Friction stir welding and additive modeling method based on in-situ measurement data

Abstract

The invention discloses a friction stir welding and additive modeling method based on in-situ measurement data. The friction stir welding and additive modeling method comprises the following steps that firstly, a finite element model is constructed; 2, collecting data to realize heat transfer simulation; 3, determining a friction parameter value interval; step 4, friction coefficient iterative calculation; 5, solving an optimal friction coefficient; and 6, performing a physical experiment by using a group of brand-new process parameters which do not participate in any optimization iteration, and performing simulation prediction on the new model. 7, converting an in-situ force parameter and a torque; and 8, adding mechanical property parameters required by stress analysis into the geometric model for stress analysis. And ninthly, boundary conditions are set for the simulation model to limit displacement of the simulation model, and stress field simulation calculation is conducted on the deposition process of friction rolling additive manufacturing. According to the method, the heat source intensity and load distribution calculation are highly matched with an actual processing scene, the prediction error of the temperature field is remarkably reduced, and accurate support is provided for process optimization.

Description

Technical Field

[0001] This invention relates to the field of aerospace welding technology, specifically to a method for thermo-mechanical coupling modeling and model parameter verification of friction stir welding and additive manufacturing based on in-situ measurement data. Background Technology

[0002] Solid-state friction stir welding (FSAM) and solid-state additive manufacturing based on friction stir have been widely applied in high-end equipment manufacturing fields such as aerospace, rail transportation, and new energy vehicles due to their core advantages such as low heat input, excellent joint / deposited part performance, and absence of melting defects. Among these, friction roll additive manufacturing (FROM-AM), a typical FROM-AM technology, has solved the discontinuous feeding problem of FSAM and FSD-AM, successfully achieving continuous material feeding into the forming area under the action of a non-consumable tool head. However, in actual printing processes, heat accumulation can cause excessively high local temperatures on the parts, leading to severe slag buildup, increased roughness, and even microcracks inside the parts due to excessive thermal stress. This deteriorates the printing quality, reduces mechanical properties, and severely impacts the lifespan of the parts. Furthermore, due to the high coefficient of thermal expansion and good thermal conductivity of aluminum alloys, significant residual stress is unavoidable after processing. Large residual stress can cause defects such as cracks, affecting the mechanical properties, dimensional stability, and reliability of components, as well as their fatigue performance. Therefore, establishing an accurate thermo-mechanical coupling model is the key to optimizing process parameters, predicting forming defects, and ensuring product performance.

[0003] However, existing thermo-mechanical coupling simulation technologies related to friction roll forming additive manufacturing still have the following problems that urgently need to be solved: First, the heat source and mechanical load formulas have poor compatibility and lack dedicated models. The heat source formulas used in existing simulations mostly follow the derivation results of friction stir welding (FSW), in which the tool head and workpiece have local point contact and the heat source exhibits a concentrated distribution characteristic. However, the friction roll forming process uses a lateral rolling method of the tool head instead of the vertical insertion method of the tool head in FSW. The motion mode and load transmission path are fundamentally different from those in friction stir welding. Existing technologies have not derived dedicated heat source and mechanical load formulas for the process characteristics of friction roll forming, resulting in a mismatch between the calculated heat source intensity and load distribution and the actual processing scenario, making it impossible to accurately describe the thermo-mechanical coupling of the friction roll forming process. First, the model suffers from several problems: 1) Inconsistent with the actual working conditions; 2) Lack of calibration for key parameter values, relying on empirical data. In the construction of existing models, core process parameters such as torque, force, and friction coefficient are often directly derived from empirical values ​​obtained through literature research, without adaptation and calibration to specific working conditions. The setting of key parameters is highly subjective, resulting in significant deviations between the model and actual working conditions, and failing to accurately reflect the thermal transfer laws of the model; 3) Simulation analysis is limited to single-field simulation, failing to achieve thermo-mechanical coupling. Currently, most related simulation studies only conduct heat transfer simulation or mechanical simulation separately, ignoring the coupling effect between the temperature field and the stress-strain field. Single-field simulation severs the above coupling relationship, failing to accurately reflect the true physical mechanism of the processing process, leading to large prediction errors for key quality indicators such as residual stress and deformation of components. Summary of the Invention

