Coupling dynamics fault twin modeling method and system of large screw pump

Abstract

The application relates to the field of electromechanical equipment maintenance, in particular to a coupling dynamics fault twin modeling method and system for a large screw pump, which comprises the following steps: dividing a large screw pump along an axial direction to obtain rotor segments, and establishing a fixed space coordinate system for the rotor segments; under the fixed space coordinate system, extracting force components of the rotor segments through a fluid simulation test, and constructing a nonlinear prediction model of the force components; performing multi-body dynamics modeling on the large screw pump to obtain a substructure dynamics model; integrating modeling on the substructure dynamics model to obtain a flow-heat-solid coupling dynamics model of the large screw pump, and performing fault simulation to obtain fault state data. The application comprehensively utilizes computational fluid dynamics modeling, data-driven modeling and multi-body dynamics modeling, realizes flow-solid-heat multi-physical field coupling modeling, and forms a systematic dynamics modeling method for the large screw pump.

Description

Technical Field

[0001] This invention relates to the field of electromechanical equipment maintenance, and in particular to a coupled dynamics fault twin modeling method and system for large screw pumps. Background Technology

[0002] Large screw pumps, with their core advantages of stable flow rate, high pressure, and adaptability to medium- and high-viscosity media, are widely used in process industries, such as petrochemicals, where they serve as core equipment for long-distance, high-pressure transportation of medium- and high-viscosity media like crude oil, heavy oil, lubricating oil, and asphalt. In the energy and power sector, they are used in the lubrication systems of large-scale thermal and hydropower units, providing stable and high-pressure lubrication for key equipment such as turbines and boiler feed pumps, ensuring safe operation of the units. In the lyocell fiber production sector, large screw pumps play a vital role in slurry transportation, polymer solution transportation, and fiber recycling and reuse. Due to their high efficiency, energy saving, ease of maintenance, and strong adaptability, large screw pumps are widely used in the fiber production industry.

[0003] In process industries, the operational stability of large screw pumps directly impacts the safety and efficiency of continuous production lines. Failures can lead to unplanned downtime, causing production line paralysis, raw material waste, product quality degradation, and even safety accidents. Large screw pumps operate in harsh environments with high temperatures, high pressures, and strong corrosion, making them prone to wear, corrosion, seal failure, and excessive vibration. Traditional periodic maintenance struggles to accurately detect early signs of these malfunctions. Intelligent operation and maintenance (O&M) systems utilize IoT sensors to collect multi-dimensional data such as vibration and temperature in real time. Combined with AI algorithms (such as machine learning and deep learning), these systems build equipment health models, predict remaining lifespan, and provide early warnings of potential failures, thus preventing unplanned downtime. Currently, one of the biggest challenges in developing intelligent O&M systems is obtaining full lifecycle data for the equipment. Utilizing digital twin technology to construct fault twin models of large screw pumps, generating massive amounts of health status data for intelligent O&M systems, is an effective method for rapidly deploying such systems.

[0004] Typical failures of large screw pumps include motor failure, gearbox failure, and pump body failure, with stator wear being the most common pump body failure. The stator of a large screw pump is usually made of rubber and is a consumable component. The ability to predict stator wear trends will greatly contribute to the development of intelligent operation and maintenance technologies for this equipment.

[0005] In the structure of a large screw pump, the sealed conveying cavity formed between the rotor and stator is the key area for material conveying. Its internal flow is complex, and the fluid pressure and forces exhibit a significant non-uniform distribution along the rotor axis. Especially when handling non-Newtonian fluids, the viscosity of the medium changes significantly with shear rate and temperature, making the axial and radial forces and their dynamic changes during rotor operation even more complex.

[0006] Existing dynamic analysis methods for screw pumps mostly employ empirical formulas, simplified mechanical models, or rigid body assumptions, and treat the fluid forces acting on the rotor as an overall equivalent, making it difficult to reflect the true force distribution along the axial direction of the helical structure. At the same time, some methods do not fully consider the rheological characteristics of non-Newtonian fluids and the coupling effect of temperature fields during fluid analysis, resulting in significant deviations between the obtained dynamic models and actual operating conditions.

[0007] To improve analytical accuracy, some studies have attempted to use three-dimensional computational fluid dynamics or fluid-structure interaction simulation to perform detailed modeling of the internal flow and rotor forces in screw pumps. However, these methods typically require high-quality mesh generation for complex helical geometries and iterative calculations under multiple operating conditions, speeds, and physical property parameters. This results in long simulation cycles and high computational resource consumption, making it difficult to meet the massive data demands of intelligent operation and maintenance systems. Furthermore, the high time and computational costs limit the engineering application of these simulation methods in the structural optimization, dynamic modeling, and operational status prediction of large screw pumps.

[0008] Currently, there are no publicly available patents for fluid-structure-thermal coupling dynamic modeling of screw pumps. Existing technologies include simplified modeling of vibration isolators at the structural dynamics level, but these do not consider the influence of complex internal flow states and temperature changes on rotor stress and dynamic response, making it difficult to reflect the non-uniform axial stress and multi-physics coupling characteristics of the rotor under non-Newtonian fluid conditions. Summary of the Invention

[0009] To address the shortcomings of existing technologies, this invention provides a coupled dynamics fault twin modeling method and system for large screw pumps.

