WO2021148761A1

WO2021148761A1 - Optimization of a method for determining shot-peening parameters by learning

Info

Publication number: WO2021148761A1
Application number: PCT/FR2021/050122
Authority: WO
Inventors: Jawad BADREDDINE; Pierre Julio RENAUD
Original assignee: Safran
Priority date: 2020-01-24
Filing date: 2021-01-25
Publication date: 2021-07-29
Also published as: FR3106676B1; FR3106676A1

Abstract

The invention relates to a method comprising the following steps: a) providing the shot-peening specifications of the part, comprising at least the areas of the part to be peened; b) providing a three-dimensional object of the part, which is modeled by finite elements using a three-dimensional mesh of the part; c) determining one or more combinations of shot-peening intensity values and coverage rates for each area.

Description

DESCRIPTION

TITLE: Optimization of a process for determining shot peening parameters by learning

Technical field of the invention

The invention relates to the operations of shot peening of parts intended for the aeronautical field, and in particular to the optimization of the setting of shot blasting machines.

State of the prior art

Shot peening is widely used in the aerospace industry. This is a mechanical surface treatment also considered a mechanical strengthening process.

This operation consists of plastically deforming a surface layer of the surface of a part by throwing balls at high speed. Upon impact, each ball forms a local recess on the surface of the part. Shot blasting operations have the following technical effects:

-improve the fatigue life of treated industrial parts,

- compensate for the reduction in service life induced by the surface treatments of the parts,

- homogenize the machined surface finish of the parts.

In order for the shot peening operation to induce homogeneous surface properties of the treated part, the surface must be completely covered by the impacts of balls. The best results are obtained with a homogeneous shot (for example in hardened and tempered steel, the hardness of which must be greater than that of the treated parts) projected as perpendicularly as possible to the surface to be treated. It is known to also use glass or ceramic balls, these balls also being called shot.

In the field of aeronautics, shot peening of prestressing is commonly applied to complex metal industrial parts such as blades, honeycombed turbine discs, one-piece bladed discs, turbine shafts, landing gear components, wheels, pinions.

The blasting intensity is quantified in Almen according to the Almen method.

The determination of the parameters of a shot peening process for a complex industrial part, that is to say the parameters entered into the machine, involves several steps.

The first step is to analyze the specification on the plan specific to the part to be treated by the shot peening process. This document, produced by engineers, includes information relating to the shot peening of the part, for example:

- the areas of the part must be shot blasted (therefore which areas should not be shot blasted), - the type of shot to be used,

- the Almen intensity to be used for each of the zones, and

- the coverage rate for each of the areas to be shot blasted.

The second step consists in choosing the first machine parameters. This step is generally carried out by specialists, able to determine, by experience, the machine parameters that can be used to meet the specifications in the plan of the part.

A third step consists in measuring, following the shot blasting of the part, the following quantities:

- coverage and intensity of shot blasting for each of the zones;

- distortions of the induced part.

This process, although not presenting any particular difficulties, relies solely on the expertise of shot blasting specialists, their tests on parts and on dummies, which poses an obvious problem of reproducibility since the operation depends on 'an operator. Moreover, this can be problematic, especially when these specialists leave companies.

Also, certain requirements, in particular those related to residual stresses, to the coverage rate, to surface roughness, or to the dimensional tolerances of the part, cannot be controlled before the part is subjected to a shot peening process. The control of these requirements can thus only be done a posteriori, thus requiring several iterations to find the optimal machine parameters with regard to the requirements of the specifications at the plan of the part.

In addition, the geometric complexity of certain areas to be blasted (for example cells in a disc, holes, internal diameter of power transmission shafts, etc.) can make the control of the blasting (Almen intensity, overlap, roughness, stresses residual) difficult, if not impossible, in these areas.

It quickly becomes apparent that the selected blasting parameters are not necessarily the optimal parameters for the desired blasting operations. The choice of the optimal parameters would require a lot of time to carry out shot blasting tests with all the possible combinations of parameters and to carry out measurements (for example of residual stresses, dimensions of the part, etc.) after the shot blasting operations. to select the combination of parameters that best meet the shot peening specifications.

This document aims, initially, to limit the shot blasting tests, allowing to deduce the good parameters of the shot blasting machine allowing a shot blasting in accordance with specifications in the plan of the part, and to propose a solution making it possible to determine the optimal parameters. to achieve optimal shot peening of the part. One Another objective of the invention is to provide a solution making it possible to capitalize on the know-how relating to the choice of machine parameters for shot blasting.

More generally, the aim of this document is to provide a simple, effective and economical solution to the drawbacks of the current technique described above.

