爆轰加载下金属柱壳膨胀破裂过程研究

采用前照明的高速分幅照相对不同壁厚的45钢柱壳在爆轰加载下的膨胀破裂过程进行了研究,实验成功拍摄到45钢柱壳前表面裂纹生成、扩展及产物泄漏的过程图像。研究表明:膨胀应变率在104s-1附近时,随着应变率的增加,45钢柱壳的各特征应变均有所增加,断裂模式由拉剪混合向单一剪切转变;45钢柱壳的膨胀破裂有一个过程,从柱壳的外壁面可以观察到裂纹生成至产物开始泄漏经历时间历程的长短与柱壳壁厚相关,柱壳壁越厚,从外壁出现裂纹到产物开始泄漏的时间间隔越长;由回收破片厚度测量推算出的膨胀断裂应变与动态实验过程中产物开始泄漏时的断裂应变基本一致。     

约束层对金属柱壳膨胀变形影响的数值模拟

对金属铝和硬质聚氨酯泡沫组合的约束层对45钢柱壳膨胀断裂性能的影响进行了数值模拟,得到三种约束条件下45钢柱壳在膨胀过程中表面的速度和位移历史剖面,对比分析了在约束层作用下45钢柱壳膨胀变形动态行为。利用高速分幅相机及中能X光机获得了多个时刻45钢柱壳的动态图像和柱壳膨胀后的工程应变,实验结果与数值模拟结果吻合。结果可对爆轰加载下约束层对45钢柱壳膨胀变形的物理规律进行定量认识。     

不同热处理条件下45钢柱壳的动态性能

以不同热处理条件下的45钢作为研究对象,利用高速摄影和金相分析来研究45钢柱壳在爆轰加载下的膨胀断裂特性,以及不同热处理条件对材料的动态断裂性能的影响。结果表明:随着回火温度的升高,45钢强度减小、静态延伸率增大,爆轰加载时柱壳表面裂纹产生及发生贯穿断裂而导致产物泄漏的时刻逐渐推迟,相应应变逐渐增加,同时回收到的破片尺寸逐渐增大;四种45钢的静态力学性能及动态断裂性能有明显差异,是因为不同热处理条件下,材料细观组织的结构、组成均发生较大变化,即材料的细观组织决定了材料的静态力学性能及动态断裂性能。     

线起爆膨胀柱壳实验加载及诊断技术

基于电爆炸丝引爆炸药、继而驱动尼龙对金属柱壳进行碰撞加载的方式,在金属柱壳中部、半柱高范围内实现了一维柱面膨胀加载。同时,基于沿轴向的加载（或径向速度）一致性和沿环向的加载（或径向速度）轴对称性,提出了一维柱面加载的有效性判据。相比于滑移爆轰加载,一维柱面加载方式具有应力状态相对简单、易简化为二维轴对称问题分析的优点。在柱壳断裂诊断方面,建立了分布式表面速度诊断方法来监测柱壳圆周范围内的初始断裂。其原理为:均匀承载壳体断裂引起的局部承载失效将导致均匀速度曲线簇出现分叉（或演化趋势变化）。与高速分幅照相诊断方法相比,分布式表面速度诊断方法可准确获取柱壳圆周范围内的初始断裂信息（含断裂时刻和断裂位置）。利用建立的线起爆膨胀柱壳实验加载和诊断技术,获得了304钢和45钢柱壳的一维柱面动态拉伸初始断裂性能数据（含断裂应变、平均应变率）,其中,45钢柱壳的断裂应变（或延展性）低于304钢柱壳的。     

20#钢动态拉伸断裂行为及其临界损伤度研究

首先对一维应变平面冲击载荷下20#钢的动态拉伸断裂行为进行了实验研究,考察了加载应力和拉伸应变率影响;然后基于损伤度函数模型,对实验进行了数值模拟研究,结果表明：模型中定义的特征物理量一断裂临界损伤度与外载荷条件（加载应力和拉伸应变率）的无关性或不敏感性;最后,采用Ls-dyna动力有限元程序并嵌入损伤度函数模型,对滑移爆轰下20#钢柱壳内爆驱动层裂进行二维数值模拟研究,结果表明：一维应变平面层裂实验和模型计算确定的损伤模型参数,也适用于描述柱壳构形复杂应力环境下的动态拉伸断裂问题．从而初步证实了断裂临界损伤度可以看作是一个表征延性金属动态拉伸断裂的内禀物性参数．     