[0010] To achieve the above objectives, in a first aspect, this invention provides a coupled dynamics fault twin modeling method for a large screw pump. The method includes the following steps: dividing the large screw pump along its axial direction to obtain rotor segments, and establishing a fixed spatial coordinate system for each rotor segment; extracting the force components of each rotor segment through fluid simulation experiments within the fixed spatial coordinate system, and constructing a nonlinear prediction model for the force components; performing multibody dynamics modeling on the large screw pump based on the nonlinear prediction model to obtain a substructure dynamic model; integrating the substructure dynamic model to obtain a fluid-thermal-structure coupled dynamic model of the large screw pump, and performing fault simulation to obtain fault state data. This invention achieves fluid-thermal-structure coupled modeling and fault simulation, significantly reducing the computational power requirements of the overall model, shortening computation time, accurately capturing the true force characteristics of the rotor, and efficiently acquiring fault data, providing reliable support for the development of intelligent operation and maintenance systems and equipment design.

[0011] Optionally, the step of extracting the force components of the rotor segment through fluid simulation experiments in the fixed spatial coordinate system includes: obtaining orthogonal test tables by setting different design factors, and constructing orthogonal experiments as the fluid simulation experiments using the central composite design method; introducing a non-Newtonian fluid dynamics model into the fluid simulation experiments, and establishing a coupled model of the flow field and temperature field by combining the energy equation; conducting the fluid simulation experiments on the large screw pump to obtain simulation results, and extracting the force components under different test conditions based on the simulation results. This invention accurately adapts to the operating conditions of screw pumps delivering non-Newtonian fluids, realistically reflects the influence of temperature on rheological properties, and the extracted force component data closely matches actual operating conditions, laying a high-precision data foundation for subsequent modeling.

[0012] Optionally, constructing the nonlinear prediction model for the force components includes: normalizing the design factors and the force components to obtain a standardized dataset; obtaining the force equations for the rotor segments based on the standardized dataset in the fixed spatial coordinate system using a generalized regression neural network structure; and using the force equations as the nonlinear prediction model. This invention accurately fits the strong nonlinear relationship between operating parameters and rotor forces, achieving high model fit and small error, and can quickly output rotor force results, avoiding complex simulation calculations and improving modeling efficiency.

[0013] Optionally, the step of obtaining a substructure dynamic model by performing multibody dynamics modeling of the large screw pump based on the nonlinear prediction model includes: connecting the rotor segments according to a massless beam model to form a rotor dynamic model, and constructing a rotor-stator dynamic model based on a geometric contact force model; obtaining a bearing dynamic model by modeling the bearings of the large screw pump using the nonlinear prediction model; obtaining a gear dynamic model by modeling the gears of the large screw pump using the nonlinear prediction model; obtaining a universal joint drive shaft dynamic model by modeling the universal joint drive shaft of the large screw pump based on the nonlinear prediction model; and using the rotor dynamic model, the rotor-stator dynamic model, the bearing dynamic model, the gear dynamic model, and the universal joint drive shaft dynamic model as the substructure dynamic model. This invention obtains the dynamic characteristics of multi-component coupling in a large screw pump with high modeling accuracy, providing a precise substructure foundation for integrated modeling and ensuring the reliability of the overall coupled model.

[0014] Optionally, the step of obtaining a bearing dynamic model by modeling the bearing of the large screw pump using the nonlinear prediction model includes: considering the bearing of the large screw pump as a first multibody dynamic system composed of an inner ring, an outer ring, and several rolling elements; obtaining the relative radial displacement and relative tangential displacement of the inner ring and the outer ring, and determining the contact deformation by combining the bearing radial clearance; when the contact deformation is greater than zero, obtaining the normal contact force between the rolling elements and the raceway; obtaining the tangential contact force based on the relative sliding velocity of the contact point and the normal contact force; and synthesizing the normal contact force and the tangential contact force to obtain the bearing dynamic model. This invention realistically reproduces the nonlinear dynamic response of the bearing, accurately characterizes its stress characteristics under different operating conditions, and achieves refined dynamic modeling of the bearing.

[0015] Optionally, the step of using the nonlinear prediction model to model the gears of the large screw pump to obtain a gear dynamics model includes: treating the transmission gear pair of the large screw pump as a second multibody dynamics system composed of a driving gear and a driven gear; for any gear pair participating in meshing at any given time, obtaining the meshing force of the gear pair along the meshing line direction, combined with the meshing deformation; obtaining the gear tangential force based on the input torque and the gear pitch circle radius; and obtaining the gear dynamics model by combining the meshing force and the gear tangential force. This invention fully restores the nonlinear contact characteristics and power transmission law of gear meshing, accurately characterizes the dynamic response of the gear pair under load and speed coupling, and improves the accuracy of gear modeling.

[0016] Optionally, the meshing deformation includes: obtaining the relative displacement of the driving gear and the driven gear in the direction of the meshing line; and obtaining the meshing deformation based on the relative displacement, combined with the tooth backlash and transmission error of the transmission gear pair. This invention accurately reflects the true deformation state during meshing, provides accurate parameters for meshing force calculation, and effectively improves the adaptability of the gear dynamics model to the actual meshing state.