Presentation of the invention

To this end, a method is proposed for optimizing shot-blasting parameters by learning for a part, the method comprising the following steps: a) providing the shot-peening specifications of the part comprising at least the areas of the part to be shot-blasted, and for each of these zones, a range of shot peening intensity values, a range of coverage rate values and a range of maximum overflow distance values; b) Provide a three-dimensional object of the part, modeled by finite elements using a three-dimensional mesh of the part; c) Determine one or more combinations of blasting intensity and coverage rate values for each zone;

For each combination: d) calculate a profile of residual stresses a ^s (z) = f {C; /) on the basis of shot-blasting intensity I and coverage rate C values, using a database; e) Calculate a residual stress tensor from the residual stress profile and as a function of the distance z to the nearest surface; f) Calculate, at each point of the three-dimensional mesh, the induced distortion, after a static balancing of the part, from the tensor of residual stresses and from the static balancing; g) Select the combinations, for each zone, for which the distortions induced at each point of the mesh conform to the dimensional tolerances of the part;

For each combination selected: h) Determine the parameters of the shot blasting machine; i) Simulate the shot peening with the shot blasting machine parameters of the selected combinations, for each zone, on the three-dimensional modeled object of the part; j) Select, for each zone, the combination with the shortest blasting time; k) Check that the deformations distortions induced following the simulation of Shot peening on the three-dimensional modeled object of the part conforms to the dimensional tolerances of the part.

This method, which can be intended to be implemented on a computer, proposes a learning method making it possible by simulation to optimize the determination of the parameterization ranges of the shot blasting machines and those from the specifications on the plane of the shot peening part. , a first and a second database.

Unlike the prior art, it is possible by simulation to know the effects induced by the shot peening operations on the treated part. Thus, it is possible to determine whether the parameters chosen comply with the requirements explained in the specifications in the plan of the part, in particular the requirements relating to the Almen intensity to which the part has been subjected, to the distortions, to the residual stresses, to the coverage rate, surface roughness, or dimensional tolerances of the part.

The simulation of all the parameters also makes it possible to select the optimum parameters with regard to the requirements explained in the specifications on the part plan.

This process therefore makes it possible to validate by calculations and simulation that the requirements of the specification in terms of the part are met at all points of the part, and this before carrying out a real test on the part. This allows a significant saving of time, since the actual tests on the part are carried out with the shot blasting parameters of the shot blasting machine validated by simulation. Thus, the actual tests on the part consist of a step of validating the parameters determined by the process.

The choice of the optimum parameters aims to determine in all the selected shot blasting ranges, to what extent these ranges can be implemented on the shot blasting machine, and the cycle time of the shot blasting machine with such parameters.

Thus, this process makes it possible to choose the best parameters of the shot blasting machine and therefore the best shot blasting strategies, while ensuring reduced distortions of the part following the shot blasting. Thus, for each of the areas to be shot peened in the part, we make sure that the optimum parameters are applied. This process translates design office sizing requirements (eg required residual stresses, etc.) into optimal shot blast parameters. Thus, this method makes it possible to take into account the distortions induced by the earlier shot peening, in particular during the dimensioning of the parts. It is therefore a question of making it possible to increase the lifespan and improve the performance of these parts.

This process also makes it possible to increase productivity, by choosing shot blasting machine parameters that reduce the occupation time of the shot blasting machine per part and the tests on parts, the latter being required in number for the production of a range. For each part produced in series, provision can also be made to keep the shot-blasting parameters, thus enabling the capitalization and perpetuation of know-how in terms of the shot-blasting operation.

This process further reduces the use of the shot blasting machine and the number of part tests to determine the shot blasting parameters required to put a series of parts into production.

Of course, this process is applicable to the shot blasting of any type of metal alloy parts.

Additionally, the shot peening specifications used in the process may include one or more of the following: i. parts of the part to be blasted; ii. parts of the part prohibited from shot blasting; iii. parts of the part for which shot blasting is admissible; iv. the type of shot; v. mechanical and dimensional characteristics of the shot; vi. a blasting intensity range for each of the parts to be blasted; vii. a range of minimum and / maximum coverage rates for each of the parts to be shot blasted; viii. minimal residual stress profile; ix. maximum permissible surface roughness; x. dimensional tolerances of the part.

These specifications are specific to each type of part. Indeed, depending on their condition of use, the operations of shot peening of the parts differ. The requirements to be fulfilled by the parts obtained as a result of the shot peening are input data to the process, so these requirements are taken into account when selecting machine parameters and then following the simulation of these machine parameters on a model. three-dimensional workpiece to ensure that some of the requirements (for example, minimum residual stress profile, part dimensional tolerances and allowable surface roughness) are met.

The areas to be blasted can be divided into sub-areas for which steps c) to j) are applied.

It is therefore a matter of defining one or more zones. Each of the zones is to be shotblasted with the same mask and with an admissible overflow and for which an admissible Almen intensity and overlap range is associated. This involves dividing the areas to be shot blast into areas compatible with the shot blast machine.