爆轰加载下柱壳断裂时刻微小差异测量方法

对于断裂时刻差异较小的对比实验,提出一种判读方法:同时采用高速摄影和干涉测速,利用高速摄影判读壳体断裂时刻应变,利用干涉测速获取壳体位移及应变曲线,两者结合得出较为精确的壳体断裂时刻差异。利用该方法得出45钢柱壳在JO-9159和JOB-9003两种炸药加载下断裂时间相差0.45 μs,钨合金柱壳在两种炸药加载下的断裂时间相差0.39 μs。同时该方法可以推广应用于单发壳体膨胀断裂实验中,更精确测定壳体的断裂时刻。     

45钢柱壳爆炸膨胀断裂的SPH模拟分析

金属柱壳爆炸膨胀断裂存在拉伸、剪切及拉剪混合等多种断裂模式,目前其物理机制及影响因素还不清晰。本文中采用光滑粒子流体动力学方法（smoothed particle hydrodynamics, SPH）对45钢柱壳在JOB-9003及RHT-901不同装药条件下的外爆实验进行了数值模拟,探讨柱壳在不同装药条件下发生的剪切断裂、拉剪混合断裂模式及其演化过程,模拟结果与实验结果一致。SPH数值模拟结果表明:在爆炸加载阶段,随着冲击波在柱壳内、外壁间来回反射形成二次塑性区,沿柱壳壁厚等效塑性应变演化呈凸形分布,壁厚中部区域等效塑性应变较内、外壁大;在较高爆炸压力（JOB-9003）作用下,柱壳断裂发生在爆轰波加载阶段,损伤裂纹从塑性应变积累较大的壁厚中部开始沿剪切方向向内、外壁扩展,形成剪切型断裂模式;而在RHT-901空心炸药加载下,虽然裂纹仍从壁厚中部开始沿剪切方向扩展,但随后柱壳进入自由膨胀阶段,未断区域处于拉伸应力状态,柱壳局部发生结构失稳,形成类似“颈缩”现象,裂纹从剪切方向转向沿颈缩区向外扩展,呈现拉剪混合断裂模式。拉伸裂纹占截面的比例与柱壳结构失稳时刻相关。可见,柱壳断裂演化是一个爆炸冲击波与柱壳结构相互作用的过程,不能简单将其作为一系列膨胀拉伸环处理。     

45钢薄壁圆柱管的冲击膨胀断裂机理研究

采用改进型霍普金森压杆实验技术,对不同膨胀断裂状态的45号钢薄壁金属圆柱管进行了冻结回收,直接观测了薄壁金属圆柱管动态膨胀断裂过程中的裂纹萌生、扩展情况以及最终断裂模式等断裂演化特征。对冻结回收样品进行的金相显微分析完整观察到了裂纹萌生、扩展直至断裂的整个膨胀断裂过程,并得出以下结论:薄壁金属圆柱管在中应变率的膨胀断裂过程中,拉伸和剪切断裂机制起主导作用。裂纹萌生于外壁面,并由外向内扩展,断裂模式随加载应变率的提高逐渐由拉剪混合向纯剪切过渡。与爆轰加载的高应变率薄壁金属圆柱管断裂过程不同的是,随加载载荷的增加,薄壁金属圆柱管的断裂逐渐由拉伸断裂向剪切断裂过渡,而非绝热剪切断裂,这种差异的产生原因尚待研究。     

爆轰加载下金属壳体膨胀断裂应变的测定方法

简要介绍了爆轰加载下金属柱壳膨胀断裂应变的传统测试方法,探索性地提出了由晶粒尺寸变化推算断裂应变的细观方法和利用锰铜计测试破片碰靶压力脉宽推算断裂应变的压力脉宽法。对几种测试方法进行分析比较,指出各方法的优点及局限性,提出利用速度历史来确定壳体断裂应变的测定方法。     

爆炸与电磁加载下无氧铜环、柱壳的断裂模式转变

考虑断面收缩率、局域断裂应变以及平均断裂应变,并基于电磁膨胀环、爆炸膨胀环(柱壳)实验平台,研究了高纯无氧铜(TU1)环及柱壳在高应变率载荷下的膨胀断裂行为。采用高速摄影技术拍摄柱壳外壁的膨胀断裂形貌演化过程,用于确定柱壳平均断裂应变;利用激光干涉测速技术获得样品径向膨胀速度历史,用于确定加载应变率;利用样品的全回收测量及微观表征,确定了无氧铜环、柱壳的局域断裂应变及断裂模式。实验表明,随着应变率的增加,TU1材料的平均断裂应变增加,断面的收缩程度加剧,并在应变率约为1.0×104 s-1附近会出现明显的断裂模式转变,断面收缩率出现量级上的跳跃,从100变化至约103,局域断裂应变呈现明显的分区现象。     

