A numerical simulation method for press riveting based on smoothed particle Galerkin method

Abstract

The present invention belongs to the technical field of numerical simulation of press riveting connection, and relates to a numerical simulation method of press riveting connection based on a smooth particle Galerkin method, comprising the following steps: S1, establishing a material model; the press riveting nut is regarded as a rigid body, and the thin plate material is regarded as a flexible body; S2, establishing two finite element models of press-in-push-out and press-in-twist-out; only a 1 / 4 model is established for the press-in-push-out, and a full model is established for the press-in-twist-out; S3, performing unit algorithm control; S4, selecting a constraint method to calculate the LS_DYNA contact degree; S5, applying boundary conditions to the press-in-push-out model and the press-in-twist-out model; S6, performing simulation tests, and comparing the experimental results.

Description

Technical Field

[0001] The invention belongs to the technical field of numerical simulation of press riveting connection, and relates to a numerical simulation method of press riveting connection based on a smooth particle Galerkin method. Background Art

[0002] With the global carbon neutrality strategy proposed, the electric vehicle industry has achieved rapid development. While the industrial chain continues to improve, the quality control of its core components has received more and more attention. The high-voltage distribution box is the electric vehicle energy distribution unit PDU (Power Distribution Unit) (working voltage 700V, current about 400A), and its internal current circuit is connected by copper or aluminum bars through bolts. Complex road excitation, accident collisions, etc. during vehicle driving can easily cause its internal connections to loosen. Loose connections will increase the contact resistance, and sparks or temperature rises can easily cause vehicle spontaneous combustion under high voltage and current conditions. Reliable connection is the key to the reliable operation of PDU products.

[0003] The main connection methods in industry are: welding, bolt connection, self-piercing riveting connection, rivet nut connection, rivet connection, etc. Welding connection is firm, but welding of different materials is difficult. Self-piercing riveting and rivets are suitable for thin plate connection and are widely used on white bodies, but the connection strength of a single riveted connection is relatively low. The connection strength of a single bolt is higher and can be disassembled. Ordinary bolt and nut connections generally do not loosen under non-vibration conditions. Vibration is inevitable during the driving of the car. In order to improve the anti-loosening performance of the bolt connection, the electric vehicle PDU adopts a connection method that combines pressure riveting nuts and bolts. The pressure riveting nut is different from ordinary nuts. It has anti-rotation teeth and anti-pull-off grooves. It is first pressed into the connected thin plate by a pressure riveting machine, and then cooperates with the bolt to tighten the connected parts. This method is stronger than the ordinary nut and bolt connection, so it is widely used in electric vehicle PDUs at home and abroad. Improving the quality of pressure riveting is the key to enhancing the anti-loosening performance of bolt connections, but because the form of pressure riveting nut and bolt connection is only applicable to the connection of thin plate components and its assembly process is complicated, its application range is relatively narrow and it is mainly used in PDUs in automobiles. Scholars at home and abroad have paid little attention to the mechanical process of pressing rivet nuts into thin plates.

[0004] Domestic and foreign research on the riveting process mainly focuses on experiments. Some scholars use two-dimensional axisymmetric models for numerical simulation, and rarely conduct three-dimensional finite element simulation analysis of the riveting process. However, three-dimensional finite element simulation is indispensable. It is the basis for bolt preload and vibration anti-loosening simulation. Based on the Taguchi method, DAUD NAINMAT et al. explored the influence of material, hole diameter and thickness on the twist-out force [4]. Lu Kailei et al. studied the riveting process of copper-clad aluminum bars and riveted nuts, and verified the research results experimentally. Based on the experimental method, Xie Jianliang et al. explored the push-out force, twist-out force and limit installation torque of different types of riveted nuts, as well as the influence of electroplating and vibration on their performance. Lu Kailei et al. used ANSYS finite element software to establish a two-dimensional parametric finite element analysis model of riveted nuts, simulated the influence of different hole diameters on the connection performance of riveted nuts, and compared them with the experimental results.

