Two-dimensional energy-gathered jet flow feature extraction method

Abstract

The invention discloses a method for extracting two-dimensional jet flow characteristic information, and aims to solve the problems that the density distribution is not uniform in the jet flow forming process and the jet flow characteristic point information cannot be directly extracted by an ALE algorithm, a regular grid point interpolation technology is adopted, jet flow and explosive areas are divided based on the density, and the jet flow grid point information is obtained. Each jet flow grid point is regarded as the center of a square and rotates around a jet flow symmetry axis to form a jet flow rotary body infinitesimal element. And the total mass and the total kinetic energy of the jet flow can be accurately obtained by summing the mass and the kinetic energy of each jet flow rotary body micro-element. According to the method, the total mass and total kinetic energy of the two-dimensional jet flow are effectively extracted, and the precision and reliability of data analysis are improved.

Description

Technical Field

[0001] The present invention belongs to the field of focused jets, and in particular relates to a two-dimensional focused jet feature extraction method. Background Art

[0002] With the rapid development of neural network technology, its significant advantages in handling complex nonlinear problems have led to its widespread application. The mapping relationship between charge structure and shaped charge jet characteristics is inherently a complex nonlinear problem, and neural networks provide a more effective and comprehensive means to deeply understand and characterize this mapping relationship.

[0003] However, neural network training requires a large amount of data. While simulating a shaped charge jet using a 3D model can yield accurate data, it consumes significant computational resources. Extracting the jet's characteristics using a 2D model is challenging. Therefore, exploring how to effectively extract shaped charge jet characteristics using a 2D model is crucial for building accurate neural network proxy models.

[0004] Lei Jingyu and Huang Hongxin respectively predicted jet velocity in "Research on the Application of Kriging Models to Predict Shaped Charge Jet Velocity," and in "A Method for Predicting Shaped Charge Jet Velocity Based on BP Artificial Neural Networks." However, simply predicting jet velocity is insufficient for assessing the destructive power of jets. To more comprehensively study the destructive effects of jets, it is crucial to incorporate additional jet characteristics. In particular, jet quality and kinetic energy are as crucial as jet velocity in jet damage analysis and cannot be ignored. Summary of the Invention

[0005] The purpose of the present invention is to propose an efficient and accurate two-dimensional focused jet feature extraction method to solve the problems of uneven density distribution in the jet forming process and the inability of the ALE algorithm to directly extract jet feature point information.

[0006] The technical solution to achieve the purpose of the present invention is: a two-dimensional focused jet feature extraction method, comprising the following steps:

[0007] Step 1: Perform numerical simulation on the shaped charge jet warhead to obtain the jet forming simulation results, and extract the jet node coordinate information, ALE domain density information, and ALE domain velocity information at a specific time point after the jet stabilizes.

[0008] Step 2: Based on the jet node coordinate information obtained in step 1, a regular grid point area is established. The regular grid point area is a rectangle, where the length is the jet length and the width is the maximum width of the jet.

[0009] Step 3: Interpolate the regular grid point area generated in step 2 with the density information of the ALE domain extracted in step 1 to obtain the density information of the regular grid points.

[0010] Step 4: Conditionally screen the density information of the regular grid points obtained in step 3 to select regular grid points with a density greater than or equal to 5 g / cm3, which are called jet grid points.

[0011] Step 5: Consider each jet grid point as the center point of a square, rotate the square around the jet symmetry axis to form a jet rotation element, calculate the mass of each jet rotation element, and sum the masses of the jet rotation elements calculated for the jet grid points to obtain the total mass of the jet.

[0012] Step 6: Interpolate the jet grid points obtained in step 4 with the velocity information of the ALE domain extracted in step 1 to obtain the velocity information of the jet grid points.

[0013] Step 7: Based on the mass of each jet rotor element obtained in step 5 and the velocity information of the jet grid point obtained in step 6, the kinetic energy of each jet rotor element is calculated. The total kinetic energy of the jet is obtained by summing the kinetic energy of each jet rotor element.

[0014] Compared with the prior art, the present invention has at least the following significant advantages:

[0015] (1) Considering the uneven density characteristics of the focused jet forming process, the jet mass and kinetic energy were extracted.

[0016] (2) Extracting jet mass and kinetic energy from two-dimensional jets significantly reduces the computational resource requirements compared to three-dimensional jets.

