Method, corresponding computer program and corresponding apparatus for providing a value of a parameter of a heat source for creating a weld between two plates

Abstract

The method comprises: - receiving a desired value of at least one spatial representation (L1, L2) of the weld (202); - determining a plurality of samples of the parameters of the heat source (104); - for each sample, determining the value of each spatial characteristic (LI, L2) of the weld (202) for this sample; performing a plurality of successive iterations of the following steps: determining, from the simulation point, a point called extrapolation point, determining a point of the function called target point at which each spatial representation (LI, L2) of the weld (202) presents a value close to the desired value, and determining the value of each spatial representation (L1, L2) of the weld (202) from the parameter values of the heat source (104) for the target point, in order to obtain a new simulation point; and - providing the parameter values of the heat source (104) for the target point obtained in the last iteration.

Description

Technical Field

[0001] The present invention relates to a method, a corresponding computer program, and a corresponding apparatus for providing parameter values ​​for a heat source used to generate a weld between two plates. Background Technology

[0002] The parameter values ​​for the heat source used to create a weld between two plates are known to be obtained in the following manner.

[0003] The user obtains a desired value for at least one spatial characteristic of the weld.

[0004] Users determine the values ​​of each spatial characteristic of the weld based on the parameter values ​​of the heat source by performing simulations on a three-dimensional mesh of the two plates. A thermomechanical solver with transient calculation capabilities is typically used for the above simulations.

[0005] Users repeat the preceding steps by manually changing the parameter values ​​of the heat source until they find a value close to the desired value for each considered spatial characteristic of the weld.

[0006] The drawback of this known method is that it requires the user to be familiar with the solver's operation, provide relevant input data, and interpret the solver's output. However, users are typically skilled in weld analysis but not in numerical simulation. Summary of the Invention

[0007] Therefore, it is desirable to provide a method for providing parameter values ​​for a heat source used to create a weld between two plates, a method that enables the overcoming of at least some of the aforementioned problems and defects.

[0008] Therefore, a method is proposed for providing parameter values ​​for a heat source used to generate a weld between two plates, characterized in that the method includes the following steps:

[0009] - Receive at least one spatial characterization of the weld;

[0010] - Determine multiple samples of the parameters of the heat source;

[0011] - For each sample, the value of each spatial representation of the weld for the sample is determined by simulation on a three-dimensional mesh of the two plates, so as to obtain a point called the simulation point, which is a function relating one or more spatial representations of the weld to the parameters of the heat source;

[0012] - Perform multiple consecutive iterations of the following steps:

[0013] • The point called the extrapolation point is determined by extrapolating the function from the simulation point.

[0014] • Determine a point, called the target point, at which each spatial representation of the weld has a value close to the desired value, and

[0015] • By performing simulations on the three-dimensional mesh of the two plates, the values ​​of each spatial representation of the weld are determined based on the values ​​of the parameters of the heat source at the target point, in order to obtain new simulation points to supplement other simulation points; and

[0016] - Provides the value of the parameter for the heat source of the target point obtained in the last iteration.

[0017] Therefore, the present invention provides a robust and reliable method for providing parameters of a heat source that does not require any specific numerical simulation knowledge and can be easily automated by a computer system.

[0018] Optionally, the method further includes the step of receiving measurement locations in the grid, wherein determining the value of each spatial characterization of the weld is achieved based on the time evolution of the temperature measured at each measurement location.

[0019] Optionally, the measurement location is located at the intersection of the grid, and the step of receiving the measurement location in the grid includes receiving at least one of the grid pitch and the grid size.

[0020] Alternatively, the method may further include the following steps:

[0021] - Receive the dimensions of at least one of the plates; and

[0022] - The grid is determined by modifying the reference grid of the two plates according to the received dimensions.

[0023] Alternatively, the received dimension is the thickness of at least one of the plates, the reference grid comprising points having coordinates along the direction of the thickness of at least one of the two plates being gridded by the reference grid, the thickness being referred to as the reference thickness, and the step of determining the grid comprising: transforming these coordinates by isotopity by a ratio equal to the ratio between the reference thickness and the received thickness.