[0004] The technical solution adopted in this invention is a method for friction stir welding and additive modeling based on in-situ measurement data, which includes the following steps:

[0005] Step 1: Build a basic framework for heat transfer simulation that matches the actual processing scenario. Through geometric modeling, material property definition, analysis step setting, and mesh generation, provide an accurate model carrier for subsequent heat transfer simulation and ensure that the simulation boundary conditions and time dimension are consistent with the actual deposition process.

[0006] S11: Based on the actual dimensions of the friction stir welded sample, a geometric model of the deposited sample was established using finite element analysis software;

[0007] S12: Assign material properties to the geometric model established based on finite element analysis software, including density, Poisson's ratio, thermal conductivity and specific heat; these thermophysical parameters will change with temperature, and the convective heat transfer coefficient needs to be set according to the corresponding actual working conditions. Define the boundary conditions according to the contact between the workpiece and the air and the workbench during the actual welding process; before welding, set the ambient temperature and the initial temperature of the test plate to 20℃.

[0008] S13: Establish an analysis procedure for friction stir welding that is consistent with the actual deposition process time, and set the analysis step time to be the same as the actual welding time;

[0009] S14: Mesh the above friction stir welded sample, and set the mesh type to an eight-node linear heat transfer hexahedral element.

[0010] Step 2: By collecting force data in situ in real time, and combining it with process parameters, a dedicated heat source model is constructed to solve the problem of poor adaptability of existing heat source formulas. This allows the heat source distribution and intensity calculation to be accurately matched with the actual working conditions of friction roller pressing additive manufacturing, providing core load input for heat transfer simulation.

[0011] S21: The three-dimensional force sensor captures the three-dimensional force signal generated by the contact between the tool and the substrate in real time. After being amplified and filtered by the three-dimensional force acquisition card, it is converted into a digital signal and transmitted to the computer to realize the in-situ real-time acquisition of force data.

[0012] S22: Collect parameters such as the angle of the contact area between the tool head and the substrate, the tool head rotation speed, the forward speed, the bottom radius of the tool head, and the height by querying the process parameter window.

[0013] S23: Collect the angle of the contact area between the tool head and the substrate, the tool head rotation speed, the forward speed, the bottom radius of the tool head, and the height through the process parameter query window.

[0014] S24: The collected process parameters are converted into a heat source model simulating the heat source distribution during friction roller additive manufacturing, specifically the frictional heat source generated by the contact between the tool head side surface and the substrate. Heat source Q 总 Set as a volumetric heat source and apply it to the working area of ​​the tool head;

[0015] ;

[0016] In the formula, 𝛳 is the angle between the tool head and the substrate contact area, and 𝜔 is the rotational speed of the tool head. Let L be the contact shear stress at the interface between the tool head and the substrate, where L is the width of the tool head and r is the radius of the tool head.

[0017] S25: Contact shear stress at the interface between the tool head and the substrate ;

[0018] ;

[0019] In the formula, The coefficient of friction is the interface between the tool head and the substrate. The downward pressure applied by the tool head perpendicular to the substrate contact interface;

[0020] S26: Transfer the heat source Q obtained in S24 to a subroutine. 总In the heat transfer model of friction stir welding, the initial position, moving path, and moving speed of the heat source are kept consistent with the actual deposition process to simulate heat transfer.

[0021] Step 3: Based on literature review and material properties, select experimental data to determine the reasonable physical constraint range of the friction coefficient. At the same time, set the benchmark value and complete the first heat transfer simulation to provide a clear parameter range and initial reference for subsequent iterative calculations, and avoid subjective bias in parameter values.