[0005] Generally, it is difficult to obtain better riveting connection quality. Its influencing factors include: geometric parameters of the riveting nut, thickness of the connecting plate, hole diameter, riveting speed during the riveting process, etc. Parameter research through experiments will consume huge manpower, material resources and time costs, so it is necessary to perform accurate numerical simulation of the riveting process. Finite element simulation of the riveting process is very difficult, which involves material nonlinearity, contact nonlinearity, large deformation and material failure. So far, material failure is still a difficult point in numerical simulation of CAE (Computer Aided Engineering). Summary of the invention

[0006] In view of the problems existing in the prior art, the present invention provides a numerical simulation method for riveting connection based on the smoothed particle Galerkin method.

[0007] The present invention is achieved through the following technical solutions:

[0008] A numerical simulation method for press riveting connection based on smoothed particle Galerkin method comprises the following steps:

[0009] S1. Establish material model; the rivet nut is regarded as a rigid body and the sheet material is regarded as a flexible body;

[0010] S2, establish two finite element models of press-in-push-out and press-in-twist-out; for press-in-push-out, only 1 / 4 model is established, while for press-in-twist-out, the whole model is established;

[0011] S3, perform unit algorithm control;

[0012] S4. Select the constraint method to calculate the LS_DYNA contact stiffness;

[0013] S5, applying boundary conditions to the compression-extrusion model and the compression-twist-out model;

[0014] S6. Conduct simulation tests and compare them with experimental results.

[0015] On the basis of the above scheme, further, in step S1, the riveting test is loaded using a universal testing machine of model CMT5105, with a loading speed of 25 mm / min, which is regarded as a quasi-static process. Therefore, the simulation model does not consider the influence of strain rate on the mechanical properties of the material; the material of the riveted nut is 0Cr18Ni9, and the material of the thin plate is T2 copper.

[0016] On the basis of the above scheme, in step S2, the modeling is further carried out by hybrid modeling of finite element unit and smooth particle. The connection between them is coupled by common node. The area below the rivet nut is discretized by smooth particle Galerkin method (SPG), and the mesh size is 0.08mm. The other areas of the thin plate are discretized by finite element mesh, and the mesh size is 0.35mm. The mesh size of the rivet nut is larger than the mesh unit in contact with it, and is set to 0.125mm.

[0017] On the basis of the above scheme, further, in step S3, the unit algorithm of the finite element mesh area is set as a constant stress calculation unit through the keyword *SECTION_SOLID; the SPG particles are controlled by the unit algorithm through the keyword *SECTION_SOLID_SPG; the parameters that need to be set for the *SECTION_SOLID_SPG card include: shape function, bond break failure mechanism and corresponding threshold value, and elongation or compression rate at bond break; the shape function adopts the Euler method; the failure mechanism of bond break adopts equivalent plastic strain, and the corresponding damage value is set to 0.55; the material elongation or compression rate is set to 1.25.

[0018] On the basis of the above scheme, further, in step S3, in view of the situation that particles scatter everywhere during the stretching process, an area is specified, and SPG calculation is performed in this area to reduce the time required for searching particles.

[0019] On the basis of the above scheme, further, in step S4, the contact between the rivet nut and the thin plate adopts the point-surface contact method, the SPG smooth particles are set as points, and the finite element units are set as surfaces; the friction in the contact area adopts the Coulomb friction model, the static friction coefficient is set to 0.1, and the dynamic friction coefficient is also set to 0.1; the constraint method calculates the contact stiffness based on the contact node mass and the global time step, which is expressed as:

[0020] (1)

[0021] Where: k is the contact stiffness, sofscl is the scaling factor set by the user, m is the node mass, and Δt is the time step.

[0022] On the basis of the above scheme, further, in step S5, the boundary conditions that need to be applied to the press-in-push-out model are: fixed constraint on the bottom surface of the thin plate, symmetry constraint on the two symmetry planes, rigid body displacement constraint on the rivet nut and rigid body displacement on the top surface of the rivet nut, wherein the rigid body displacement of the rivet nut has 6 degrees of freedom, and the degrees of freedom along the direction of the rivet force need to be released; the rigid body displacement of the top surface of the rivet nut needs to be divided into two stages: in the first stage, a rigid body displacement of 1.195 mm is applied downward, and in the second stage, an upward displacement of 2.375 mm is applied, which just pushes the rivet nut out of the thin plate.

