-
本文中所用的样品为单基发射药、双基发射药、三基发射药和单晶冰糖,如表 1所示。单基发射药以石墨作为光泽剂,添加剂为2号中定剂(C15H16N2O1)。双基发射药的添加剂为钝感剂二氧化钛(TiO2)。单晶冰糖的成分为蔗糖,化学式为C12H22O11。本文中所用单基发射药及三基发射药为七孔粒状,双基发射药为管状。为了保证激光烧蚀面的平整,提高样品的统一度。对发射药样品进行了预处理,处理过程为:(a)沿轴线方向,并且避开孔结构对发射药进行纵切;(b)将纵切面在砂纸上轻轻打磨,使表面平整;(c)该纵切面作为激光烧蚀的平面;(d)每次实验进行前都用一发脉冲激光对样品表面杂质进行清理。
表 1 样品配方
Table 1. Sample formula
sample formula(mass fraction) single base propellant nitrocotton(NC) 0.97(nitrogen content 0.126),additive 0.03 double base propellant nitrocotton(NC) 0.55(nitrogen content 0.126),nitroglycerin(NG) 0.43,additive 0.02 triple base propellant nitrocotton(NC) 0.25(nitrogen content 0.126),nitroglycerin(NG) 0.30,Hexogen(RDX) 0.45 monocrystal rock sugar sucrose 1.00 -
基于脉冲激光作用的发射药反应行为及能量性能参数快速检测的实验平台如图 1所示。实验平台主要装置包括:激光器SGR-60,Beamtech(输出波长1064 nm,最大能量约3 J),输出激光与含能材料作用;激光器Dava-100,Beamtech(输出波长1064 nm,最大能量约100 mJ),用于搭建激光阴影诊断系统;增强型电荷耦合器件(intensified charge-coupled device,ICCD)iStar DH334T,Andor(像素1024×1024,最短曝光时间约2 ns);电荷耦合器件(charge-coupled device,CCD)EOS 750D,Canon(互补金属氧化物半导体图像感应器,约2420万pixel);激光能量计,表头NOVA Ⅱ,Ophir,探头PE50BF-DIF-C,Ophir(探测波长范围:0.19 μm~2.2 μm,能量范围:200 μJ~10 J,损伤阈值:4 J/cm2),用于监测激光能量。图中,L为平凸透镜;BS(beam splitter)为分束镜;R为反射镜;BE(beam expander)为激光扩束镜;EM(energy meter)为激光能量计;PD(photon detector)为光电探测器;OSC(oscilloscope)为示波器;AG(attenuator group)为能量衰减器。
脉冲激光(1064 nm)小部分能量被分光镜BS1反射至光电二极管用于时序监测,绝大部分能量透过分光镜BS1经衰减器衰减后再经第二分光镜BS2分光,部分光束(能量比例已知)反射至能量计进行能量检测,其余部分透过分光镜BS2后经透镜L1聚焦烧蚀样品,在样品表面形成等离子体并引起含能材料发生反应。等离子体的发光图像由ICCD相机搭配镜头直接测量。诊断激光束(1064 nm)由诊断激光器产生,经扩束后光斑覆盖待测区域,通过待测区域后经透镜聚焦,从而在装有1064 nm带通滤波片的CCD相机中成像,诊断激光束的传输同样需经过衰减和分光以确保安全和进行时序监测。
-
上述研究表明,较大激光光斑尺寸会引起更多质量发射药的反应,使得不同发射药在冲击波及等离子体特性上的差异更加显著。本节中在400 mJ/1474 μm激光参数条件下获得脉冲激光作用下3种发射药的冲击波传播距离,并结合支持向量机(support vector machines, SVM)机器学习方法,围绕单基、双基及三基发射药建立了脉冲激光作用下冲击波特征速度与火药力、爆热及爆温的定标关系。
-
基于上述实验诊断平台,在激光能量400 mJ、光斑直径1474 μm条件下获取单、双、三基发射药脉冲激光作用过程冲击波阴影图像,并同步监测每发激光的能量。基于不同时刻的激光阴影图像,采用高斯滤波、Canny边缘提取等提取出冲击波前沿位置,从而得到包含1 μs~20 μs共20个散点的冲击波传播距离-时间数据。利用最小二乘法对散点数据进行4次多项式拟合后得到冲击波传播距离变化曲线,对该曲线求导即可得到冲击波速度随时间的变化。将冲击波速度-时间拟合曲线在零时刻取点,可得到该发射药样品的冲击波特征速度。LASEM技术是基于该冲击波特征速度建立与含能材料火药力、爆热及爆温等的定标关系[14]。需要指出的是,冲击波特征速度虽然在数学推导意义上为冲击波的起始速度,但其并无实际的物理意义,仅作为描述脉冲激光作用过程快速变化量的参数。
图 6为3种发射药冲击波速度散点及其拟合曲线。由曲线y1、y2、y3的截距可得到特征冲击波速度:单基发射药(1296 m/s)、双基发射药(1256 m/s)、三基发射药(1492 m/s)。由图中误差棒分布可知,单基及三基发射药的冲击波速度抖动较小,而双基发射药在起始阶段(0 μs~5 μs)冲击波速度计算误差较大,从而对特征冲击波速度的截取造成影响。实际研究中,每次重复实验的脉冲激光能量存在微小抖动,对冲击波位置曲线及速度曲线的拟合存在显著影响。
