Method and system for dynamic monitoring of hot forming of duplex stainless steel based on digital model

Abstract

The application discloses a duplex stainless steel hot forming dynamic monitoring method and system based on a digital model, and relates to the technical field of stainless steel hot forming.The method comprises the following steps: determining the geometric shape of a duplex stainless steel pipe fitting, and constructing a digital coupling model of heat-power-phase change; detecting the duplex stainless steel pipe fitting and reconstructing a field based on the digital coupling model according to a front-end distributed monitoring array, and determining hot forming data; triggering a dynamic decision maker for hierarchical connection judgment and adjustment guide generation, performing adjustment simulation and parameter control conversion based on the digital coupling model, and determining hot forming control data; and performing feedback adjustment of the hot forming process of the duplex stainless steel pipe fitting according to the hot forming control data.The technical problem of unstable forming quality caused by the lack of a dynamic control mechanism in the hot forming process of the duplex stainless steel pipe fitting in the prior art is solved, and the technical effect of improving the hot forming quality of the duplex stainless steel pipe fitting is achieved.

Description

Technical Field

[0001] This invention relates to the field of stainless steel hot forming technology, and specifically to a dynamic monitoring method and system for dual-phase stainless steel hot forming based on a digital model. Background Technology

[0002] Duplex stainless steel is widely used in critical fields such as petrochemicals, nuclear power equipment, and marine engineering due to its excellent strength, toughness, corrosion resistance, and high-temperature performance. However, duplex stainless steel faces challenges during hot forming, including sensitivity to the austenite-ferrite phase ratio, complex phase transformation processes, and embrittlement. These issues result in a narrow forming process window and significant control difficulties. Traditional hot forming methods rely heavily on preset process parameters and empirical adjustments, lacking real-time sensing and dynamic feedback control of the temperature field, stress field, and microstructure evolution during the actual forming process. This makes it difficult to ensure consistent forming quality and reliable structural performance. Summary of the Invention

[0003] This application provides a method and system for dynamic monitoring of duplex stainless steel hot forming based on a digital model, which solves the problem that the existing technology lacks a dynamic control mechanism for the hot forming process of duplex stainless steel pipe fittings.

[0004] The first aspect of this application provides a method for dynamic monitoring of hot forming of duplex stainless steel based on a digital model, the method comprising: The geometry of duplex stainless steel pipe fittings is determined, and a digital coupling model of thermo-mechanical-phase transformation is constructed. Based on a front-end distributed monitoring array, the duplex stainless steel pipe fittings are inspected and the field is reconstructed based on the digital coupling model to determine hot forming data, which includes first inspection data and second simulation data. For the hot forming data, a dynamic decision-maker is triggered to perform hierarchical judgment and adjustment guidance generation, and adjustment simulation and parameter control conversion based on the digital coupling model are executed to determine hot forming control data. The hierarchical judgment elements include at least thermal uniformity, embrittlement temperature threshold, stress path, and ferrite phase. Feedback adjustment of the hot forming process of the duplex stainless steel pipe fittings is performed based on the hot forming control data.

[0005] A second aspect of this application provides a dynamic monitoring system for the hot forming of duplex stainless steel based on a digital model, the system comprising: Model building module: Determines the geometry of duplex stainless steel pipe fittings and constructs a digitally coupled model of thermo-mechanical-phase transformation; Detection module: Based on a front-end distributed monitoring array, detects the duplex stainless steel pipe fittings and reconstructs the field based on the digitally coupled model to determine hot forming data, wherein the hot forming data includes first detection data and second simulation data; Decision module: For the hot forming data, triggers a dynamic decision-maker to perform hierarchical judgment and adjustment guidance generation, executes adjustment simulation and parameter control conversion based on the digitally coupled model, and determines hot forming control data, wherein the hierarchical judgment elements include at least thermal uniformity, embrittlement temperature threshold, stress path, and ferrite phase; Feedback adjustment module: Performs feedback adjustment of the hot forming process of duplex stainless steel pipe fittings based on the hot forming control data.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages: First, the geometry of the duplex stainless steel pipe fitting is determined, and a digital coupling model of thermo-mechanical-phase transformation is constructed. Next, based on a front-end distributed monitoring array, the duplex stainless steel pipe fitting is inspected and its field reconstructed using the digital coupling model to determine hot forming data, which includes first inspection data and second simulation data. Then, based on the hot forming data, a dynamic decision-maker is triggered to perform hierarchical judgment and regulation guidance generation, executing regulation simulation and parameter control conversion based on the digital coupling model to determine hot forming control data. The hierarchical judgment elements include at least thermal uniformity, embrittlement temperature threshold, stress path, and ferrite phase. Finally, feedback regulation of the hot forming process of the duplex stainless steel pipe fitting is performed based on the hot forming control data. This solves the technical problem of unstable forming quality caused by the lack of a dynamic control mechanism in the hot forming process of duplex stainless steel pipe fittings in existing technologies, achieving the technical effect of improving the hot forming quality of duplex stainless steel pipe fittings. Attached Figure Description

[0007] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0008] Figure 1 A schematic diagram of the dynamic monitoring method for hot forming of duplex stainless steel based on a digital model provided in this application embodiment; Figure 2 A schematic diagram of the structure of a dynamic monitoring system for hot forming of duplex stainless steel based on a digital model, provided in an embodiment of this application.

