Compressor stability enhancement method based on modal energy migration direction control

Abstract

The present invention discloses a compressor stabilization method based on the regulation of the modal energy migration direction. The present invention integrates the impeller blade parameterization, diffuser parameterization and casing geometry parameterization methods, expands the physical design space, takes into account the coupling effect between the casing and the impeller and diffuser, and expands the limits of traditional optimization methods. The instantaneous dynamic modal decomposition method and the gradient variation hybrid optimization algorithm are used to implement multi-objective optimization of the compressor, with maximizing the inherent modal energy ratio and minimizing the characteristic modal energy ratio as the optimization goals, and finally obtains the optimal design scheme for the stabilization effect. The present invention more clearly analyzes the stabilization mechanism of casing processing and provides a new method for compressor stabilization design.

Description

Technical Field

[0001] The present invention relates to the technical field of impeller machinery optimization design, and more specifically, to a compressor expansion and stabilization design method. Background Art

[0002] The compressor is a key component of turbocharging technology and plays an irreplaceable role in the development of modern national defense and national economy. Over the past decades, with the pursuit of high-performance engines, the compressor has been developing towards high speed, high pressure ratio and small volume. The speed of the compressor has been increasing, resulting in complex flow phenomena inside, aggravating the flow instability of the compressor and narrowing its stable operating range sharply. Therefore, the research on the stability expansion of the turbocharger centrifugal compressor is very important.

[0003] At present, there are two main methods to expand the stable operating margin of compressors: active control and passive control. Active control technology requires real-time monitoring of compressor instability precursor wave signals, and the development and evolution speed of compressor instability disturbances is extremely fast, so it is very difficult to use in industrial applications and is still in the laboratory stage. Passive stabilization only requires simple changes to the existing geometry to achieve the purpose of stabilization. In the 1970s, researchers discovered that casing processing can stabilize compressors, and it has the advantages of simple and reliable structure, easy processing, and good stabilization effect. It has become one of the most widely used stabilization technologies.

[0004] However, the introduction of casing treatment to improve the compressor stability margin is usually accompanied by additional efficiency and pressure ratio losses. Limited by the physical boundary limits of the casing design space and the limitations of the understanding of the stabilization mechanism, a universal design method that can achieve the stabilization effect and clarify the stabilization mechanism to improve the compressor stability margin has not yet been found. Summary of the invention

[0005] The present invention aims to develop an optimization method that can more clearly analyze the stabilization mechanism of casing treatment and break through the space limit of traditional optimization design from the perspective of coupling analysis of the sudden increase / migration of unsteady flow modal energy and the change of unsteady flow structure in the flow field, and provide a compressor stabilization method based on the regulation of modal energy migration direction.

[0006] The present invention provides a compressor stabilization method based on modal energy migration direction control, which is performed according to the following steps:

[0007] (1) According to the geometric parameters of the compressor blade and casing, the parametric modeling method is determined; the hub and shroud profiles in the centrifugal impeller meridian plane are fitted using a cubic B-spline interpolation curve, and the blades are fitted with a second-order continuously differentiable cubic spline interpolation curve at the mid-arc control points in the B2 B plane, and the blade thickness is combined for parametric modeling; the casing is described by calculating the coordinates of all control points in the meridian plane through the corresponding relationship given the design parameters of the meridian plane;

[0008] (2) Through single parameter variation range analysis, under the condition that the compressor geometry does not suffer from severe distortion and deformation, the influence of a single design variable within the variation range on the compressor performance curve is calculated, and key design variables that improve the compressor performance by more than 4% in the impeller, diffuser, and casing treatment are selected, such as the impeller blade installation angle; the variation range in which the design variable always has a positive impact on the performance is taken as the design space;

[0009] (3) The optimal Latin hypercube sampling method is used to generate a sample set so that the sample points are evenly distributed in the entire design space. The number of initial samples in the sample set is generally related to the number of design variables, and 10 times the number of design variables can be selected;

[0010] (4) For each design scheme in the initial sample set, the unsteady Reynolds average method is used for numerical simulation; the computational domain is discretized using a structured grid, the blade surface is O-type grid, and other regions are H-type grid. The NS equations are discretized based on the finite volume method, and the SST model is selected as the turbulence model. The total temperature and total pressure conditions are given at the inlet boundary, and the design flow rate and near-stall flow rate are used as boundary conditions at the outlet to obtain the unsteady flow field of the compressor design condition and near-stall condition (the above description is for the entire unsteady numerical simulation method);

