Method and system for rapidly predicting detonation velocity of single particle energetic material under laser loading

Abstract

The application provides a laser loading single-particle energetic material blast speed rapid prediction method and system, uses a pulse laser to load single-particle energetic material to generate a shock wave, uses an enhanced sCMOS camera to collect a shock wave image, calculates a shock wave characteristic speed through shock wave position information in the shock wave image, and establishes a linear relationship model through a partial least square regression of the shock wave characteristic speed and a known macroscopic blast speed, so that the blast speed rapid prediction of the energetic material is realized; the application consumes a very small amount of sample, eliminates the interaction between particles, is high in safety, can realize the excitation of only one single-particle energetic material by each laser pulse, and only needs a particle order of energetic material to complete the test, so that the application has important significance for improving the safety of energetic material testing and testing the performance of an ultra-small amount of energetic material even a few single-particle order of energetic material.

Description

Technical Field

[0001] This invention belongs to the field of detonation performance testing technology for energetic materials, and particularly relates to a method and system for rapid prediction of detonation velocity of laser-loaded single-particle energetic materials. Background Technology

[0002] Energetic materials are an important energy source in weapon systems. The rapid detection of their explosive performance parameters directly affects the research and development of energetic materials. Detonation velocity, as a key parameter for measuring the rate at which energetic materials release energy, is directly related to the intensity and effect of the explosion. Testing detonation performance is crucial for guiding the research and development of new energetic materials and evaluating the safety of storage and transportation.

[0003] Traditional detonation performance testing is constrained by charge density and critical diameter, resulting in high charge consumption and high risk. New materials are constantly emerging, but due to limited production capacity, they cannot meet the critical diameter and density requirements, making it impossible to test key performance indicators.

[0004] In recent years, detection techniques based on laser-loaded micro-explosion particle swarms have shown promise in addressing this issue. However, this method inevitably introduces factors such as particle size differences, interparticle interface effects, and variations in ablation volume and thermal conductivity, leading to insufficient detection reliability. Furthermore, some novel energetic materials, in their initial synthesis, contain only a few or a single particle, failing to meet the milligram-level minimum detection dose requirement for micro-explosion particle swarm detection methods. Summary of the Invention

[0005] To address the problems of traditional detonation velocity testing methods, such as high consumption of explosives, complex experimental operations, and high risks, as well as the unreliability of particle group micro-explosion methods due to particle size differences and interface effects, and the inability to test the detonation velocity of certain novel energetic materials with only a few particles, this invention provides a method and system for rapid prediction of the detonation velocity of laser-loaded single-particle energetic materials. This method involves capturing images of the shock wave generated by a single laser-loaded energetic material particle, extracting the shock wave location information to calculate the characteristic velocity matrix of the shock wave, and then using an intelligent algorithm to establish a model to predict the detonation velocity.

[0006] A laser-loaded single-particle energetic material detonation velocity rapid prediction system includes a timing controller, a shock wave image acquisition subsystem, an optical processing subsystem, and a single-particle energetic material selection subsystem and a LIBS subsystem with a shared optical path.

[0007] The single-particle energetic material selection subsystem is used to select an imaging position for placing the single-particle energetic material. The imaging position satisfies the following conditions: the laser focusing position of the LIBS subsystem is at the same height as the position with the clearest image in the single-particle energetic material selection subsystem, and the laser focusing spot of the LIBS subsystem at the imaging position is smaller than the area of ​​the single-particle energetic material.

[0008] The LIBS subsystem is used to emit lasers to excite single-particle energetic materials located at the imaging position.

[0009] The timing controller is used to control the triggering time of the LIBS subsystem and the shock wave image acquisition subsystem, so that the LIBS subsystem is triggered to emit laser first, and the shock wave image acquisition subsystem acquires the shock wave image of the single particle energetic material after it is excited at different delays. The triggering time of the shock wave image acquisition subsystem is later than the triggering time of the LIBS system, and the difference between the two triggering times is the delay, until the shock wave images of each type of single particle energetic material at all delay times are acquired.

[0010] The optical processing subsystem is used to establish a linear prediction model of the shock wave characteristic velocity and detonation velocity of a single-particle energetic material based on shock wave images with different delays.

[0011] Furthermore, the single-particle energetic material selection subsystem includes a first visible light source 8, a first dichroic mirror 4, a second dichroic mirror 5, a microscope objective 6, and a CMOS camera 9;

[0012] The first visible light source 8 is used to provide an illumination light source for imaging the sample surface. The illumination light source is reflected by the first dichroic mirror 4 and transmitted through the second dichroic mirror 5 and the microscope objective 6 to illuminate the sample surface placed on the sample stage 7. The illumination light source illuminates the sample surface to form a sample optical signal. The sample optical signal is transmitted through the microscope objective 6 and reflected by the second dichroic mirror 5 to enter the CMOS camera 9. The CMOS camera 9 is used to convert the sample optical signal into a sample image.

[0013] Furthermore, the LIBS subsystem includes a pulsed laser 1, a reflector 2, a concave lens 3-1, a plano-convex lens 3-2, a first dichroic mirror 4 for reflection, a second dichroic mirror 5, and a microscope objective 6;

[0014] The pulsed laser 1 is used to provide a pulsed laser source, wherein the pulsed laser source is reflected by the mirror 2 in sequence, and transmitted by the concave lens 3-1 and the plano-convex lens 3-2 to obtain a divergent pulsed laser; the divergent pulsed laser is transmitted through the first dichroic mirror 4 and the second dichroic mirror 5, and finally the divergent pulsed laser is focused by the microscope objective 6, and finally focused onto the single energetic material on the sample stage 7.