基于SHPB装置的膨胀圆柱管实验技术

采用改进的SHPB实验装置对45钢薄壁金属圆柱管进行了膨胀断裂加载,完成了不同变形程度 (覆盖圆柱管整个变形及断裂过程)圆柱管的冻结回收实验。回收样品可用于断裂机理研究分析,通过数值模 拟辅助分析和实验数据拟合得到了圆柱管凸起处的径向应变、应变率和环向拉伸应力。通过在圆柱管端粘贴 应变片监测断裂信号的方式,准确判断了圆柱管凸起处发生断裂的时间,以及径向断裂应变、应变率和环向拉 伸断裂应力,在102~104s-1应变率范围内,SHPB实验装置可用于研究金属圆柱管膨胀断裂。     

Fracture behavior of explosively loaded spherical molded steel shells

Experimental and numerical works are made to study the fracture of explosively loaded spherical molded steel shells. The first series of experiments included three sawdust recovery shots to save fragments for examination. In this series, detonation was initiated from the center of the sphere. Results of the experiments show that two types of fractures are observed in spherical shells: radial and shear as in cylindrical shells. Spall fracture is also observed in spherical shells. To assist understanding of the experimental results, a computer simulation of expanding shells is performed to provide information on the stress, strain, strain rate and position of each element of the shell wall as a function of time after detonation. For t=7.5 μs after detonation, triaxial non-uniform strain prevails in the middle of the thickness of the wall. The peak of the stress equals to 6.5 GPa and exceeds the spall strength of carbon steel. In the second series of experiments, spall fracture is suppressed by displacing the point of detonation initiation from the center to periphery of HE charge.     

超强钢18NiC250在不同加载速率下的断裂韧性

为了揭示国产超强钢18NiC250的强度、断裂韧性随加载速率的变化规律,利用电子万能试验机和Hopkinson压杆,测试其在0.001～2 000 s-1的塑性流动应力应变曲线及在10-1~106 MPam1/2/s的断裂韧性,同时对断裂破坏机理进行了微观分析。结果表明:该材料的强度对加载速率不敏感,即流动应力基本保持在1.9 GPa;而断裂韧性很敏感,当加载速率由10-1 MPam1/2/s增大到106 MPam1/2/s时,断裂韧性降低了38.2%,断裂模式由韧窝断裂转变为解理断裂。     

Brittle fracture of precompressed steel as affected by hydrostatic pressure,temperature and strain concentration

Highly precompressed 1020 HR steel, 0.65 prestrain at 400°F (204°C), tested in nominally uniform tension at ?80°F (?62°C) fractures at about 110,000 psi (760 MN/m²) with less than 0.02 plastic strain. Yet the addition of a hydrostatic pressure of less than 7000 psi (48 MN/m²) converts this visually brittle fracture to a ductile one with appreciable necking. The explanation of this surprising experimental result is shown to follow directly and simply from the combination of a tensile stress criterion of fracture, strain concentration and the low tangent modulus of the stress-strain curve in tension beyond the Bauschinger transition region of a few percent of plastic strain. Temperature dependence and strain-rate dependence of brittle fracture similarly are predictable in an almost trivial manner from the appropriate stress-strain curves for different amounts of precompression. So also is the amazingly high ductility or fracture toughness of the most complex of perforated or notched statically loaded structures of mild steel in an undamaged or fully annealed state.     

Constitutive modeling of the dynamic fracture of smooth tensile bars

A recently developed viscoplastic-damage type of constitutive theory for high strain-rate flow processes and ductile fracture is used to model the deformation and fracture of dynamically loaded smooth cylindrical tensile bars. The analysis assumes polycrystalline materials which usually contain microvoids with an average density of the order of 10⁶ per cm³ that are dispersed homogeneously throughout. It is shown that for dynamically imposed loading that produce nominal strain rates ranging between 5 × 10² − 5 × 10³ sec ⁻¹, the inhomogeneous fields of stress and deformation caused by wave propagation and wave reflection induce necking at different locations along the gauge section, depending upon the strain-rate imposed. This occurs without imposition of any geometrical or material irregularity to preposition the location of the necking. The imposed rate of strain is also shown to affect the magnitude of the strain at which necking initiates, as well as the strain required for fracture.     