[0017] Optionally, the step of modeling the universal joint drive shaft of the large screw pump based on the nonlinear prediction model to obtain the universal joint drive shaft dynamic model includes: considering the universal joint drive shaft of the large screw pump as a third multibody dynamic system composed of a main shaft, a connecting rod shaft, and a screw rotor; the universal joint drive shaft includes a driving shaft and a driven shaft, and obtaining the kinematic relationship between the driving shaft and the driven shaft; and establishing the universal joint drive shaft dynamic model using the kinematic relationship as the angular coupling relationship. This invention fully captures the characteristics in torque transmission, improves the fit of the drive shaft dynamic model, and provides an accurate dynamic foundation for the transmission system of the whole machine coupling model.

[0018] Optionally, the step of integrating and modeling the substructure dynamics model to obtain the fluid-thermal-structure coupled dynamics model of the large screw pump, and performing fault simulation to obtain fault state data, includes: integrating and modeling the substructure dynamics model using kinematic pairs to form the fluid-thermal-structure coupled dynamics model; obtaining the fluid-thermal-structure coupled dynamics model simulating the fault state by changing the three-dimensional model of the substructure of the large screw pump, and performing simulation to obtain the fault state data. This invention simulates various faults, and the fault data obtained through simulation is rich and accurate, providing high-quality data support for equipment fault early warning, vibration analysis, and intelligent operation and maintenance research and development.

[0019] Secondly, this invention provides a coupled dynamics fault twin modeling system for large screw pumps. The system executes the coupled dynamics fault twin modeling method for large screw pumps provided by this invention. The system includes an input device, an output device, a processor, and a memory, which are interconnected. The memory stores a computer program, which includes program instructions, and the processor is configured to call the program instructions. This invention, through high-performance hardware collaboration, quickly adapts to the modeling needs of different operating conditions, providing efficient system support for intelligent operation and maintenance of screw pumps. Attached Figure Description

[0020] Figure 1 This is a flowchart of a coupled dynamics fault twin modeling method for a large screw pump according to an embodiment of the present invention; Figure 2 This is a schematic diagram of a large screw pump according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the rotor segmentation of a large screw pump according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the forces acting on a rotor segment according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the generalized regression neural network structure according to an embodiment of the present invention; Figure 6 This is a schematic diagram of the universal drive shaft modeling in an embodiment of the present invention; Figure 7 This is a framework diagram of a coupled dynamics fault twin modeling system for a large screw pump according to an embodiment of the present invention; Attachment markings: Motor 21, Reducer 22, Coupling 23, Shaft sealing system 24, Housing 25, Universal drive shaft 26, Stator 27, Rotor 28, Main shaft 61, Connecting rod shaft 62, Screw rotor 63. Detailed Implementation

[0021] Specific embodiments of the present invention will now be described in detail. It should be noted that the embodiments described herein are for illustrative purposes only and are not intended to limit the invention. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the invention. However, it will be apparent to those skilled in the art that these specific details are not necessary to practice the invention. In other instances, well-known circuits, software, or methods have not been specifically described to avoid obscuring the invention.

[0022] Throughout this specification, references to "an embodiment," "an embodiment," "an example," or "an example" mean that a particular feature, structure, or characteristic described in connection with that embodiment or example is included in at least one embodiment of the invention. Therefore, the phrases "in an embodiment," "in an embodiment," "an example," or "an example" appearing in various places throughout the specification do not necessarily refer to the same embodiment or example. Furthermore, specific features, structures, or characteristics can be combined in one or more embodiments or examples in any suitable combination and / or sub-combination. Moreover, those skilled in the art will understand that the illustrations provided herein are for illustrative purposes and are not necessarily drawn to scale.

[0023] Please see Figure 1 One embodiment of the present invention provides a twin modeling method for coupled dynamic faults of a large screw pump, the method comprising the following steps: S1. Divide the large screw pump along the axial direction to obtain rotor segments, and establish a fixed spatial coordinate system for the rotor segments.

[0024] Please see Figure 2 The diagram shows a schematic of a large screw pump. The large screw pump mainly includes a motor 21, a reducer 22, a coupling 23, a shaft sealing system 24, a housing 25, a universal drive shaft 26, a stator 27, and a rotor 28.

[0025] During the operation of the screw pump, the surface of the rotor 28 is subjected to multi-source hydrodynamic loads generated by the internal fluid. These loads mainly include the normal pressure of the high-pressure fluid acting on the helical surface and the tangential shear force generated by the flow of high-viscosity media. From a spatial distribution perspective, since the pressure in the sealed cavity inside the pump increases gradually from the inlet to the outlet along the axial direction, the fluid load borne by the rotor 28 exhibits a significant gradient distribution characteristic in the axial direction. From a geometric perspective, the helical groove structure causes the point of application of the resultant force of the fluid action to change periodically with the rotation angle of the rotor 28, thereby forming a radial unbalanced force and a corresponding bending moment component on the rotor 28 that varies with the rotation angle.

[0026] The aforementioned stress states exhibit non-uniform distribution characteristics in both the axial and circumferential directions. If traditional overall equivalent load or center-point concentrated load modeling methods are used, it will be difficult to simultaneously describe the axial load distribution and its phase characteristics as a function of the helix angle, thus limiting the accurate characterization of the local bending deformation and axial deformation behavior of the rotor 28. Especially under fluid-solid-thermal coupling conditions, the local differences in fluid viscosity caused by temperature variations further affect the pressure load distribution, making it difficult for the simplified equivalent load model to reflect the potential gap changes and contact risks between the rotor 28 and the stator 27.