[0015] The different distances between the concave lens 3-1 and the plano-convex lens 3-2 result in different divergence angles of the divergent pulsed laser. With different divergence angles, the height of the laser focusing position obtained when the microscope objective 6 focuses the divergent pulsed laser varies. The smaller the distance, the larger the divergence angle and the lower the height of the laser focusing position. Conversely, the larger the distance, the smaller the divergence angle and the higher the height of the laser focusing position.

[0016] Furthermore, the method for obtaining the imaging position is as follows:

[0017] Adjust the distance between the sample stage 7 and the microscope objective 6 until the sample image in the CMOS camera 9 is clearest;

[0018] Adjust the distance between the concave lens 3-1 and the plano-convex lens 3-2 to change the height of the laser focusing position obtained when the microscope objective 6 focuses the diverging pulsed laser, until the laser focusing position is at the same height as the clearest position of the sample image in the CMOS camera 9.

[0019] Furthermore, after obtaining the imaging position, the imaging position is marked using the following method:

[0020] A marker is placed on the sample stage, and pulsed laser 1 sends a pulsed laser to create an ablation pit on the marker. The three-dimensional coordinates of the ablation pit are then marked as the three-dimensional coordinates of the imaging position.

[0021] Furthermore, the shock wave image acquisition subsystem includes a second visible light source 10, a first convex lens 11, a slit 12, a second convex lens 13, a third convex lens 14, an aperture 15, a filter 16, a lens 17, and an enhanced sCMOS camera 18; meanwhile, a single-particle energetic material is located between the second convex lens 13 and the third convex lens 14, and the region where the plasma generated after the single-particle energetic material is excited and the shock wave are located is a disturbance region;

[0022] The second visible light source 10 is used to provide a background light source for the shock wave disturbance generated after a single particle of energetic material is excited. The background light source passes through the first convex lens 11, the slit 12, and the second convex lens 13 in sequence to form parallel light. The parallel light is deflected after passing through the disturbance area. The deflected light passes through the third convex lens 14, the aperture 15, the filter 16, and the lens 17 in sequence before being incident on the enhanced sCMOS camera 18, where a shock wave image is formed.

[0023] Furthermore, the delay of the shock wave image acquisition subsystem is 0.5μs, 1.0μs, 1.5μs, 2.0μs, 2.5μs, 3.0μs, 4.0μs, 5.0μs, 6.0μs, 7.0μs, 8.0μs, 9.0μs, 10.0μs, 15.0μs, and 20.0μs after the single energetic particle is excited.

[0024] A method for rapid prediction of the detonation velocity of a laser-loaded single-particle energetic material based on a laser-loaded single-particle energetic material rapid prediction system includes the following steps:

[0025] S1: For each energetic material, two sets of shock wave images were acquired after excitation with different delays, including 15 different delays;

[0026] S2: For each energetic material, the characteristic velocity vector of the shock wave in each of the two sets of shock wave images is obtained. Specifically, the method for obtaining the characteristic velocity vector of the shock wave in any set of shock wave images for each energetic material is as follows:

[0027] S21: Obtain the position information of the shock wave front in the shock wave images of different delays in the current group for each energetic material;

[0028] S22: Obtain the average shock wave radius for each delay, wherein the method for obtaining the average shock wave radius for the nth delay is as follows:

[0029] The major axis radius is obtained by elliptic fitting of the position information of the shock wave front at the nth delay. minor axis radius and major axis radius Corresponding to the lateral propagation radius, minor axis radius Corresponding to the longitudinal propagation radius, and the major axis radius With minor axis radius The average radius is denoted as the shock wave radius.

[0030] S23: Obtain the average propagation velocity of the shock wave between each pair of delays, where the average propagation velocity of the shock wave between the nth delay and the (n+1)th delay is... The calculation method is as follows:

[0031]

[0032] in, The average radius of the (n+1)th delay, t n+1 For the (n+1)th delay, t n This is the nth delay;

[0033] S24: Obtain the average transverse velocity and average longitudinal propagation velocity of the shock wave between each pair of delays, where the average transverse velocity of the shock wave between the nth delay and the (n+1)th delay is... and the average longitudinal propagation velocity of the shock wave The calculation method is as follows:

[0034]

[0035] in, The major axis radius corresponding to the nth delay. The minor axis radius corresponding to the nth delay;

[0036] S25: Obtain the shock wave propagation distance fitting curve corresponding to the current group of shock wave images using the Dewey model. Then, differentiate the shock wave propagation distance fitting curve at each of the 15 delays to obtain the shock wave propagation fitting velocity v at each of the 15 delays. n ;

[0037] S26: Construct the shock wave characteristic velocity vector I of the current group of shock wave images. m as follows:

[0038]

[0039] Where v0 is the average propagation velocity of the shock wave with a delay of 0.5 μs, which is obtained by dividing the radius of the shock wave with a delay of 1 μs by 1 μs; m = 1, 2;

[0040] S3: Construct the shock wave characteristic velocity matrix q of the current energetic material from the two sets of shock wave characteristic velocity vectors corresponding to different delays.