[0023] The boundary conditions that need to be applied to the press-in-twist-out model are: fixed constraint on the bottom surface of the thin plate, rigid displacement constraint on the rivet nut, and rigid displacement on the top surface of the rivet nut. The rigid displacement of the rivet nut has 6 degrees of freedom, and the degrees of freedom along the rivet force direction and the torsion direction need to be released. The rigid displacement of the top surface of the rivet nut needs to be divided into two stages: the first stage applies a downward rigid displacement of 1.195 mm, and the second stage applies an angular displacement of 10°.

[0024] Beneficial technical effects of the present invention:

[0025] The present invention applies the smooth particle Galerkin method to simulate the failure and damage of materials in the riveting connection and the riveting quality test. Compared with the finite element method based on the unit erosion technology, it does not delete the unit, but adopts the key breaking mode. The element is not deleted after failure, so the conservation of mass and momentum can be maintained. From the comparison of the curves and eigenvalues ​​obtained by simulation and experiment, it can be seen that the method has obtained very good results in both qualitative and quantitative aspects for the riveting process and the riveting quality test experiment. At the same time, the stress-strain diagram obtained by simulation can also clearly reflect the phenomenon that the material cold flows into the anti-pull groove of the riveting nut, the material is flattened during the push-out process, and the tooth groove is destroyed during the twisting-out process. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a schematic diagram of the structure of the rivet nut and the thin plate to be connected;

[0027] Figure 2 It is a schematic diagram of the riveting process;

[0028] Figure 3 Schematic diagram of the launch test;

[0029] Figure 4 Schematic diagram of the twist-out test;

[0030] Figure 5 This is the finite element mesh of the unit deletion method, that is, the simulation result diagram;

[0031] Figure 6 This is the riveting force-displacement curve obtained by the unit deletion method;

[0032] Figure 7 Schematic diagram of stress concentration and stress singular distribution area;

[0033] Figure 8 It is a bond breaking model diagram;

[0034] Fig. 9 Break the graph for multiple bonds;

[0035] Fig.10 This is the stress-strain curve of T2 copper;

[0036] Fig.11 This is a physical picture of T2 copper sheet and rivet nut;

[0037] Fig.12 This is the finite element model diagram of the press-in-pull-out 1 / 4;

[0038] Fig.13 It is a full model diagram of press-in-twist-out;

[0039] Fig.14 This is the boundary condition diagram for the push-in-push-out model;

[0040] Fig.15 This is the boundary condition diagram of the compression-torsion model;

[0041] Fig.16 It is the riveting force-displacement curve;

[0042] Fig.17 is the push-out force-displacement curve;

[0043] Fig.18 It is a torque-torsion angle curve diagram;

[0044] Fig.19 To simulate the equivalent strain diagram of different displacements during the riveting process;

[0045] Fig. 20 It is the equivalent stress diagram at different displacements during the riveting process;

[0046] Fig.21 It is the equivalent strain diagram at different displacements during the push-out process;

[0047] Fig. 22 It is the equivalent stress diagram at different displacements during the twisting process. DETAILED DESCRIPTION

[0048] Example embodiments will now be described more fully with reference to the accompanying drawings.

[0049] 1. Riveting process and riveting quality test

[0050] The model of the rivet nut is S-M6-2. The structure of the rivet nut and the connected sheet is as follows: Figure 1As shown in the figure, the prefabricated hole diameter of the thin plate should be slightly larger than the diameter of the anti-pull groove. The material of the rivet nut is mostly carbon steel, and the thin plate is mostly aluminum or copper, and the strength difference between the two is large. The nut is almost not deformed during the riveting process, but the deformation near the plate hole is larger.