[0027] Please see Figure 3 The diagram shows a segmented view of a large screw pump rotor; from left to right, these are the first, second, third, fourth, and fifth segments.

[0028] In this embodiment, a segmented modeling method based on axial discretization is adopted. By dividing the rotor 28 along the axial direction into multiple local units (preferably divided into 8-12 segments per lead) as rotor segments, the fluid load continuously distributed on the spiral surface is discretized into a set of force vectors with clear spatial positional relationships.

[0029] Please see Figure 4 The diagram shows a force diagram of a segmented rotor; including... , and The force component in the direction.

[0030] Specifically, a fixed spatial coordinate system is established for each rotor segment to facilitate the extraction of each rotor segment in the fixed spatial coordinate system. , and The force component in the direction.

[0031] It should be noted that the above modeling method can retain the phase information of the load as the rotor 28 rotates while maintaining the characteristics of the axial pressure gradient, thus providing a consistent distributed load input for the calculation of the swing response, vibration characteristics and nonlinear contact behavior in the subsequent dynamic analysis of the rotor 28.

[0032] S2. Under the fixed spatial coordinate system, extract the force components of the rotor segment through fluid simulation experiments, and construct a nonlinear prediction model of the force components.

[0033] Specifically, S2 includes the following steps: S21. Under the fixed spatial coordinate system, extract the force components of the rotor segment through fluid simulation test.

[0034] In this embodiment, in order to establish a mathematical model of rotor segments using a data-driven approach, an experimental design is first carried out to construct a fluid simulation experiment, and the force components are obtained as experimental data through the fluid simulation experiment.

[0035] Taking the fault twin modeling of a large screw pump for conveying lyocell fiber porridge as an example, the orthogonal experiment was designed using the Central Composite Design (CCD) method as a fluid simulation experiment for computational fluid dynamics (CFD) simulation.

[0036] Specifically, using the large screw pump's rotational speed, inlet pressure, medium temperature, average diameter of the stator 27 inner cavity, and rotor 28 phase as design factors, a 5-factor, 3-level orthogonal experimental table was designed, comprising 54 sets of tests. Fluid simulation experiments were conducted according to the orthogonal experimental table by setting different design factors. Then, the force components of the fluid action on each rotor segment in its three orthogonal directions were extracted from the simulation results. , and This allows us to obtain a dataset showing how the force components of the rotor segments change under test conditions.

[0037] In this embodiment, considering the common use of large screw pumps to transport non-Newtonian fluids with temperature-dependent characteristics, a non-Newtonian fluid dynamics model is introduced into the fluid simulation experiment, and a coupled model of the flow field and temperature field is established in combination with the energy equation.

[0038] The key to fluid modeling and simulation is to set up a reasonable non-Newtonian fluid dynamics model. Common non-Newtonian fluid dynamics models include the Ostwald-De Wale model, the Carreau-Yasuda model, and the Cross model.

[0039] The Ostwald-De Wale model is applicable to pseudoplastic or dilatant fluids at a wide range of shear deformation rates and is the most widely used non-Newtonian fluid model in engineering applications. This model has been widely used in the simulation of Lyocell fiber solutions. The equations of the Ostwald-De Wale model satisfy the following relationships: in, Shear viscosity, This is the fluid consistency index. Shear strain rate This is the power-law exponent.

[0040] It should be noted that when When the model reflects the characteristics of shear-thickened dilatant fluids (such as starch solutions, sucrose solutions, paint solutions, etc.); when When the model reflects the properties of shear-thinned pseudoplastic fluids (such as lyocell fiber solutions, tomato sauce, etc.); when At that time, the model reflects the characteristics of Newtonian fluids.

[0041] However, the Ostwald-De Wale model does not directly consider the effect of temperature on the fluid. Therefore, the Arrhenius equation is introduced to account for the effect of temperature on the rheological properties of the solution. The Arrhenius equation satisfies the following relationship: in, Shear viscosity, These are characteristic constants. It is a natural constant. For activation energy, The gas constant is... This refers to absolute temperature.

[0042] In fluid simulation experiments, a non-Newtonian fluid setting that takes temperature changes into account is achieved by simultaneously employing the Ostwald-De Wale model and the Arrhenius equation.

[0043] S22. Construct a nonlinear prediction model for the force components.

[0044] In this embodiment, considering the complexity of the internal flow field of the screw pump, the force components of the rotor segments ( , and There is a strong nonlinear coupling relationship between the model and design factors, and neural network models have a stronger mapping ability; a data-driven General Regression Neural Network (GRNN) is used to obtain... , and Nonlinear prediction model.

[0045] During the modeling process, the input variables (design factors) and output variables (force components of rotor segments) are subjected to Min-Max normalization to eliminate the influence of dimensions.

[0046] Specifically, the input and output data of different dimensions are normalized using the Min-Max Normalization method (Min-Max for short), thus normalizing the original data. Mapped to the range [0,1] It satisfies the following relationship: in, To standardize the dataset, For the original dataset, The minimum value in the original dataset. This represents the maximum value in the original dataset.