[0022] S31: Screen experimental data on friction coefficients that are relevant to the materials used in the experiment and to solid-state additive manufacturing. Through statistical analysis of these data, a reasonable range of friction coefficient µ values ​​[0.1, 0.5] is initially determined as the constraint range for friction parameters in subsequent optimization of friction stir welding.

[0023] S32: Take the minimum value of the above-mentioned friction coefficient µ range as the reference value of the friction coefficient. In the constructed finite element geometric model, assign this reference value to the contact shear stress at the interface between the tool head and the substrate, and run a complete heat transfer simulation to simulate the actual processing process.

[0024] Step 4: Starting from the baseline value, by setting loop variables, adding analysis steps and binding formal parameters, the friction coefficient is implemented in an ordered incremental iteration within the constraint range, and multiple rounds of heat transfer simulation are completed, providing sufficient simulation data support for establishing the mapping relationship between "friction coefficient and simulation error".

[0025] S41: Add the same number of analysis steps as the number of iterations to the geometric model analysis step block. The objects selected in each analysis step are the same, and the rest remain unchanged.

[0026] S42: Set the friction coefficient µ as a cyclic variable. (Base value) Starting from [0.1, 0.5], a total of 20 iterations are performed within the interval [0.1, 0.5]. The µ value is automatically updated and the solver is invoked at each analysis step.

[0027] S43: Bind the friction coefficient µ, which needs to be incremented, to the formal parameter. Set: for each solution step completed, i.e., KSTEP increments by 1, the friction coefficient µ increases by 0.02 increments. Friction coefficient µ = +Δµ×(KSTEP-1), The initial value is 0.1, and Δµ is the increment for each iteration. Each time DLOAD is called in the solution step, the subroutine automatically updates the friction coefficient according to the rules, realizing the iterative loop of "solution-increment-resolution".

[0028] Step 5: By comparing the simulated temperature curve with the measured temperature curve, the mean square error is calculated to quantify the simulation deviation, the correspondence between the friction coefficient and the error is established, and finally the optimal friction coefficient that minimizes the error is selected to ensure that the core process parameters are adapted to the specific working conditions and improve the model prediction accuracy.

[0029] S51: The thermocouple probe is placed in the thermocouple pre-set hole and comes into direct contact with the workpiece to be processed. The temperature data is transmitted wirelessly to the receiving end through the temperature acquisition card, and then uploaded to the computer by the receiving end, realizing the in-situ real-time acquisition of temperature data.

[0030] S52: After the simulation is complete, extract the temperature-time history data of the nodes that perfectly correspond to the spatial coordinates of the thermocouple placement points in the experiment from the results file. Compare this simulated temperature curve with the experimentally measured temperature curve and calculate the mean square error (MSE). As a core evaluation indicator, the simulation deviation under initial parameters is quantified, thereby establishing a mapping relationship between the "friction coefficient and simulation error".

[0031] MSE ( )= ;

[0032] In the formula, MSE ( ) is the mean square error of the temperature at node i, and n is the number of iterations; It is the node temperature obtained from the experiment; This is the node temperature obtained from the i-th simulation;

[0033] S53: After all 20 iterations are completed, compare the values ​​of all MSE(T). Let the objective function E(µ) = MSE(T). The friction coefficient µ that minimizes the objective function value is defined as the optimal friction coefficient under the current verification conditions. .

[0034] Step Six: To prove This is not an "overfit" to specific process parameters; independent verification is necessary. A completely new set of process parameters, not involved in any of the aforementioned optimization iterations, was used for physical experiments: the tool head rotation speed was adjusted to 1200 r / min, the forward speed to 240 mm / min, and the downward pressure to 0.2 mm. All other experimental conditions (material, tool head size, thermocouple placement, etc.) remained consistent with previous settings. Subsequently, Substitute the data into the model and perform simulation predictions. If the predicted temperature field matches the new experimental data well (error less than 5%), then the verified data is proven. It has good generalization ability.