[0009] Figure labeling: Model building module 11, detection module 12, decision-making module 13, feedback adjustment module 14. Detailed Implementation

[0010] This application provides a dynamic monitoring method and system for the hot forming of duplex stainless steel based on a digital model, which solves the problem of the lack of dynamic control mechanism in the hot forming process of duplex stainless steel pipe fittings in the prior art.

[0011] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0012] It should be noted that the terms "comprising" and "having" are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to these processes, methods, products, or devices.

[0013] Example 1, as Figure 1 As shown, this application provides a method for dynamic monitoring of duplex stainless steel hot forming based on a digital model, wherein the method includes: The geometry of duplex stainless steel pipe fittings was determined, and a digital coupled model of thermo-mechanical-phase transformation was constructed.

[0014] In this embodiment, structural dimensional data of duplex stainless steel pipe fittings are obtained by importing CAD drawings, including parameters such as the outer diameter, wall thickness, length, bending angle, and connection interface shape. Based on this, a three-dimensional geometric model of the duplex stainless steel pipe fittings is constructed using a finite element modeling tool (such as Abaqus). Next, based on the thermo-mechanical-phase transformation multiphysics coupling relationship, a material constitutive model is loaded onto the geometric model. This material constitutive model includes temperature- and phase composition-related physical parameters such as thermal conductivity, specific heat capacity, Young's modulus, Poisson's ratio, latent heat of phase transformation, coefficient of thermal expansion, and yield strength. Further, heat conduction equations, elastoplastic mechanics equations, and phase transformation evolution equations coupled with the above geometric model are constructed. The phase transformation evolution equations employ a microstructure evolution model based on the time-temperature-phase composition (TTP) curve, describing the phase transformation path and rate between austenite and ferrite, forming a dynamic evolution relationship. The above-mentioned thermo-mechanical-phase transformation coupling equations are solved numerically to form a digital coupling model, which enables synchronous simulation of the temperature field, stress-strain field and microstructure phase composition evolution of duplex stainless steel pipe fittings during the hot forming process.

[0015] Based on the front-end distributed monitoring array, the duplex stainless steel pipe fittings are inspected and the field is reconstructed based on the digital coupling model to determine the thermoforming data, wherein the thermoforming data includes first inspection data and second simulation data.

[0016] During the thermoforming process, a front-end distributed monitoring array is deployed at key parts of the hot-working equipment. This monitoring array includes a temperature sensor array, a stress / strain sensor array, an acoustic emission sensor array, and a visual monitoring module. Through multi-point sensor deployment, it collects real-time data on the temperature distribution, stress distribution, material deformation behavior, and microstructure evolution of the duplex stainless steel pipe during the forming process, forming raw monitoring data sets such as thermal field point clouds, force field point clouds, and acoustic field point clouds. Subsequently, the collected monitoring data is input into a pre-constructed digital coupling model. The digital coupling model performs feature transformation and coupling reconstruction of the spatial field: using the temperature vector as the main variable, coupling the stress vector and phase composition variables, and performing the first stage of model initialization and simulation calculation under the thermo-mechanical-phase transition coupling relationship, outputting the first detection data reflecting the current actual thermoforming state. Simultaneously, based on the thermoforming conditions at the current sampling time (such as heat source input, cooling rate, forming rate, etc.), the digital coupling model undergoes a second stage of thermal field initialization. Combined with the embedded time-series phase transition model, the thermal field evolution and microstructure phase transition behavior at future time points are deduced, forming predictive second simulation data.

[0017] Furthermore, the inspection of duplex stainless steel pipe fittings and the field reconstruction based on the aforementioned digital coupling model include: Through distributed monitoring, thermal field point clouds, force field point clouds, and acoustic field point clouds are determined; feature transformation and coupling field reconstruction based on temperature vector, stress vector, and ferrite phase state are performed on the thermal field point clouds, force field point clouds, and acoustic field point clouds, and the digital coupling model is initialized to determine the first detection data; wherein, with temperature as the independent variable and stress and ferrite phase state as the dependent variables, a time-series phase transformation relationship of duplex stainless steel is established, and the time-series phase transformation relationship is embedded in the digital coupling model.

[0018] The surface temperature distribution, mechanical stress response, and acoustic emission characteristics of duplex stainless steel pipe fittings during the hot forming process are collected using a monitoring array. The monitoring array works synchronously and uses data fusion processing technology to form thermal field point clouds (representing temperature distribution), force field point clouds (representing stress and strain distribution), and acoustic field point clouds (characterizing acoustic responses during microcracks, plastic deformation, etc.), thus constituting a multiphysics data representation space.