[0011] (5) The transient dynamic modal decomposition method is used to decompose the unsteady flow field of the compressor to obtain discrete frequency modes and their energy proportions. By analogy with the resonance problem, the modes are divided into two categories: one is the inherent modes that represent the inherent dynamic characteristics of the system, such as the blade passing frequency (BPF) and rotor passing frequency (RPF) generated by the sweep of the compressor blades and their multiples. These modes have a large energy proportion in the full operating range; the other is the characteristic modes that represent the response of the compressor internal flow system to external disturbances, which are usually multiple low-frequency small-scale modes excited by near-stall conditions. As the flow rate decreases, the energy proportion usually increases gradually;

[0012] (6) Construct a Kriging model to obtain the mapping relationship between the modal energy ratio and the design variables. Use the gradient mutation optimization algorithm to implement the optimization, with maximizing the inherent modal energy and minimizing the characteristic modal energy as the objective function, obtain the Pareto frontier solution set, select the geometry corresponding to some optimal solutions for numerical simulation calculations, obtain the performance curve, and finally obtain the optimal design scheme for stabilization effect.

[0013] Compared with the prior art, the present invention has the following beneficial effects:

[0014] 1. The compressor stabilization design method based on modal energy migration direction control of the present invention integrates the impeller blade parameterization, diffuser parameterization and casing geometry parameterization methods, expands the physical design space, considers the coupling between the casing and the impeller and diffuser, and expands the limit of the traditional optimization method. At the same time, the gradient mutation optimization algorithm based on the NSGA-II algorithm is used to implement the optimization, which significantly accelerates the convergence process of the multi-objective optimization algorithm and reduces the computational cost.

[0015] 2. The compressor stabilization design method based on the control of the modal energy migration direction of the present invention relies on the unsteady flow field for optimization design, clarifies the impact of changing the design parameters on the unsteady flow structure, and has a clearer physical mechanism and high robustness. Based on the instantaneous dynamic modal decomposition method, the optimization target in the traditional method is replaced by the modal energy ratio from efficiency and pressure ratio, and the mapping relationship between the design parameters and the modal energy ratio is constructed using the kriging model. From the perspective of the coupled analysis of the modal energy surge / migration and the unsteady flow structure change of the flow field, the stabilization mechanism of the casing treatment can be analyzed more clearly, providing new ideas and directions for the compressor stabilization design theory and method.

[0016] The present invention has the following beneficial effects:

[0017] The compressor stabilization method based on the regulation of the modal energy migration direction of the present invention integrates the impeller blade parameterization, diffuser parameterization and casing geometry parameterization methods, expands the physical design space, takes into account the coupling effect between the casing and the impeller and diffuser, and extends the limits of traditional optimization methods; on the other hand, based on the calculation results of the unsteady flow field, the optimization indicators in the traditional design are replaced by the modal energy ratio from efficiency and pressure ratio, and the stabilization mechanism of the casing treatment is more clearly analyzed, providing a new method for the compressor stabilization design theory and method. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 Shown is a schematic diagram of the geometric structure of the self-circulating casing selected in a specific embodiment;

[0019] Figure 2 The figure shows the comparison of modal energy proportions of three casing treatments in design conditions;

[0020] Figure 3 The figure shows the comparison of modal energy proportions of three casing treatments in near-stall conditions;

[0021] Figure 4 Shown is the flow chart of the gradient mutation hybrid optimization algorithm;

[0022] Figure 5 Shown is a comparison of isentropic efficiency curves for three types of casing treatments;

[0023] Figure 6 Shown is a comparison of the total pressure ratio curves of three casing treatments. DETAILED DESCRIPTION

[0024] The compressor stabilization method based on modal energy migration direction control of the present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0025] The specific process of the present invention is as follows:

[0026] (1) According to the geometric parameters of the compressor blade and casing, the parametric modeling method is determined; the hub and shroud profiles in the centrifugal impeller meridian plane are fitted using a cubic B-spline interpolation curve, and the blades are fitted with a second-order continuously differentiable cubic spline interpolation curve at the mid-arc control points in the B2 B plane, and the blade thickness is combined for parametric modeling; the casing is described by calculating the coordinates of all control points in the meridian plane through the corresponding relationship given the design parameters of the meridian plane;

