A method for optimizing the parameters of a nozzle atomization model

Abstract

The present invention discloses a method for optimizing nozzle atomization model parameters, including: conducting atomization experiments to measure atomization parameters, recording the atomization parameters to form an experimental database, and then obtaining an empirical model or semi-empirical model related to the atomization parameters; according to the model and several groups of experimental conditions, calculating the characteristics of the oil mist field, which are defined as the first characteristics, and then adopting numerical simulation, using the same experimental conditions, to obtain the characteristics of the oil mist field under simulation, which are defined as the second characteristics; according to the difference between the first characteristics and the second characteristics, iteratively adjusting the model parameters in the numerical simulation to obtain atomization parameters comparable to the empirical or semi-empirical model, and obtaining the final nozzle atomization model; the present invention avoids large errors between the data obtained after numerical calculation using the Fluent default atomization model and the experiments, and provides reliable atomization parameters for combustion numerical calculation.

Description

Technical Field

[0001] The present invention relates to the field of nozzle atomization models, and particularly to a method for optimizing nozzle atomization model parameters. Background Art

[0002] In liquid rocket engines and aero-engines, as one of the key components, the nozzle can atomize fuel oxidants and transport them to a suitable area, playing a decisive role in the combustion performance of the combustion chamber.

[0003] The atomization process includes primary atomization and secondary atomization. Among them, primary atomization is affected by parameters such as the nozzle geometric structure, and the secondary atomization process is affected by the external flow field. However, when using the default nozzle atomization model in Fluent for numerical calculation, it cannot be guaranteed that an oil mist field comparable to experimental data can be obtained. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a method for optimizing nozzle atomization model parameters to solve the problem that the default nozzle atomization model in Fluent cannot guarantee an oil mist field comparable to experimental data. The method of the present invention can obtain a suitable nozzle atomization model by optimizing the atomization experimental data, and finally an oil mist field comparable to the measured value can be obtained in numerical calculation.

[0005] To achieve the above purpose, the present invention adopts the following technical solutions:

[0006] A method for optimizing nozzle atomization model parameters, the method comprising:

[0007] Step S1: First, perform atomization experiment measurement on the nozzle to obtain its atomization parameters, then record the atomization parameters to form an experimental database, then obtain an empirical model or semi-empirical model related to the atomization parameters, and finally dynamically update the empirical model or semi-empirical model according to the experimental database to obtain a first model;

[0008] Step S2: First, according to the first model obtained in step S1 and several groups of experimental conditions, calculate the characteristics of the nozzle oil mist field, which are defined as the first characteristics. Then, by means of CFD numerical simulation, using the same several groups of experimental conditions, obtain the characteristics of the nozzle oil mist field under the simulation results, which are defined as the second characteristics;

[0009] Step S3: Iteratively adjust the model parameters in the numerical simulation according to the differences between the first feature and the second feature to obtain atomization parameters comparable to those of the first model, and fix the atomization model corresponding to these atomization parameters to obtain the final nozzle atomization model. Among them, in the iterative adjustment process, Latin hypercube sampling is selected to form parameter combinations to obtain an optimization method for the nozzle atomization model parameters.

[0010] Further, in step S1, when conducting the atomization experiment of the nozzle, measure the nozzle through a high-speed camera and a Malvern particle size analyzer to obtain its atomization parameters under different working conditions. Among them, the atomization parameters include: particle diameter, atomization cone angle, mass flow rate, and penetration depth.

[0011] Further, in step S3, when iteratively adjusting the model parameters in the numerical simulation, consider the influence of the external flow field on the secondary atomization of the nozzle, specifically including: adding a correction factor to the experimental database.

[0012] Further, in step S3, the method of selecting Latin hypercube sampling to form parameter combinations to obtain an optimization method for the nozzle atomization model parameters specifically includes:

[0013] Step S301: For the selection of model parameters in the numerical simulation, use Latin hypercube sampling to cross-combine to form a variation range covering the target field. Each group of variation parameters in the multiple groups of variation parameters respectively forms its corresponding nozzle atomization model for numerical calculation to obtain each group of variation parameters and their corresponding nozzle parameters.

[0014] Step S302: Evaluate the errors of the multiple groups of nozzle parameters calculated in step S301 and the characteristics of the oil mist field given by the first model under the same working conditions to obtain multiple groups of actual error values.