Step b) may further include the following steps:

- Carry out a two-dimensional mesh of the surface of the modeled object of the part. Step k) can include the following steps:

- Calculate for each point of the three-dimensional finite element mesh of step b), the distance z to the nearest surface;

- Calculate for each point of the two-dimensional mesh the residual stress profile o ^r _x ^e x according to z from the combination of parameters of the shot blasting machine;

- Calculate the initial residual stress tensor s ° (z) for each mesh point, according to the combination of parameters of the shot blasting machine;

- Calculate the static balancing of the part by finite elements from the three-dimensional modeled object of the part and the elastic properties of a component material of the part;

- Calculate, for each point of the three-dimensional mesh, the induced distortion, after a static balancing of the part, from the residual stress tensor and the static balancing.

Since shot peening involves distortions and deformations of the part, step k) consists of calculations in order to determine for any point in the volume of the part the residual stresses resulting from the simulated shot blasting operation. These calculations help to obtain a model of the deformed part following the simulated shot peening. The residual stress tensor can be determined using a third database, relating the residual stress tensor to the parameters of the shot blast machine. This third database can also enable the capitalization of the company's know-how. The identification of stress profiles can be based on charts, in which the residual stress profile is related to the blasting conditions. The calculation of the stress tensor resulting from the simulated shot peening operation is carried out for each point of the mesh in the volume of the part.

The modeling of the post-shot peening part also takes into consideration the static balancing of the part following the shot peening, and therefore allows a three-dimensional model to be obtained, where the distortions induced by the shot peening are thus taken into account. following shot blasting.

This step then makes it possible to visualize a prediction of the strains induced by the shot peening with the shot peening parameters selected and used for the simulation. It is then possible, without carrying out actual shot peening, to visualize the extent to which the shot peening parameters induce distortions, and whether these are acceptable with respect to the tolerances acceptable for the shot peened part.

In addition, step c) can use a statistical selection such as, for example, the Monte-Carlo algorithm.

This document also relates to a computer program comprising instructions for implementing the method as described above. Such a program thus makes it possible to simulate the shot peening of a part with shot blasting parameters chosen in line with the shot peening requirements specific to the part to be shot peened, to visualize the distortions and geometric deformations of the part taking into consideration the material. component of the part to be blasted, and whether or not to validate the blasting parameters if the geometric deformations are within the accepted tolerances.

Also, before any part test, the program makes it possible to select the optimal blasting parameters, that is to say allowing a faster blasting, that is to say having the shortest blasting time.

Thus, during the shot peening tests on the part, the parameters of the shot blasting machine are therefore a priori chosen so that the shot blasting meets all the requirements in terms of the part, and so that the shot blasting is as fast as possible.

This therefore makes it possible to simulate the shot blasting parameters and to determine, before shot blasting a part in real conditions, whether these shot blasting parameters comply with the shot blasting specifications.

Thus, unlike the prior art, shot peening in real conditions on a part is carried out only with parameters which may be, depending on the simulation, in line with the shot peening specifications. This therefore reduces the number of tests on a real part, and therefore effectively reduces the number of iterations required to determine the correct shot peening parameters for a given part by a person with no shot peening experience.

Therefore, it is also possible through this program to share knowledge relating to shot peening of parts, so that those in charge of shot blasting can draw on the experience of others, including simulations. parameters can be stored in memory.

This document also relates to a data storage medium on which the aforementioned computer program is recorded.

In addition, the present document relates to a CAD system comprising a processor coupled to a memory and a graphical user interface and able to communicate with a shot blasting machine, such as the aforementioned program is stored in memory. Brief description of the figures

[Fig. 1] represents a flowchart of the process according to this document.

[Fig. 2] shows an example of shot blasting specifications for an example part to be shot blasted.

[Fig. 3] shows a detailed flowchart of the process according to this document and distributed over three parts A, B and C which when they are arranged side by side form the detailed flowchart of the process, part B being interposed between part A and part. vs. [Fig. 4], part A of which represents a three-dimensional modeling of the part to be shot peening and part B represents a three-dimensional modeling of a portion of the part to be shot peened.

[Fig. 5], part A of which is a three-dimensional representation of a shot peening simulation of the part to be shot blasted.

[Fig. 6], part A of which represents a map of the Almen intensity observed by the surface of the portion of the part for which the shot peening is simulated for the parameters of the machine's shot peening and part B of which represents a map of the coverage rate observed by the area of the portion of the part for which the shot peening is simulated for the machine shot peening parameters.