[0051] The riveting process has three steps, such as Figure 2 As shown: first, put the rivet nut into the prefabricated hole of the connected thin plate; second, apply pressure through a riveting machine or a universal testing machine. Under the action of pressure, the material around the plate hole undergoes plastic deformation and gradually develops into fractures, and the fractured material partially flows into the anti-pull groove of the rivet nut; third, the entire teeth of the rivet nut are embedded in the thin plate, and more material cold flows into the anti-pull groove. When the teeth of the rivet nut are embedded in the thin plate, the nut will be difficult to rotate, and the material coldly flowing into the anti-pull groove will also prevent the nut from moving up and down. The loosening of the bolted connection structure is mostly due to the reduction of the bolt preload, and the reduction of the preload is related to the relative movement or rotation between the bolt and the nut. Therefore, the rivet nut will enhance the anti-loosening performance of the bolted connection.

[0052] The quality of the press riveting can be tested by push-out test and twist-out test, such as Figure 3 As shown in the figure, the push-out test is to push the rivet nut out of the mounting plate by applying pressure on the testing machine, thereby measuring the push-out force and displacement curve. The test process requires first screwing the bolt into the nut from the back, and ensuring that the gap between the bolt head and the plate surface is large enough, then placing the test plate mounting surface on the support sleeve, and finally pushing the rivet nut out through the testing machine. The twist-out test is to determine the maximum twist-out torque of the rivet nut by applying torque on the rivet nut, such as Figure 4 The test process requires placing a washer between the nut and bolt to prevent the bolt neck from contacting the rivet nut threads, and then using a digital torque wrench to apply a torque load to the bolt.

[0053] 2. Advantages of the SPG bond fracture failure model and its simulation of material failure

[0054] There is a lot of material damage during the riveting process. Traditional finite element simulation of material failure and damage usually uses unit deletion. The unit deletion standard is specific and the results are very sensitive to the standard. It will not only make the calculation domain discontinuous and cause stress singularity, but also lead to non-physical responses, such as reduced force or torque response, or fail to obtain the correct failure form. The simulation results of the riveting process using unit deletion are as follows: Figure 5 As shown in the figure. Although the mesh size of the riveting area is 0.08mm, the unit below the teeth of the riveted nut cannot be avoided. The unit calculation result is too large due to the stress singularity, resulting in the mesh below the teeth being completely deleted. The actual physical process is that the material below the teeth continues to be compressed and deformed after breaking and flows into the anti-tension groove of the nut. The riveting force-displacement curve obtained by deleting the unit is shown in the figure. Figure 6As shown in the figure, the force in the initial stage will increase with the increase of displacement because the material is still in the linear elastic stage. As some units are deleted, the force required for deformation oscillates repeatedly without continuing to increase, which is inconsistent with the physical process. It can be seen that the unit deletion finite element method cannot simulate this process.

[0055] The 1 / 4 geometric model and contact area distribution of the rivet nut and the thin plate are as follows Figure 7 (a) and (b), the area moves downward due to pressure. During the downward pressure process, one side of the material is compressed and the other side is pulled, which is a shear process. The surface where the displacement occurs is called the shear surface, and this surface is the surface where stress concentration occurs and the actual fracture position of the material, such as Figure 7 (c) shows. Stress singularity is a problem caused by finite element analysis. All units near stress concentration will produce stress singularity, and the area where it occurs is obviously larger than the stress concentration. Stress singularity exists in all the compressed areas below the contact position between the rivet nut and the thin plate, such as Figure 7 As shown in (d), these regions will be deleted during the unit deletion simulation.

[0056] Other methods for simulating material failure and damage include: extended finite element method, cohesion unit method, node splitting method, etc. These are only applicable to crack propagation. In addition, the adaptive grid method is suitable for metal processing simulation with large deformation but no fracture. They are not suitable for large-area material loss removal problems. Smoothed Particle Hydrodynamics (SPH) was originally used to simulate fluids and later expanded to the solid field. Although it is also a particle method, its simulation accuracy in the solid field is lower than that of the SPG method. The SPG method can also perform multi-stage analysis, for example: the first step is to simulate the riveting process, and the second step is to analyze the push-out, twist-out, and bolt pre-tightening of the nut.