[0035] Step 7: Transform the force and torque data collected in situ in the early stage into a mechanical load model that meets the simulation requirements, clarify the calculation relationship between surface pressure, tangential force and torque, provide accurate mechanical load input for subsequent stress-strain simulation, and achieve matching between the load model and the actual processing force transmission law.

[0036] S71: The force and torque data collected in step two are converted into a mechanical load model simulating the friction roller additive manufacturing process. The downward pressure applied to the tool head's action area is simplified and considered as a surface pressure P uniformly distributed on the intersecting cross section and a tangential force distributed along the circumference. Torque described .

[0037] ;

[0038] ;

[0039] ;

[0040] In the formula, is the radius of the tool head, L is the length of the tool head; r is the distance from the integration point to the center of the tool head; x and z are the coordinates of the integration point. Mz is the downward pressure of the tool head in the direction perpendicular to the substrate contact interface, and Mz is the tool head torque.

[0041] Step 8: Establish a mechanical load model, adding the mechanical performance parameters required for stress analysis to the geometric model of the stress analysis, such as Poisson's ratio, elastic modulus, coefficient of thermal expansion, and yield strength. The constitutive model describes the material constitutive structure. The mechanical analysis model selects the C3D8R element, which is an 8-node hexahedral element used for three-dimensional mechanical analysis. During the calculation, the reduced integration method and hourglass control are used for this element.

[0042] Step 9: Apply the mechanical load model of the tool head to the mechanical analysis using the DLOAD subroutine, including the surface pressure P uniformly distributed on the intersecting cross-section and the tangential force distributed along the circumference. In the mechanical analysis model, the initial position, movement path, and movement speed of the mechanical load are kept consistent with the actual deposition process. In the temperature field obtained in step three of the stress-strain simulation coupling, boundary conditions are set to restrict the displacement of the simulation model. After the deposition process is completed, the constraints are gradually released, while the rest remains consistent with the heat transfer model, resulting in a finite element model for stress field simulation calculations of the deposition process in friction roller additive manufacturing.

[0043] In summary, the shortcomings of existing technologies in parameter calibration, coupling mechanisms, and formula adaptability make it difficult for the prediction accuracy and reliability of the friction roller pressure-thermal coupling model to meet engineering requirements. Therefore, there is an urgent need to propose a model establishment and verification method based on in-situ measurement data to address these technical challenges.

[0044] 1. Improve model adaptability and accuracy: In response to the characteristics of lateral rolling of friction roller pressing tool head, a special heat source and mechanical load formula is established to solve the problem of poor adaptability of existing technical formulas. This makes the calculation of heat source intensity and load distribution highly matched with the actual processing scenario, accurately restoring the essence of thermo-mechanical coupling.

[0045] 2. Modeling based on in-situ measurement data and improving the reliability of model parameters: Heat source model and mechanical load model are established by in-situ force and torque, and the optimal friction coefficient is determined by combining in-situ temperature measurement data with iterative calibration and independent verification. This avoids the subjective bias of parameters caused by relying on empirical data, ensures that the core parameters are adapted to specific working conditions, and significantly reduces the prediction error of temperature field.

[0046] 3. Achieve realistic thermo-mechanical coupling simulation: Break through the limitations of single-field simulation, construct a complete model of temperature field and stress-strain field coupling, accurately reflect the physical mechanism of the processing, improve the reliability of prediction of key quality indicators such as residual stress and deformation of deposited parts, and provide precise support for process optimization. Attached Figure Description

[0047] Figure 1 This is a flowchart of the in-situ parametrically constrained friction roller additive manufacturing simulation method described in this invention.

[0048] Figure 2 This is a flowchart of the model parameter verification process described in this invention. Detailed Implementation

[0049] The following describes the invention in further detail with reference to the accompanying drawings, using the establishment of a thermo-mechanical coupling model and the verification of model parameters based on in-situ measurement data from friction roller additive manufacturing as an example:

[0050] Step 1: Construct a finite element model for heat transfer;

[0051] S11: Establish a geometric model of the deposited sample using finite element analysis software based on the actual dimensions of the sample.