[0041]

[0042] S4: The final shock wave characteristic velocity matrix Q for each energetic material is obtained by averaging the shock wave characteristic velocity vectors I1 and I2 corresponding to each energetic material. k k is the label for each energetic material, and the shock wave characteristic velocity matrix for all energetic materials is Q:

[0043]

[0044] S5: By fitting the shock wave characteristic velocity matrix Q using an intelligent algorithm based on partial least squares regression, a linear prediction model for the shock wave characteristic velocity and detonation velocity of a single-particle energetic material is obtained.

[0045] Furthermore, the method for obtaining the shock wave front of any shock wave image is as follows:

[0046] Calculate the gradient value matrix of the shock wave image along the positive horizontal direction, negative horizontal direction, negative vertical direction, and positive vertical direction to obtain the gradient map in four directions;

[0047] The gradient maps in each direction are smoothed and filtered separately. Then, based on the Laida criterion, significant noise points with gradient intensity fluctuations greater than a set multiple of the standard deviation are directly removed from the smoothed gradient maps by using a gradient intensity threshold, thus obtaining smoothed gradient maps in each direction.

[0048] Opening and closing operations are performed on the smooth gradient maps in each direction. The opening operation is used to eliminate secondary noise points where the connectivity does not meet the set requirements, and the closing operation is used to eliminate pulse noise points inside the shock wave. Then, the smooth gradients after each opening and closing operation are denoised by a median filter to obtain the first denoised maps in each direction with noise removal.

[0049] The first-order denoised gradient maps corresponding to the positive and negative gradient maps in the horizontal direction and the first-order denoised gradient maps corresponding to the positive and negative gradient maps in the vertical direction are stitched together to obtain a first-order stitched gradient map. The dividing line of the horizontal stitching is determined by the estimated value of the laser incident line position, and the dividing line of the vertical stitching is determined by the estimated value of the height position of the sample cover glass surface.

[0050] Calculate the gradient value matrix of the first-order stitched gradient map along the positive horizontal direction, negative horizontal direction, negative vertical direction, and positive vertical direction to obtain the second-order gradient maps in four directions;

[0051] The quadratic gradient maps in four directions are stitched together to obtain a quadratic stitched gradient map;

[0052] The secondary denoising algorithm is used to perform secondary denoising on the secondary stitched gradient map, and the secondary denoised gradient map is used as the final shock wave front.

[0053] Beneficial effects:

[0054] 1. This invention provides a rapid prediction system for the detonation velocity of a single-particle energetic material loaded with a laser. It utilizes a pulsed laser to load a single-particle energetic material to generate a shock wave, and an enhanced sCMOS camera to acquire images of the shock wave. The characteristic velocity of the shock wave is calculated using the shock wave's position information from the images. A linear relationship model is established between the characteristic velocity and the known macroscopic detonation velocity using partial least squares regression, enabling rapid prediction of the detonation velocity of the energetic material. This invention consumes very little sample, eliminates inter-particle interactions, and is highly safe. It allows each laser pulse to excite only one single-particle energetic material, requiring only a particle-scale amount of energetic material for testing. This is of great significance for improving the safety of energetic material testing and for performance testing of energetic materials with ultra-small quantities or even several single-particle scales.

[0055] 2. This invention provides a method for rapid prediction of the detonation velocity of a single-particle energetic material loaded by laser. It provides a detailed modeling process for the linear prediction model of the shock wave characteristic velocity and detonation velocity of a single-particle energetic material, which can realize rapid prediction of the detonation velocity of energetic materials. This is of great significance for improving the safety of energetic material testing and for performance testing of energetic materials with ultra-small charge or even several single-particle quantities. Attached Figure Description

[0056] Figure 1A schematic diagram of a laser-loaded rapid detonation velocity prediction system for single-particle energetic materials provided by the present invention;

[0057] 1—Pulsed laser; 2—Mirror; 3-1—Concave lens; 3-2—Plan-convex lens; 3—Beam expander; 4—First dichroic mirror; 5—Second dichroic mirror; 6—Microscope objective; 7—Sample stage; 8—First visible light source; 9—CMOS camera; 10—Second visible light source; 11—First convex lens; 12—Slit; 13—Second convex lens; 14—Third convex lens; 15—Aperture; 16—Filter; 17—Lens; 18—Enhanced sCMOS camera; 19—Timing controller. Detailed Implementation

[0058] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0059] A laser-loaded single-particle energetic material detonation velocity rapid prediction system includes a DG535 timing controller, a shock wave image acquisition subsystem, an optical processing subsystem, and a single-particle energetic material selection subsystem and a LIBS subsystem with a shared optical path.

[0060] The single-particle energetic material selection subsystem is used to select an imaging position for placing the single-particle energetic material. The imaging position satisfies the following conditions: the laser focusing position of the LIBS subsystem is at the same height as the position with the clearest image in the single-particle energetic material selection subsystem, and the laser focusing spot of the LIBS subsystem at the imaging position is smaller than the area of ​​the single-particle energetic material.

[0061] The LIBS subsystem is used to emit lasers to excite single-particle energetic materials located at the imaging position.

[0062] The timing controller is used to control the triggering time of the LIBS subsystem and the shock wave image acquisition subsystem, so that the LIBS subsystem is triggered to emit laser first, and the shock wave image acquisition subsystem acquires the shock wave image of the single particle energetic material after it is excited at different delays. The triggering time of the shock wave image acquisition subsystem is later than the triggering time of the LIBS system, and the difference between the two triggering times is the delay, until the shock wave images of each type of single particle energetic material at all delay times are acquired.