[0024] Alternatively, the modified reference grid can be selected from a set of reference grids.

[0025] Alternatively, the set of reference meshes has been used in advance in the corresponding reference simulation of the weld, which is verified by comparing it with the actual weld produced.

[0026] Alternatively, the step of determining the sample may be performed by pseudo-random sampling, for example by a Latin hypercube.

[0027] A computer program capable of being downloaded from a communication network and / or stored on a computer-readable medium is also proposed, characterized in that the computer program includes instructions for performing the steps of the method according to the invention when the program is executed on a computer.

[0028] A device for providing parameter values ​​for a heat source used to generate a weld between two plates is also proposed, characterized in that the device comprises:

[0029] - An interface module designed to receive a desired value of at least one spatial representation of the weld;

[0030] - A sampling module, the sampling module being designed to determine multiple samples of the parameters of the heat source;

[0031] - Spatial characterization module, which is designed to: for each sample, use simulation on a three-dimensional mesh of the two plates to determine the value of each spatial characterization of the weld for the sample, so as to obtain points called simulation points that correlate one or more spatial characterizations of the weld with parameters of the heat source;

[0032] - An extrapolation module, designed to determine a point called the extrapolation point of the function by extrapolation from the simulation point; and

[0033] - A search module designed to determine points, called target points, of the function, at which each spatial representation of the weld has a value close to the desired value.

[0034] The spatial characterization module is further adapted to: use simulation on the three-dimensional mesh of the two plates to determine the value of each spatial characterization of the weld based on the value of the heat source parameter for the target point, so as to obtain new simulation points to supplement other simulation points. Attached Figure Description

[0035] The invention will be better understood by reading the following description, which is given by way of example only and with reference to the accompanying drawings, in which:

[0036] [ Figure 1 ] Figure 1 It is a three-dimensional view of two plates joined together and the heat source used to create a weld between the two plates.

[0037] [ Figure 2 ] Figure 2 yes Figure 1 The two plates, once the weld seam is formed, are viewed in three dimensions.

[0038] [ Figure 3 ] Figure 3 This is a simplified illustration of a device according to the present invention, the device being used to provide Figure 1 and Figure 2 The parameter values ​​of the heat source,

[0039] [ Figure 4 ] Figure 4 This is a block diagram illustrating the steps of a method according to the invention, the method being used to manufacture... Figure 2 welds,

[0040] [ Figure 5 ] Figure 5 yes Figure 1 and Figure 2 A 3D view of the two plates, showing the 3D mesh of the two plates.

[0041] [ Figure 6 ] Figure 6 yes Figure 1 , Figure 2 and Figure 5 A three-dimensional view of the two plates, showing the measurement grid, and

[0042] [ Figure 7 ] Figure 7 yes Figure 6 The front view of the measurement grid. Detailed Implementation

[0043] Reference Figure 1 The illustration shows two plates P1 and P2 to be welded together by a weld seam. In the described example, plates P1 and P2 are coplanar and have the same thickness e, with their respective straight edges adjacent to each other to form a joint J between them. A weld seam is formed at the joint J by advancing a nozzle 102, which emits a heat source 104 toward the joint J on one side (referred to as the positive side) of the two plates P1 and P2. The other side is referred to as the negative side. In the described example, the nozzle 102 and the heat source 104 are intended to advance at a constant welding speed VS.

[0044] The formed weld seam is Figure 2 As shown in, Figure 2 In the figure, the weld is indicated by reference numeral 202. As can be seen in the figure, weld 202 has a width L1 perpendicular to weld J on the positive side, called the positive side width, and the weld also has a width L2 on the negative side, called the negative side width.

[0045] Reference Figure 3 An example of a device 300 according to the invention will now be described. The device 300 is designed to provide parameter values ​​for a heat source 104 for forming a weld 202 between two plates P1 and P2.