[0052] S12: Assign material properties to the established finite element model, including density, Poisson's ratio, thermal conductivity and specific heat. These thermophysical parameters will change with temperature. Set the ambient temperature and the initial temperature of the test plate to 20℃.

[0053] S13: Establish a friction roll additive manufacturing analysis step that matches the actual deposition process time.

[0054] S14: Mesh the above sample with a mesh size of 0.3mm×0.3mm×0.3mm and set the mesh type to eight-node linear heat transfer hexahedral element (DC3D8).

[0055] Step 2: Collect in-situ force parameters to achieve heat transfer simulation.

[0056] S21: The three-dimensional force sensor captures the three-dimensional force signal generated by the contact between the tool and the substrate in real time. After being amplified and filtered by the three-dimensional force acquisition card, it is converted into a digital signal and transmitted to the computer to realize the in-situ real-time acquisition of force data.

[0057] S22: Collect parameters such as the angle of the contact area between the tool head and the substrate, the tool head rotation speed, the forward speed, the bottom radius of the tool head, and the height by querying the process parameter window.

[0058] S23: The collected process parameters are converted into a heat source model simulating the heat source distribution during the friction roll additive manufacturing process, i.e., the frictional heat source generated by the contact between the tool head side surface and the substrate. The heat source is set as a volumetric heat source and applied to the tool head's working area.

[0059]

[0060] In the formula, The angle between the tool head and the substrate contact area. The rotational speed of the tool head. Let L be the contact shear stress at the interface between the tool head and the substrate, where L is the width of the tool head and r is the radius of the tool head.

[0061] S24: Contact shear stress at the interface between the tool head and the substrate .

[0062] ;

[0063] In the formula, Let P be the coefficient of friction and P be the downward force.

[0064] S25: Apply the heat source Q obtained in S23 to the model through the DFLUX subroutine. The initial position, moving path, and moving speed of the heat source are kept consistent with the actual deposition process to perform heat transfer simulation.

[0065] Step 3: Determine the range of friction parameters.

[0066] S31: Based on literature review, experimental data on friction coefficients related to "6061-T6 aluminum alloy" and solid-state additive manufacturing were selected. Through statistical analysis of these data, a reasonable range of friction coefficient µ [0.1, 0.5] was initially determined as the physical constraint range for subsequent optimization.

[0067] S32: Take the minimum value of the above interval ( =0.1) is used as the reference value for the friction coefficient. In the constructed finite element geometric model, this reference value is assigned to the contact shear stress at the interface between the tool head and the substrate, and a complete heat transfer simulation is run to simulate the actual processing process.

[0068] Step 4: Iterative calculation of friction coefficient.

[0069] S41: Add the same number of analysis steps as the number of iterations to the geometric model analysis step block. The objects selected in each analysis step are the same, and the rest remain unchanged.

[0070] S42: Set the friction coefficient µ as a cyclic variable. (Base value) Starting from [0.1, 0.5], a total of 20 iterations are performed within the interval [0.1, 0.5] (i.e., µ = 0.1, 0.12, 0.14, ...). The µ value is automatically updated and the solver is invoked at each analysis step.

[0071] S43: Bind the friction coefficient µ, which needs to be incremented, to the formal parameter, and set "for each solution step completed (KSTEP increments by 1), the friction coefficient µ increases by 0.02 increments", friction coefficient µ = +Δµ×(KSTEP-1)( The initial value is 0.1, and Δµ is the increment for each iteration. Each time DLOAD is called in the solution step, the subroutine automatically updates the friction coefficient according to the rules, realizing the iterative cycle of "solution-increment-resolution".

[0072] Step 5: Solving for the optimal friction coefficient.