为了尽量减小激光能量抖动对特征冲击波速度计算的影响,获取更为可靠、准确的定标基础数据,进一步基于每次实验对应激光能量参数的监测数据,通过SVM[20]的回归算法对获取的冲击波传播距离进行修正,修正流程如图 7所示。
流程中,通过固定时间参数t0为1 μs~20 μs来获取修正后的各时刻冲击波位置散点数据。激光能量参数E0取最大值与最小值的中间值,从而在算法上去除能量抖动影响。
基于修正后的冲击波传播距离数据计算得到特征冲击波,图 8为经激光能量参数修正后的3种发射药冲击波速度散点及其拟合曲线。由曲线y1、y2、y3的截距可得到修正后的特征冲击波速度:单基发射药(1047 m/s)、双基发射药(982.5 m/s)、三基发射药(1140 m/s)。
图 8 发射药冲击波速度及拟合曲线(修正后)
Figure 8. Shockwave velocity and fitting curve of propellants (after correction)
综上所述,在400 mJ/1474 μm激光条件下获得的单、双、三基发射药冲击波特征速度如表 2所示。其在一定程度上体现了发射药的能量性能,下面进一步将其作为关键参数,建立其与发射药火药力、爆热及爆温的实验定标关系。
表 2 发射药冲击波特征速度
Table 2. Shockwave characteristic velocity of propellants
propellant shockwave characteristic velocity/(m·s-1) shockwave characteristic velocity(after correction)/(m·s-1) single base 1296 1047 double base 1256 982.5 triple base 1492 1140 -
由单基发射药与单晶冰糖脉冲激光作用过程冲击波及等离子体特性研究可知,发射药脉冲激光作用冲击波速度与火药力、爆热及爆温的关联性在于:火药力衡量的气体做功能力、爆热及爆温衡量的能量释放及转化水平,对化学反应产物生成速率和气体运动速度产生影响,从而与脉冲激光作用过程的冲击波传播速度存在关联。本文中所采用的3种发射药的火药力、爆热及爆温参数如表 3所示,其中火药力、爆热为实测值,爆温为计算值,实测数据基于GJB770B-2005试验方法测得。
表 3 发射药性能参数
Table 3. Propellant property parameters
propellant explosive force/(kJ·kg-1) explosive heat/(kJ·kg-1) explosive temperature/K single base 1028 3876 2997 double base 931.0 3188 2580 triple base 1226 4812 3618 基于表 2和表 3数据,分别建立冲击波特征速度及修正的冲击波特征速度与发射药能量性能参数的线性定标模型,定标结果如图 9~图 11所示。未修正时,冲击波特征速度对于火药力、爆热及爆温的线性定标效果较好,决定系数R2分别为0.9718、0.9255、0.9380,但在95%的置信度下,都存在置信区间(confi-dence interval,CI)较大的问题。由图中红色点线可知,修正后的冲击波特征速度对于火药力、爆热及爆温的线性定标效果得到进一步提高,决定系数R2分别为0.9912、0.9998、0.9999,均优于未经激光能量修正的定标模型,并且在95%的置信度下,修正后的定标模型置信区间都有所减小,以爆热和爆温的定标关系最为显著。
表 4和表 5为3种发射药火药力、爆热、爆温参数的预测值及误差。可见,基于冲击波特征速度建立的线性定标模型,对发射药性能参数的预测值误差均小于7%,在误差允许的条件下可实现火药力、爆热及爆温参数的有效预测。并且由表 5可知,经过激光能量参数修正,定标模型对火药力的预测误差减小至不高于2%,对爆热、爆温的预测误差减小至不高于0.5%,模型预测准确度显著提高。
表 4 发射药性能参数预测
Table 4. Prediction of propellants property parameters
propellant explosive force/(kJ·kg-1) explosive heat/(kJ·kg-1) explosive temperature/K measurement prediction error measurement prediction error calculation prediction error single base 1028 1000 2.7% 3876 3636 6.2% 2997 2857 4.7% double base 931.0 954.0 2.5% 3188 3388 6.3% 2580 2697 4.5% triple base 1226 1231 0.4% 4812 4853 0.9% 3618 3642 0.7% 表 5 发射药性能参数预测(修正后)
Table 5. Prediction of propellants property parameters (after correction)