[0047] Please see Figure 5 The diagram shows a schematic of a generalized regression neural network structure, with the input layer containing 5 neurons. The model layer receives 31 sets of simulation data as input; the model layer has the same number of samples (54) and calculates the kernel function weights; the summation layer performs the summation of the denominator in parallel. ) and summation of the molecule ( Two calculations; the output layer has one neuron (for each calculation). , and (Separate modeling).

[0048] The core of GRNN is a weighted average based on a kernel function (usually a Gaussian kernel function). The Gaussian kernel function is used to calculate the similarity (kernel weights) between the point to be predicted and each training sample, satisfying the following relationship: in, For kernel function weights, Let be the input vector of the point to be predicted. It is an exponential function. For the first The input vector of each training sample, A device for representing vectors. This is a smoothing factor.

[0049] GRNN predictions It is the output of all training samples The weighted average satisfies the following relationship: in, for The corresponding predicted output, This represents the total number of samples in the training set. For sample index, For the first The actual output value of each training sample. These are the kernel function weights.

[0050] In this embodiment, leave-one-out cross-validation is used to determine the optimal smoothing factor and evaluate the model's generalization ability. The steps are as follows: Iterate N times. Each time, remove one sample as the test set (Ntest=1), and use the remaining N-1 samples as the training set. Predict the test sample using the trained model and record the error. Root Mean Square Error (RMSE) is used as the core indicator to measure the quality of the smoothing factor and evaluate the model's accuracy. The smoothing factor value that minimizes RMSE is the optimal parameter. RMSE satisfies the following relationship: in, The root mean square error, This represents the total number of cross-validations. For the index of the number of tests, For the first The true value of this test For the first The predicted value for this test.

[0051] Furthermore, the coefficient of determination is used. The goodness of fit of the model is evaluated to satisfy the following relationship: in, As the coefficient of determination, This represents the total number of cross-validations. For the index of the number of tests, For the first The true value of this test For the first The predicted value for this test, For all true values The average value.

[0052] In this embodiment, cross-validation is used to optimize the smoothing factor to balance the model's approximation accuracy and generalization ability. The final model is evaluated using mean squared error (MSE) and coefficient of determination (COP). The trained GRNN model Achieve a value of 0.99 or higher, with the relative error controlled within 5%.

[0053] Furthermore, through the above modeling, the force equations for the rotor segments in three directions under a fixed spatial coordinate system will be obtained, which respectively satisfy the following relationships: in, The force equations in the axial direction are as follows: For coefficients, For rotational speed, For the temperature of the medium, For inlet pressure, The average diameter of the inner cavity of stator 27. It is the rotor 28 phase.

[0054] It should be noted that the coefficient , , The results are determined by the least squares method from the GRNN output samples to characterize the combined effects of different operating parameters and their coupling terms on the radial and axial forces of rotor 28.

[0055] S3. Based on the nonlinear prediction model, perform multibody dynamics modeling on the large screw pump to obtain the substructure dynamics model.

[0056] In this embodiment, secondary development is performed using the Automatic Dynamic Analysis of Mechanical Systems (Adams) software to obtain the obtained data. , and The nonlinear prediction model is embedded within the software. It can be directly called during the dynamic modeling of large screw pumps. , and The nonlinear prediction model enables flow-thermal coupling force modeling of rotor segments.

[0057] Specifically, S3 includes the following steps: S31. Connect the rotor segments according to the massless beam model to form a rotor dynamics model, and construct a rotor-stator dynamics model based on the geometric contact force model.

[0058] In this embodiment, the rotor segments are connected by a massless beam model in the software, thus forming the entire rotor dynamics model. The rotor segments and stator 27 are connected by a nonlinear contact relationship established through a geometric contact force model in the software, thus forming the rotor-stator dynamics model.

[0059] S32. The bearing dynamics model is obtained by modeling the bearing of the large screw pump using the nonlinear prediction model.

[0060] In this embodiment, the bearing used is a cylindrical roller bearing. To address the problem that the main shaft bearing 61 is prone to vibration and nonlinear response under the coupled action of continuous load and speed during the operation of a large screw pump, the screw pump main shaft bearing 61 is regarded as a first multibody dynamic system composed of an inner ring, an outer ring, and several rolling elements. The dynamic characteristics of the main shaft bearing 61 are described by establishing a nonlinear contact force model between the rolling elements and the inner and outer raceways.

[0061] For the The normal contact force between a rolling element and the raceway is established based on Hertz nonlinear contact theory. When the contact deformation... When the force is greater than zero, the normal contact force satisfies the following relationship: in, For normal contact force, Here is the Hertz equivalent contact stiffness. For contact deformation, To access the nonlinear exponent, is the contact damping coefficient.

[0062] It should be noted that the contact stiffness coefficient is calculated based on the equivalent elastic modulus and equivalent radius of curvature of the rolling element and raceway materials; the contact damping coefficient is selected based on the equivalent mass and damping ratio to reflect the energy dissipation characteristics of the screw pump spindle bearing 61 during actual operation; when the ball and raceway are in point contact, the contact nonlinearity index... When the rollers and raceways are in contact, the contact nonlinearity index... .