[Fig. 7] comprises two parts, part A illustrating the residual stress field after shot peening (balanced in the part) and part B illustrating the deformation field of the part after shot peening (comparison between the initial and deformed geometry (factor 20 on the deformations), this last part B illustrates the displacement before and after shot peening

[Fig. 8] represents a diagram of the steps for calculating the residual stresses induced by the shot blasting operations.

Detailed description of the invention

Figure 1 illustrates the general principle of the method according to this document is to optimize a method for determining, by learning, shot blasting parameters of the shot blasting machine.

The shot blasting parameters of the shot blasting machine are so-called machine parameters, and can be for example:

- distance of the nozzle from the part (nozzle),

- angle of the shot jet,

- distance traveled by the nozzle (d working),

- pressure of the jet, in particular the pressure of the compressed air which makes it possible to project the jet through the shot-blasting nozzle;

- flow rate of the shot at the nozzle outlet,

- nozzle movement speed,

- number of cycles, that is to say the number of round trips of the nozzle in relation to the treated area.

The shot peening resulting from a combination of these machine parameters of the shot blasting machine, can be characterized, for example, by the Almen intensity of the shot peening, the coverage rate, the residual stress profile in the part.

The part to be shot-blasted can be any part, made of metal or of a metal alloy, which may or may not be integrated into an aircraft. In the field of aeronautics, parts present in turbomachines, landing gear or even in power transmission equipment can be shot peened.

This part can present symmetries, in this case, these geometrical symmetries are exploited in the course of the process.

The first step A consists in providing the shot peening specifications of the part comprising at least one area of the part to be shot peening, and for each of these areas, a range of shot peening intensity values, a range of coverage rate values and a range of maximum overflow distance values.

Shot peening specifications bring together all of the information needed to shot peen a part so that it is usable and within dimensional tolerances. In particular, the specifications ensure the shot peening of the part does not induce a problem with the mechanical behavior of the latter. An example shot blasting specification 2 is shown in Figure 2. As can be seen, the specification includes information such as: i. parts 4, 6 of the piece to be shot blast. ii. parts 10 of the part prohibited from shot blasting; iii. parts of the part for which shot blasting is admissible; iv. the type of shot; v. mechanical and dimensional characteristics of the shot; vi. a blasting intensity range for each of the parts to be blasted; vii. a range of minimum and / maximum coverage rates for each of the parts to be shot blasted; viii. minimal residual stress profile; ix. maximum permissible surface roughness; x. dimensional tolerances of the part.

In the illustrative example of the process, the part to be shot peening is a turbine disk, constituting a sealing labyrinth. In the parts of the part to be shot blast, different parts can be distinguished in the example of Figure 2. The first 4 and second 6 types of part designate areas to be shot blast of different Almen intensities. A third type 8 of part designates the zones, which should not be shot blast, but for which shot rebounds are authorized. The parts 10 of the part prohibited for shot blasting are parts where even shot rebounds are not tolerated. When shot blasting the part, so-called tight protections around these parts are used.

The type of shot can be steel, ceramic or glass balls of more or less variable diameters. The choice of shot combined with the shot peening intensity then impacts the surface roughness obtained following shot peening. The coverage rate is also specified: it is the ratio of the shot-blasted area to the area to be shot-blasted. To ensure complete shot blasting of the part, the minimum coverage rate is generally chosen from the range of values between [1.25; 1.5].

For a combination of a given Almen intensity and recovery rate, several combinations of shot peening machine parameters are possible, and lead to similar shot peening results. The object of this process is to provide the combination of shot peening parameters, chosen from several combinations of shot peening parameters to achieve shot peening in accordance with specifications, reducing shot peening time and therefore allowing optimized shot peening. The object of the following stages, B, C, D, E, F and G, is to determine the states of admissible stresses for couples of values of intensity Almen and rate of recovery, in conformity with the specifications in the plan, at the end shot blasting. At the end of these steps, it is possible to determine which torques imply distortions of the part conforming to the dimensional tolerances of the part specified in the requirements in the drawing.

In the second step B, a three-dimensional model of the part is provided, produced by finite elements using a three-dimensional mesh of the part and subdividing the three-dimensional model into a plurality of areas. This can also be done using a two-dimensional model of the part. The model obtained will then be used to perform the shot peening simulation for given combinations of parameters.

This second step consists in particular in providing a model modeled according to the finite element method. The use of the finite element method is necessary to allow quantification of the deformations of the part following the shot peening simulation.

For this, a three-dimensional model of the part 12 (which can be obtained via a computer-aided design tool) is imported into software using the finite element method, such as, for example, the ABAQUS® software. The three-dimensional model 12 of the disc here considered as the part to be shot peening is illustrated in FIG. 4A. As can be seen, the disc 12 is an annular part, having a central symmetry.