propellant explosive force/(kJ·kg-1) explosive heat/(kJ·kg-1) explosive temperature/K measurement prediction error measurement prediction error calculation prediction error single base 1028 1044 1.6% 3876 3863 0.3% 2997 3004 0.2% double base 931.0 921.0 1.1% 3188 3196 0.3% 2580 2576 0.2% triple base 1226 1219 0.6% 4812 4818 0.1% 3618 3615 0.1%
基于脉冲激光的发射药参数快速检测方法
Rapid detection method of propellant parameters based on pulsed laser
-
摘要: 为了验证基于脉冲激光作用的发射药能量性能参数快速测量方法的有效性,采用实验方法研究了脉冲激光作用下单、双、三基发射药和单晶冰糖的冲击波及等离子体特性,得到了激光参数对脉冲激光作用过程的影响规律以及单基发射药与单晶冰糖的冲击波、等离子体图像差异。结果表明,单基发射药的冲击波及等离子体膨胀特性主要受其化学反应释能的影响,而冰糖主要受激光辐照度的影响;采用支持向量机回归算法,建立了单、双、三基发射药激光作用下的冲击波特征速度与其火药力、爆热、爆温的线性定标模型,得到的决定系数R2值分别为0.9912、0.9998、0.9999。该方法对发射药火药力、爆热及爆温有较好的预测能力,为脉冲激光与含能材料作用的研究提供了参考。Abstract: In order to verify the applicability of the rapid measurement method of energy performance parameters of propellant based on pulsed laser ablation, a pulsed laser interaction platform with energetic materials and an experimental platform for transient diagnoses such as shockwave and plasma image were firstly established. Based on this platform, the shockwave and plasma characteristics of single, double, and triple base propellant and monocrystal rock sugar under pulsed laser ablation were studied experimentally. The influence of laser parameters on the ablation process of pulse laser and the difference of shockwave and plasma image between the single base propellant and monocrystal rock sugar was obtained. The results show that the shockwave and plasma expansion characteristics of single base propellant are mainly affected by the chemical reaction energy release of energetic substances, while that of rock sugar is mainly affected by laser irradiance. Furthermore, taking single, double, and triple base propellants as samples, based on shockwave propagation distance parameters and laser energy parameters, the linear calibration models of shockwave characteristic velocity and its explosive force, explosive heat, and explosive temperature were established by using support vector machines regression algorithm. The determinate coefficient R2 values were 0.9912, 0.9998, and 0.9999, respectively. The results show that this method predicts the explosive force, explosive heat, and explosive temperature of the propellant well and provides a reference for the study of the interaction between pulsed laser and energetic materials.