[0057] The SPG method is a true gridless method that integrates nodes. Each node searches for other nodes in its neighborhood. If other nodes can be found, they are considered to be connected by bonds and the corresponding shape functions are established. If no other nodes can be found, it is considered that there is no bond connection between the nodes and the shape function is zero. Figure 8 As shown in (a), assume that the search area of ​​1 node is a circle, and there are bond connections between 1-2, 1-5, 1-6, 1-7, and 1-8 in the circular area. When the specified bond breaking index is reached, the bond is considered broken, such as Figure 8 (b) shows that nodes are equivalent to atoms and bonds are equivalent to chemical bonds. The material failure is simulated by bond breaking. After the bond breaks, a free surface is formed, and the nodes still exist and participate in the subsequent mechanical process.

[0058] There are two indicators of bond breaking: the first is the elongation or shortening rate of the distance between the node and other nodes. The second is the material failure criteria used in the traditional finite element method, such as equivalent plastic strain, first principal stress, maximum shear strain, third principal strain, damage value, etc.

[0059] From the above analysis, it can be seen that the SPG method can well simulate the failure fracture of the shear surface during the riveting process by means of key breaking, such as Fig. 9 As shown, the particles on both sides of the fracture surface do not interfere with each other, and the compressed particles will continue to deform under the extrusion effect and be squeezed into the anti-pull groove reserved in the rivet nut.

[0060] 3. Establish a three-dimensional simulation model for riveting and its quality testing

[0061] 3.1 Material model

[0062] The riveting test uses a universal testing machine of model CMT5105 for loading, and the loading speed is 25mm / min. It can be regarded as a quasi-static process, so the simulation model does not consider the influence of strain rate on the mechanical properties of the material. The material of the riveted nut is 0Cr18Ni9, which is a low-carbon stainless steel with high tensile strength, and the sheet material is T2 copper. The nut is almost not deformed during the riveting overload, and can be regarded as a rigid body in the simulation model. The thin plate undergoes a large deformation and is regarded as a flexible body. The true stress-strain curve of T2 copper under quasi-static compression conditions at room temperature is shown in Figure 2. Fig.10 As shown, the multilinear isotropic strengthening material model *MAT_PIECEWISE_LINEAR_PLASTICITY is used in LS_DYNA.

[0063] 3.2 Finite element mesh

[0064] In the project, it is only necessary to press the rivet nut into the plate. However, in order to study the quality of the rivet, it is necessary to perform the withdrawal and twisting tests after the rivet is pressed in. Therefore, it is necessary to establish two finite element models, namely the press-in-pull-out and press-in-twist-out. The size of the connected sheet is 60mm*60mm*3mm. Fig.11 As shown in the figure, it is a geometrically symmetrical structure. To study the pressing and pushing process, only 1 / 4 model can be established to reduce the calculation time, such as Fig.12 If we want to study the pressing and twisting process, we need to build a full model because the direction of the torque vector is not on the geometric symmetry plane, as shown in Fig.13 In addition, according to Saint-Venant's principle, the entire thin plate is subjected to less pressure and deformation in the area away from the action of the rivet nut, so it is not necessary to model the entire plate during finite element analysis.

[0065] Although the SPG method has many advantages, its calculation takes more time. Therefore, the finite element unit and smooth particle hybrid modeling method is used in modeling, and the connection between them is coupled by common nodes. The area below the rivet nut is discretized by the SPG method, and the mesh size is 0.08mm. The other areas of the thin plate are discretized by finite element mesh, with a mesh size of 0.35mm. The mesh size of the rivet nut using the rigid body model can be slightly larger than the mesh unit in contact with it, set to 0.125mm.

