other and said limit (L), the compression zone (29) extending beyond said limit (L); - choice of the determined length (I) such that the hollow area (15) covers the entire traction area (27).

6. A method according to claim 4 or 5, wherein the determination step comprises the following substeps: - simulation of plastic deformations generated on the internal surface (9) of the ends of conduits (11, 13) by butt welding of said ends of conduits (11, 13) to each other in the absence of the prior steps of removal and filling; - determination on the internal surface (9) of a limit (L') between a deformation zone (31) of the internal surface (9) exhibiting in the simulation a plastic deformation greater than a determined level, and a slightly deformed zone (33) of the internal surface (9) exhibiting in the simulation a plastic deformation less than said determined level, the deformation zone (31) extending between a weld (30) of the ends of conduits (11, 13) to each other and said limit (L'), the slightly deformed zone (33) extending beyond said limit (L'); - choice of the determined length (I) such that the hollow zone (15) covers the entire deformation zone (31).

7. A method according to any one of the preceding claims, wherein the welding step comprises the following substeps: - machining of the free edges (14) of the ends of conduits (11, 13), the free edges (14) defining between them a chamfer (19) when the ends of conduits (11, 13) are placed end-to-end; - filling the chamfer (19) with said stainless steel.

8. A method according to any one of the preceding claims, wherein the ends of conduits (11, 13) are welded to each other by an original weld (5), the method comprising, before the removal step, a step of separating the ends of conduits (11, 13) and removing the original weld (5).

9. Piping comprising two conduits (1, 3) welded to each other according to the method of any one of claims 1 to 8.

Images