Depending on whether the geometry of the part is suitable or not, it is possible to model only a portion of part 12 to simplify the modeling. In the example, the part being a disc, the symmetry of revolution of the part makes it possible to simplify the problem by considering only a portion 14 of it. A three-dimensional model 14 of the portion considered is then used to perform a mesh, as illustrated in FIG. 4B. This method then applies to any axisymmetric part.

Once the three-dimensional model of this portion 14 has been obtained, the three-dimensional model is enriched with data defining the elastic properties of the material in which the disc is designed as well as the boundary conditions of the portion. Data related to the elasticity of the material can be, for example, the Young's modulus and the Poisson's ratio of the material used. As for the boundary conditions, the symmetry, the clamping and the forces to which the part is subjected are taken into account.

A mesh of the part portion 14 is subsequently calculated on the surface of the part portion, so as to define at least 5 elements, preferably 10.

A set, representing the mesh, comprising the nodes as well as the surface elements constituting the mesh of the portion 14, is recorded.

In parallel, or subsequently, a file is generated by the software using the finite element method, including balancing calculations.

In the particular case where a portion 14 of part 12 exhibiting symmetries has been modeled, step B further comprises a step aimed at establishing an axisymmetric two-dimensional model of part 12. In particular, in this step, a two-dimensional model of the part is supplied as input to the modeling software according to the finite element method. As for the first part of the second step B, the two-dimensional model is then used to produce a mesh of the part, dividing the surface of the part into sub-surfaces. The whole constituting the mesh, that is to say the nodes and the surface elements are recorded. In addition, the two-dimensional model is enriched with data relating to the elastic properties of the constituent materials of the part. Data related to the elasticity of the material can be, for example, the Young's modulus and the Poisson's ratio of the material used. As for the boundary conditions, the symmetry, the clamping and the forces to which the part is subjected are taken into consideration.

Also, at the end of this second step B, the surface of the part 12 to be shot peened is divided into zones, hereinafter called macro-zones. In particular, each macro-zone to be blasted is considered individually, and in the case where the macro-zones to be blasted have a large area, they are divided into elementary zones for which instructions for the blast machine are determined individually. It is of course conceivable that the part 12 has only one zone to be blasted with one or more macro-zones and one or more elementary zones.

The third step C of the method comprises determining the admissible stress states for each of the macro-zones.

For each of the macro-zones, the minimum stress profile to be respected, the range of minimum and maximum Almen currents, as well as the range of minimum and maximum coverage rates are identified from the requirements of the room plan.

From these data, a statistical extraction is carried out simultaneously in the ranges of Almen intensities and of recovery rate, specified in the specification in the plan. This Statistical drawing makes it possible to form pairs of Almen intensity / coverage rate (called intensity / coverage pairs hereinafter in the document) admissible for each macro-zone of the part. For example, the statistical selection can be carried out with the Monte-Carlo algorithm.

For each of the intensity / recovery pairs, a test matrix is generated. This test matrix is a digital test plan to be carried out, that is to say a list of calculations to be carried out to determine the post-shot peening states of matter for each of these pairs of parameters.

Thus, subsequently, in step D, a state of residual stresses a ^s (z) is associated with each test matrix. This is either calculated using the Math formula. 1, or determined using a first database, linking an intensity / recovery pair with a state of residual stresses. This first database can be a first numerical or experimental chart.

[Math 1]

As can be seen, the computation of the stress states is carried out under the assumption of a semi-infinite mass, which means that the stress state at a point of the part depends only on the depth of this point at the surface of the room.

Thus, in the sixth step E of the method, for each test matrix, that is to say for each intensity / overlap pair, a distance z is calculated from each point of integration of GAUSS of the mesh to a surface element of the closest mesh. The surface elements are noted E _n for n elementary zones. GAUSS integration points are an integral part of the finite element method. At these GAUSS integration points, several differential equations are solved (by integration) so as to obtain the residual stresses which are induced by shot peening and which are responsible for the distortions of parts. Figure 8 illustrates the method of calculating the distance to the surface of the integration points of the mesh according to the finite element method. This is done for each macro-zone defined previously, and this in parallel.

The distance z, from each element 34 of the part, corresponds to the shortest distance to the surface among the surface elements Ei, E2, E3 of the mesh located nearby. In the example illustrated, the distance z is determined using the surface element E2, the distances zi and Z2 being greater than the distance Z3. The distance z is then the distance between the orthogonal projection of a point 36 of the element 34 in the part and the point 36 of the element 34 in the part as can be seen in figure 8. From this distance, the residual stress profile o ^r _x ^e _x ^s at each point z is determined from the stress states for the given intensity / overlap pair determined in step D.

In the fifth step E, an initial stress tensor is calculated for each point of the volume of the part to be shot peened.

This can be done by two mathematical methods. It is conceivable to carry it out with other known mathematical methods applicable to the present case.