[0046] In the described example, device 300 is a computer system including a processing unit 304 (e.g., a microprocessor) and a main memory 306 (e.g., random access memory (RAM)) accessible to the processing unit 304. Computer system 302 also includes a mass storage 308 (e.g., a local or remote hard disk, accessible via a communication network), in which a computer program 310 containing instructions for the processing unit 304 is stored. This computer program 310 is loaded into the main memory 306 so that the processing unit 304 can execute its instructions. The instructions of the computer program 310 are organized into software modules, which will be described below.

[0047] Alternatively, all or part of these modules can be implemented as hardware modules that do not involve computer programs, i.e., in the form of electronic circuits (e.g., micro-leads).

[0048] Database 312 is also stored in mass storage 308. Database 312 provides the melting temperature and performance characteristics of each of the various materials.

[0049] In addition, one or more reference simulations 314 are stored in mass storage 308. Each reference simulation 314 includes an input model (described below) for solver 317 to simulate a weld between two plates. Specifically, each input model includes a mesh of the plates as the subject of the simulation. The input model also includes parameters of one or more materials forming these plates and parameters of the heat sources used in the simulation, such as those detailed below. Each reference simulation 314 also includes the results of the simulation, particularly the parameter values ​​of the weld obtained through the simulation, such as those detailed below.

[0050] The device 300 also includes a human-machine interface 316, which includes, for example, an output device and an input device, the output device being, for example, a display device (e.g., a screen), and the input device being, for example, a keyboard and / or a mouse.

[0051] The modules of computer program 310 will now be described in detail. (Refer to...) Figure 4 Describe in more detail the functions performed by these modules.

[0052] First, the computer program 310 includes a solver 317. It is at least a thermal solver with transient calculation capabilities. In the described example, it is a thermomechanical solver. The solver 317 is designed to simulate the formation of a weld between two plates over time based on an input model, i.e., designed to provide at least the temperature evolution over time at each point of the mesh of the input model (which is provided to the solver).

[0053] The computer program 310 also includes an interface module 318, which is designed to receive information from a user, for example, through a human-machine interface 316.

[0054] The computer program 310 also includes an initialization module 320, which is configured to prepare input data to be provided to the solver 317.

[0055] The computer program 310 also includes a sampling module 322, which is designed to provide samples of the parameters of the heat source 104 in addition to the input data determined by the initialization module 320 and to provide them sequentially to the solver 317.

[0056] Computer program 310 also includes measurement module 326, which is designed to provide the temperature evolution over time at at least a plurality of measurement locations predefined by interface module 328 based on the output of solver 317.

[0057] Computer program 310 also includes a spatial characterization module 328, which is designed to provide a spatial characterization of weld 202 based on the evolution of temperature over time at at least the measurement location.

[0058] Therefore, for each sample, points (called simulation points) are obtained that correlate the spatial representations L1 and L2 of weld 202 with the parameters PS, GS, and VS of heat source 104.

[0059] The computer program 310 also includes an extrapolation module 330, which is designed to provide extrapolation points based on the simulation points.

[0060] The computer program 310 also includes a search module 332, which is designed to search for the target point of the function, that is, the point that makes the value of weld 202 close to the expected value received by the interface module 318. The search module 332 is also designed to provide the solver 317 with the parameter values ​​of the heat source 104 of the found target point in order to obtain a new simulation point.

[0061] Reference Figure 4 An example of a method 400 for manufacturing welds according to the present invention will now be described.

[0062] In step 402, a simulation of the weld between a pair of plates is performed.

[0063] In step 404, the simulation is compared with the actual weld in order to select the effective simulation (i.e., the simulation that faithfully reflects the actual weld).

[0064] In step 406, at least one of the simulations selected in step 404 is stored in mass storage 308 as a reference simulation 314.

[0065] In step 408, the interface module 318 receives the expected values ​​of the parameters of the weld 202 between boards P1 and P2 from the user.