[0066] 3.3 Unit Algorithm

[0067] The unit algorithm of the finite element mesh area is set to a constant stress calculation unit through the keyword *SECTION_SOLID. The unit algorithm of SPG particles is controlled by the keyword *SECTION_SOLID_SPG. The main parameters that need to be set for the *SECTION_SOLID_SPG card include: shape function, bond failure mechanism and corresponding threshold value, stretching rate or compression rate when bond breaks, etc. There are three shape functions: Update Lagrangian method, Euler method, and pseudo-Lagrangian method. The similarities between them are that they are all methods for setting particles in the search neighborhood. The differences are: Update Lagrangian method will change the shape and size of the search neighborhood according to the deformation situation, and it is more suitable for the situation where the plastic strain is not particularly high; the size and shape of the search neighborhood of the Euler method do not change, but the search range will move accordingly with the displacement of the particles, and it is more suitable for the situation where the plastic strain is particularly large or the material is damaged over a large area; the pseudo-Lagrangian method is the same as the Update Lagrangian method, and its search neighborhood will change with the deformation, but the difference is that the particles in its field can only enter but not exit, and it is mainly used in grinding, etc. The plastic strain of the compressed unit during the riveting process is particularly large, and a large area of ​​material will be torn on the shear surface, so the shape function is more suitable for the Euler method. The failure mechanism of bond breaking uses equivalent plastic strain, and the corresponding damage value is set to 0.55. Another standard for bond breaking is that the material stretching or compression rate is set to 1.25. In addition, in view of the situation where particles splash everywhere during the stretching process, an area can be specified, and SPG calculation is performed in this area, which can reduce the time required to search for particles.

[0068] 3.4 Contact Algorithm

[0069] The contact between the rivet nut and the thin plate adopts the point-to-surface contact method, the SPG smooth particles are set as points, and the finite element unit is set as a surface. The Coulomb friction model is used for the friction in the contact area, and the static friction coefficient is set to 0.1. In order to increase convergence and reduce unnecessary data oscillations, the dynamic friction coefficient is also set to 0.1. There are two methods for calculating the contact stiffness in LS_DYNA: the penalty function method and the constraint method. The penalty function method is the default method. It uses the size and material properties of the contact segment to determine the contact stiffness, which is more suitable for situations where the stiffness of the two contact materials is not much different. The constraint method calculates the contact stiffness based on the contact node mass and the global time step, expressed as:

[0070] (1)

[0071] Where: k is the contact stiffness, sofscl The scaling factor set by the user, m is the node mass and Δt is the time step.

[0072] From formula (1), it can be seen that the contact stiffness calculated by the constraint method is independent of the material. The stiffness of the rivet nut and the thin plate is quite different, so the constraint method is selected as the algorithm for contact stiffness.

[0073] 3.5 Boundary conditions

[0074] The boundary conditions that need to be applied to the press-in-push-out model are: fixed constraint on the bottom surface of the thin plate, symmetry constraint on the two symmetry surfaces, rigid displacement constraint on the rivet nut, and rigid displacement on the top surface of the rivet nut. Fig.14 As shown. The rigid body displacement of the rivet nut has 6 degrees of freedom, and the degrees of freedom along the direction of the rivet force need to be released. The rigid body displacement of the top surface of the rivet nut needs to be divided into two stages: the first stage applies a downward rigid body displacement of 1.195mm, and the second stage applies an upward displacement of 2.375mm, just to push the rivet nut out of the thin plate.

[0075] The boundary conditions that need to be applied to the press-in-twist-out model are: fixed constraint on the bottom surface of the thin plate, rigid displacement constraint on the rivet nut, and rigid displacement on the top surface of the rivet nut, such as Fig.15 As shown. The rigid body displacement of the rivet nut has 6 degrees of freedom, and the degrees of freedom along the rivet force direction and the torsion direction need to be released. The rigid body displacement of the top surface of the rivet nut needs to be divided into two stages: the first stage applies a downward rigid body displacement of 1.195mm, and the second stage applies an angular displacement of 10°.

[0076] 4. Comparison with experimental results

[0077] The relationship between the riveting force and the displacement of the riveted nut during the riveting process can be expressed by the riveting force-displacement curve. The riveting force-displacement curves obtained by simulation and experiment are shown in Fig.16As shown, the two are highly consistent. The height of the riveted nut teeth is 0.9mm. At this time, the riveted nut is just pressed into the thin plate. The riveting force measured in the test is 8.9kN, and the simulation value is 9.1KN, which is 2.24% larger than the experimental value. The displacement of the riveted nut at the end of riveting is 1.195mm. At this time, the riveting force is the largest. The riveting force measured in the experiment is 18.5kN, and the simulation value is 17.8kN, which is 3.78% smaller than the experimental value. It should be noted that the riveting force changes suddenly at the 0.9mm position. This is because the contact area between the nut and the thin plate suddenly increases at this time, so it takes a greater force to increase the displacement of the nut.

