CN-121999945-A - Aluminum alloy corrosion prediction method based on digital twin and grating sensing

CN121999945ACN 121999945 ACN121999945 ACN 121999945ACN-121999945-A

Abstract

The application relates to the technical field of structural health monitoring and nondestructive testing, and discloses an aluminum alloy corrosion prediction method based on digital twin and grating sensing. And judging the heat conduction state by using the environmental heat change rate index, and controlling the digital twin model to execute an asynchronous iteration correction strategy, wherein the local equivalent heat conduction coefficient is updated by using the temperature residual error in an unsteady heat conduction stage, and the local equivalent elastic modulus is updated by using the relative transmissibility in a quasi-static heat balance stage. And finally, the system calculates the weight based on the thermal and mechanical relative errors, performs weighted fusion on the damping ratio of the converged heat conductivity coefficient and the elastic modulus, and calculates and outputs continuous corrosion depth distribution. According to the application, by decoupling the multiple physical field parameters by utilizing the natural temperature change characteristics, the accurate quantitative evaluation of the corrosion damage of the aluminum alloy structure under the unknown external load condition is realized.

Inventors

Cen Yuanyao
Liao Guangmeng
ZHU YUQIN
LIU CONG
HE QIONGYAO
LI LINGJIE
HE JIANXIN

Assignees

重庆大学
中国兵器装备集团西南技术工程研究所

Dates

Publication Date: 20260508
Application Date: 20260211

Claims (10)