[0063] Tangential contact force The main sources include Coulomb friction and viscous damping effects, which can be expressed as a function of the normal contact force and are related to the relative sliding velocity at the contact point. Their mathematical expression satisfies the following relationship: in, For tangential contact force, This is the equivalent coefficient of friction of the bearing contact surface. For normal contact force, Tangential damping coefficient, The relative sliding velocity of the contact point in the tangential direction.

[0064] It should be noted that the equivalent friction coefficient of the bearing contact surface is used to characterize the Coulomb friction characteristics under rolling / micro-slipping conditions; the tangential damping coefficient is used to describe the speed-related damping effect caused by the lubricating oil film and material internal friction.

[0065] In this embodiment, the amount of contact deformation The geometric relationship determined by the relative radial and tangential displacements of the inner and outer rings of the 61 bearing on the main shaft of the large screw pump, as well as the radial clearance of the bearing, satisfies the following relationship: in, For contact deformation, The relative radial displacement of the inner and outer rings of the main spindle 61 bearing. This represents the instantaneous angular position of the rolling element during the rotation of the main shaft 61. The relative tangential displacement of the inner and outer rings of the main spindle 61 bearing. This refers to the radial clearance of the bearing.

[0066] It should be noted that the instantaneous angular position of the rolling element during the rotation of the main shaft 61 changes with time and is determined by the angular velocity of the cage.

[0067] Furthermore, by synthesizing the contact forces of all rolling elements in the radial and tangential directions, the overall force on the screw pump spindle 61 bearing and its dynamic response over time can be obtained, thereby realizing a unified modeling of the nonlinear dynamic behavior of the spindle 61 bearing under different loads, speeds and clearances, and obtaining the bearing dynamic model.

[0068] Meanwhile, the contact force model of the bearing components is also embedded into the software through secondary software development and called during modeling.

[0069] S33. The gear dynamics model of the large screw pump is obtained by using the nonlinear prediction model.

[0070] In this embodiment, to address the problem of vibration and nonlinear dynamic response easily generated by the transmission gear pair under continuous load and speed coupling during the operation of a large screw pump, the transmission gear pair of the screw pump is regarded as a second multibody dynamic system composed of a driving gear and a driven gear. The dynamic transmission characteristics of the gear pair are described by establishing a nonlinear contact force model in the gear meshing process.

[0071] For any meshing gear pair at any given moment, the contact force along the meshing line is established based on nonlinear elastic contact theory and combined with meshing deformation. The meshing force satisfies the following relationship: in, For meshing force, This is the gear meshing stiffness coefficient. This is the amount of meshing deformation. is the meshing damping coefficient.

[0072] It should be noted that the gear meshing stiffness coefficient varies with the meshing position and is determined based on the gear material parameters, tooth geometry, and meshing state; the meshing damping coefficient is used to characterize the energy dissipation characteristics during gear meshing and is selected based on the equivalent mass and damping ratio.

[0073] Furthermore, gear tangential force It is mainly determined by the input torque and is related to the gear pitch circle radius and meshing state. Its expression satisfies the following relationship: in, For gear tangential force, The input torque acting on the gear shaft, The pitch circle diameter of the gear. Let be the pitch circle radius of the gear.

[0074] In this embodiment, the amount of meshing deformation Determined by the relative displacement of the driving gear and driven gear in the direction of the meshing line, and taking into account the effects of tooth backlash and transmission error of the gear pair, their geometric relationship satisfies the following relationship: in, This is the amount of meshing deformation. This represents the displacement of the driving gear in the direction of the meshing line. The displacement of the driven gear in the direction of the meshing line. This represents the overall transmission error of the gear pair. This refers to the tooth flank clearance.

[0075] Furthermore, a gear dynamics model is obtained by combining meshing force and gear tangential force.

[0076] It should be noted that the contact force model between gears is also embedded into the software through secondary software development and called during modeling.

[0077] S34. Based on the nonlinear prediction model, the universal drive shaft of the large screw pump is modeled to obtain the dynamic model of the universal drive shaft.

[0078] Please see Figure 6 The diagram shows a schematic model of a universal drive shaft; three core components, namely the main shaft 61, connecting rod shaft 62, and screw rotor 63, are arranged in series along the central axis: the main shaft 61 at the left end rotates at a speed of The input power has an angle with the axis of the connecting rod shaft 62. It is connected to the connecting rod shaft 62 in the middle through the coupling 23 to realize the transmission of torque at a variable angle. The connecting rod shaft 62 rotates at a speed of The rotor operates, and then connects to the right-side screw rotor 63 via another set of couplings 23, ultimately driving the screw rotor 63 to a speed of [missing information]. Output rotational motion.

[0079] In this embodiment, addressing the issue of periodic speed fluctuations and additional loads easily introduced into the torque transmission process of the universal joint drive shaft 26 under conditions of installation deviation and shaft misalignment in the transmission system during the operation of a large screw pump, the universal joint drive shaft 26 in the screw pump transmission system is considered as a third multibody dynamic system composed of the main shaft 61, connecting rod shaft 62, and screw rotor 63. A dynamic model considering angular coupling is established to describe the dynamic transmission characteristics of the universal joint drive shaft 26. When an angle exists in the universal joint drive shaft 26, such as... Figure 6 As shown, the angular velocity and angular acceleration between the driving shaft and the driven shaft exhibit a nonlinear relationship, satisfying the following relationship: in, The angular velocity of the driven shaft, For the angular velocity of the driving shaft, The angle between the axis of main spindle 61 and the axis of connecting rod 62. It is the instantaneous rotation angle of the active universal drive shaft 26 axis.