The first method is the so-called "Eigen Strain" method, taking into account the inhomogeneity of thermal expansion, following shot blasting operations.

The plastic strain induced by shot peening is calculated for each GAUSS point of the mesh. This calculation is made from the following equation:

[Math 2]

Then, the calculation of the initial state to be introduced into the part consists in determining the equivalent stress tensor. The initial residual stress tensor s ° (z) is calculated for each mesh point, corresponding to the GAUSS integration points, according to the combination of shot blasting machine parameters, according to the following equation:

[Math 3] s ° ( _z ) = - 2nd ^r _z )

[Math 4]

This calculation of the initial residual stress tensor s ° (z) is carried out using a second database, in which the shot peening control parameters (shot peening intensity, shot peening intensity, etc.) are linked to the shot peening intensity. semi-infinite solid residual stress profile (a _xx ^res (z)). This second database also serves to capitalize on the company's know-how, in particular shot blasting professionals. Once these calculations have been carried out, a file is generated, so as to contain all the information necessary for the calculation of the static balancing of the part following the shot peening, that is to say the plastic deformation induced by the shot peening as well as the initial state.

The second method that can be implemented is the method of thermal fields. Thus, the elastic balancing of the part, in other words the final residual stresses, can be determined. A mechanical elastic balancing calculation is performed to determine the final residual stresses, by the finite element method, using Math Equation 5, the generic formula for the final residual stress tensor.

[Math 5]

The sixth step F consists in calculating, at each point of the three-dimensional mesh, the induced distortion, after a static balancing of the part, from the residual stress tensor and the static balancing.

This step aims to determine to what extent, for each test die, a shot peening with the intensity / overlap pair in question would impact the geometry of the part, two methods can be used to calculate the elastic balancing. Using the Math equation. 5, one calculates the relative displacement of the points Z in the volume relating to the given macro-zone.

Once, the final position of the points of the volume relative to the macro-zone is obtained, the final dimensional, for the test matrix for a given intensity / overlap couple and a given macro-zone, is analyzed and compared with the dimensional tolerances of the part supplied in the specifications to the plan in a step G.

If the dimensional tolerances are not respected, then another test matrix is chosen, that is to say another current / recovery pair, and steps E, F and G are repeated, until a matrix of test resulting in distortions respecting the dimensional tolerances of the specifications on the plane, for the given macro-zone. Once a test matrix is identified for a given macro-area, it is saved to a file. For each of the elementary zones, all the test matrices resulting in distortions respecting the specifications on the plane are thus recorded in a file.

Once all the test matrices have been identified for a given macro-zone, another macro-zone is then considered, for which steps B, C, DE, F and G are carried out, so as to identify all the test matrices in agreement with the specifications on the plan. In the same way, a test matrix, whose induced distortions are in agreement with the dimensional tolerances of the specification on the plane, is identified for each of the elementary zones. Each time a test matrix is identified, it is saved in the file.

Thus, once all the test matrices have been identified for each of the elementary zones, the file includes all the intensity / overlap pairs to carry out the shot peening of the part considered. In the eighth step H, from a second database linking the intensity / overlap pairs to the machine parameters of the shot blasting machine, for each of the elementary zones, the parameters of the shot blasting machine are extracted.

At this stage, work instructions are defined for each macro-zone, specified in the plan and machine and geometric constraints linked to the shot peening of the part. For example, a macro-zone can be the subject of three unit work instructions, suitable for shot peening elementary zones of the macro-zone. There is then one work instruction per elementary zone. Each work instruction, for a given elementary zone, may correspond to different machine parameters of the shot blasting machine, which must, however, respect the specifications in the plan and the machine and trade constraints linked to shot blasting (for example, the accessibility of the jet, the shot peening angle, etc.).

For each of the work instructions:

The second database is used to determine from the current / overlap couple, and other information from the specification to the drawing such as dimensional tolerances, influencing machine parameters of the shot blasting machine (Pressure; Mass flow rate of media; angle of incidence; processing time; robot speed ...).

For example, for a work instruction relating to an elementary zone of a given macro zone, the list of admissible blasting parameters {a, (angle of the jet); d, (distance of the nozzle from the part); Q, (flow rate of the shot at the nozzle outlet); P, (jet pressure); V, (nozzle movement speed); N _c (number of cycles); dt, · (working distance traveled by the nozzle); etc.} is identified, where / represents the number of allowable combinations, that is, the number of test matrix.

- The combination of admissible shot blasting parameters with the shortest processing time (shot peening time) is retained (and which also respects machine and business constraints).

Thus, for each of the work instructions, combinations of shot blasting parameters are kept and then simulated during the ninth step. The set of shot blasting combinations kept is thus the optimum machine parameters of the shot blasting machine for each of the work instructions. At the end of this step there is an optimal combination of parameters per work instruction.