1. The aluminum alloy corrosion prediction method based on digital twin and grating sensing is characterized by comprising the following steps of: Establishing a digital twin model of an aluminum alloy structure, wherein the digital twin model comprises a transient heat conduction model and a heat engine coupling mechanical model which are mutually coupled, and takes a local equivalent heat conductivity coefficient and a local equivalent elastic modulus as model parameters to be inverted; Synchronously acquiring temperature data and strain data of a monitoring point sensor group and a reference point sensor group through a distributed optical fiber sensing network arranged on the surface of the aluminum alloy structure; calculating an environmental thermal change rate index according to the time change characteristics of the temperature data acquired by the reference point sensor group; According to a comparison result of the environmental thermal change rate index and a preset threshold value, controlling the digital twin model to perform asynchronous iterative switching between a thermal resistance dominant correction mode and a stiffness dominant correction mode, and driving the digital twin model by using synchronously acquired temperature data and strain data, and respectively updating the local equivalent heat conductivity coefficient and the local equivalent elastic modulus in the digital twin model until a global convergence condition is met; And extracting the converged local equivalent heat conductivity coefficient and local equivalent elastic modulus, calculating the corrosion depth of the aluminum alloy structure through weighted mapping, and outputting.
2. The aluminum alloy corrosion prediction method based on digital twin and grating sensing according to claim 1, wherein the monitoring point sensor group and the reference point sensor group both comprise a temperature grating and a strain grating; the synchronous acquisition monitoring point sensor group and the temperature data and the strain data of the reference point sensor group comprise: respectively acquiring the wavelength drift amount of the temperature grating and the wavelength drift amount of the strain grating in each sensor group; converting the wavelength drift amount of the temperature grating into an actual measured temperature by using a temperature sensitivity coefficient, and taking the actual measured temperature as the temperature data; performing temperature compensation on the wavelength drift amount of the strain grating by using the actually measured temperature, and calculating actually measured total strain; And subtracting a thermal expansion component from the measured total strain based on the linear thermal expansion coefficient of the aluminum alloy material and the measured temperature, calculating to obtain the measured mechanical strain caused by the mechanical load only, and taking the measured mechanical strain as the strain data.
3. The aluminum alloy corrosion prediction method based on digital twinning and grating sensing according to claim 1, wherein the establishing a digital twinning model of an aluminum alloy structure comprises: Discretizing an aluminum alloy structure based on a finite element method, and constructing a transient heat conduction equation set containing an integral heat conduction matrix and a mechanical balance equation set containing an integral rigidity matrix; wherein the overall heat conduction matrix is constructed by utilizing the local equivalent heat conductivity coefficient, and the overall rigidity matrix is constructed by utilizing the local equivalent elastic modulus; And applying the node temperature field calculated by the transient heat conduction equation set to the mechanical balance equation set as a thermal equivalent load by adopting a sequential coupling strategy, and solving the mechanical balance equation set to obtain a thermally induced stress response.
4. The aluminum alloy corrosion prediction method based on digital twin and grating sensing according to claim 3, The digital twin model for building the aluminum alloy structure further comprises the step of applying boundary conditions, and specifically comprises the following steps: mapping the measured temperature of the reference point sensor group to a heat exchange boundary of the digital twin model as a forced temperature boundary condition of the transient heat conduction model; And applying zero displacement constraint on the constraint boundary of the digital twin model as a displacement boundary condition of the thermo-mechanical coupling mechanical model.
5. The aluminum alloy corrosion prediction method based on digital twinning and grating sensing according to claim 1, wherein the steps of calculating an environmental thermal change rate index and switching modes comprise: calculating the absolute value of the derivative of the measured temperature of the reference point sensor group with respect to time as the environmental thermal change rate index; When the environmental thermal change rate index is larger than the preset threshold, judging that the aluminum alloy structure is in an unsteady heat conduction stage, and executing the heat resistance dominant correction mode; And when the environmental thermal change rate index is smaller than or equal to the preset threshold, judging that the aluminum alloy structure is in a quasi-static heat balance stage, and executing the rigidity leading correction mode.
6. The aluminum alloy corrosion prediction method based on digital twin and grating sensing according to claim 5, wherein the specific steps of executing the thermal resistance dominant correction mode are: locking the local equivalent elastic modulus in the digital twin model to remain unchanged; Driving the digital twin model to simulate a transient temperature field by using the measured temperature of the reference point sensor group to obtain a simulated temperature of the position of the monitoring point; constructing a thermal target functional, wherein the thermal target functional characterizes an error between the measured temperature of the monitoring point sensor group and the simulation temperature; And iteratively updating the local equivalent thermal conductivity in the digital twin model by minimizing the thermal target functional.
7. The aluminum alloy corrosion prediction method based on digital twin and grating sensing according to claim 5, wherein the specific steps of executing the stiffness dominant correction mode are: Locking the local equivalent heat conductivity coefficient in the digital twin model to remain unchanged; Applying a virtual unit load on the digital twin model, performing statics simulation, and respectively calculating virtual mechanical strain at the position of the monitoring point and virtual mechanical strain at the position of the reference point; Calculating the ratio of the actual mechanical strain of the monitoring point relative to the actual mechanical strain of the reference point by utilizing the actual measurement data to serve as the actual measurement relative transmissibility, and calculating the ratio of the virtual mechanical strain of the monitoring point relative to the virtual mechanical strain of the reference point by utilizing the simulation data to serve as the simulation relative transmissibility; Constructing a mechanical target functional, wherein the mechanical target functional characterizes an error between the measured relative transfer rate and the simulated relative transfer rate; and iteratively updating the local equivalent elastic modulus in the digital twin model by minimizing the mechanical target functional.
8. The aluminum alloy corrosion prediction method based on digital twinning and grating sensing according to claim 1, wherein the until a global convergence condition is satisfied comprises: Tracking the update courses of the local equivalent heat conductivity coefficient and the local equivalent elastic modulus respectively; calculating the relative change rate of the current updating parameter relative to the previous historical state, and taking the relative change rate as a relative convergence norm; And when the relative convergence norm of the local equivalent heat conductivity coefficient and the relative convergence norm of the local equivalent elastic modulus are simultaneously smaller than a preset global convergence tolerance and the iteration times of the two reach the lower limit of the minimum iteration steps, judging that the global convergence condition is met.
9. The aluminum alloy corrosion prediction method based on digital twinning and grating sensing according to claim 8, wherein the calculating the corrosion depth of the aluminum alloy structure by weighted mapping comprises: taking the value of the thermal target functional when the global convergence condition is met as thermal residual, and calculating thermal relative errors based on the thermal residual; Taking the numerical value of the mechanical target functional when the global convergence condition is met as a mechanical residual error, and calculating a mechanical relative error based on the mechanical residual error; calculating a thermal component weight coefficient and a mechanical component weight coefficient based on the reciprocal relation between the thermal relative error and the mechanical relative error; reading a reference heat conductivity coefficient and a reference elastic modulus of the aluminum alloy structure in an initial nondestructive state; Calculating the damping ratio of the converged local equivalent thermal conductivity coefficient relative to the reference thermal conductivity coefficient and the damping ratio of the converged local equivalent elastic modulus relative to the reference elastic modulus; And weighting and summing the attenuation ratio by using the thermal component weight coefficient and the mechanical component weight coefficient to obtain the local corrosion depth of each unit.
10. The aluminum alloy corrosion prediction method based on digital twinning and grating sensing of claim 9, wherein the outputting step further comprises: Extracting local corrosion depths and geometric coordinates of all units, and constructing a continuous corrosion depth distribution function by using an interpolation algorithm; Mapping the continuous corrosion depth distribution function into a visual image to generate a full-field corrosion damage cloud picture; and extracting a maximum corrosion depth value in the continuous corrosion depth distribution function, and triggering an alarm signal when the maximum corrosion depth value exceeds a preset safety tolerance threshold.