[0080] It should be noted that the instantaneous rotation angle of the active rotation around the axis of the universal joint drive shaft 26 is used to characterize the phase characteristics of the change of the angular velocity of the screw rotor 63 with time during the universal joint drive process.

[0081] In this embodiment, when the drive shaft has an angular velocity During rotation, the angular velocity of the driven shaft As the transmission angle changes periodically, its kinematic relationship can be determined by the geometric constraints of the universal joint. Based on this kinematic relationship, during torque transmission, the dynamic torque experienced by the universal drive shaft 26 is not only related to the average transmitted torque, but also to the speed fluctuation and the inertial characteristics of the shaft system. Thus, periodic excitation is introduced into the system to construct a dynamic model of the universal drive shaft.

[0082] In an optional embodiment, the dynamic model of the universal joint drive shaft can be directly constructed using existing software modules.

[0083] S4. Integrate the substructure dynamic model to obtain the fluid-thermal-solid coupled dynamic model of the large screw pump, and perform fault simulation to obtain fault state data.

[0084] In this embodiment, after completing the multibody dynamics modeling of rotor 28, bearings, gears, and universal joint drive shaft 26, the dynamic models of each substructure are connected by kinematic pairs within Adams software to form a fluid-thermal-structure coupled dynamics model of a large screw pump. The output shaft of motor 21 is connected to the input shaft of gearbox via a fixed joint; the gear output shaft is connected to the main shaft 61 via a fixed shaft; the main shaft 61 is connected to the connecting rod shaft 62 via a universal joint; and the connecting rod shaft 62 is connected to the screw rotor 63 via a universal joint. These connections are standard operating settings within the software and will not be elaborated further.

[0085] Furthermore, a fluid-structure-thermal coupled dynamic model of a large screw pump can be established based on the aforementioned modeling method. This method is based on the three-dimensional models of the substructures of bearings, gears, stator 27, and rotor 28. Therefore, by changing the three-dimensional models of these components—such as creating grooves on the outer and inner rings of bearings, the surface of gears, or changing the average dimensions of the inner cavity of stator 27—wear failures of the corresponding components can be simulated. Changing the shape of these components, such as modeling shaft-like parts or rotor 28 as a bent shape, can simulate deformation failures and imbalance failures of these components. Other failures that can be simulated by changing the three-dimensional model can also be simulated. Simulation using the fluid-structure-thermal coupled dynamic model of the simulated failure states can obtain failure state data under various failure conditions.

[0086] The large-scale screw pump flow-solid-thermal coupled dynamic model established using the above method can reduce the computational power requirements of the whole machine model, shorten the calculation time, meet the equipment design and development requirements, and obtain high-fault data for further development of intelligent operation and maintenance system.

[0087] Please see Figure 7 In one optional embodiment, the present invention provides a coupled dynamics fault twin modeling system for large screw pumps. The system includes an input device, an output device, a processor, and a memory, all interconnected. The memory stores a computer program comprising program instructions, and the processor is configured to invoke the program instructions to execute specific steps as described in the relevant embodiments of the coupled dynamics fault twin modeling method for large screw pumps provided by the present invention. The coupled dynamics fault twin modeling system for large screw pumps provided by the present invention has a complete structure, is objective and stable, and enhances the overall applicability and practical application capability of the present invention.

[0088] In summary, the present invention provides a coupled dynamics fault twin modeling method and system for large screw pumps. Based on finite element modeling, data-driven modeling, and multibody dynamics modeling, it transforms the forces acting on the rotor 28 into a mathematical model and embeds it into the overall multibody dynamics model. This avoids finite element calculations during whole-machine simulation, significantly saving computation time. By selecting and using linear flexibility theory based on structural and operating characteristics, it avoids unnecessary nonlinear calculations, reduces the computational requirements of the whole-machine model, shortens computation time, meets equipment design and development needs, and overcomes the difficulty of obtaining fault data for large screw pumps. The method of this invention is easy to understand, computationally simple, requires less work, and is convenient for engineering applications, providing a theoretical foundation and technical support for the further development of the electromechanical equipment maintenance field.

[0089] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A coupled dynamics fault twin modeling method for a large screw pump, characterized in that, Includes the following steps: The large screw pump is divided into rotor segments along the axial direction, and a fixed spatial coordinate system is established for the rotor segments. Under the fixed spatial coordinate system, the force components of the rotor segment are extracted through fluid simulation experiments, and a nonlinear prediction model of the force components is constructed. The force equations of the rotor segments are obtained using a generalized regression neural network structure, and the force equations are used as the nonlinear prediction model. The force equations satisfy the following relationship: in, for Force equations in the axial direction, for Force equations in the axial direction, for Force equations in the axial direction, For coefficients, For rotational speed, For the temperature of the medium, For inlet pressure, The average diameter of the stator cavity. For rotor phase; Based on the nonlinear prediction model, a multibody dynamics model of the large screw pump is obtained to obtain a substructure dynamics model. The fluid-thermal-solid coupled dynamic model of the large screw pump is obtained by integrating the dynamic model of the substructure and performing fault simulation to obtain fault state data.