[0015] Step S303: Fit and screen or interpolate the multiple groups of actual error values to obtain atomization parameters comparable to those of the first model.

[0016] The beneficial effects of the present invention are:

[0017] The present invention can combine an optimization method based on experimental data and an empirical / semi-empirical model to obtain optimized nozzle atomization model parameters. Based on these parameters for numerical calculation, an oil mist field comparable to the measured value can be finally obtained, avoiding large errors between the data obtained by numerical calculation using the Fluent default atomization model and the experiment, and providing reliable atomization parameters for combustion numerical calculation. Description of the Drawings

[0018] Figure 1Particle size distribution diagram of the transverse jet experiment with a direct injection nozzle in Example 1;

[0019] Figure 2 Schematic diagram of the CFD domain in Example 1;

[0020] Figure 3 In Example 1, based on Figure 1 which is used as the initial value and substituted into Fluent, the calculated particle size distribution diagram is obtained;

[0021] Figure 4 In Example 1, after optimizing the atomization parameters based on the atomization experiment data and substituting them into Fluent for calculation, the obtained particle size distribution diagram;

[0022] Figure 5 In Example 1, the comparison diagram of the experimental results, the results after experimental calculation, and the results after optimizing the experimental data; Specific implementation mode

[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0024] Example 1

[0025] Refer to Figures 1 - 5 In this Example 1, a method for optimizing the parameters of a nozzle atomization model is provided, which specifically includes:

[0026] Step S1: First, conduct an atomization experiment on the nozzle to measure its atomization parameters, then record the atomization parameters to form an experimental database. At the same time, obtain an empirical model or semi-empirical model related to the atomization parameters. Finally, combine the experimental data and the empirical model or semi-empirical model to obtain a more substantial and perfect result, and the model corresponding to this result is the first model.

[0027] Specifically, in this embodiment, when conducting the atomization experiment on the nozzle, use a high-speed camera and a Malvern particle size analyzer to measure the nozzle to obtain its atomization parameters under different working conditions. Among them, the atomization parameters include: particle size (SMD), atomization cone angle, mass flow rate, and penetration depth. While solving the data management problem, it can facilitate the dynamic update of the empirical / semi-empirical model based on the experimental data. Enrich the experimental database based on the previous literature research and the conclusions of others.

[0028] Step S2: First, based on the first model obtained in Step S1 and several groups of experimental conditions, calculate the characteristics of the nozzle oil mist field, which are defined as the first characteristics. Then, using the CFD numerical simulation method and the same several groups of experimental conditions, obtain the characteristics of the nozzle oil mist field under the simulation results, which are defined as the second characteristics;

[0029] Specifically, since there is a certain gap between the finally obtained oil mist field by numerical calculation and the results of the empirical / semi-empirical model, after analyzing the data, it is decided whether to stop the calculation or adjust the atomization model parameters and recalculate. Therefore, an iterative relationship and an objective optimization method between the oil mist field simulation results and the nozzle atomization model also need to be established, that is, by iterating the simulation results to achieve atomization parameters comparable to the experiments and save the corresponding atomization model in the database.

[0030] Step S3: According to the difference between the first characteristics and the second characteristics, iteratively adjust the model parameters in the numerical simulation to obtain atomization parameters comparable to the first model, and fix the atomization model corresponding to the atomization parameters to obtain the final nozzle atomization model. Among them, in the iterative adjustment, the Latin hypercube sampling is selected to form parameter combinations to obtain an optimization method for the nozzle atomization model parameters.

[0031] Specifically, considering the characteristic that the numerical simulation takes a long time and the commonly used gradient descent method may not be suitable for fine-tuning the atomization model parameters, the Latin hypercube sampling is selected to form parameter combinations to obtain an optimization method for the nozzle atomization model parameters, which specifically includes:

[0032] Step S301: For the selection of model parameters in the numerical simulation, the Latin hypercube sampling cross-combination is used to form a reasonable change interval covering the target field. Each group of change parameters in the multiple groups of change parameters respectively forms its corresponding nozzle atomization model for numerical calculation to obtain each group of change parameters and their corresponding nozzle parameters, such as the atomization cone angle, SMD distribution, etc.;

[0033] Specifically, in this embodiment, the selected parameters have a relatively wide distribution range, such as 0.1 MPa - 2 MPa.