[0019] Feature transformation processing is performed on thermal field point clouds, force field point clouds, and acoustic field point clouds: temperature vector distribution features are extracted from the thermal field point clouds, the force field point clouds are converted into spatial stress component vectors, and the acoustic field point clouds are mapped to ferrite phase transformation activity intensity indices through spectral and time-domain analysis, thus forming temperature vectors, stress vectors, and ferrite phase characteristic data. These three types of feature mapping data are input into a pre-constructed digital coupling model as the first initialization boundary and initial conditions. Based on this, the digital coupling model performs simulations according to the constructed heat conduction equation, elastoplastic mechanics equation, and embedded microstructure phase transformation evolution equation (i.e., time-series phase transformation relationship). The time-series phase transformation relationship uses temperature as the independent variable and stress state and ferrite phase content as dependent variables. The time-temperature-stress path mapping model describes the evolution trend of austenite and ferrite, capturing the phase transformation critical point, transformation rate, and microstructural response behavior. Through the above initialization and calculation process, the first stage model perception result is output, namely the first detection data. The first detection data reflects the actual temperature field distribution, stress path evolution and microstructure characteristics in the current thermoforming process.

[0020] Furthermore, after determining the first detection data, the following includes: First hot forming conditions are determined, wherein the first hot forming conditions are the hot forming conditions for collecting data at lower time nodes; based on the hot forming conditions, the digital coupling model is initialized with a thermal field, and the stress and ferrite phase state are digitally dynamically simulated using the time-series phase transformation relationship to determine second simulation data; the first detection data and the second simulation data are time-series cascaded as hot forming data.

[0021] After completing the initial initialization of the digital coupling model and acquiring the first detection data, the next lower time node thermoforming conditions in the current thermoforming process are further collected, i.e., the operating condition information immediately following the first detection data acquisition time point. The first thermoforming conditions include, but are not limited to, dynamic process parameters such as: current heat source power and heating rate, loading rate of the forming equipment, cooling method and cooling rate, contact state between the workpiece and the mold, ambient temperature, and initial microstructure of the material.

[0022] Based on the first hot forming condition, a digital coupling model is invoked to reinitialize the thermal field. During this process, using the first hot forming condition as the boundary input, and combining it with the embedded time-series phase transformation relationship model, the stress evolution path and dynamic change process of the ferrite phase content in the duplex stainless steel pipe fitting within the next time window are simulated and predicted. A set of multiphysics evolution results for the predicted time period is output as the second simulation data. Finally, the first detection data and the second simulation data are time-series concatenated and integrated according to the time series of the hot forming process to form the hot forming data.

[0023] For the hot forming data, a dynamic decision-maker is triggered to perform hierarchical judgment and adjustment guidance generation, and to perform adjustment simulation and parameter control conversion based on the digital coupling model to determine the hot forming control data. The hierarchical judgment elements include at least thermal uniformity, embrittlement temperature threshold, stress path and ferrite phase.

[0024] Based on the acquired hot forming data, the system triggers a dynamic decision-maker to perform multi-level, multi-factor comprehensive judgments on the current and predicted hot forming process status, and generates targeted adjustment guidelines. The dynamic decision-maker operates based on a multi-level decision architecture. The decision process first executes a cascaded judgment mechanism to identify key anomalies or optimization opportunities in the current hot forming state. Specifically, the dynamic decision-maker uses a first decision tree as the main judgment architecture, which consists of multiple sequentially cascaded binary classification judgment layers, including a thermal uniformity judgment layer, an embrittlement temperature threshold judgment layer, a stress path judgment layer, and a ferrite phase content judgment layer. These judgment layers are performed sequentially, with the output of each layer serving as the input condition for the next layer, achieving multi-factor cascaded and linked judgments. After completing all level judgments, the dynamic decision-maker comprehensively evaluates the overall judgment results through a subsequent judgment layer to determine whether industrial control strategy intervention or optimization adjustment is necessary.

[0025] If the judgment result triggers the adjustment condition, the system enters the adjustment guidance generation stage. Based on the deviation factors identified in the judgment, target adjustment guidance information is formed, mainly including stress path adjustment suggestions and ferrite phase state control targets, i.e., stress-microstructure control targets. Subsequently, the digital coupling model is retrieved, and its stress and ferrite phase fields are reinitialized. Adjustment simulation calculations are performed for the adjustment guidance target. Furthermore, the controllable parameters of the hot forming equipment are combined to establish a parameter control matrix, complete the parameter control conversion mapping of the adjustment parameters, and finally output a set of hot forming control data that can be used for equipment driving.

[0026] Furthermore, the construction of the dynamic decision-maker before triggering it includes: Construct a first decision tree, which includes a cascaded judgment layer; using the first decision tree as the judgment architecture and the field adjustment guidance as the generation architecture, retrieve monitoring and recording data and perform data integration based on cascaded judgment and adjustment guidance, use the data as sample data to perform architecture adversarial convergence training, and determine the dynamic decision-maker.