[0066] Specifically, the parameters of weld 202 received by interface module 318 include the expected value of at least one spatial representation of weld 202. In the described example, the spatial representation is the positive side length L1 and the negative side length L2 of weld 202. Alternatively, the spatial representation of weld 202 may be the area of ​​the transverse surface of weld 202. In the described example, other parameters of weld 202 include one or more of the following parameters: the thickness e of plates P1 and P2, one or more materials forming plates P1 and P2, and the welding speed VS.

[0067] In step 410, the initialization module 320 retrieves the melting temperature and performance characteristics of each material received in step 408 from the database 312.

[0068] In step 412, interface module 318 receives a selection of reference simulation 314 from the user. This allows the user to select a reference simulation whose input model closely approximates the desired weld 202.

[0069] Alternatively, interface module 318 receives the geometric features of boards P1 and P2 and / or parameters of one or more materials forming boards P1 and P2 and / or welds 202 (e.g., parameters detailed above). Initialization module 320 then selects a reference simulation 314 that most closely approximates the received information.

[0070] In step 414, initialization module 320 retrieves the reference simulation mesh selected in step 412. This mesh is referred to below as the reference mesh and is denoted by M*.

[0071] In step 415, in order to determine the grid M of plates P1 and P2 (e.g. Figure 5 As shown), initialization module 320 modifies reference mesh M* based on at least one dimension of plates P1 and P2 (i.e., the thickness e received in step 408 in the described example). Points in reference mesh M* have corresponding coordinates along the thickness (referred to as reference thickness) direction of at least one of the two plates meshed by reference mesh M*. Therefore, determining mesh M includes transforming these coordinates in the reference thickness direction by isotopy at a ratio equal to the ratio between the reference thickness and thickness e:

[0072] [Formula.1]

[0073]

[0074] Where Z* is the coordinate of a point in the reference mesh M* along the reference thickness direction, and Z is the coordinate of the same point in the mesh M.

[0075] Preferably, the coordinates of points in the reference grid M* remain unchanged in other orthogonal directions, and are therefore the same as those in grid M.

[0076] Using a reference mesh M* to determine mesh M allows for the benefit of feedback from the simulations already performed. In particular, starting with a reference mesh M* that has already yielded good results increases the chances of mesh M fully converging, i.e., enabling the solver 317 to provide a representative mapping of the actual weld.

[0077] In step 416, interface module 318 receives measurement positions in the reference grid M from the user. These measurement positions constitute "virtual sensors" that are not necessarily located at points on grid M. Preferably, the measurement positions are located on a grid transverse to weld 202.

[0078] The grid is like Figure 6 As shown in the figure, the grid is denoted by G. The grid G ​​has intersections, and the virtual sensor is located at at least a portion of the intersections. Preferably, the height of the grid G ​​is at least equal to the thickness e of the plates P1 and P2. In embodiments using the grid G, the interface module 118 receives parameters of the grid G, for example, from a user. The parameters of the grid are, for example, one or more of the following parameters: horizontal pitch PH (parallel to the front and / or rear), grid width LG (perpendicular to the front and / or rear), and vertical pitch PV (perpendicular to the front and / or rear).

[0079] return Figure 4 In step 418, sampling module 322 determines multiple sample parameters of heat source 104. These parameters include at least one of the following: power parameter PS of heat source 104, multiple parameters GS characterizing the three-dimensional geometry of heat source 104, and welding speed VS. Therefore, each sample groups the values ​​of parameters PS, GS, and VS of heat source 104. Preferably, this determination of samples is performed by pseudo-random sampling. In the described example, pseudo-random sampling is a Latin hypercube. In this case, preferably, at least 30 samples are selected.

[0080] In step 420, for each sample, computer program 310 determines the values ​​of each spatial representation L1, L2 of weld 202 for the parameters PS, GS, VS of heat source 104 for the sample under consideration. More precisely, this determination uses a simulation performed on mesh M by solver 317. Thus, step 420 is able to obtain points (referred to as simulation points) for each sample that correlate the spatial representations L1, L2 of the weld with the parameters PS, GS, VS of heat source 104 as a function.

[0081] In the described example, step 420 includes steps 422, 424, and 426.

