Session Q3: ID-7: Dynamic Properties of Steels

Spall Experiments On Stainless Steel 21-6-9 Varying Pulse Lengths and Longitudinal Stress

The material addressed in this research is 21-6-9, a 200 series stainless steel alloy which has so far found applications in aviation, demolition, motor-vehicle design and nuclear reactor containment. The shock Hugoniot, HEL and spall strength of this material has been determined in gas-gun driven uni-axial strain experiments. The spall signature has been determined in experiments varying input pulse lengths and maximum stress. The material shows a decrease of spall strength with pulse duration and an increase with longitudinal stress. Post loading microstructure analysis has revealed both brittle and ductile failure and a change from dislocation to twinning deformation mechanisms at a stress between 2.5 and 6.5 GPa.   

Ductile damage evolution assessment in high purity copper and stainless steel subjected to different shock-loading profiles using cohesive modeling

At continuum level, the triaxiality of the stress state has the major effect to reduce the ductility that the material can exhibit. As a result of this, ductile metals under uniaxial strain loading condition, such as that experienced in flyer plate impact experiment, should show a very limited ductility that is often in contrast with the energy dissipation estimated from the pull-back signal amplitude. In this paper, it is proposed that the developing of damage in form of voids causes a dramatic drop of the stress triaxiality giving the material the ability to sustain much larger deformation and consequent plastic work dissipation. Thus, a numerical model using cohesive elements has been developed to account for this dissipation in the energy balance. The model has been used to investigate the damage evolution and plastic strain accumulation in plate impact experiments with different loading profiles, for both high purity copper and stainless steel, in order to achieve a better comprehension of the processes related to spall fracture. Comparisons of numerical results with experimental data, available in term of both velocity profiles and optical metallography of soft-recovered samples, seem to confirm the capability of the proposed model to correctly capture distinctive features of spall event.   

Experimental study and structural scaling analysis of damage-failure transition in shocked Armco-iron

Ballistic set-up (125mm gas gun) coupled with VISAR registration system has bee developed for soft recovery plate impact test to investigate defect induced structure evolution and damage failure transition. Experiments were carried out at different impactor velocity and thickness of Armco-iron and Vanadium specimens. 3D New View profilometry and AFM study were used for structure investigation and roughness correlation technique was developed. It was found that for material far from spall surface the Hurst exponent is close to 0.5 which corresponds weakly correlated defects. In vicinity of spall surface the Hurst exponent is 0.6-0.7. More correlated behavior along several spatial scales can be considered as precursor of damage-failure transition.   

Vortex Structures in the Shock-deformed Armor Steels

Several kinds of armor steel were tested under uniaxial strain conditions within impact velocity range from 250 to 400 m/s. Using optical and REM microscopy, the post shocked specimens were studied to reveal the kinematical mechanisms of dynamic deformation at the mesoscale. In all the specimens, the cross-section of specimens was found to be filled with rotational cells of very complex space morphology. Each rotation cell consists of central core of 1-2 $\mu $m in diameter and family of petals surrounding the core, so the space configuration of eddy is closely remands a fan of total size 6-7 $\mu $m. During the deformation, the petals move around the core providing the vortical motion of rotation as a whole. Dependence of rotational cell density on the strain rate changes non-monotonously, maximum density corresponds to maximum macrohardness and maximum of spall-strength of steel.   

Computational Comparisons of Homogeneous and Statistical Descriptions of AeroMet100 Steel Subjected to Explosive Loading

Computational continuum codes can provide many details on the response of metals to explosive loading. However, most ``production'' level calculations use a homogeneous description of the metal. This representation is incorrect and a statistically compensated Johnson-Cook fracture model has been implemented into several computational continuum codes in an attempt to improve the predictive capability for fragment distributions from explosively loaded shells. This analysis employs a distribution of failure strains from experimental data to determine the parameters for the statistically compensated Johnson-Cook fracture model. Then the parameterized model was used in calculations of several experiments where fragments from explosively loaded shells were soft recovered. The results were analyzed and compared back to baseline homogeneous calculations.   

Identification of Two Constitutive Models for a Low Alloy Structural Steel S355K2G3

An experimental characterization of the dynamic response of a low strength structural steel (S355K2G3) has been investigated using various experimental techniques performed at THIOT INGENIERIE impact shock physics test facility: Taylor impacts, hat shaped shear tests, dynamic tensile tests and plate impact experiments. This paper presents simulations of these experiments performed in a wide range of strain-rate conditions. The objective is to simulate impact, blast and explosion on a whole structure. Impacts generate huge gradients of stress and strain because of shock waves propagation and interaction on free surfaces. Blast and explosion have a tendency to induce large strains. Based on the overall experiments, it is proposed to identify two classical constitutive models (Johnson-Cook and Zerilli-Amstrong) integrated in two non-linear explicit finite element hydrocodes ANSYS-AUTODYN and LS-DYNA. Both hydrocodes take into account various modes of fracture from single conventional fracture thresholds such as critical strain value or hydro tensile failure model to more complex modes such as Johnson Holmquist damage failure model. Choice between models is discussed with respect to material nature and to solicitations the material is subjected to.   

Experimental Study on Dynamic Mechanical Properties of 30CrMnSiNi2A Steel.

Under dynamic conditions, the strain-rate dependence of material response and high levels of hydrostatic pressure cause the material behavior to be significantly different from what is observed under quasi-static condition. The curves of stress and strain of 30CrMnSiNi2A steel in different strain rates are obtained with SHPB experiments. Metallographic analyses show that 30CrMnSiNi2A steel is sensitive to strain rate, and dynamic compression leads to shear failure with the angle 45$^{\circ}$ as the small carbide which precipitates around grain boundary changes the properties of 30CrMnSiNi2A steel. From the SHPB experiments and quasi-static results, the incomplete Johnson-Cook model has been obtained: $\sigma =[1587+382.5({\bar {\varepsilon}^p)}^{0.245}][1+0.017\ln \dot {\varepsilon }^\ast ]$, which can offer parameters for theory application and numerical simulation.   

A combined technique for measuring Hugoniot and interfacial temperature of preheated metals

A convenient method was developed to perform a combined shock-Hugoniot and interfacial temperature measurement of metals over initial temperature range of 300-1000 K. Experimental details in our investigation are described of (i) a resistive heater placed around the metal specimen to generate a controllable, stable heating source and (ii) a fiber-optic probe with an optical lens coupling system and thirteen 62.5 $\mu $m diameter silica fibers to carry out non-contact measurements for shock velocities and interfacial thermal temperatures of preheated metal. Using shock experimental results of tantalum initially heated to 773 K, a best linear fit of shock velocity to particle velocity gives the coefficients of $U_{S}$(km/s)=1.540(km/s) +1.883$U_{p}$ (km/s) between 248 and 307 GPa, and is obviously 3{\%}-5{\%} below Hugoniot measurements from a room temperature initial state. And obtained interfacial temperatures are in agreement with theoretical calculations of pressure-temperature (P-T) curve. It is indicated that our method is practicable for measuring Hugoniot and shock temperature of preheated metal, which could provide an important approach for studying the temperature effect of shocked metals.