[0027] A first decision tree is constructed as the core judgment architecture of the dynamic decision-maker. The first decision tree contains multiple cascaded judgment layers. Each layer performs a binary classification decision judgment on specific hot forming state elements. The judgment results are passed to the next judgment layer in sequence to achieve layer-by-layer filtering and refined judgment. Among them, the cascaded judgment layers include thermal uniformity layer, embrittlement temperature threshold layer, stress path layer and ferrite phase distribution layer.

[0028] After completing the design of the first decision tree architecture, the system uses it as the judgment architecture and the regulation-oriented model as the generation architecture, forming a combined decision framework that integrates judgment and generation capabilities. Specifically, it retrieves monitoring data from historical periods, including recorded temperature, stress, phase evolution data, and corresponding regulation feedback results. The monitoring data is structured according to the characteristics of thermoforming state and the regulation execution results, and input as sample data into the combined architecture. During training, an adversarial convergence training mechanism is introduced, that is, adversarial information constraints are introduced between the judgment architecture and the generation architecture. On the one hand, the judgment architecture needs to perform accurate hierarchical classification output on the input thermoforming data; on the other hand, the generation architecture generates regulation guidance based on these judgment results and attempts to simulate its response trend in the coupled model. Through continuous backpropagation optimization, the two parts converge in the adversarial process, ultimately determining a dynamic decision-maker that combines generalization ability and regulation adaptability. After training, the dynamic decision-maker has the ability to perform structured identification of real-time monitoring data, anomaly feature localization, regulation strategy generation, and execution effect prediction.

[0029] Furthermore, using the first decision tree as the judgment framework, it includes: A first classification layer is constructed based on thermal uniformity, a second classification layer is constructed based on embrittlement temperature threshold, a third classification layer is constructed based on stress path, and a fourth classification layer is constructed based on ferrite phase distribution. The classification layers are cascaded to determine the first decision tree, wherein each classification layer performs a binary classification judgment. A post-judgment layer is introduced, wherein the post-judgment layer performs a judgment on whether to perform industrial control adjustment under the comprehensive output of the first decision tree. The judgment architecture is built based on the first decision tree and the post-judgment layer.

[0030] The first classification layer extracts indicators such as spatial temperature difference and heat flux density fluctuation from the temperature distribution in the thermal field point cloud data to determine whether there are abnormally concentrated or overheated areas in the thermal field. By setting temperature difference thresholds and fluctuation ranges, a binary classification judgment is achieved between "thermally uniform" and "thermally non-uniform" states.

[0031] The second classification layer determines whether there is a region or time period within the material's known embrittlement-sensitive range based on the temperature time trajectory during the thermoforming process. If the temperature of the detected area remains within the embrittlement range for an extended period, it is classified as "risk of embrittlement exists"; otherwise, it is classified as "no risk of embrittlement."

[0032] The third classification layer analyzes the continuity, rate of change, and concentration trend of the stress loading path based on the time-series trajectory of the stress vector extracted from the force field point cloud. If there are stress abrupt changes, path discontinuities, or excessively large gradients, it is judged as "stress path anomaly"; otherwise, it is "stress path stable".

[0033] The fourth classification layer, combined with the microstructure evolution module in the digital model, extracts the ferrite phase ratio and distribution uniformity of the current or predicted region. If the phase state deviates from the material processing window or there is significant unevenness in the microstructure distribution, it is judged as "abnormal phase distribution"; otherwise, it is "stable phase distribution".

[0034] Each classification layer outputs a "normal / abnormal" label through binary classification, and the output of the upper layer is used as the input condition for the lower layer in a cascading manner, thereby realizing multi-factor comprehensive identification of the thermoforming state. The first, second, third, and fourth classification layers together constitute the main body of the first decision tree.

[0035] Building upon the aforementioned judgment process, a post-judgment layer is further introduced. This post-judgment layer performs strategy judgment on the comprehensive output of the first decision tree, determining whether to trigger subsequent industrial control adjustment processes. The post-judgment layer introduces weighting factors to weight and synthesize the four-layer classification results, assessing whether the system has reached the control threshold (e.g., multiple anomalies occurring simultaneously, or a critical element deviating significantly). If the judgment is "trigger adjustment," an adjustment request is output; if "no adjustment is needed," the current adjustment process is terminated.

[0036] By integrating the first decision tree with the subsequent judgment layer to build a complete judgment architecture, the hierarchical discrimination and control requirements assessment of multi-dimensional states during the thermoforming process are realized.

[0037] Furthermore, triggering the dynamic decision-maker to perform hierarchical judgments and adjustment guidance generation includes: Using the first decision tree, a cascaded judgment based on layer feature recognition is performed on the thermoforming data to determine the cascaded judgment result; the subsequent judgment layer performs a comprehensive judgment on the cascaded judgment result, and if the generation architecture is not triggered, the detection process is terminated.