[0017] (3) The present invention can be used to extract jet mass and kinetic energy in batches, supporting the generation of large-scale data sets, thereby improving the training efficiency and generalization ability of neural network models. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 Flowchart of the two-dimensional jet feature extraction method.

[0019] Figure 2 Construct a schematic diagram for a region of regular grid points.

[0020] Figure 3 Construct a schematic diagram for the jet grid points.

[0021] Figure 4 Schematic diagram of the jet rotating body. DETAILED DESCRIPTION

[0022] Combine Figure 1The two-dimensional jet feature information extraction method described in this invention is based on regular grid point interpolation, dividing the explosive region and the jet region by density threshold, thereby accurately obtaining the feature information of the jet grid points. By constructing the jet rotation micro-element, the mass and kinetic energy of the jet rotation micro-element are calculated, and the total mass and total kinetic energy of the jet are obtained by summing the jet rotation micro-element. This method can not only effectively reduce computational costs, but also provide a high-quality large-scale data set for neural network-based proxy model training, thereby promoting the research and application of the characteristics of shaped charge jets. The steps are as follows:

[0023] Step 1: Perform numerical simulation on the shaped charge jet warhead to obtain the jet forming simulation results, and extract the jet node coordinate information, ALE domain density information, and ALE domain velocity information at a specific time point after the jet stabilizes.

[0024] The jet node coordinate information comes from the calculation results of LS_DYNA; the first column is the node number of the jet, the second column is the x-coordinate of the jet node, the third column is the y-coordinate of the jet node, and the fourth column is the z-coordinate of the jet node. For two-dimensional jets, z is always zero.

[0025] The density information of the ALE domain comes from the calculation results of LS_DYNA; the first column is the node number of the ALE domain, and the second column is the density information of the node.

[0026] The velocity information results of the ALE domain are derived from the calculation results of LS_DYNA. The first column is the node number of the ALE domain, the second column is the velocity in the x-direction of the ALE domain, the third column is the velocity in the y-direction of the ALE domain, and the fourth column is the velocity in the z-direction of the ALE domain. For a two-dimensional jet, the velocity in the z-direction is always zero.

[0027] Step 2: Based on the jet node coordinate information obtained in step 1, a regular grid point area is established. The regular grid point area is rectangular and arranged at regular intervals. The upper and lower intervals between grid points are consistent with the left and right intervals. The interval range is 0.005-0.01 cm. The length is the jet length, and the width is the maximum width of the jet. Figure 2 Schematic diagram of the regular grid point area.

[0028] Step 3: Interpolate the regular grid point area generated in step 2 with the density information of the ALE domain extracted in step 1 to obtain the density information of the regular grid points.

[0029] Step 4: Distinguish the explosive area and the jet area based on the density information of the regular grid points obtained in step 3, where the explosive density range is 1 to 2 g / cm 3 , the jet density is 8.96g / cm 3 ; Using 5g / cm 3As the density threshold, it can effectively distinguish the explosive area from the jet area. 3 The regular grid points of , which are called jet grid points; Figure 3 Schematic diagram of the jet grid points.

[0030] Step 5: Consider each jet grid point as the center point of a square, and rotate the square around the jet symmetry axis to form a jet rotation body microelement as shown in the following figure: Figure 4 As shown, the details are as follows:

[0031] When the jet symmetry axis is outside the square formed with the jet grid point as the center, the jet rotation body infinitesimal element is a ring.

[0032] When the jet symmetry axis coincides with the side of a square centered on the jet grid point, the jet rotation body element is a cylinder.

[0033] Calculate the mass of each jet rotation unit and sum the masses of the jet rotation units calculated at the jet grid points to obtain the total mass of the jet. The total mass of the jet is calculated as follows:

[0034]

[0035] Where M is the total mass of the jet, ρ i is the density of the i-th jet rotation unit, V i is the volume of the i-th jet rotation unit, c is the distance between adjacent jet grid points, that is, the height of the jet rotation unit, x i is the distance from the i-th jet grid point to the center of the jet symmetry axis, and n is the total number of jet grid points.