Strain-rate history and temperature effects on the torsional-shear behavior of a mild steel

Previous investigations on the effects of strain-rate and temperature histories on the mechanical behavior of steel are briefly reviewed. A study is presented on the influence of strain rate and strain-rate history on the shear behavior of a mild steel, over a wide range of temperature Experiments were performed on thin-walled tubular specimens of short gage length, using a torsional split-Hopkinson-bar apparatus adapted to permit quasi-static as well as dynamic straining at different temperatures. The constant-rate behavior was first measured at nominal strain rates of 10^?3 and 10³ s^?1 for ?150, ?100, ?50, 20, 200 and 400°C. Tests were then carried out, at the same temperatures, in which the strain rate was suddenly increased during deformation from the lower to the higher rate at various large values of plastic strain. The increase in rate occurred in a time of the order of 20 μs so that relatively little change of strain took place during the jump. The low strain-rate results show a well-defined elastic limit but no yield drop, a small yield plateau is found at room temperature. The subsequent strain hardening shows a maximum at 200°C, when serrated flow occurs and the ductility is reduced. The high strain-rate results show a considerable drop of stress at yield. The post-yield flow stress decreases steadily with increasing temperature, throughout the temperature range investigated. At room temperature and below, the strain-hardening rate becomes negative at large strains. The adiabatic temperature rise in the dynamic tests was computed on the assumption that the plastic work is entirely converted to heat. This enabled the isothermal dynamic stress-strain curves to be calculated, and showed that considerable thermal softening took place. The initial response to a strain-rate jump is approximately elastic, and has a magnitude which increases with decrease of testing temperature; it is little affected by the amount of prestrain. At 200 and 400° C, a yield drop occurs after the initial stress increment. The post-jump flow stress is always greater than that for the same strain in a constant-rate dynamic test, the strain-hardening rate becoming negative at large strains or low testing temperature. This observed effect of strain-rate history cannot be explained by the thermal softening accompanying dynamic deformation. These and other results concerning total ductility under various strain-rate and temperature conditions show that strain-rate history strongly affects the mechanical behavior of the mild steel tested and, hence, should be taken into account in the formulation of constitutive equations for that material.     

High-strain-rate compressive behavior of a rigid polyurethane foam with various densities

The dynamic compressive stress-strain behavior of a rigid polyurethane foam with four values of density (78, 154, 299, and 445 kg/m³) has been determined in the strain-rate range of 1000–5000 s⁻¹. A pulse shaping technique was used with a split Hopkinson pressure bar to ensure homogeneous deformation in the foam specimens under dynamic compression. Dynamic stress equilibrium in the specimen was monitored during each experiment using piezoelectric force transducers mounted close to the specimen end-faces. Quasi-static experiments were also performed to demonstrate rate effects. Experimental results show that both the quasistatic and the dynamic stress-strain curves of the foam exhibit linear elasticity at small strains until a peak is reached. After the peak, the stress-strain curves have a plateau region followed by a densification region. The peak stress is strain-rate sensitive and depends on the square of the foam density.     

Chain scission in transient extensional flow kinetics and molecular weight dependence

Recent experimental investigations have shown that large macromolecules can be fully stretched and fractured in an extensional flow. In this situation, the critical strain-rate for bond scission was found to depend on molecular weight as (M_W)⁻², in agreement with the theoretical predictions of the bead---rod model. One of the conditions prerequisite to full chain extension is that the residence time at the appropriate strain-rate must be much larger than the terminal relaxation time of the macromolecule. If this requirement is not fulfilled, chain fracture could still occur at sufficiently high strain-rate, but in a partially uncoiled state. In the present studies we have measured the critical strain-rate for chain scission in transient extensional flow of extremely sharp PS fractions dissolved in dekalin at the θ-temperature. The molecular weight range investigated varied from 2.86 × 10⁶ to 426,000. The critical strain-rate for chain scission was found experimentally to scale as (M_W)^−0.95 instead of (M_W)⁻² as predicted for stagnant extensional flow. Our results are in good accord with a recent theory for rupture of partly extended coils. Even in the partially uncoiled state, the degraded macromolecules showed a remarkable propensity for chain halving, indicating that midchain scission in flow is a general property that is not uniquely reserved to the fully extended chain.