[0038] After determining the hot forming data, a dynamic decision-maker is triggered to initiate the identification and control process for the current hot forming state. Specifically, the first decision tree is invoked, using the hot forming data as input, to extract key physical features layer by layer, including thermal uniformity, the time period corresponding to the embrittlement temperature range, the stress path change trend, and the ferrite phase distribution structure, which are then classified into four classification layers in the decision tree. Each classification layer performs a binary classification judgment based on the specific features, outputting a "normal / abnormal" label, which is then passed to the next layer as an auxiliary decision condition, ultimately forming a complete cascaded judgment result at the end of the decision tree.

[0039] The system inputs the cascaded judgment result to the subsequent judgment layer. The subsequent judgment layer performs a comprehensive evaluation of the multi-layer judgment results, using preset feature weights, combination patterns, or rule expressions to determine whether the current thermoforming state triggers the control requirement. The evaluation method may include: if multiple judgment layers output abnormal states or the abnormal value of a certain high-weight judgment item (such as embrittlement temperature) significantly exceeds the threshold, then "Trigger generation architecture" is output; otherwise, "Generation architecture not triggered" is output.

[0040] If the generation architecture is not triggered, it means that the current thermoforming state of the system is within an acceptable range and there is no need for adjustment. In this case, the current detection process will terminate and the system will enter the sampling and monitoring cycle of the next time node to avoid redundant calculations and misoperations.

[0041] Furthermore, if the generation architecture is triggered, a target adjustment guide is generated, wherein the target adjustment guide is the phase control requirement of stress and ferrite phase state; according to the target adjustment guide, the field initialization of stress-ferrite phase state is performed on the digital coupling model, and the temperature is digitally dynamically simulated according to the time-series phase transition relationship to determine the third simulation data.

[0042] If the generation architecture is determined to be triggered, a target adjustment guide is generated. The target adjustment guide is the optimization target of the stress state and ferrite phase distribution in the current duplex stainless steel hot forming process, including but not limited to the target stress distribution path, stress gradient control range, ferrite phase ratio window value, and phase distribution uniformity requirements.

[0043] Based on the target-oriented regulation, a digital coupling model is invoked to initialize the stress-ferrite phase field. Specifically, according to the stress regulation requirements and the target ferrite phase distribution in the target-oriented regulation, the stress field and phase transformation modules in the digital coupling model are configured with field constraints, and the initial stress distribution matrix and phase distribution function boundary conditions are reset to ensure that the model has a solution basis that reflects the target regulation direction. Based on the above initialized field, the system uses the temperature vector as the main driving variable, based on the embedded time-series phase transformation relationship, to simulate the stress state evolution and phase evolution process in conjunction with the simulation. Coupled numerical calculations are performed at multiple time nodes to obtain the simulation prediction results, i.e., the third simulation data. This data reflects the dynamic evolution trend of the stress-phase-temperature three fields that the material may reach under the target-oriented regulation control.

[0044] Furthermore, the parameter control conversion is performed to determine the thermoforming control data, including: Based on the processing equipment group for duplex stainless steel pipe fittings, a parameter control matrix is ​​determined; based on the parameter control matrix, the target adjustment guide and the third simulation data are converted into parameter control data to determine the thermoforming control data.

[0045] Based on the current processing equipment configuration for duplex stainless steel pipe fittings, the system identifies the control equipment groups involved in the hot forming process, such as induction heaters, rolling devices, temperature control devices, and mold cooling units. Based on the relationship between equipment process parameters and control variable responses, a parameter control matrix is ​​constructed to establish the mapping function between equipment control commands and target physical quantities (temperature, stress, phase state). The parameter control matrix is ​​a multi-dimensional mapping matrix, where element values ​​represent the degree of influence or sensitivity coefficient of different equipment control quantities (such as heating power, temperature control range, cooling rate, strain rate, etc.) on the thermal field, stress field, and phase transformation field.

[0046] Using the target adjustment guidance and third-party simulation data as inputs to the parameter control transformation, the parameter control matrix is ​​invoked to calculate the operational quantities that should be applied to each type of control equipment. By solving the optimal matching function, it is ensured that the control parameters can best approximate the physical response required by the simulation. The output results are the thermoforming control data, including the control target values ​​of specific equipment, such as the target heating power curve, temperature control time schedule, roll load adjustment scheme, and mold cooling rate.

[0047] The thermoforming process of duplex stainless steel pipe fittings is adjusted based on the thermoforming control data.

[0048] The hot forming control data is synchronously sent to each processing equipment module, driving them to perform corresponding temperature adjustment, stress control, and cooling management operations. During the control process, the latest thermal, mechanical, and phase state data are obtained in real time by calling the front-end monitoring array, and feedback closed-loop adjustment is performed by comparing the target adjustment guidance with the third simulation data. If there is a deviation between the monitoring feedback and the simulation response, the control parameters are dynamically corrected and the control process is iteratively optimized, thereby achieving dynamic and precise control of the duplex stainless steel hot forming process and improving the forming quality and phase stability.

[0049] Furthermore, the feedback adjustment of the hot forming process of duplex stainless steel pipe fittings is performed based on the aforementioned hot forming control data, including: Based on the thermoforming control data, the thermoforming process of duplex stainless steel pipe fittings is dynamically adjusted; the monitoring array is simultaneously triggered to perform feedback adjustment monitoring with the target adjustment guide and the third simulation data as the response target.