[0063] The optical processing subsystem is used to establish a linear prediction model of the shock wave characteristic velocity and detonation velocity of a single-particle energetic material based on shock wave images with different delays.

[0064] Specifically, such as Figure 1As shown, the single-particle energetic material selection subsystem includes a first visible light source 8, a first dichroic mirror 4, a second dichroic mirror 5, a microscope objective 6, and a CMOS camera 9.

[0065] The first visible light source 8 is used to provide an illumination light source for imaging the sample surface. The illumination light source is reflected by the first dichroic mirror 4 and transmitted through the second dichroic mirror 5 and the microscope objective 6 to illuminate the sample surface placed on the sample stage 7. The illumination light source illuminates the sample surface to form a sample optical signal. The sample optical signal is transmitted through the microscope objective 6 and reflected by the second dichroic mirror 5 to enter the CMOS camera 9. The CMOS camera 9 is used to convert the sample optical signal into a sample image.

[0066] Optionally, the first visible light source 8 is a white LED, the first dichroic mirror 4 and the second dichroic mirror 5 transmit a wavelength of 1064nm and reflect a wavelength of 200-1000nm, and the microscope objective 6 has a magnification of 5x and a numerical aperture of 0.14.

[0067] The LIBS subsystem is based on LIBS technology. It uses a pulsed laser 1 with a wavelength of 1064nm, a pulse width of 9ns, a single-pulse laser energy of 10-120mJ@1064nm that is adjustable, a divergence angle of less than 0.5mrad, and a pulse repetition frequency of 1-10Hz as the excitation source to provide injection energy. The output energy of the pulsed laser 1 is adjusted to control the laser energy at the focal point of the focusing lens at 6mJ. Specifically, the LIBS subsystem includes a pulsed laser 1, a reflector 2, a concave lens 3-1, a plano-convex lens 3-2, a first dichroic mirror 4 for reflection, a second dichroic mirror 5, and a microscope objective 6.

[0068] The pulsed laser 1 provides a pulsed laser source, which is reflected sequentially by mirror 2 and transmitted through concave lens 3-1 and plano-convex lens 3-2 to obtain a diverging pulsed laser. Mirror 2 is used to change the optical path direction of the 1064nm laser, making the laser perpendicular to the horizontal plane with a reflectivity greater than 99.5%. Simultaneously, the concave lens 3-1 and plano-convex lens 3-2 are used in combination to change the laser divergence angle, expand the laser spot area to avoid damage to subsequent optical lens groups, and ultimately change the distance between the actual focal point of the laser and the microscope objective 6. After transmission through the first dichroic mirror 4 and the second dichroic mirror 5, the divergent pulsed laser is finally focused by the microscope objective 6 onto the single-particle energetic material on the sample stage 7. The excited single-particle energetic material generates high-temperature and high-pressure plasma and a rapid exothermic chemical reaction, and the surrounding air is violently compressed to generate an air shock wave. The first dichroic mirror 4 is used to change the optical path direction of the first visible light source 8 so that it illuminates the sample surface and transmits 1064nm laser light. The second dichroic mirror 5 is used to change the optical path direction of the sample's optical signal so that it enters the CMOS camera 9 and transmits 1064nm laser light.

[0069] The different distances between the concave lens 3-1 and the plano-convex lens 3-2 result in different divergence angles of the divergent pulsed laser. With different divergence angles, the height of the laser focusing position obtained when the microscope objective 6 focuses the divergent pulsed laser varies. The smaller the distance, the larger the divergence angle and the lower the height of the laser focusing position. Conversely, the larger the distance, the smaller the divergence angle and the higher the height of the laser focusing position.

[0070] Optionally, the microscope objective 6 is a Mitutoyo model with a magnification of 5x and a numerical aperture of 0.14; the sample stage 7 is adjustable in three dimensions (XYZ) for placing the sample. The CMOS camera 9 uses a planar array CMOS sensor with an exposure time of 0.3-1000ms.

[0071] Based on this, the method for obtaining the imaging position is as follows:

[0072] The distance between the sample stage 7 and the microscope objective 6 is changed until the sample image in the CMOS camera 9 is clearest. The distance between the concave lens 3-1 and the plano-convex lens 3-2 is adjusted to change the height of the laser focusing position obtained when the microscope objective 6 focuses the diverging pulsed laser, until the laser focusing position is at the same height as the position where the sample image in the CMOS camera 9 is clearest. That is, the single-particle energetic material selection subsystem and the LIBS subsystem are focused on the same plane. The laser spot corresponding to the horizontal position of this plane is the smallest, so that only a single particle is excited.

[0073] After obtaining the imaging location, the imaging location can be marked. The marking method is as follows:

[0074] A marker is placed on the sample stage. The pulsed laser 1 emits a single laser pulse, creating an ablation pit on the marker. The three-dimensional coordinates of the ablation pit are then used as the three-dimensional coordinates of the imaging position. In other words, before placing a single energetic particle, a laser pulse is emitted to locate the ablation pit in the field of view of the CMOS camera 9. The location of the ablation pit is the laser focusing point. The sample stage 7 is moved to place a single energetic particle at the laser focusing point in the field of view of the CMOS camera 9. Then, a laser pulse is emitted to excite the single energetic particle.