The ninth step of process I aims to simulate the optimal shot blasting machine parameters retained. The simulation consists in simulating the projection of a shot jet at each point of said zone and deducing therefrom a surface condition at each point comprising information on Almen intensity and coverage rate.

First, the simulation volumes of the shot 22 jet are defined. These volumes serve to delimit the space in which the shot (the shot peening balls) can be simulated. The dimensions and geometry of these volumes are chosen so as to encompass the elementary zones of the part 12 and the volume swept by the shot jet 22 following the trajectory of the nozzle 24. Several volumes are created as a function of the shot peening strategies and of the geometry of the part 12, in other words of the number and position of the elementary zones defined in step A with respect to each other. The volumes are created using a copy of the three-dimensional model from step two B. Once the volumes are defined, a surface mesh is produced using the finite element method. The information relating to the mesh (that is to say the vertices and surface elements delimited by segments connecting the nodes) of the different volumes are stored in different files.

Secondly, the projection of a shot jet is simulated at each point of said zone and deduce therefrom at each point a surface condition 24 comprising information on Almen intensity and coverage rate.

For this, for each of the elementary zones defined previously and for each work instruction, the parameters obtained from the shot blasting machine using the abacus are used as input data for the simulation of the shot blasting machine instruction for a given area. Thus, in this step, the current work instruction is precisely simulated for the elementary zone considered with the parameters retained at the start of the step. A user interface has been developed to visualize the simulation, as shown in figure 5.

It can therefore be seen in FIG. 5 that the trajectory of each of the balls of the shot of the jet 22 of the shot blasting machine for each instruction for each of the elementary zones of the part is thus simulated.

Although not shown, it is possible to obtain a representation of the recovery rate for the area considered.

For each elementary zone (that is to say for each work instruction), therefore, a shot blast simulation is carried out for each of the test dies (and therefore of the identified shot blast parameters).

A next step in the process is to combine the surface states of the elementary areas of the part in order to obtain an Almen intensity and coverage rate map for the three-dimensional model.

The simulation of the shot peening work instructions being carried out successively or in parallel for each of the elementary zones, when all the simulations for each of the elementary zones are carried out, all the data are consolidated to analyze and visualize the Almen intensities as well as the recovery rate obtained for the optimized parameters of the shot blasting machine. FIG. 6 represents the recombination of the simulation results for each of the elementary zones. To facilitate the analysis, a mapping 30 of the Almen intensity and a mapping 32 of the coverage rate on a two-dimensional section of the three-dimensional model of the part 12 is carried out. With each color corresponding to a range of Almen Intensity or Coverage Rate, it can be easily seen whether the Almen Intensity and / or Coverage Rate meet specification requirements. Compliance or non-compliance criteria can also be displayed to the user, based on simulation results and plan requirements for the area in question.

Such mapping is instructive, because it makes it possible to obtain additional information, in particular on the overlap for the junctions separating the elementary zones.

Thus, it is checked locally to verify whether the simulated Almen intensity and the coverage rate comply with the shot peening intensity and coverage rate ranges included in the shot peening specifications. In other words, it is validated that the selected shot blasting machine parameters allow shot peening of the part according to the shot peening requirements explained in the specifications.

If locally, the Almen intensities and / or the coverage rate are not in line with the shot peening specifications, the work instructions concerned for the elementary areas in which the Almen intensity values and / or the coverage rate are not compliant are identified. In this case, for each work instruction, the combination of machine parameters of the shot blasting machine with the lowest processing time is chosen. The simulation is then carried out, and as long as the Almen intensity values and / or recovery rate do not comply with the requirements of the plan. The combination of the parameters with the shortest processing time and the values of Almen currents and simulated recovery rates in line with the requirements of the plan is then selected.

Once all Almen intensities and coverage rate locally meet the shot peening specifications, for the machine parameter combinations of the shot blasting machine, the eleventh step K is performed.

The eleventh step K consists in determining the three-dimensional model of the part 12 following the shot peening, taking into account the deformations induced by the shot peening.

For that initially, the stress tensor is calculated starting from the cartography. The stress tensor is then applied to each of the constituent points of part 12.

This is done using a stress tensor that is calculated from the mapping. The stress tensor is then applied to each of the constituent points of part 12.

As for the fifth step, the computation of the stress tensor is carried out under the assumption of a semi-infinite mass, which means that the stress tensor does not depend on that of the depth z. Computation is carried out similarly to step F, except for the computation of the initial residual stress tensor.

Indeed, the calculation of the initial residual stress tensor s ° (z) is carried out using a third database, in which the machine parameters of the shot blasting machine are linked to the semi-infinite massive residual stress profile ( a _xx ^res (z)). This third database also serves as a capitalization of the company's know-how, in particular shot blasting professionals. This calculation is followed by the mechanical calculation of static balancing, as detailed in step F.

