Yehuda Partom

MODELING FAILURE WAVES IN GLASS

Failure waves in glass were first observed in planar impact tests in 1991. Since then they have been further investigated for various glasses and at different laboratories. Kanel et al. introduced a failure wave in a wave code by assuming its existence and propagation from the outset. We propose a model in which the failure wave forms and propagates as part of the overall dynamics. Our model includes two main components: • An equation that relates damage accumulation rate to the damage parameter gradient. • Degradation of the shear modulus as damage accumulates. We incorporated the model in a special purpose 1D plane Lagrange wave code and reproduced data obtained at the Cambridge Impact Laboratory for soda-lime and borosilicate glasses. The agreement we get is quite good.

Porous compaction as the mechanism causing the Hugoniot Elastic Limit

Abstract Unlike in metals, stress and particle velocity histories of shock waves in ceramic materials show a typical ramping above the Hugoniot Elastic Limit (HEL). Under the assumption that the HEL signifies the beginning of failure of the material, this ramping has been described using a viscoplastic model. However, there is ample experimental evidence to rule out plastic flow at the HEL-level of stress. Alternatively, moduli degradation can be used to model the ramping, but it seems improbable that the moduli would degrade significantly under increasing pressure. The proposed micro-mechanical mechanism for the HEL, and for the ramping beyond the HEL, is based on the process of porous compaction due to pressure above a threshold pressure pcrush. Although the pore volume of high-grade ceramics is quite small, it is sufficient to cause the ramping observed in Hugoniot measurements from the seventies. Additional evidence for the effect of porosity on the HEL stress has been given in recent years. Here, this experimental evidence is supported by simulations using a simple model for porous compaction. The simulations suggest that porous compaction is the main micro-mechanical mechanism causing the ramping and other features related to the HEL.

Revisiting Shock Initiation Modeling of Homogeneous Explosives

Shock initiation of homogeneous explosives has been a subject of research since the 1960s, with neat and sensitized nitromethane as the main materials for experiments. A shock initiation model of homogeneous explosives was established in the early 1960s. It involves a thermal explosion event at the shock entrance boundary, which develops into a superdetonation that overtakes the initial shock. In recent years, Sheffield and his group, using accurate experimental tools, were able to observe details of buildup of the superdetonation. There are many papers on modeling shock initiation of heterogeneous explosives, but there are only a few papers on modeling shock initiation of homogeneous explosives. In this article, bulk reaction reactive flow equations are used to model homogeneous shock initiation in an attempt to reproduce experimental data of Sheffield and his group. It was possible to reproduce the main features of the shock initiation process, including thermal explosion, superdetonation, input shock overtake, overdriven detonation after overtake, and the beginning of decay toward Chapman-Jouget (CJ) detonation. The time to overtake (TTO) as function of input pressure was also calculated and compared to the experimental TTO.

Detonation Products EOS by Specifying Gamma (V) for the Principal Isentrope

The standard way of defining an equation of state (EOS) for detonation products is (1) choose a function Ps(V) for the pressure along the principal isentrope, with enough adjustable parameters; (2) integrate it to obtain the internal energy Es(V); (3) determine the parameters from available data (Chapman Jouget (CJ) state and cylinder expansion test); (4) refer a Gruneisen EOS to this principal isentrope. Using this approach, (1) most of the adjustable parameters have no physical meaning; (2) they are determined simultaneously; and (3) changing one of them requires changing the others. Instead, we define the principal isentrope by choosing a function for the adiabatic gamma γs(V). We show that this has the following advantages over the standard approach: (1) the parameters have physical meaning; (2) they can be determined by a recursive process; (3) the influence of changes in the parameters to cylinder expansion results is obvious.