[0075] The shock wave image acquisition subsystem includes a second visible light source 10, a first convex lens 11, a slit 12, a second convex lens 13, a third convex lens 14, an aperture 15, a filter 16, a lens 17, and an enhanced sCMOS camera 18. Simultaneously, a single-particle energetic material is located between the second convex lens 13 and the third convex lens 14, and the region containing the plasma and shock wave generated after the single-particle energetic material is excited is a disturbed region. The second visible light source 10 uses a green LED light source as the imaging light source, with a power of 20W, a light-emitting center size of 5×5mm, and a luminous flux of 1800-2000lm. The wavelength range is 520-525nm; the first convex lens 11 is a cemented doublet achromatic lens that focuses the light from the second visible light source 10; the second convex lens 13 is a cemented doublet achromatic lens with a focal length of 150mm; the third convex lens 14 is a cemented doublet achromatic lens with a focal length of 100mm; the filter 16 has a center wavelength of 520nm, a half-width of 10nm, and a peak transmittance of 85%; the lens 17 is a built-in lens of the camera; the enhanced sCMOS camera 18 has a frame rate of 200MHz, a resolution of 2560×2160 pixels, a pixel size of 6.5μm, and a sensor size of 16.6×14.0mm;

[0076] The second visible light source 10 is used to provide a background light source for the plasma and shock wave disturbance generated after the single-particle energetic material is excited. The background light source forms parallel light after passing through the first convex lens 11, the slit 12 and the second convex lens 13 in sequence. After passing through the disturbance area, the parallel light forms a shock wave with uneven light intensity distribution. The shock wave passes through the third convex lens 14, the aperture 15, the filter 16 and the lens 17 in sequence and is incident on the enhanced sCMOS camera 18, forming a shock wave image in the enhanced sCMOS camera 18.

[0077] In other words, the slit 13 focuses the background light source emitted by the two visible light sources 10 into an approximate point light source, which is placed at the front focal plane of the second convex lens 13. The emitted light beam becomes parallel light after passing through the second convex lens 13. The parallel light passes through the microfluidic field region disturbed by the shock wave and is converged by the third convex lens 14. An aperture 15 is placed at the focal point of the third convex lens 14, and then a filter 16 is placed to reduce the strong light effect brought by the plasma's own radiation. The filter 16 has a center wavelength of 520nm, a half bandwidth of 10nm, and a peak transmittance of 85%. Finally, the shock wave image is presented in the enhanced sCMOS camera 18 through the lens 17.

[0078] It should be noted that the timing controller controls the triggering time of the LIBS subsystem and the shock wave image acquisition subsystem, which actually controls the delay between the pulsed laser 1 and the enhanced sCMOS camera 18. The shock wave image acquisition subsystem captures shock wave images with different delays after laser emission. For each sample, each set of shock wave images selects 15 delays: 0.5μs, 1.0μs, 1.5μs, 2.0μs, 2.5μs, 3.0μs, 4.0μs, 5.0μs, 6.0μs, 7.0μs, 8.0μs, 9.0μs, 10.0μs, 15.0μs, and 20.0μs.

[0079] Therefore, it can be seen that, Figure 1 The connection relationships of the laser-loaded single-particle energetic material detonation velocity rapid prediction system shown can be summarized as follows:

[0080] The mirror 2, concave lens 3-1, plano-convex lens 3-2, first dichroic mirror 4, second dichroic mirror 5, microscope objective 6, and sample stage 7 are arranged vertically from top to bottom. The mirror surfaces of mirror 2, first dichroic mirror 4, and second dichroic mirror 5 are all placed at a 45° angle to the horizontal plane. The planes of concave lens 3-1, plano-convex lens 3-2, microscope objective 6, and sample stage 7 are placed parallel to the horizontal plane. The planes of first convex lens 11, slit 12, second convex lens 13, third convex lens 14, aperture 15, filter 16, and lens 17 are all placed perpendicular to the horizontal plane. The plane of sample stage 7 is placed horizontally to the horizontal plane and located on the microscope objective. The best focal position for the clearest image is found at mirror 6, which is also the laser focusing position. The slit 12 is located at the focal point of the second convex lens 13. The visible light passing through the second convex lens 13 forms parallel light. The aperture 15 is located at the focal point of the third convex lens 14. The parallel light is focused after being converged by the third convex lens 14. The enhanced sCMOS camera 18 and its lens 17 are placed in a suitable position after the filter 16. The suitable standard is that the image is clear and the size is appropriate. The timing controller 19 connects the pulsed laser 1 and the enhanced sCMOS camera 18 to control the delay between the two and capture the delayed shock wave image.

[0081] After the shock wave image is captured, this invention extracts the position information of the shock wave leading edge from the image and constructs a shock wave feature matrix, thereby establishing a linear prediction model for the shock wave characteristic velocity and detonation velocity of a single-particle energetic material. Therefore, based on a laser-loaded single-particle energetic material detonation velocity rapid prediction system, this invention provides a laser-loaded single-particle energetic material detonation velocity rapid prediction method, including the following steps:

[0082] S1: For each energetic material, two sets of shock wave images were acquired after excitation with different delays. There are 15 different delays. For example, the triggering time of the shock wave image acquisition subsystem is 0.5μs, 1.0μs, 1.5μs, 2.0μs, 2.5μs, 3.0μs, 4.0μs, 5.0μs, 6.0μs, 7.0μs, 8.0μs, 9.0μs, 10.0μs, 15.0μs, and 20.0μs after the single energetic material is excited.