[0082] In step 422, the solver 317 receives the parameter values ​​of the heat source 104 for the current sample, the mesh M, the welding speed VS, and the melting temperature and performance characteristics of each material in plates P1 and P2. The solver 317 then performs a thermal simulation of the welding at least by solving the following thermal equation (e.g., the Fourier equation, where q represents heat here):

[0083] [Formula.2]

[0084] T: Temperature

[0085] λ: Thermal conductivity

[0086] ρ: density

[0087] Cp: ​​Specific heat

[0088] In one embodiment, in addition to thermal simulations, the simulation performed can also be a mechanical simulation.

[0089] Solver 317 then provides a mapping of weld 202 as output. This mapping indicates the evolution of temperature (and possibly displacement) over time at each point in the mesh M.

[0090] During step 424, the measurement module 326 determines the time evolution of temperature and displacement (if applicable) at each measurement location based on the mapping provided by the solver 317.

[0091] In step 426, the spatial characterization module 328 determines at least one spatial characterization of the weld 202 based on at least a portion of the time evolution of at least a portion of the measurement locations, namely, the front side dimension L1 and the back side dimension L2 in the example described. In the described example, the spatial characterization module 328 determines the highest temperature reached at each measurement location and compares this highest temperature with the melting temperature of the material at that measurement location. Then, the spatial characterization module 328 determines the number of consecutive measurement locations on the front side of plates P1, P2 where the highest temperature exceeds the melting temperature, and derives the front side length L1 of the weld 202. For example, this front side length L1 is considered to be equal to the determined number (minus 1) multiplied by the horizontal pitch PH of the grid G. Similarly, in the described example, the spatial characterization module 328 determines the number of consecutive measurement locations on the back side of plates P1, P2 where the highest temperature exceeds the melting temperature, and derives the back side length L2 of the weld 202. For example, this back side length L2 is equal to the determined number (minus 1) multiplied by the horizontal pitch PH of the grid G.

[0092] Figure 7 The diagram shows region 702 of grid G, which categorizes intersections where the highest temperature exceeds the melting temperature, and region 704, which categorizes intersections where the highest temperature is still below the melting temperature.

[0093] The use of virtual sensors enables computer program 310 to determine spatial representations L1 and L2 in a stable manner from one determination to the next.

[0094] When the cross-sectional area of ​​weld 202 is used as a spatial characterization, the spatial feature module 328 determines the area, for example, based on the number of measurement locations where the highest temperature exceeds the melting temperature (i.e., the number of intersections of grid G ​​included in region 702), the horizontal pitch PH, and the vertical pitch PV.

[0095] In other embodiments, the evolution of displacement over time at at least some measurement locations can also be used to determine the spatial characterization of weld 202.

[0096] Back Figure 4 Therefore, a set of simulation points is obtained at the end of step 420.

[0097] Then, method 400 includes multiple successive iterations of steps 428, 430, and 432.

[0098] During step 428, the extrapolation module 330 determines the point called the extrapolation point of the function by extrapolating from the simulation point.

[0099] In step 430, the search module 332 determines a point of the function called the target point, at which each spatial characterization L1, L2 of the weld 202 has a value close to the expected value.

[0100] In step 432, the spatial characterization module 328 determines the values ​​of each spatial characterization L1, L2 of the weld 202 based on the values ​​of the parameters PS, GS, VS of the heat source 104 for the target point (e.g., in the same manner as described above). Thus, a new simulation point for the function is obtained, which combines the values ​​of the parameters PS, GS, VS of the heat source 104 for the target point with the values ​​of each spatial characterization L1, L2 of the weld 202 obtained in step 432. This new simulation point complements the other simulation points used for the next iteration.

[0101] Preferably, the repetition of previous steps 428, 430, and 432 is stopped when a predetermined condition is met (e.g., after a predetermined number of iterations), or when the values ​​of the spatial representations L1 and L2 obtained in the last iteration are very close to the values ​​obtained in the previous iteration.

[0102] In step 434, computer program 310 provides the values ​​of parameters PS, GS, and VS of the heat source 302 at the target point obtained in the last iteration of steps 428, 430, and 432.