At the end of this calculation, we can then obtain the modeling of the strain induced following the simulated shot peening, and then visualize it as illustrated in Figure 7.

The initial shape 40 of the part before shot peening is illustrated and the shape of the part after shot peening 42 is also illustrated.

The twelfth step L consists in dimensionally comparing the post-shot peening three-dimensional model to the three-dimensional model of the part.

A color code can identify / visualize to what extent these deformations are acceptable with regard to the tolerances imposed for the part in question.

If the dimensional tolerances do not conform, the process is carried out again from the eighth step H, for the areas for which the tolerances are not respected.

At the end of these steps, the process is repeated so as to generate the entire range of parameters for which the shot peening specification requirements are met. In addition, the final range of parameters is optimized: this is the range of parameters for which the processing time will be the shortest, while still meeting the requirements of the shot peening specifications.

Following the progress of the detailed in this document, a final step aims to validate the range of shot peening parameters, obtained following the execution of the process, on a real part. Thus, following the simulation, shot peening in real conditions is carried out on different parts in order to validate the different ranges of parameters of the shot blasting machine obtained.

This allows validation on the experimental results that comply with the shot peening requirements specified for the part in question.

The method as described in the document is intended to be implemented and implemented on a computer.

In particular, a computer program comprising instructions for implementing the method can thus be developed. This program can thus be recorded on a memory of a computer or on a computer readable data storage medium, allowing the process to be carried out on a computer. Also, the present document relates to a CAD system comprising a processor coupled to a memory and a graphical user interface and able to communicate with a shot blasting machine, such that the program is stored in memory. Thus, the process is carried out on a computer and once the parameters of the shot blasting machine are identified, these can be sent to the shot blast machine, so as to test these parameters in real conditions with regard to the shot peening requirements of the part.

The simulation thus makes it possible to test, in real shot peening conditions, only shot blasting machine parameters which have a good chance of allowing shot peening in accordance with the shot peening specifications. This saves time, and also saves money, thanks to the simulation.

Claims

1. A method of optimizing shot-blasting parameters by learning for a part, the method comprising the following steps: a) Providing the shot-peening specifications of the part comprising at least the areas of the part to be shot-peened, and for each of these areas , a range of shot peening intensity values, a range of coverage rate values and a range of maximum overflow distance values; b) Provide a three-dimensional object of the part, modeled by finite elements using a three-dimensional mesh of the part; c) Determine one or more combinations of blasting intensity and coverage rate values for each zone;

For each combination: d) calculate a profile of residual stresses a ^s (z) = / (C; /) from values of shot peening intensity I and coverage rate C, using a basis of data; e) Calculate a residual stress tensor from the residual stress profile and as a function of the distance z to the nearest surface; f) Calculate, at each point of the three-dimensional mesh, the induced distortion, after a static balancing of the part, from the tensor of residual stresses and from the static balancing; g) Select the combinations, for each zone, for which the distortions induced at each point of the mesh conform to the dimensional tolerances of the part;

For each combination selected: h) Determine the parameters of the shot blasting machine; i) Simulate the shot peening with the shot blasting machine parameters of the selected combinations, for each zone, on the three-dimensional modeled object of the part; j) Select, for each zone, the combination with the shortest blasting time; k) Check that the deformations distortions induced following the shot peening simulation on the three-dimensional modeled object of the part comply with the dimensional tolerances of the part.

2. The method of claim 1, wherein step a) consists of providing one or more of the following:

. parts of the part to be blasted; i. parts of the part prohibited from shot blasting; ii. parts of the part for which shot blasting is admissible; iv. the type of shot; v. mechanical and dimensional characteristics of the shot; vi. a range of blasting intensity for each of the parts to be blasted; vii. a range of minimum and / maximum coverage rates for each of the parts to be shot blasted; viii. minimal residual stress profile; ix. maximum permissible surface roughness; x. dimensional tolerances of the part.

3. The method of claim 1 or 2, wherein the areas to be shot blast are divided into sub-areas for which steps c) to j) are applied.

4. Method according to one of claims 1 to 3, wherein step b) further comprises the following steps:

- Create a two-dimensional mesh of the surface of the modeled object of the part

5. The method of claim 4, wherein step k) comprises the following steps:

6. Method according to one of claims 1 to 5, wherein step c) uses a statistical selection such as for example the Monte-Carlo algorithm.

7. Computer program comprising instructions for implementing the method according to one of claims 1 to 6, when said program is executed on a computer.

8. Data storage medium on which is recorded the computer program according to claim 7.

9. CAD system comprising a processor coupled to a memory and a graphical user interface and able to communicate with a shot blasting machine, such that the program according to claim 7 is stored in memory.