[0050] Specifically, based on hot forming control data, dynamic adjustment is performed on the hot forming process of duplex stainless steel pipe fittings. This includes: adjusting the heating power distribution and temperature zone of the heating system; adjusting the cooling flow rate and time window of the controlled cooling system; and linking the process parameters such as the strain rate and mold contact pressure of the forming system to ensure that the hot processing process continuously approaches the expected targets for thermal uniformity, stress path, and ferrite phase state, avoiding problems such as local overheating, microstructure embrittlement, or forming cracks. Secondly, the monitoring array is simultaneously triggered. Utilizing a distributed sensor network deployed at the front end of the hot forming equipment, based on the target adjustment guidance and the third simulation data as monitoring reference targets, feedback adjustment monitoring of the current hot processing state is performed. This includes collecting thermal field point clouds, force field point clouds, and sound field point clouds, and extracting temperature vectors, stress vectors, and ferrite phase evolution characteristics. The collected data is input into a digital coupling model, combined with a preset temporal phase transition relationship, to determine the degree of deviation between the current forming state and the control target. If necessary, the dynamic decision-maker is re-triggered to generate the next round of adjustment path, thereby constructing an adaptive closed-loop control mechanism.

[0051] In summary, the embodiments of this application have at least the following technical effects: First, the geometry of the duplex stainless steel pipe fitting is determined, and a digital coupling model of thermo-mechanical-phase transformation is constructed. Next, based on a front-end distributed monitoring array, the duplex stainless steel pipe fitting is inspected and its field reconstructed using the digital coupling model to determine hot forming data, which includes first inspection data and second simulation data. Then, based on the hot forming data, a dynamic decision-maker is triggered to perform hierarchical judgment and regulation guidance generation, executing regulation simulation and parameter control conversion based on the digital coupling model to determine hot forming control data. The hierarchical judgment elements include at least thermal uniformity, embrittlement temperature threshold, stress path, and ferrite phase. Finally, feedback regulation of the hot forming process of the duplex stainless steel pipe fitting is performed based on the hot forming control data. This solves the technical problem of unstable forming quality caused by the lack of a dynamic control mechanism in the hot forming process of duplex stainless steel pipe fittings in existing technologies, achieving the technical effect of improving the hot forming quality of duplex stainless steel pipe fittings.

[0052] Example 2 is based on the same inventive concept as the dynamic monitoring method for duplex stainless steel hot forming based on a digital model in the previous examples, such as... Figure 2 As shown, this application provides a dynamic monitoring system for the hot forming of duplex stainless steel based on a digital model, wherein the system includes: Model building module 11: Determines the geometric shape of the duplex stainless steel pipe fitting and constructs a digital coupling model of thermo-mechanical-phase transformation; Detection module 12: Based on the front-end distributed monitoring array, detects the duplex stainless steel pipe fitting and reconstructs the field based on the digital coupling model to determine the hot forming data, wherein the hot forming data includes first detection data and second simulation data; Decision module 13: For the hot forming data, triggers a dynamic decision-maker to perform hierarchical judgment and adjustment guidance generation, executes adjustment simulation and parameter control conversion based on the digital coupling model, and determines the hot forming control data, wherein the hierarchical judgment elements include at least thermal uniformity, embrittlement temperature threshold, stress path and ferrite phase; Feedback adjustment module 14: Performs feedback adjustment of the hot forming process of the duplex stainless steel pipe fitting according to the hot forming control data.

[0053] Furthermore, the decision module 13 is used to perform the following methods: Construct a first decision tree, which includes a cascaded judgment layer; using the first decision tree as the judgment architecture and the field adjustment guidance as the generation architecture, retrieve monitoring and recording data and perform data integration based on cascaded judgment and adjustment guidance, use the data as sample data to perform architecture adversarial convergence training, and determine the dynamic decision-maker.

[0054] Furthermore, the decision module 13 is used to perform the following methods: A first classification layer is constructed based on thermal uniformity, a second classification layer is constructed based on embrittlement temperature threshold, a third classification layer is constructed based on stress path, and a fourth classification layer is constructed based on ferrite phase distribution. The classification layers are cascaded to determine the first decision tree, wherein each classification layer performs a binary classification judgment. A post-judgment layer is introduced, wherein the post-judgment layer performs a judgment on whether to perform industrial control adjustment under the comprehensive output of the first decision tree. The judgment architecture is built based on the first decision tree and the post-judgment layer.

[0055] Furthermore, the detection module 12 is used to perform the following method: Through distributed monitoring, thermal field point clouds, force field point clouds, and acoustic field point clouds are determined; feature transformation and coupling field reconstruction based on temperature vector, stress vector, and ferrite phase state are performed on the thermal field point clouds, force field point clouds, and acoustic field point clouds, and the digital coupling model is initialized to determine the first detection data; wherein, with temperature as the independent variable and stress and ferrite phase state as the dependent variables, a time-series phase transformation relationship of duplex stainless steel is established, and the time-series phase transformation relationship is embedded in the digital coupling model.