2. The coupled dynamics fault twin modeling method for large screw pumps according to claim 1, characterized in that, The extraction of the force components of the rotor segments through fluid simulation experiments in the fixed spatial coordinate system includes: Orthogonal experimental tables were obtained by setting different design factors, and orthogonal experiments were constructed using the central composite design method as the fluid simulation experiments. In the fluid simulation experiment, a non-Newtonian fluid dynamics model was introduced, and a coupled model of the flow field and temperature field was established by combining the energy equation. The fluid simulation test was conducted on the large screw pump to obtain simulation results, and the force components under different test conditions were extracted based on the simulation results.

3. The coupled dynamics fault twin modeling method for large screw pumps according to claim 2, characterized in that, The construction of the nonlinear prediction model for the force components includes: The design factors and the force components are normalized to obtain a standardized dataset; Under the fixed spatial coordinate system, based on the standardized dataset, the force equations of the rotor segments are obtained using a generalized regression neural network structure; The force equation is used as the nonlinear prediction model.

4. The coupled dynamics fault twin modeling method for large screw pumps according to claim 1, characterized in that, The process of performing multibody dynamics modeling on the large screw pump based on the nonlinear prediction model to obtain the substructure dynamic model includes: The rotor segments are connected according to the massless beam model to form a rotor dynamics model, and a rotor-stator dynamics model is constructed based on the geometric contact force model. The bearing dynamics model of the large screw pump is obtained by modeling the bearing using the nonlinear prediction model. The gear dynamics model of the large screw pump is obtained by using the nonlinear prediction model to perform gear modeling. Based on the nonlinear prediction model, the universal drive shaft of the large screw pump is modeled to obtain the dynamic model of the universal drive shaft. The rotor dynamics model, the rotor-stator dynamics model, the bearing dynamics model, the gear dynamics model, and the universal joint drive shaft dynamics model are used as the substructure dynamics model.

5. The coupled dynamics fault twin modeling method for large screw pumps according to claim 4, characterized in that, The process of obtaining a bearing dynamic model by modeling the bearing of the large screw pump using the nonlinear prediction model includes: The bearing of the large screw pump is considered as a first multibody dynamic system consisting of an inner ring, an outer ring, and several rolling elements; Obtain the relative radial and relative tangential displacements of the inner and outer rings, and determine the contact deformation by combining the bearing radial clearance; When the contact deformation is greater than zero, the normal contact force between the rolling element and the raceway is obtained; Based on the relative sliding velocity at the contact point, the tangential contact force is obtained by combining the normal contact force; The bearing dynamics model is obtained by synthesizing the normal contact force and the tangential contact force.

6. The coupled dynamics fault twin modeling method for large screw pumps according to claim 4, characterized in that, The process of using the nonlinear prediction model to model the gears of the large screw pump to obtain a gear dynamics model includes: The transmission gear pair of the large screw pump is regarded as a second multibody dynamics system consisting of a driving gear and a driven gear; For any pair of teeth engaged at any given moment, the meshing force of the pair is obtained along the meshing line direction, taking into account the amount of meshing deformation. The gear tangential force is obtained based on the input torque and the gear pitch circle radius; The gear dynamics model is obtained by combining the meshing force and the gear tangential force.

7. The coupled dynamics fault twin modeling method for large screw pumps according to claim 6, characterized in that, The meshing deformation includes: In the direction of the meshing line, the relative displacement of the driving gear and the driven gear is obtained; The meshing deformation is obtained based on the relative displacement, combined with the tooth backlash and transmission error of the transmission gear pair.

8. The coupled dynamics fault twin modeling method for large screw pumps according to claim 4, characterized in that, The process of modeling the universal joint drive shaft of the large screw pump based on the nonlinear prediction model to obtain the universal joint drive shaft dynamic model includes: The universal drive shaft of the large screw pump is regarded as a third multibody dynamic system consisting of the main shaft, connecting rod shaft and screw rotor; The universal drive shaft includes a drive shaft and a driven shaft, and the kinematic relationship between the drive shaft and the driven shaft is obtained; The dynamic model of the universal joint drive shaft is established using the kinematic relationship as the angular coupling relationship.

9. The coupled dynamics fault twin modeling method for large screw pumps according to claim 1, characterized in that, The integrated modeling of the substructure dynamics model yields the fluid-thermal-structure coupled dynamics model of the large screw pump, and fault simulation is performed to obtain fault state data, including: The substructure dynamics models are integrated and modeled using kinematic pairs to form the fluid-thermal-solid coupled dynamics model. The fluid-thermal-solid coupled dynamic model simulating the fault state is obtained by modifying the three-dimensional model of the substructure of the large screw pump, and the fault state data is obtained by simulation.

10. A coupled dynamics fault twin modeling system for a large screw pump, characterized in that, The system includes an input device, an output device, a processor, and a memory, which are interconnected. The memory stores a computer program, which includes program instructions. The processor is configured to invoke the program instructions to execute the coupled dynamics fault twin modeling method for a large screw pump as described in any one of claims 1-9.

Images