[0034] Step S302: Evaluate the errors of the multiple groups of nozzle parameters obtained through calculation in Step S301 and the oil mist field characteristics given by the first model under the same working conditions to obtain multiple groups of actual error values;

[0035] Step S303: Fit and screen or interpolate the multiple groups of actual error values to obtain atomization parameters comparable to the first model

[0036] Specifically, since the external flow field has a secondary atomization effect on the nozzle during the actual flow process, it is necessary to iteratively adjust the model parameters in the numerical simulation in step S3 and then add a correction factor to the experimental database after comparing and analyzing the influence of the external flow field on the secondary atomization of the nozzle.

[0037] In this embodiment, in order to verify the advancement and correctness of the method, a simulation experiment was carried out. The specific results are as follows: Figures 1 - 5 As shown, the particle size distribution results of the direct nozzle lateral jet experiment ( Figure 1 ) is the reference data. After optimizing it using the above steps, the optimization results are brought into Fluent for numerical calculation.

[0038] After analyzing the calculation results, the particle size distribution diagram is obtained. Figure 4 It can be seen that the particle size distribution diagram of the optimized result after Fluent calculation is more similar to the experimental measurement result than the result obtained by directly bringing the original experimental data into Fluent calculation, indicating that the optimization result and the experimental result are comparable.

[0039] The matters not described in detail in the present invention are all known technologies to those skilled in the art.

[0040] The preferred specific embodiments of the present invention are described in detail above. It should be understood that a person skilled in the art can make many modifications and changes based on the concept of the present invention without creative work. Therefore, any technical solution that can be obtained by a person skilled in the art through logical analysis, reasoning or limited experiments based on the concept of the present invention on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A method for optimizing the parameters of a nozzle atomization model, characterized in that, The method includes: Step S1: First, conduct an atomization experiment on the nozzle to measure its atomization parameters, then record the atomization parameters to form an experimental database, then obtain an empirical model or semi-empirical model related to the atomization parameters, and finally dynamically update the empirical model or semi-empirical model according to the experimental database to obtain a first model; Step S2: First, according to the first model obtained in Step S1 and several groups of experimental conditions, calculate the characteristics of the nozzle oil mist field, which is defined as the first characteristic. Then, by means of CFD numerical simulation, using the same several groups of experimental conditions, obtain the characteristics of the nozzle oil mist field under the simulation results, which is defined as the second characteristic; Step S3: According to the difference between the first characteristic and the second characteristic, iteratively adjust the model parameters in the numerical simulation to obtain atomization parameters comparable to the first model, and fix the atomization model corresponding to the atomization parameters to obtain the final nozzle atomization model. Among them, in the iterative adjustment, select to use Latin hypercube sampling to form a parameter combination to obtain an optimization method for the nozzle atomization model parameters; In Step S3, the selection of using Latin hypercube sampling to form a parameter combination to obtain an optimization method for the nozzle atomization model parameters specifically includes: Step S301: For the selection of model parameters in the numerical simulation, use Latin hypercube sampling cross-combination to form a variation range covering the target field. Each group of variation parameters in the multiple groups of variation parameters respectively forms its corresponding nozzle atomization model for numerical calculation to obtain each group of variation parameters and its corresponding nozzle parameters; Step S302: Evaluate the errors of the multiple groups of nozzle parameters calculated in Step S301 and the characteristics of the oil mist field given by the first model under the same working conditions to obtain multiple groups of actual error values; Step S303: Fit and screen or interpolate the multiple groups of actual error values to obtain atomization parameters comparable to the first model.

2. The method for optimizing the nozzle atomization model parameters according to claim 1, characterized in that In Step S1, when conducting the atomization experiment on the nozzle, measure the nozzle through a high-speed camera and a Malvern particle size analyzer to obtain its atomization parameters under different working conditions. Among them, the atomization parameters include: particle size, atomization cone angle, mass flow rate, and penetration depth.

3. The method for optimizing the nozzle atomization model parameters according to claim 1, characterized in that, In Step S3, when iteratively adjusting the model parameters in the numerical simulation, consider the influence of the external flow field on the secondary atomization of the nozzle, which specifically includes: adding a correction factor to the experimental database.

Images