[0073] S51: The thermocouple probe is placed in the pre-set thermocouple hole, making direct contact with the workpiece to be processed. The temperature data is transmitted wirelessly to the receiver via the temperature acquisition card, and then uploaded to the computer by the receiver, realizing in-situ real-time acquisition of temperature data.

[0074] S52: After the simulation is completed, extract the temperature-time history data of the nodes that perfectly correspond to the spatial coordinates of the thermocouple placement points in the experiment from the results file. Compare this simulated temperature curve with the experimentally measured temperature curve and calculate the mean square error (MSE). As a core evaluation indicator, the simulation deviation under initial parameters is quantified, thereby establishing a mapping relationship between the "friction coefficient and simulation error".

[0075] MSE ( )= ;

[0076] In the formula, MSE ( ) is the mean square error of the temperature at node i, and n is the number of iterations; It is the node temperature obtained from the experiment; This is the node temperature obtained from the i-th simulation;

[0077] S53: After all 20 iterations are completed, compare the values ​​of all MSE(T). Let the objective function E(µ) = MSE(T). The friction coefficient µ that minimizes the objective function value is defined as the optimal friction coefficient under the current verification conditions. .

[0078] Step Six: To prove This is not simply "overfitting" to specific process parameters; independent validation is essential. Physical experiments should be conducted using a completely new set of process parameters (such as different deposition rates or rotational speeds) that have not been involved in any of the aforementioned optimization iterations. Subsequently, Substitute the data into the model and perform simulation predictions. If the predicted temperature field matches the new experimental data well (error less than 5%), then the verified data is proven. It has good generalization ability.

[0079] Step 7: Conversion of in-situ force parameters and torque.

[0080] S71: The force and torque data collected in step two are converted into a mechanical load model simulating the friction roller additive manufacturing process. The downward pressure applied to the tool head's action area is simplified and considered as a surface pressure P uniformly distributed on the intersecting cross section and a tangential force distributed along the circumference. Torque described

[0081] ;

[0082] ;

[0083] ;

[0084] In the formula, is the radius of the tool head, L is the length of the tool head; r is the distance from the integration point to the center of the tool head; x and z are the coordinates of the integration point. Mz is the downforce, and Mz is the tool head torque.

[0085] Step 8: Establish a mechanical load model, adding the mechanical performance parameters required for stress analysis to the geometric model of the stress analysis, such as Poisson's ratio, elastic modulus, coefficient of thermal expansion, and yield strength. The constitutive model describes the material constitutive structure. The mechanical analysis model selects the C3D8R element, which is an 8-node hexahedral element used for three-dimensional mechanical analysis. During the calculation, the reduced integration method and hourglass control are used for this element.

[0086] Step 9: Apply the mechanical load model of the tool head to the mechanical analysis using the DLOAD subroutine, including the surface pressure P uniformly distributed on the intersecting cross-section and the tangential force distributed along the circumference. In the mechanical analysis model, the initial position, movement path, and movement speed of the mechanical load are kept consistent with the actual deposition process. In the temperature field obtained in step three of the stress-strain simulation coupling, boundary conditions are set to restrict the displacement of the simulation model. After the deposition process is completed, the constraints are gradually released, while the rest remains consistent with the heat transfer model, resulting in a finite element model for stress field simulation calculations of the deposition process in friction roller additive manufacturing.