-
Key words:
- laser technique /
- rapid detection /
- pulsed laser ablation /
- propellant /
- shockwave
-
表 1 样品配方
Table 1. Sample formula
sample formula(mass fraction) single base propellant nitrocotton(NC) 0.97(nitrogen content 0.126),additive 0.03 double base propellant nitrocotton(NC) 0.55(nitrogen content 0.126),nitroglycerin(NG) 0.43,additive 0.02 triple base propellant nitrocotton(NC) 0.25(nitrogen content 0.126),nitroglycerin(NG) 0.30,Hexogen(RDX) 0.45 monocrystal rock sugar sucrose 1.00 表 2 发射药冲击波特征速度
Table 2. Shockwave characteristic velocity of propellants
propellant shockwave characteristic velocity/(m·s-1) shockwave characteristic velocity(after correction)/(m·s-1) single base 1296 1047 double base 1256 982.5 triple base 1492 1140 表 3 发射药性能参数
Table 3. Propellant property parameters
propellant explosive force/(kJ·kg-1) explosive heat/(kJ·kg-1) explosive temperature/K single base 1028 3876 2997 double base 931.0 3188 2580 triple base 1226 4812 3618 表 4 发射药性能参数预测
Table 4. Prediction of propellants property parameters
propellant explosive force/(kJ·kg-1) explosive heat/(kJ·kg-1) explosive temperature/K measurement prediction error measurement prediction error calculation prediction error single base 1028 1000 2.7% 3876 3636 6.2% 2997 2857 4.7% double base 931.0 954.0 2.5% 3188 3388 6.3% 2580 2697 4.5% triple base 1226 1231 0.4% 4812 4853 0.9% 3618 3642 0.7% 表 5 发射药性能参数预测(修正后)
Table 5. Prediction of propellants property parameters (after correction)
propellant explosive force/(kJ·kg-1) explosive heat/(kJ·kg-1) explosive temperature/K measurement prediction error measurement prediction error calculation prediction error single base 1028 1044 1.6% 3876 3863 0.3% 2997 3004 0.2% double base 931.0 921.0 1.1% 3188 3196 0.3% 2580 2576 0.2% triple base 1226 1219 0.6% 4812 4818 0.1% 3618 3615 0.1% -
[1] 李强, 闫光虎, 张玉成, 等. NG含量对改性单基发射药燃烧性能的影响[J]. 火炸药学报, 2012, 35(1): 73-76. LI Q, YAN G H, ZHANG Y Ch, et al. Effect of NG content on burning performance of modified sing-base gun propellant[J]. Chinese Journal of Explosives & Propellants, 2012, 35(1): 73-76(in Chin-ese). [2] THOMAS J C, MORROW G R, DILLIER C A M, et al. Comprehensive study of ammonium perchlorate particle size/concentration effects on propellant combustion[J]. Journal of Propulsion and Power, 2019, 36(1): 1-6. [3] 郝海霞, 裴庆, 赵凤起, 等. 固体推进剂激光点火性能研究综述[J]. 含能材料, 2009, 17(4): 491-498. HAO H X, PEI Q, ZHAO F Q, et al. Summarization of laser ignition characteristics of solid propellant[J]. Chinese Journal of Energetic Materials, 2009, 17(4): 491-498 (in Chinese). [4] 张陆, 王霆威, 王晓军, 等. 激光敏感型含能配合物类起爆药研究进展[J]. 含能材料, 2022, 30(4): 385-395. ZHANG L, WANG T W, WANG X J, et al. Review on laser sensitive energetic complex primary explosives[J]. Chinese Journal of Energetic Materials, 2022, 30(4): 385-395 (in Chinese). [5] 刘彦汝, 孙杰, 金波, 等. 360 nm紫外激光辐照下HMX晶体的微观结构变化[J]. 含能材料, 2021, 29(12): 1208-1215. LIU Y R, SUN J, JIN B, et al. Microstructure changes of HMX crystals irradiated by 360 nm UV laser[J]. Chinese Journal of Energetic Materials, 2021, 29(12): 1208-1215 (in Chinese). [6] 伍俊英, 刘嘉锡, 杨利军, 等. 不同频率飞秒激光脉冲序列加工炸药过程安全性的数值计算[J]. 