[0056] Furthermore, the detection module 12 is used to perform the following method: First hot forming conditions are determined, wherein the first hot forming conditions are the hot forming conditions for collecting data at lower time nodes; based on the hot forming conditions, the digital coupling model is initialized with a thermal field, and the stress and ferrite phase state are digitally dynamically simulated using the time-series phase transformation relationship to determine second simulation data; the first detection data and the second simulation data are time-series cascaded as hot forming data.

[0057] Furthermore, the decision module 13 is used to perform the following methods: Using the first decision tree, a cascaded judgment based on layer feature recognition is performed on the thermoforming data to determine the cascaded judgment result; the subsequent judgment layer performs a comprehensive judgment on the cascaded judgment result, and if the generation architecture is not triggered, the detection process is terminated.

[0058] Furthermore, the decision module 13 is used to perform the following methods: If the generation architecture is triggered, a target adjustment guide is generated, wherein the target adjustment guide is the phase control requirement of stress and ferrite phase state; according to the target adjustment guide, the field initialization of stress-ferrite phase state is performed on the digital coupling model, and the temperature is digitally dynamically simulated according to the time-series phase transition relationship to determine the third simulation data.

[0059] Furthermore, the decision module 13 is used to perform the following methods: Based on the processing equipment group for duplex stainless steel pipe fittings, a parameter control matrix is ​​determined; based on the parameter control matrix, the target adjustment guide and the third simulation data are converted into parameter control data to determine the thermoforming control data.

[0060] Furthermore, the feedback adjustment module 14 is used to perform the following method: Based on the thermoforming control data, the thermoforming process of duplex stainless steel pipe fittings is dynamically adjusted; the monitoring array is simultaneously triggered to perform feedback adjustment monitoring with the target adjustment guide and the third simulation data as the response target.

[0061] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0062] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0063] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A method for dynamic monitoring of duplex stainless steel hot forming based on a digital model, characterized in that, The method includes: Determine the geometry of duplex stainless steel pipe fittings and construct a digital coupled model of thermo-mechanical-phase transformation; Based on the front-end distributed monitoring array, the duplex stainless steel pipe fittings are inspected and the field is reconstructed based on the digital coupling model to determine the thermoforming data, wherein the thermoforming data includes first inspection data and second simulation data. For the aforementioned hot forming data, a dynamic decision-maker is triggered to perform hierarchical judgment and adjustment guidance generation, execute adjustment simulation and parameter control conversion based on a digital coupling model, and determine the hot forming control data. The hierarchical judgment elements include at least thermal uniformity, embrittlement temperature threshold, stress path, and ferrite phase. The thermoforming process of duplex stainless steel pipe fittings is adjusted based on the thermoforming control data. The inspection of duplex stainless steel pipe fittings and the field reconstruction based on the aforementioned digital coupling model include: The thermal field point cloud, force field point cloud, and sound field point cloud are determined through distributed monitoring. The temperature vector distribution features of the thermal field point cloud are extracted, the force field point cloud is converted into a spatial stress component vector, and the sound field point cloud is mapped into a ferrite phase transformation activity intensity index through spectrum and time domain analysis, forming temperature vector, stress vector and ferrite phase characteristic data. The three types of characteristic data are input into a pre-constructed digital coupling model as the first initialization boundary and initial conditions. The digital coupling model is solved and simulated based on the constructed heat conduction equation, elastoplastic mechanics equation and embedded time-series phase transition relationship, and outputs the first detection data. The first detection data reflects the actual temperature field distribution, stress path evolution and microstructure characteristics in the current thermoforming process. Among them, temperature is used as the independent variable and stress and ferrite phase state are used as dependent variables to establish the time-series phase transformation relationship of duplex stainless steel, and the time-series phase transformation relationship is embedded in the digital coupling model. After determining the first detection data, the following is included: Determine the first thermoforming condition, wherein the first thermoforming condition is the operating condition information immediately following the first detection data acquisition time point; Using the first hot forming condition as the boundary input, and combining the embedded time-series phase transformation relationship model, the stress evolution path and dynamic change process of ferrite phase content of duplex stainless steel pipe fittings in the next time window are simulated and predicted, and a set of multiphysics field evolution results for the predicted time period are output as the second simulation data. The first detection data and the second simulation data are cascaded in time to form thermoforming data.

2. The method for dynamic monitoring of duplex stainless steel hot forming based on a digital model as described in claim 1, characterized in that, Before triggering the dynamic decision-maker, the construction of the dynamic decision-maker includes: Construct a first decision tree, wherein the first decision tree consists of multiple cascaded decision layers; Using the first decision tree as the judgment architecture and the field adjustment orientation as the generation architecture, the monitoring and recording data are retrieved and integrated based on cascaded judgment and adjustment orientation. This data is then used as sample data to perform adversarial convergence training to determine the dynamic decision-maker.