Claims

1. A method for friction stir welding and additive manufacturing modeling based on in-situ measurement data, characterized in that, The method includes the following steps: Step 1: Build a basic framework for heat transfer simulation that matches the actual processing scenario. Through geometric modeling, material property definition, analysis step setting and mesh generation, provide a model carrier for subsequent heat transfer simulation and ensure that the simulation boundary conditions and time dimension are consistent with the actual deposition process. Step 2: By collecting force data in situ in real time, and combining it with process parameters to construct a heat source model, the heat source distribution and intensity calculation are accurately matched with the actual working conditions of friction roller pressing additive manufacturing, providing core load input for heat transfer simulation; Step 3: Screen experimental data, determine the reasonable physical constraint range of friction coefficient, set the benchmark value and complete the first heat transfer simulation to provide a clear parameter range and initial reference for subsequent iterative calculations, and avoid subjective bias in parameter values; Step 4: Starting from the baseline value, by setting loop variables, adding analysis steps and binding formal parameters, the friction coefficient is implemented in an ordered incremental iteration within the constraint range, and multiple rounds of heat transfer simulation are completed, providing simulation data support for establishing the friction coefficient-simulation error mapping relationship; Step 5: By comparing the simulated temperature curve with the measured temperature curve, the mean square error is calculated to quantify the simulation deviation, the correspondence between the friction coefficient and the error is established, and finally the optimal friction coefficient that minimizes the error is selected to ensure that the core process parameters are adapted to the specific working conditions and improve the model prediction accuracy. Step Six: Perform physical experiments to verify the process parameters using a completely new set of process parameters that have not participated in any of the aforementioned optimization iterations; Step 7: Transform the force and torque data collected in situ into a mechanical load model that meets the simulation requirements, clarify the calculation relationship between surface pressure, tangential force and torque, provide mechanical load input for subsequent stress and strain simulation, and achieve the matching between the load model and the actual processing force transmission law; Step 8: Establish a mechanical load model and add the mechanical performance parameters required for stress analysis to the geometric model of stress analysis; Step 9: Apply the mechanical load model of the tool head to the mechanical analysis using the DLOAD subroutine, including the surface pressure P uniformly distributed on the intersecting cross-section and the tangential force distributed along the circumference. The mechanical load is applied to the mechanical analysis model, and the initial position, movement path, and movement speed of the mechanical load are kept consistent with the actual deposition process. In the temperature field obtained in the third step of stress-strain simulation coupling, boundary conditions are set to limit the displacement of the simulation model. After the deposition process is completed, the constraints are gradually released, and the rest are kept consistent with the heat transfer model. The finite element model is then used to perform stress field simulation calculations on the deposition process of friction roller additive manufacturing.

2. The method for friction stir welding and additive modeling based on in-situ measurement data according to claim 1, characterized in that, Step one includes: S11: Based on the actual dimensions of the friction stir welded sample, a geometric model of the deposited sample was established using finite element analysis software; S12: Assign material properties to the geometric model established based on finite element analysis software, including density, Poisson's ratio, thermal conductivity and specific heat; S13: Establish an analysis procedure for friction stir welding that is consistent with the actual deposition process time, and set the analysis step time to be the same as the actual welding time; S14: Mesh the above friction stir welded sample and set the mesh type to an eight-node linear heat transfer hexahedral element.

3. The method for friction stir welding and additive modeling based on in-situ measurement data according to claim 1, characterized in that, Step two includes: S21: The three-dimensional force sensor captures the three-dimensional force signal generated by the contact between the tool and the substrate in real time. After being amplified and filtered by the three-dimensional force acquisition card, it is converted into a digital signal and transmitted to the computer to realize the in-situ real-time acquisition of force data. S22: Collect parameters such as the angle of the contact area between the tool head and the substrate, the tool head rotation speed, the forward speed, the bottom radius of the tool head, and the height through the process parameter query window; S23: Collect the angle of the contact area between the tool head and the substrate, the tool head rotation speed, the forward speed, the bottom radius of the tool head, and the height by querying the process parameter window; S24: The collected process parameters are converted into a heat source model simulating the heat source distribution during the friction roll additive manufacturing process, i.e., the frictional heat source generated by the contact between the tool head side surface and the substrate; heat source Q 总 Set as a volumetric heat source and apply it to the working area of ​​the tool head; ； In the formula, 𝛳 is the angle between the tool head and the substrate contact area, and 𝜔 is the rotational speed of the tool head. Let L be the contact shear stress at the interface between the tool head and the substrate, and let r be the width of the tool head and r be the radius of the tool head. S25: Contact shear stress at the interface between the tool head and the substrate ; ； In the formula, The coefficient of friction is the interface between the tool head and the substrate. The downward pressure applied by the tool head perpendicular to the substrate contact interface; S26: Transfer the heat source Q obtained in S24 to a subroutine. 总 In the heat transfer model of friction stir welding, the initial position, moving path, and moving speed of the heat source are kept consistent with the actual deposition process to simulate heat transfer.