[0103] In step 436, weld 302 is manufactured by heat source 104 parameterized according to values ​​provided by computer program 310.

[0104] Clearly, the method described above enables the acquisition of parameters for the heat source used to create the desired weld.

[0105] It should also be noted that the present invention is not limited to the embodiments described above. In fact, it will be apparent to those skilled in the art that various modifications can be made to the embodiments described above based on the teachings disclosed herein.

[0106] In the detailed description of this invention above, the terminology used should not be construed as limiting the invention to the embodiments set forth in this specification, but should be interpreted to include all equivalents that could be contemplated by a person skilled in the art by applying their general knowledge to the teachings of this disclosure.

Claims

1. A method (400) for generating a weld (202) between two plates (P1, P2) by a heat source (104), said heat source (104) having at least one of the following parameters: a power parameter (PS) of said heat source (104), a plurality of parameters (GS) characterizing the three-dimensional geometry of said heat source (104), and a welding speed (VS), characterized in that, The method includes the following steps: - Receive (408) the expected values ​​of the spatial characterization (L1, L2) of the weld (202); - Determine multiple samples of the parameters of the heat source (104) as described in (418); - For each sample, the values ​​of the spatial representation (L1, L2) of the weld (202) for the sample are determined (420) by simulation on the three-dimensional mesh (M) of the two plates (P1, P2) in order to obtain the simulation points of the function relating the spatial representation (L1, L2) of the weld (202) to the parameters of the heat source (104); - Perform multiple consecutive iterations of the following steps: • The extrapolation point of the function (428) is determined by extrapolation from the simulation point. • Determine the target point of the function (430), at which the spatial representation (L1, L2) of the weld (202) has a value close to the expected value, and • By performing simulations on the three-dimensional mesh (M) of the two plates (P1, P2), the values ​​of the spatial representation (L1, L2) of the weld (202) are determined (432) based on the values ​​of the parameters of the heat source (104) for the target point, so as to obtain new simulation points to supplement other simulation points; - Provide the values ​​of the parameters of the heat source (104) at the target point obtained in the last iteration (434); - Parameterize the heat source (104) according to the provided values; and - The weld is generated by the heat source (104) parameterized according to the provided values.

2. The method (400) according to claim 1, further comprising the step of receiving (416) the measured position in the three-dimensional mesh (M), wherein, The values ​​of the spatial characterization (L1, L2) of the weld (202) are determined based on the time evolution of the temperature obtained at each measurement location.

3. The method (400) according to claim 2, wherein, The measurement position is located at the intersection of the grid (G), and the step of receiving (416) the measurement position in the three-dimensional grid (M) includes receiving at least one of the pitch (PH, PV) of the grid (G) and the size (LG) of the grid (G).

4. The method (400) according to any one of claims 1 to 3, further comprising the following step: - Receive (408) the dimension (e) of at least one of the two plates (P1, P2); as well as - The three-dimensional mesh (M) is determined by modifying the reference mesh of the two plates according to the received dimensions (e).

5. The method (400) according to claim 4, wherein, The received dimension (e) is the thickness of at least one of the two plates (P1, P2), wherein the reference grid includes points having coordinates in the direction of the thickness of at least one of the two plates meshed by the reference grid, the thickness being called the reference thickness, and wherein the step of determining (415) the three-dimensional grid (M) includes: transforming these coordinates by isotopity with a second ratio equal to a first ratio of the reference thickness to the received thickness (e).

6. The method (400) according to claim 5, wherein, The modified reference mesh is selected from a set of reference meshes.

7. The method (400) according to claim 6, wherein, The set of reference meshes has been used in advance in the corresponding reference simulation (314) of the weld, which is verified by comparing it with the actual weld (404) respectively.

8. The method (400) according to any one of claims 1 to 3, wherein, The steps to determine the (420) sample are performed by pseudo-random sampling.

9. The method (400) according to claim 8, wherein, The steps to determine the (420) sample are performed using the Latin hypercube sampling method.

Images