含能材料, 2021, 29(3): 192-201. WU J Y, LIU J X, YANG L J, et al. Numerical calculation of the safety of processing explosives with femtosecond laser sequence with different frequencies[J]. Chinese Journal of Energetic Materials, 2021, 29(3): 192-201. [7] GOTTFRIED J L. Laser-induced plasma chemistry of the explosive RDX with various metallic nanoparticles[J]. Applied Optics, 2012, 51(7): B13-B21. doi: 10.1364/AO.51.000B13 [8] KIMBLIN C, TRAINHAM R, CAPELLE G A, et al. Characterization of laser-induced plasmas as a complement to high-explosive large-scale detonations[J]. AIP Advances, 2017, 7(9): 095208. doi: 10.1063/1.4999793 [9] HAUER M, FUNK D J, LIPPERT T, et al. Time resolved study of the laser ablation induced shockwave[J]. Thin Solid Films, 2004, 453/454(4): 584-588. [10] LUSSELL F C D, HARMON R S, MCNESBY K L, et al. Laser-induced breakdown spectroscopy analysis of energetic materials[J]. Applied Optics, 2003, 42(30): 6148-6152. doi: 10.1364/AO.42.006148 [11] ZHANG Zh, WANG A, WU J, et al. Spatial confinement effects of bubbles produced by laser ablation in liquids[J]. AIP Advances, 2019, 9(12): 125048. doi: 10.1063/1.5127261 [12] 葛一凡, 陆旭, 刘玉柱. 基于激光诱导击穿光谱和神经网络的蛋壳研究[J]. 激光技术, 2022, 46(4): 532-537. GE Y F, LU X, LIU Y Zh. Research on eggshell via laser-induced breakdown spectroscopy and neural network[J]. Laser Technology, 2022, 46(4): 532-537(in Chinese). [13] GOTTFRIED J L. Influence of exothermic chemical reactions on laser-induced shock waves[J]. Physical Chemistry Chemical Physics, 2014, 16(39): 21452-21466. doi: 10.1039/C4CP02903H [14] GOTTFRIED J L. Laboratory-scale method for estimating explosive performance from laser-induced shock waves[J]. Propellants, Explosives, Pyrotechnics, 2015, 40(5): 674-681. doi: 10.1002/prep.201400302 [15] COLLINS E S, GOTTFRIED J L. Laser-induced deflagration for the characterization of energetic materials[J]. Propellants, Explosives, Pyrotechnics, 2017, 42(6): 592-602. doi: 10.1002/prep.201700040 [16] KALAM S A, MURTHY N L, MATHI P, et al. Correlation of molecular, atomic emissions with detonation parameters in femtosecond and nanosecond LIBS plasma of high energy materials[J]. Journal of Analytical Atomic Spectrometry, 2017, 32(8): 1535-1546. doi: 10.1039/C7JA00136C [17] BISS M M, BROWN K E, TILGER C F. Ultra-high fidelity laser-induced air shock from energetic materials[J]. Propellants, Explosives, Pyrotechnics, 2020, 45(3): 396-405. doi: 10.1002/prep.201900130 [18] 王茜蒨, 赵宇, 卢小刚, 等. 激光诱导击穿光谱与拉曼光谱技术在危险物检测中的研究进展[J]. 光谱学与光谱分析, 2017, 37(8): 2430-2434. WANG X Q, ZHAO Y, LU X G, et al. Progress in laser induced breakdown spectroscopy and Raman spectroscopy for hazardous material detection[J]. Spectroscopy and Spectral Analysis, 2017, 37(8): 2430-2434 (in Chinese). [19] 郭文灿, 郑贤旭, 张旭, 等. 含铝炸药在激光烧蚀下的发射光谱分布及瞬态温度测量[J]. 含能材料, 2018, 26(8): 671-676. GUO W C, ZHENG X X, ZHANG X, et al. Emission spectrum distribution and transient temperature measurement of aluminized explosives under laser ablation[J]. Chinese Journal of Energetic Materials, 2018, 26(8): 671-676 (in Chinese). [20] VAPNIK V. The nature of statistical learning theory[M]. Berlin, Germany: Springer, 2000: 15-174.