3. The method for dynamic monitoring of duplex stainless steel hot forming based on a digital model as described in claim 2, characterized in that, Using the first decision tree as the judgment framework, it includes: The first classification layer is constructed based on thermal uniformity, the second classification layer is constructed based on embrittlement temperature threshold, the third classification layer is constructed based on stress path, and the fourth classification layer is constructed based on ferrite phase distribution. The classification layers are cascaded to determine the first decision tree, wherein each classification layer performs binary classification judgment. A post-decision layer is introduced, wherein the post-decision layer performs the determination of whether to perform industrial control adjustment under the comprehensive output of the first decision tree; The judgment architecture is constructed using the first decision tree and the subsequent judgment layer.

4. The method for dynamic monitoring of duplex stainless steel hot forming based on a digital model as described in claim 1, characterized in that, Trigger the dynamic decision-making unit to perform hierarchical judgment and adjustment guidance generation, including: Using the first decision tree, a cascaded judgment based on layer feature recognition is performed on the thermoforming data to determine the cascaded judgment result; The detection process is terminated if the generation architecture is not triggered by the comprehensive judgment result of the post-judgment layer.

5. The method for dynamic monitoring of duplex stainless steel hot forming based on a digital model as described in claim 4, characterized in that, If the generation architecture is triggered, a target adjustment guide is generated, wherein the target adjustment guide is the phase control requirement of stress and ferrite phase state; Based on the target adjustment guidance, the field initialization of the stress-ferrite phase state of the digital coupling model is performed, and the temperature is digitally dynamically simulated according to the time-series phase transition relationship to determine the third simulation data.

6. The method for dynamic monitoring of duplex stainless steel hot forming based on a digital model as described in claim 5, characterized in that, Perform parameter conversion to determine thermoforming control data, including: Determine the parameter control matrix based on the processing equipment group for duplex stainless steel pipe fittings; Based on the parameter control matrix, the target adjustment guidance and the third simulation data are converted into parameter control data to determine the thermoforming control data.

7. The method for dynamic monitoring of duplex stainless steel hot forming based on a digital model as described in claim 6, characterized in that, Based on the aforementioned hot forming control data, feedback adjustment of the hot forming process of duplex stainless steel pipe fittings is performed, including: Based on the aforementioned thermoforming control data, the thermoforming process of duplex stainless steel pipe fittings is dynamically adjusted. The monitoring array is synchronously triggered to perform feedback adjustment monitoring with the target adjustment guidance and the third simulation data as the response target.

8. A dynamic monitoring system for duplex stainless steel hot forming based on a digital model, characterized in that, For implementing the dynamic monitoring method for hot forming of duplex stainless steel based on a digital model as described in any one of claims 1-7, the system comprises: Model building module: Determines the geometry of duplex stainless steel pipe fittings and constructs a digital coupled model of thermo-mechanical-phase transformation; Detection module: Based on the front-end distributed monitoring array, the duplex stainless steel pipe fittings are detected and the field is reconstructed based on the digital coupling model to determine the thermoforming data, wherein the thermoforming data includes first detection data and second simulation data; Decision module: For the hot forming data, trigger the dynamic decision-maker to perform hierarchical judgment and adjustment guidance generation, execute adjustment simulation and parameter control conversion based on digital coupling model, and determine the hot forming control data. The hierarchical judgment elements include at least thermal uniformity, embrittlement temperature threshold, stress path and ferrite phase. Feedback adjustment module: Performs feedback adjustment of the hot forming process of duplex stainless steel pipe fittings based on the hot forming control data; The detection module is used to perform the following method: The thermal field point cloud, force field point cloud, and sound field point cloud are determined through distributed monitoring. The temperature vector distribution features of the thermal field point cloud are extracted, the force field point cloud is converted into a spatial stress component vector, and the sound field point cloud is mapped into a ferrite phase transformation activity intensity index through spectrum and time domain analysis, forming temperature vector, stress vector and ferrite phase characteristic data. The above three characteristic data are input into the pre-constructed digital coupling model as the first initialization boundary and initial conditions. The digital coupling model is solved and simulated based on the constructed heat conduction equation, elastoplastic mechanics equation and embedded time-series phase transition relationship, and outputs the first detection data. The first detection data reflects the actual temperature field distribution, stress path evolution and microstructure characteristics in the current thermoforming process. Among them, temperature is used as the independent variable and stress and ferrite phase state are used as dependent variables to establish the time-series phase transformation relationship of duplex stainless steel, and the time-series phase transformation relationship is embedded in the digital coupling model. After determining the first detection data, the following is included: Determine the first thermoforming condition, wherein the first thermoforming condition is the operating condition information immediately following the first detection data acquisition time point; Using the first hot forming condition as the boundary input, and combining the embedded time-series phase transformation relationship model, the stress evolution path and dynamic change process of ferrite phase content of duplex stainless steel pipe fittings in the next time window are simulated and predicted, and a set of multiphysics field evolution results for the predicted time period are output as the second simulation data. The first detection data and the second simulation data are cascaded in time to form thermoforming data.

Images