Description

Aluminum alloy corrosion prediction method based on digital twin and grating sensing Technical Field The invention relates to the technical field of structural health monitoring and nondestructive testing, in particular to an aluminum alloy corrosion prediction method based on digital twin and grating sensing. Background The aluminum alloy material is widely applied to the manufacturing of bearing structures in the fields of aerospace, rail transit, ocean engineering and the like due to high specific strength and good processability. However, such structures are often subjected to environmental wet and hot alternation and chemical medium attack during long-term service, and electrochemical corrosion occurs. Corrosion not only results in a reduction of the effective thickness of the matrix material, but also results in a loose oxide layer, thereby causing a dual degradation of the structural local thermal conductivity and mechanical stiffness. If the corrosion degree cannot be timely and accurately estimated, structural failure and even serious safety accidents can be caused. The existing aluminum alloy corrosion detection means mainly comprise traditional nondestructive detection technologies such as ultrasonic detection, eddy current detection and X-ray imaging. These methods generally require equipment to be shut down and operated off-line, have long detection periods and high labor costs, and are difficult to meet the requirements of real-time state sensing of critical structures. In recent years, an on-line monitoring technology based on a strain gauge or a fiber bragg grating is gradually applied, and the basic principle is to back-push internal damage by monitoring the strain response of the surface of a structure. However, in an actual service environment, the aluminum alloy structure is often under the coupling action of a variable temperature field and dynamic mechanical load. The existing online monitoring and inversion method has certain limitation in processing the complex working conditions. On the one hand, conventional strain inversion algorithms typically rely on known external load magnitudes as boundary conditions, but during flight or navigation, the aerodynamic or hydrodynamic loads to which the structure is subjected are random and unknown, which makes it difficult to achieve stiffness identification based on absolute deformation. On the other hand, the corrosion-induced change in physical parameters has a multi-physical field characteristic, i.e., a decrease in thermal conductivity occurs simultaneously with an attenuation in elastic modulus. The prior art often addresses the heat conduction problem independently of the mechanical deformation problem or attempts to invert all parameters simultaneously in one system of equations. Because thermal expansion deformation caused by a temperature field and elastic deformation caused by mechanical load are mutually overlapped in a sensor signal, under the condition of lacking an accurate decoupling mechanism, non-uniqueness of non-convergence or solution of a calculation process is easily caused by directly carrying out multi-parameter synchronous inversion, and signal change caused by environmental temperature change and structural performance degradation caused by corrosion damage are difficult to accurately distinguish. Disclosure of Invention Aiming at the defects of the prior art, the invention provides the aluminum alloy corrosion prediction method based on digital twin and grating sensing, which solves the problem that the aluminum alloy structure corrosion damage is difficult to accurately decouple and quantitatively evaluate through online monitoring data under the conditions of unknown external load and coupling of a heat engine in the prior art. In order to achieve the above purpose, the invention is realized by the following technical scheme: the first aspect of the invention provides an aluminum alloy corrosion prediction method based on digital twin and grating sensing, which comprises the following steps: First, a digital twin model of the aluminum alloy structure is established. Discretizing the aluminum alloy structure by using a finite element method to construct a transient heat conduction model and a thermal-mechanical coupling mechanical model. The transient heat conduction model comprises an overall heat conduction matrix constructed by local equivalent heat conductivity coefficients, and the mechanical balance equation set comprises an overall stiffness matrix constructed by local equivalent elastic modulus. And applying a node temperature field calculated by the transient heat conduction equation set to the mechanical balance equation set as a thermal equivalent load by adopting a sequential coupling strategy, and calculating the thermal stress response of the structure. And secondly, synchronous acquisition and calculation of the multi-physical-field data are executed. And the wavelength drift amount is respectively ob