The V1V2 EOS for Detonation Products

Many equations of state (EOS) for detonation products have been proposed and used. Some of them are in analytical form and some in tabular form. The most popular is the Jones-Wilkins-Lee (JWL) EOS. One of the main parameters of a products EOS is the so-called adiabatic gamma along its main isentrope (γs). For JWL EOSs γs(V) varies in a nonmonotonic way. Going down from the CJ point along the main isentrope, it first increases to create a hump, and then, as V goes to infinity, gamma decreases to perfect gas-like behavior with gamma around 1.3. But according to Davis [1], γs(V) should decrease monotonically with V. Accordingly, in this article we investigate the following: (1) Is the hump in γs(V) necessary? and (2) Is it possible to construct a products EOS with a monotonic γs(V) that is consistent with experimental data? We find that (1) it is possible to construct a products EOS without a hump in γs(V); and (2) without a hump in γs(V) there are not enough degrees of freedom to reproduce cylinder test data.

On the HEL and the “Ramping” above HEL

Unlike in metals, stress and particle velocity histories of shock waves in ceramic materials show a typical “ramping” above the Hugoniot Elastic Limit (HEL). Under the assumption that the HEL signifies the beginning of failure of the material, this “ramping” has been described by viscoplasticity or by moduli degradation. However, there is ample experimental evidence to rule out plastic flow at the HEL level of stress, and it seems improbable that the moduli would degrade significantly during the compressive phase of a plate‐impact. The proposed micro‐mechanical mechanism for the HEL, and for the ramping beyond the HEL, is based on the process of porous compaction due to pressure above a threshold pressure pcrush. Although the pore volume of high‐grade ceramics is quite small, it is sufficient to cause the ramping observed in Hugoniot measurements from the seventies. Additional evidence for the effect of porosity on the HEL stress has been given in recent years. The experimental evidence is supported by si...

Failure diameter of confined explosive rods

We use our hydroreactive code to compute the failure diameter (FD) of confined PBX-9502 rods. First we calibrate our reaction model to match the standard pop-plot. Next we validate the calibration by reproducing the experimentally determined unconfined FD. Then we compute FD as a function of confinement thickness (CT) for three confinement materials: light (PMMA), medium (aluminum) and heavy (copper). We find that light confinement has no effect. For medium and heavy confinements we find, as expected, that FD decreases when CT increases. But for each material there is a lower limit to FD no matter how thick the confinement. We compare our results to data by Ramsay, obtained in slab symmetry, and agreement is reasonable.

On the hydrodynamic limit of long rod penetration

According to the quasi-steady-state approximation of long rod penetration, the total penetration efficiency should approach the so called hydrodynamic limit at high impact velocities. However, both experiments and computer simulations show that this is not the case. The total penetration efficiency usually overshoots the hydrodynamic limit at velocities above approximately 3 km/s. We ran computer simulations to investigate the significance of the hydrodynamic limit. We found that the overshoot occurs during the end (or terminal) phase. The end phase starts upon completion of the rod erosion. The kinetic energy trapped in the residual rod and in the target at this stage can cause significant additional deepening of the crater. As the overshoot of the hydrodynamic limit is an edge effect, it is expected to decrease as the aspect ratio of the rod increases.

Prediction of low-velocity-impact ignition threshold of energetic materials by shear-band mesoscale simulations

ABSTRACT When energetic materials are subjected to a low-velocity impact, they can develop viscoplastic localization, also known as an adiabatic shear band (ASB), leading to unwanted ignition. In this work, the simulation of shear band formation at mesoscale resolution has been incorporated in an ignition model. The simulation results suggest that a low-velocity impact induces the formation of an ASB, leading to ignition. Good agreement was found between the ignition model predictions and available impact test data. The ability of the model to predict the ignition threshold solely on the basis of mesoscale simulations demonstrates its superiority over existing empirical models.

Modeling shear band interaction in 1D torsion

When two shear bands are being formed at close distance from each other they interact, and further development of one of them may be quenched down. As a result there should be a minimum distance between shear bands. In the literature there are at least three analytical models for this minimum distance. Predictions of these models do not generally agree with each other and with test results. Recently we developed a 1D numerical scheme to predict the formation of shear bands in a torsion test of a thin walled pipe. We validated our code by reproducing results of the pioneering experiments of Marchand and Duffy, and then used it to investigate the mechanics of shear localization and shear band formation. We describe our shear band code in a separate publication, and here we use it only as a tool to investigate the interaction between two neighboring shear bands during the process of their formation. We trigger the formation of shear bands by specifying two perturbations of the initial strength. We vary the p...