[0083] S2: For each energetic material, the characteristic velocity vector of the shock wave in each of the two sets of shock wave images is obtained. Specifically, the method for obtaining the characteristic velocity vector of the shock wave in any set of shock wave images for each energetic material is as follows:

[0084] S21: Obtain the position information of the shock wave front in the shock wave images of different delays in the current group for each energetic material;

[0085] S22: Obtain the average shock wave radius for each delay, wherein the method for obtaining the average shock wave radius for the nth delay is as follows:

[0086] The major axis radius is obtained by elliptic fitting of the position information of the shock wave front at the nth delay. minor axis radius and major axis radius Corresponding to the lateral propagation radius, minor axis radius Corresponding to the longitudinal propagation radius, and the major axis radius With minor axis radius The average radius is denoted as the shock wave radius.

[0087] S23: Obtain the average propagation velocity of the shock wave between each pair of delays, where the average propagation velocity of the shock wave between the nth delay and the (n+1)th delay is... The calculation method is as follows:

[0088]

[0089] in, The average radius of the (n+1)th delay, t n+1 For the (n+1)th delay, t nThis is the nth delay;

[0090] S24: Obtain the average transverse velocity and average longitudinal propagation velocity of the shock wave between each pair of delays, where the average transverse velocity of the shock wave between the nth delay and the (n+1)th delay is... and the average longitudinal propagation velocity of the shock wave The calculation method is as follows:

[0091]

[0092] in, The major axis radius corresponding to the nth delay. The minor axis radius corresponding to the nth delay;

[0093] S25: Obtain the shock wave propagation distance fitting curve corresponding to the current group of shock wave images using the Dewey model. Then, differentiate the shock wave propagation distance fitting curve at each of the 15 delays to obtain the shock wave propagation fitting velocity v at each of the 15 delays. n ;

[0094] In other words, the evolution of the shock wave front over time can be fitted using the Dewey model, with the following fitting formula:

[0095]

[0096] In the formula, a, b, c, and d are fitting coefficients, R is the shock wave propagation distance, and t is the shock wave propagation time; differentiating the fitted curve yields the shock wave propagation fitting velocity v at 15 different times. n Divide the shock wave radius at time 1 μs by 1 μs to obtain the average propagation velocity v0 of the shock wave at time 0.5 μs.

[0097] S26: Combine the above shock wave velocity data in a specific order to construct the shock wave characteristic velocity vector I of the current group of shock wave images. m as follows:

[0098]

[0099] Where v0 is the average propagation velocity of the shock wave with a delay of 0.5 μs, which is obtained by dividing the radius of the shock wave with a delay of 1 μs by 1 μs; m = 1, 2;

[0100] S3: Construct the shock wave characteristic velocity matrix q of the current energetic material from the two sets of shock wave characteristic velocity vectors corresponding to different delays.

[0101]

[0102] S4: The final shock wave characteristic velocity matrix Q for each energetic material is obtained by averaging the shock wave characteristic velocity vectors I1 and I2 corresponding to each energetic material. k k is the label for each energetic material, and the shock wave characteristic velocity matrix for all energetic materials is Q:

[0103]

[0104] S5: By fitting the shock wave characteristic velocity matrix Q using an intelligent algorithm based on partial least squares regression, a linear prediction model of the shock wave characteristic velocity and detonation velocity of a single-particle energetic material is obtained, realizing the macroscopic detonation velocity prediction of a single-particle energetic material based on laser-loaded induced shock wave.

[0105] Furthermore, the method for obtaining the shock wave front of any shock wave image is as follows:

[0106] Calculate the gradient value matrix of the shock wave image along the positive horizontal direction, negative horizontal direction, negative vertical direction, and positive vertical direction to obtain a more obvious gradient map of the shock wave front in four directions;

[0107] The gradient maps in each direction are smoothed by filtering. Then, based on the Laida criterion, significant noise points with gradient intensity fluctuations greater than a set multiple, such as m times the standard deviation, are directly removed from the smoothed gradient maps by using a gradient intensity threshold. This yields the smoothed gradient maps in each direction. The value of m is selected based on the actual image conditions.

[0108] Opening and closing operations are performed on the smooth gradient maps in each direction. The opening operation is used to eliminate secondary noise points where the connectivity does not meet the set requirements, and the closing operation is used to eliminate pulse noise points inside the shock wave. Then, the smooth gradients after each opening and closing operation are denoised by a median filter to obtain the first denoised maps in each direction with noise removal.

[0109] The first-order denoised gradient maps corresponding to the positive and negative gradient maps in the horizontal direction and the first-order denoised gradient maps corresponding to the positive and negative gradient maps in the vertical direction are stitched together to obtain a first-order stitched gradient map. The dividing line of the horizontal stitching is determined by the estimated value of the laser incident line position, and the dividing line of the vertical stitching is determined by the estimated value of the height position of the sample cover glass surface.

[0110] Calculate the gradient value matrix of the first-order stitched gradient map along the positive horizontal direction, negative horizontal direction, negative vertical direction, and positive vertical direction to obtain the second-order gradient map in four directions; the second-order gradient calculation is to find the edge lines of the optimized first-order stitched gradient map.