[0078] The relationship between the push-out force and the displacement of the rivet nut during the push-out process can be expressed by the push-out force-displacement curve. The push-out force-displacement curves obtained by simulation and experiment are shown in Fig.17 As shown in the figure, the simulation extreme point is earlier than the experimental value, and the agreement is good. The push-out force measured in the test is 3.15kN, and the simulation value is 3.135kN, which is slightly smaller than the experimental value.

[0079] The tool used in the twist-out test is a digital torque wrench, which can measure the maximum applied torque, but cannot obtain a curve of the torque changing with time. The maximum applied torque measured in the test is 26.63N‧m, and the simulation value is 24.12N‧m. The simulation value is 9.4% smaller than the test value, and the difference is slightly larger than that in the riveting and pushing-out stages. There are two possible reasons: first, the accumulated errors between the particle method and the experimental results in the riveting stage; second, the failure and damage in the twist-out process itself are too serious, and the material of the thin plate and the teeth of the rivet nut are all twisted and all the SPG particles are destroyed. The relationship between the twist-out force and the angular displacement of the rivet nut during the twist-out process can be expressed by the twist-out torque-torsion angle curve. The simulated twist-out torque-torsion angle curve is shown as follows: Fig.18 shown.

[0080] The equivalent strain at different displacements during the riveting process is obtained by simulation as follows: Fig.19 As shown. When the rivet nut is displaced by 0.2mm, the tooth groove shape of the rivet nut has been formed on the thin plate. The equivalent plastic strain is larger near the tooth groove due to stress concentration, and the phenomenon of material cold flow into the anti-pull groove is not obvious. When the displacement is 0.9mm, the tooth groove depth becomes larger, and the compressed material has mostly cold flowed into the anti-pull groove under the action of pressure, forming free surfaces on both sides of the key break. At 1.195mm, the material has completely filled the anti-pull groove, which shows that SPG can simulate the fracture of the material well through key breaking.

[0081] The equivalent stress at different displacements during the riveting process is obtained by simulation. Fig. 20 As shown in the figure, the area with larger concentration value of the thin plate will gradually move downward during the riveting process, and the area with larger equivalent stress will gradually diffuse as the displacement increases, but it is smaller in the original riveting area, which conforms to the Saint-Venant principle.

[0082] The equivalent strain at different displacements during the push-out process is as follows: Fig.21 As shown in the figure, the SPG particles gradually withdraw from the anti-pull groove, and the inner wall of the prefabricated hole of the thin plate is gradually flattened. This phenomenon well explains why the push-out force gradually decreases.

[0083] The equivalent strain of different angular displacements during the twisting process is as follows: Fig. 22 As shown, it can be seen that: the angular displacement applied to the rivet nut is in the clockwise direction, the pressure-bearing side surface of the tooth groove on the thin plate is gradually shortened, and most of the anti-torsion tooth groove on the thin plate has disappeared after twisting one tooth.

[0084] The present invention applies the smooth particle Galerkin method to simulate the failure and damage of materials in the riveting connection and the riveting quality test. Compared with the finite element method based on the unit erosion technology, it does not delete the unit, but adopts the key breaking mode. The element is not deleted after failure, so the conservation of mass and momentum can be maintained. From the comparison of the curves and eigenvalues ​​obtained by simulation and experiment, it can be seen that the method has obtained very good results in both qualitative and quantitative aspects for the riveting process and the riveting quality test experiment. At the same time, the stress-strain diagram obtained by simulation can also clearly reflect the phenomenon that the material cold flows into the anti-pull groove of the riveting nut, the material is flattened during the push-out process, and the tooth groove is destroyed during the twisting-out process.