[0027] (2) Through single parameter variation range analysis, under the condition that the compressor geometry does not suffer from severe distortion and deformation, the influence of a single design variable within the variation range on the compressor performance curve is calculated, and key design variables that improve the compressor performance by more than 4% in the impeller, diffuser, and casing treatment are selected, such as the impeller blade installation angle; the variation range in which the design variable always has a positive impact on the performance is taken as the design space;

[0028] (3) The optimal Latin hypercube sampling method is used to generate a sample set so that the sample points are evenly distributed in the entire design space. The number of initial samples in the sample set is generally related to the number of design variables, and 10 times the number of design variables can be selected;

[0029] (4) For each design scheme in the initial sample set, the unsteady Reynolds average method is used for numerical simulation; the computational domain is discretized using a structured grid, the blade surface is O-type grid, and other regions are H-type grid. The NS equations are discretized based on the finite volume method, and the SST model is selected as the turbulence model. The total temperature and total pressure conditions are given at the inlet boundary, and the design flow rate and near-stall flow rate are used as boundary conditions at the outlet to obtain the unsteady flow field of the compressor design condition and near-stall condition (the above description is for the entire unsteady numerical simulation method);

[0030] (5) The transient dynamic modal decomposition method is used to decompose the unsteady flow field of the compressor to obtain discrete frequency modes and their energy proportions. By analogy with the resonance problem, the modes are divided into two categories: one is the inherent modes that represent the inherent dynamic characteristics of the system, such as the blade passing frequency (BPF) and rotor passing frequency (RPF) generated by the sweep of the compressor blades and their multiples. These modes have a large energy proportion in the full operating range; the other is the characteristic modes that represent the response of the compressor internal flow system to external disturbances, which are usually multiple low-frequency small-scale modes excited by near-stall conditions. As the flow rate decreases, the energy proportion usually increases gradually;

[0031] (6) Construct a Kriging model to obtain the mapping relationship between the modal energy ratio and the design variables. Use the gradient mutation optimization algorithm proposed by the inventors, take maximizing the inherent modal energy and minimizing the characteristic modal energy as the objective function, obtain the Pareto frontier solution set, select the geometry corresponding to some optimal solutions for numerical simulation calculations, obtain the performance curve, and finally obtain the optimal design scheme for stabilization effect.

[0032] Example:

[0033] 1) This example is a turbocharger compressor. Figure 1 The self-circulating casing geometry selected in this optimization process is given. According to the process in step (2), the influence of a single design variable on the compressor performance curve within the variation range is calculated. Based on the experience of previous related studies, a total of 6 design variables are selected in the impeller and diffuser, including blade sweep, trailing edge blade tip installation angle, etc., as well as 11 design variables that can fully describe the casing geometry (such as Figure 1 As shown in the figure, 7 key design variables that affect the compressor performance by more than 4% in this example are selected from a total of 17 design variables, including blade sweep, trailing edge blade tip installation angle, diffuser blade trailing edge radius, diffuser blade installation angle, casing suction slot width, suction slot axial position, and return slot incident angle.

[0034] 2) Through univariate analysis, the specific range of parameter changes is determined (with the original parameters as the benchmark value): the sweep degree changes in the range of 1° to 10°; the trailing edge blade tip installation angle changes in the range of -10° to 25°; the diffuser blade trailing edge radius changes in the range of -10mm to 0mm; the diffuser blade angle changes in the range of -1.5° to 2.5°; the casing suction groove width changes in the range of -1.5mm to 1.5mm; the suction groove axial position changes in the range of -5.2mm to 5.8mm; the reflow groove incident angle changes in the range of -15° to 25°.

[0035] 3) Use the method introduced in step (3) to generate 70 samples in the design space, which can ensure that the sample points are evenly distributed in the design space. Perform unsteady numerical simulation on these samples according to the grid generation method and boundary conditions introduced in step (4) to obtain unsteady flow field results. After the pressure at the monitoring point remains stable or fluctuates periodically, select the calculation results within a fluctuation range to ensure that the flow field is accurate and reliable;

[0036] 4) Use the method in step (5) to perform modal decomposition on the unsteady flow field of the compressor to obtain discrete frequency modes and their energy proportions. Figure 2 and 3 The modal energy proportions of the compressors that introduce solid wall casings and reference casings in this example under design conditions and near-stall conditions are given respectively. The blade passing frequency (BPF) and their frequency multiples generated by the sweep of the compressor blades have a large energy proportion in the entire operating range and are called natural modes; at the same time, the small-scale low-frequency modes with frequencies lower than the BPF gradually increase their energy proportion as the flow rate decreases, and are considered to be characteristic modes representing the response of the compressor internal flow system to external disturbances, corresponding to the unstable flow structure generated under near-stall conditions. After completing the calculation of all samples, a kriging model is constructed to obtain the mapping relationship between the modal energy proportion and the design variables. The cross-validation method is used to verify the accuracy of the surrogate model. The predicted values ​​of the performance indicators are well compared with the calculated values, and the root mean square errors (RMSE) are all less than 0.1. It can be considered that the surrogate model in this example has a high mapping accuracy and can be used for subsequent optimization.