[0111] The quadratic gradient maps in four directions are stitched together to obtain a quadratic stitched gradient map;

[0112] The secondary denoising of the concatenated gradient map is performed using the Random Sample-Side Consensus (RANSAC) algorithm, and the denoised secondary gradient map is used as the final shock wave front. Specifically, the RANSAC method for secondary denoising is as follows:

[0113] Take any three points from the quadratic gradient result image and perform circle fitting to obtain the center coordinates and radius. Calculate the distance from the remaining points to the center and compare it with the fitted radius. Record the number of points less than 2σ. Repeat the above steps k times (k is the number of non-zero pixels in the quadratic gradient result image). Select the center coordinates and radius of the circle with the most points as the parameters of the fitted circle. Calculate the distance from the remaining points to the center of the final fitted circle and compare it with the radius of the final fitted circle. Remove points greater than 3σ as noise. The quadratic gradient image after secondary noise reduction is the final shock wave front.

[0114] Furthermore, the method of using the laser-loaded single-particle energetic material detonation velocity rapid prediction system provided by this invention is summarized as follows:

[0115] Turn on the second visible light source 10 and the enhanced sCMOS camera 18, and adjust the second visible light source 10, lens groups 11, 13, 14, 16, 17, slit 12, and aperture 15 in the shock wave image acquisition system to be coaxial until a clear shock wave image can be captured.

[0116] Adjust the optical axis of the single-particle energetic material selection system and the optical lens group 2-5 of LIBS and the microscope objective 6 to be aligned, and turn on the first visible light source 8 and the CMOS camera 9.

[0117] Adjust the LIBS system and the distance between the microscope objective 6 and the sample stage 7 until the image in the CMOS camera 9 is clearest.

[0118] Adjust the distance between concave lens 3-1 and plano-convex lens 3-2 to change the size of the laser spot and the divergence angle until the laser is focused on the surface of the sample stage 7 when the image of the CMOS camera 9 is clearest.

[0119] Before placing the sample, a laser pulse is emitted to leave an ablation pit. The location of this ablation pit is the laser focusing position. The sample is placed, and the horizontal position of the sample stage 7 is moved. At the same time, the real-time image in the CMOS camera 9 is observed until a certain particle of the energetic material is located at the laser focusing position.

[0120] After the sample stage 7 is fixed in position, a laser pulse is emitted. The single energetic material at the laser focal point is excited to generate plasma and compress the surrounding gas environment to generate a shock wave. When the uniform parallel light passing through the area between the second convex lens 13 and the third convex lens 14 passes through the shock wave disturbance area with uneven density distribution, the parallel light will be deflected downward in the horizontal direction after passing through the shock wave disturbance area because the density of the shock wave front is higher. The brightness changes with the derivative of the refractive index with respect to the spatial distance in the shock wave disturbance space. Finally, the light intensity distribution on the image-enhanced sCMOS camera 18 is uneven, thus forming a shock wave image.

[0121] The timing controller 19 controls the enhanced sCMOS camera 18 to capture shock wave images at different delay times after the pulsed laser 1 emits a pulsed laser. Each sample captures shock wave images at 15 delay times: 0.5μs, 1.0μs, 1.5μs, 2.0μs, 2.5μs, 3.0μs, 4.0μs, 5.0μs, 6.0μs, 7.0μs, 8.0μs, 9.0μs, 10.0μs, 15.0μs, and 20.0μs.

[0122] After capturing a series of shock wave images at different time delays, the position information of the shock wave front in each image is extracted: In the shock wave image, there is a dark area behind the shock wave front and a bright area in front of it, with peaks and troughs appearing successively in terms of intensity. The shock wave front is the point where the gray value changes the fastest between the peaks and troughs, that is, the point where the maximum gradient is located. The shock wave front identification is to extract the point with the maximum gray gradient. The shock wave image is processed by gradient calculation, noise reduction, stitching, secondary gradient calculation, and secondary noise reduction. The secondary gradient map after secondary noise reduction is the final shock wave front.

[0123] The characteristic velocity matrix Q of the shock wave was calculated.

[0124] A linear model for predicting macroscopic detonation velocity of energetic materials at the particle size level is established by using a smart algorithm based on partial least squares regression. Ultimately, this enables safe, rapid, and ultra-small charge prediction of detonation velocity of energetic materials at the particle size level.

[0125] In summary, this invention provides a method and system for rapid prediction of the detonation velocity of a single-particle energetic material loaded with a laser. It utilizes a pulsed laser to load a single-particle energetic material to generate a shock wave, and an enhanced sCMOS camera to acquire images of the shock wave. The characteristic velocity of the shock wave is calculated using the shock wave's position information from the images. A linear relationship model is established between the characteristic velocity and the known macroscopic detonation velocity using partial least squares regression, enabling rapid prediction of the detonation velocity of the energetic material. This invention consumes very little sample, eliminates inter-particle interactions, and is highly safe. It allows each laser pulse to excite only one single-particle energetic material, requiring only a particle-scale amount of energetic material for testing. This is of great significance for improving the safety of energetic material testing and for performance testing of energetic materials with ultra-small quantities or even several single-particle scales.

[0126] Of course, the present invention may have other various embodiments. Without departing from the spirit and essence of the present invention, those skilled in the art can make various corresponding changes and modifications according to the present invention, but these corresponding changes and modifications should all fall within the protection scope of the appended claims.