4. The method for friction stir welding and additive modeling based on in-situ measurement data according to claim 1, characterized in that, Step three includes: S31: Screen experimental data on friction coefficients that are relevant to the materials used in the experiment and to solid-phase additive manufacturing; through statistical analysis of these data, determine a range of friction coefficient µ values ​​[0.1, 0.5] as the constraint range for friction parameters of friction stir welding to be optimized in the future; S32: Take the minimum value of the above friction coefficient µ range as the reference value of the friction coefficient; in the constructed finite element geometric model, assign this reference value to the contact shear stress at the interface between the tool head and the substrate, run a complete heat transfer simulation, and simulate the actual processing process.

5. The method for friction stir welding and additive modeling based on in-situ measurement data according to claim 1, characterized in that, Step four includes: S41: Add an analysis step to the geometric model analysis step block with the same number of iterations. The objects selected in each analysis step are the same, and the rest remain unchanged. S42: Set the friction coefficient µ as a cyclic variable; based on the reference value. Starting from [0.1, 0.5], a total of 20 iterations are performed within the interval [0.1, 0.5] with a step size of 0.02; the µ value is automatically updated and the solver is invoked in each analysis step; S43: Bind the friction coefficient µ, which needs to be incremented, to the formal parameter. Set: for each solution step completed, i.e., KSTEP increments by 1, the friction coefficient µ increases by 0.02 increments. Friction coefficient µ = +Δµ×(KSTEP-1), The initial value is 0.1, and Δµ is the increment for each iteration. When DLOAD is called in each solution step, the subroutine automatically updates the friction coefficient according to the rules, realizing the iterative cycle of solution-increment-re-solution.

6. The method for friction stir welding and additive modeling based on in-situ measurement data according to claim 1, characterized in that, Step five includes: S51: The thermocouple probe is placed in the thermocouple pre-set hole and comes into direct contact with the workpiece to be processed. The temperature data is transmitted wirelessly to the receiving end through the temperature acquisition card, and then uploaded to the computer by the receiving end to realize the in-situ real-time acquisition of temperature data. S52: After the simulation is completed, extract the temperature-time history data of the nodes that perfectly correspond to the spatial coordinates of the thermocouple placement points in the experiment from the result file; compare this simulated temperature curve with the experimentally measured temperature curve, and calculate the mean square error (MSE). As a core evaluation indicator, the simulation deviation under initial parameters is quantified; and the mapping relationship between friction coefficient and simulation error is established. MSE( )= ; In the formula, MSE ( ) is the mean square error of the temperature at node i, and n is the number of iterations; It is the node temperature obtained from the experiment; This is the node temperature obtained from the i-th simulation; S53: After all 20 iterations are completed, compare the values ​​of all MSE(T). Let the objective function E(µ) = MSE(T). The friction coefficient µ that minimizes the objective function value is defined as the optimal friction coefficient under the current verification conditions. .

7. The method for friction stir welding and additive modeling based on in-situ measurement data according to claim 1, characterized in that, Step seven includes: S71: The force and torque data collected in step two are converted into a mechanical load model simulating the friction roller additive manufacturing process. The downward pressure applied to the tool head's action area is simplified and considered as a surface pressure P uniformly distributed on the intersecting cross section and a tangential force distributed along the circumference. Torque described ; ； ； ； In the formula, is the radius of the tool head, L is the length of the tool head; r is the distance from the integration point to the center of the tool head; x and z are the coordinates of the integration point. Mz is the downward pressure of the tool head in the direction perpendicular to the substrate contact interface, and Mz is the tool head torque.

Images