[0036] Step 6: interpolate the jet grid points obtained in step 4 with the velocity information of the ALE domain extracted in step 1 to obtain the velocity information of the jet grid points;

[0037] Step 7: Based on the mass of each jet rotor element obtained in step 5 and the velocity information of the jet grid point obtained in step 6, the kinetic energy of each jet rotor element is calculated. The total kinetic energy of the jet is obtained by summing the kinetic energy of each jet rotor element. The calculation formula for the total kinetic energy of the jet is as follows:

[0038]

[0039] Where E is the total kinetic energy of the jet, v i is the velocity of the i-th jet rotating body element.

Claims

1. A method for extracting characteristic information from a two-dimensional jet, characterized in that: The specific steps are as follows: Step 1: numerically simulate the shaped charge jet warhead to obtain the jet forming simulation results, and extract the jet node coordinate information, ALE domain density information, and ALE domain velocity information at a specific time point after the jet stabilizes; Step 2: Based on the jet node coordinate information obtained in step 1, a regular grid point area is established. The regular grid point area is a rectangle, where the length is the jet length and the width is the maximum width of the jet; Step 3: interpolate the regular grid point area generated in step 2 with the density information of the ALE domain extracted in step 1 to obtain the density information of the regular grid points; Step 4: Conditionally filter the density information of the regular grid points obtained in step 3 and select the points with a density greater than or equal to 5g / cm 3 The regular grid points of , which are called jet grid points; Step 5: Consider each jet grid point as the center point of a square, rotate the square around the jet symmetry axis to form a jet rotation unit, calculate the mass of each jet rotation unit, and sum the jet rotation unit masses calculated for the jet grid points to obtain the total mass of the jet; Step 6: interpolate the jet grid points obtained in step 4 with the velocity information of the ALE domain extracted in step 1 to obtain the velocity information of the jet grid points; Step 7: Based on the mass of each jet rotor element obtained in step 5 and the velocity information of the jet grid point obtained in step 6, the kinetic energy of each jet rotor element is calculated. The total kinetic energy of the jet is obtained by summing the kinetic energy of each jet rotor element.

2. The two-dimensional jet feature extraction method according to claim 1, characterized in that: In step 1, the injection molding simulation results include but are not limited to jet node coordinate information, density information of the ALE domain, and velocity information of the ALE domain.

3. The two-dimensional jet feature extraction method according to claim 1, characterized in that: In step 2, the regular grid point area is rectangular and is arranged at regular intervals. The upper and lower intervals between the grid points are consistent with the left and right intervals, and the interval range is 0.005 to 0.01 cm.

4. The two-dimensional jet feature extraction method according to claim 1, characterized in that: In step 5, the square is rotated around the jet symmetry axis to form a jet rotational body microelement, as follows: When the jet symmetry axis is outside the square formed by the jet grid point as the center, the jet rotation body infinitesimal element is a ring; When the jet symmetry axis coincides with the side of a square centered on the jet grid point, the jet rotation body element is a cylinder.

5. The two-dimensional jet feature extraction method according to claim 1, characterized in that: The explosive area and the jet area are distinguished based on the density information of the regular grid points obtained in step 3, where the explosive density range is 1 to 2 g / cm 3 , the jet density is 8.96g / cm 3 ; Using 5g / cm 3 As a density threshold, it can effectively distinguish the explosive area from the jet area.

6. The two-dimensional jet feature extraction method according to claim 1, characterized in that: In step 5, each jet grid point is regarded as the center point of a square. The square is rotated around the jet symmetry axis to form a jet rotation unit. The mass of each jet rotation unit is calculated. By summing the masses of the jet rotation units, the total mass of the jet is obtained, as follows: Where M is the total mass of the jet, ρ i is the density of the i-th jet rotation unit, V i is the volume of the i-th jet rotation unit, c is the distance between adjacent jet grid points, that is, the height of the jet rotation unit, x i is the distance from the i-th jet grid point to the center of the jet symmetry axis, and n is the total number of jet grid points.

7. The two-dimensional jet feature extraction method according to claim 1, characterized in that: In step 7, the kinetic energy of each jet rotor element obtained in step 5 is calculated based on the mass of each jet rotor element and the velocity information of the jet grid point obtained in step 6. The total kinetic energy of the jet is obtained by summing the kinetic energy of each jet rotor element, as follows: Where E is the total kinetic energy of the jet, v i is the velocity of the i-th jet rotating body element.

Images