Claims

1. A method for rapid prediction of the detonation velocity of a laser-loaded single-particle energetic material using a laser-loaded rapid prediction system, characterized in that, The prediction system includes a timing controller, a shock wave image acquisition subsystem, an optical processing subsystem, and a partial common-path single-particle energetic material selection subsystem and a LIBS subsystem. The single-particle energetic material selection subsystem is used to select an imaging position for placing the single-particle energetic material. The imaging position satisfies the following conditions: the laser focusing position of the LIBS subsystem is at the same height as the position with the clearest image in the single-particle energetic material selection subsystem, and the laser focusing spot of the LIBS subsystem at the imaging position is smaller than the area of ​​the single-particle energetic material. The LIBS subsystem is used to emit lasers to excite single-particle energetic materials located at the imaging position. The timing controller is used to control the triggering time of the LIBS subsystem and the shock wave image acquisition subsystem, so that the LIBS subsystem is triggered to emit laser first, and the shock wave image acquisition subsystem acquires the shock wave image of the single particle energetic material after it is excited at different delays. The triggering time of the shock wave image acquisition subsystem is later than the triggering time of the LIBS subsystem, and the difference between the two triggering times is the delay, until the shock wave images of each type of single particle energetic material at all delay times are acquired. The optical processing subsystem is used to establish a linear prediction model of the shock wave characteristic velocity and detonation velocity of a single-particle energetic material based on shock wave images with different delays. The prediction method includes the following steps: S1: For each energetic material, two sets of shock wave images were acquired after excitation with different delays, including 15 different delays; S2: For each energetic material, the characteristic velocity vector of the shock wave in each of the two sets of shock wave images is obtained. Specifically, the method for obtaining the characteristic velocity vector of the shock wave in any set of shock wave images for each energetic material is as follows: S21: Obtain the position information of the shock wave front in the shock wave images of different delays in the current group for each energetic material; S22: Obtain the average value of the shock wave radius corresponding to each delay, where the... n The method for obtaining the average radius of the delayed shock wave is as follows: For the n The major axis radius is obtained by elliptic fitting of the position information of the delayed shock wave front. minor axis radius And the major axis radius Corresponding to the lateral propagation radius, minor axis radius Corresponding to the longitudinal propagation radius, and the major axis radius With minor axis radius The average radius is denoted as the shock wave radius. ; n =1,2,…,15; S23: Obtain the average propagation velocity of the shock wave between each pair of delays, where the... n The average propagation velocity of the shock wave between the nth delay and the (n+1)th delay The calculation method is as follows: in, For the first n +1 delay radius average, For the first n +1 delay, For the first n One delay; S24: Obtain the average transverse velocity and average longitudinal propagation velocity of the shock wave corresponding to each of the two delays, where the first... n The average transverse velocity of the shock wave between the nth delay and the (n+1)th delay and the average longitudinal propagation velocity of the shock wave The calculation method is as follows: in, For the first n The major axis radius corresponding to each delay For the first n The minor axis radius corresponding to each delay; S25: Obtain the shock wave propagation distance fitting curve corresponding to the current group of shock wave images using the Dewey model. Then, differentiate the shock wave propagation distance fitting curve at each of the 15 delays to obtain the shock wave propagation fitting velocity at each of the 15 delays. ; S26: Construct the shock wave characteristic velocity vector of the current group of shock wave images. as follows: in, The average propagation velocity of the shock wave with a 0.5 μs delay is obtained by dividing the radius of the shock wave with a 1 μs delay by 1 μs. ; S3: Construct the shock wave characteristic velocity matrix q of the current energetic material from the two sets of shock wave characteristic velocity vectors corresponding to different delays. S4: The characteristic velocity vector of the shock wave corresponding to each energetic material. and After averaging, the final shock wave characteristic velocity matrix Q for each energetic material is obtained. k k is the label for each energetic material, and the shock wave characteristic velocity matrix for all energetic materials is Q: Q= S5: By fitting the shock wave characteristic velocity matrix Q using a smart algorithm based on partial least squares regression, a linear prediction model for the shock wave characteristic velocity and detonation velocity of a single-particle energetic material is obtained. The method for obtaining the shock wave front of any shock wave image is as follows: Calculate the gradient value matrix of the shock wave image along the positive horizontal direction, negative horizontal direction, negative vertical direction, and positive vertical direction to obtain the gradient map in four directions; The gradient maps in each direction are smoothed and filtered separately. Then, based on the Laida criterion, significant noise points with gradient intensity fluctuations greater than a set multiple of the standard deviation are directly removed from the smoothed gradient maps by using a gradient intensity threshold, thus obtaining smoothed gradient maps in each direction. Opening and closing operations are performed on the smooth gradient maps in each direction. The opening operation is used to eliminate secondary noise points where the connectivity does not meet the set requirements, and the closing operation is used to eliminate pulse noise points inside the shock wave. Then, the smooth gradients after each opening and closing operation are denoised by a median filter to obtain the first denoised maps in each direction with noise removal. The first-order denoised gradient maps corresponding to the positive and negative gradient maps in the horizontal direction and the first-order denoised gradient maps corresponding to the positive and negative gradient maps in the vertical direction are stitched together to obtain a first-order stitched gradient map. The dividing line of the horizontal stitching is determined by the estimated value of the laser incident line position, and the dividing line of the vertical stitching is determined by the estimated value of the height position of the sample cover glass surface. Calculate the gradient value matrix of the first-order stitched gradient map along the positive horizontal direction, negative horizontal direction, negative vertical direction, and positive vertical direction to obtain the second-order gradient maps in four directions; The quadratic gradient maps in four directions are stitched together to obtain a quadratic stitched gradient map; The secondary denoising algorithm is used to perform secondary denoising on the secondary stitched gradient map, and the secondary denoised gradient map is used as the final shock wave front.

Images