[0037] 5) Optimize using step (6). During the optimization process, the gradient mutation hybrid optimization algorithm is used to implement the multi-objective optimization of the compressor. The process is shown in the attached figure. Figure 4 As shown in the figure, the maximization of the natural mode energy ratio and the minimization of the characteristic mode energy ratio are taken as optimization objectives, the Pareto frontier solution set is obtained, the geometry corresponding to some optimal solutions is selected for numerical simulation calculation, the performance curve is obtained, and finally the optimal design scheme for the expansion stabilization effect is obtained. Figure 2 , 3It can be seen that under the design condition and near-stall condition, the natural modal energy proportion of the coupled optimized casing is greater than that of the reference casing; and as the flow rate decreases, the low-frequency modal energy proportion of the coupled optimized casing is significantly lower than that of the reference casing, which means that it has improved the influence of some unsteady unstable flow structures and suppressed the potential possibility of instability. Figure 5 , 6 As shown in the figure, the performance of the casing after coupling optimization is better than that of the benchmark casing in the full range of operating conditions, which broadens the stability of the compressor.

[0038] The above is only an introduction to the preferred embodiments of the present invention. It should be noted that, for those skilled in the art, several modifications and substitutions can be made without departing from the principle of the present invention, and these modifications and substitutions should also be regarded as the protection scope of the present invention.

Claims

1. A compressor stabilization method based on modal energy migration direction control, characterized in that: The following steps are involved: (1) According to the geometric parameters of the compressor blade and casing, the parametric modeling method is determined. The hub and shroud profiles in the meridian plane of the centrifugal impeller are fitted using a cubic B-spline interpolation curve. The blade uses a second-order continuously differentiable cubic spline interpolation curve to fit the mid-arc control points in the B2B plane, and the blade thickness is combined for parametric modeling. The casing is described by calculating the coordinates of all control points in the meridian plane through the corresponding relationship by giving the design parameters of the meridian plane. (2) Through single parameter variation range analysis, under the condition that the compressor geometry does not suffer from serious distortion and deformation, the influence of a single design variable within the variation range on the compressor performance curve is calculated, and the key design variables that affect the compressor performance by more than 4% in the impeller, diffuser, and casing treatment are selected, including blade sweep, trailing edge blade tip installation angle, diffuser blade trailing edge radius, diffuser blade installation angle, casing suction slot width, suction slot axial position, and return slot incident angle; the variation range in which the design variable always has a positive impact on the performance is taken as the design space; (3) Perform optimal Latin hypercube sampling in the design space to generate an initial sample set so that the sample points are evenly distributed in the entire design space; (4) For each design scheme in the initial sample set generated in step (3), the unsteady Reynolds average method is used for numerical simulation. The NS equations are discretized based on the finite volume method. The SST model is selected as the turbulence model. The inlet boundary is given as the total temperature and total pressure conditions. The outlet is given as the design flow rate and the near-stall flow rate as boundary conditions to obtain the unsteady flow field of the compressor design condition and the near-stall condition. (5) The instantaneous dynamic modal decomposition method is used to decompose the unsteady flow field of the compressor to obtain discrete frequency modes and their energy proportions. By analogy with the resonance problem, the modes are divided into two categories: inherent modes, i.e., the inherent dynamic characteristics of the system, and characteristic modes, i.e., the response of the compressor internal flow system to external disturbances. The initial population fitness is obtained: the inherent modal energy proportion and the characteristic modal energy proportion. (6) Construct a Kriging model to obtain the mapping relationship between the energy proportions of the inherent modes and characteristic modes and the key design variables selected in step (2), and use a gradient mutation optimization algorithm to implement optimization. The objective function is to maximize the inherent mode energy and minimize the characteristic mode energy, and finally obtain the optimal design scheme for the stabilization effect.

Images