Claims

1. A numerical simulation method for press riveting based on smoothed particle Galerkin method, characterized in that: The steps include: S1. Establish material model; the rivet nut is regarded as a rigid body and the sheet material is regarded as a flexible body; S2, establish two finite element models of press-in-push-out and press-in-twist-out; for press-in-push-out, only 1 / 4 model is established, while for press-in-twist-out, the whole model is established; S3, perform unit algorithm control; S4. Select the constraint method to calculate the LS_DYNA contact stiffness; S5, applying boundary conditions to the compression-extrusion model and the compression-twist-out model; S6. Conduct simulation test and compare with experimental results; In step S2, the modeling is performed by hybrid modeling of finite element units and smooth particles. The connections between them are coupled by common nodes. The area below the rivet nut is discretized by smooth particle Galerkin method (SPG) with a mesh size of 0.08 mm. The other areas of the thin plate are discretized by finite element mesh, with a mesh size of 0.35 mm; the mesh size of the rigid body model of the rivet nut is larger than the mesh unit in contact with it, which is set to 0.125 mm; In step S3, the unit algorithm of the finite element mesh area is set to a constant stress calculation unit through the keyword *SECTION_SOLID; the SPG particles are controlled by the unit algorithm through the keyword *SECTION_SOLID_SPG; the parameters that need to be set for the *SECTION_SOLID_SPG card include: shape function, bond breaking failure mechanism and corresponding threshold value, and elongation or compression rate at bond breaking; the shape function adopts the Euler method; the bond breaking failure mechanism adopts equivalent plastic strain, and the corresponding damage value is set to 0.55; the material elongation or compression rate is set to 1.

25.

2. The method for numerical simulation of press riveting connection based on smoothed particle Galerkin method according to claim 1 is characterized in that: In step S1, the riveting test is loaded using a universal testing machine of model CMT5105 at a loading speed of 25 mm / min, which is regarded as a quasi-static process. Therefore, the simulation model does not consider the influence of strain rate on the mechanical properties of the material. The material of the riveted nut is 0Cr18Ni9, and the material of the thin plate is T2 copper.

3. The method for numerical simulation of press riveting connection based on smoothed particle Galerkin method according to claim 1 is characterized in that: In step S3, in view of the situation that particles scatter everywhere during the stretching process, an area is specified, and SPG calculation is performed only in this area to reduce the time required for searching particles.

4. The method for numerical simulation of press riveting connection based on smoothed particle Galerkin method according to claim 1, characterized in that: In step S4, the contact between the rivet nut and the thin plate adopts the point-surface contact mode, the SPG smooth particles are set as points, and the finite element units are set as surfaces; the friction in the contact area adopts the Coulomb friction model, the static friction coefficient is set to 0.1, and the dynamic friction coefficient is also set to 0.1; the constraint method calculates the contact stiffness based on the contact node mass and the global time step, which is expressed as: (1) Where: k is the contact stiffness, sofscl is the scaling factor set by the user, m is the node mass, and Δt is the time step.

5. The method for numerical simulation of press riveting connection based on smoothed particle Galerkin method according to claim 1, characterized in that: In step S5, the boundary conditions that need to be applied to the press-in-push-out model are: fixed constraint on the bottom surface of the thin plate, symmetry constraint on the two symmetry planes, rigid body displacement constraint on the rivet nut, and rigid body displacement on the top surface of the rivet nut. The rigid body displacement of the rivet nut has 6 degrees of freedom, and the degrees of freedom along the direction of the rivet force need to be released; the rigid body displacement of the top surface of the rivet nut needs to be divided into two stages: in the first stage, a downward rigid body displacement of 1.195 mm is applied, and in the second stage, an upward displacement of 2.375 mm is applied, just to push the rivet nut out of the thin plate; The boundary conditions that need to be applied to the press-in-twist-out model are: fixed constraint on the bottom surface of the thin plate, rigid body displacement constraint on the rivet nut, and rigid body displacement on the top surface of the rivet nut. The rigid body displacement of the rivet nut has 6 degrees of freedom, and the degrees of freedom along the direction of the rivet force and the torsion direction need to be released. The rigid body displacement of the top surface of the rivet nut needs to be divided into two stages: the first stage applies a downward rigid body displacement of 1.195 mm, and the second stage applies an angular displacement of 10°.

Images