The work discussed in this report was supported by a Campus Fellowship LDRD. The report contains three papers that were published by the fellowship recipient and these papers form the bulk of his dissertation. They are reproduced here to satisfy LDRD reporting requirements.
The PANDA code is used to build tabular equations of state (EOS) for titanium and the alloy Ti4Al6V. Each EOS includes solid-solid phase transitions, melting, vaporization, and thermal electronic excitation. Separate EOS tables are constructed for the solid and fluid phases, and the PANDA phase transition model is used to construct a single multiphase table. The model explains a number of interesting features seen in the Hugoniot data, including an anomalous increase in shock velocity, recently observed near 200 GPa in Ti6Al4V. These new EOS tables are available for use with the CTH code and other hydrocodes that access the CTH database.
The PANDA code is used to construct tabular equations of state (EOS) for four metals-- beryllium, nickel, tungsten and gold. Each EOS includes melting, vaporization, and thermal electronic excitation. Separate EOS tables are constructed for the solid and fluid phases, and the PANDA phase transition model is used to construct a multiphase EOS table for each metal. These new EOS tables are available for use with the CTH code and other hydrocodes that access the CTH database.
The specific problem to be addressed in this work is the secondary combustion that arises from shock-induced mixing in volumetric explosives. It has been recognized that the effects of combustion due to secondary mixing can greatly alter the expansion of gases and dispersal of high-energy explosive. Furthermore, this enhanced effect may be a tailored feature for the new energetic material systems. One approach for studying this problem is based on the use of Large Eddy Simulation (LES) techniques. In this approach, the large turbulent length scales of motion are simulated directly while the small scales of turbulent motion are explicitly treated using a subgrid scale (SGS) model. The focus of this effort is to develop a SGS model for combustion that is applicable to shock-induced combustion events using probability density function (PDF) approaches. A simplified presumed PDF combustion model is formulated and implemented in the CTH shock physics code. Two classes of problems are studied using this model. The first is an isolated piece of reactive material burning with the surrounding air. The second problem is the dispersal of highly reactive material due to a shock driven explosion event. The results from these studies show the importance of incorporating a secondary combustion modeling capability and the utility of using a PDF-based description to simulate these events.
Abstract not provided.
Twelve experiments were conducted to determine the effect of water filled targets on the penetration of tungsten long rods in terms of their residual mass and integrity. CTH hydrocode calculations were performed for each of the experiments to ensure that the erosion and breakup of the tungsten projectiles could be accurately reproduced. The CTH hydrocode predictions correlation well with the experimental results in most cases. Only 8% of the variance is unexplained. The slip interface between the rod and water was approximated in one of two ways: (1) using the CTH BLINT option in 2-D or (2) using a standard Eulerian mixed cells treatment. Results indicate that a 3-D BLINT algorithm is critical to predicting rod residual lengths. The authors were unable to reproduce rod fracture that occurred in every experiment where the water column exceeded 25 cm in length. The authors feel that this is due to a change in rod material properties during penetration, and continue to investigate the issue.
The CTH Eulerian hydrocode, and the SPHINX smooth particle hydrodynamics (SPH) code were used to model a shock tube, two long rod penetrations into semi-infinite steel targets, and a long rod penetration into a spaced plate array. The results were then compared to experimental data. Both SPHINX and CTH modeled the one-dimensional shock tube problem well. Both codes did a reasonable job in modeling the outcome of the axisymmetric rod impact problem. Neither code correctly reproduced the depth of penetration in both experiments. In the 3-D problem, both codes reasonably replicated the penetration of the rod through the first plate. After this, however, the predictions of both codes began to diverge from the results seen in the experiment. In terms of computer resources, the run times are problem dependent, and are discussed in the text.
Advances in Engineering Software
This paper presents an overview of an explicit message-passing paradigm for a Eulerian finite volume method for modeling solid dynamics problems involving shock wave propagation, multiple materials, and large deformations. Three-dimensional simulations of high-velocity impact were conducted on the IBM SP2, the SGI Power Challenge Array, and the SGI Origin 2000. The scalability of the message-passing code on distributed-memory and symmetric multiprocessor architectures is presented and compared to the ideal linear performance.
This paper presents experimental data and computational modeling for a well-defined glass material. The experimental data cover a wide range of strains, strain rates, and pressures that are obtained from quasi-static compression and tension tests, split Hopkinson pressure bar compression tests, explosively driven flyer plate impact tests, and depth of penetration ballistic tests. The test data are used to obtain constitutive model constants for the improved Johnson-Holmquist (JH-2) brittle material model. The model and constants are then used to perform computations of the various tests.
The Hypervelocity Experimental Research Database (HERD) described in this paper was developed to aid researchers with code validation for impacts that occur at velocities faster than the testable regime. Codes of concern include both hydrocodes and fast-running analytical or semi-empirical models used to predict the impact phenomenology and damage that results to projectiles and targets. There are several well documented experimental programs that can serve as benchmarks for code validation; these are identified and described. Recommendations for further experimentation (a canonical problem) to provide validation data are also discussed.
The UNIX LANs in 1500 are experiencing explosive growth. The individual departments are creating LANs to address their particular needs; however, at the same time, shared software tools between the departments are becoming more common. It is anticipated that users will occasionally need access to various department software and/or LAN services, and that support personnel may carry responsibilities which require familiarization with multiple environments. It would be beneficial to users and support personnel if the various department environments share some basic similarities, allowing somewhat transparent access. This will become more important when departments share specific systems, as 1510 and 1550 have proposed with an unclassified UNIX system. Therefore, standards/conventions on the department LANs and the central site systems have to be established to allow for these features. it should be noted that the goal of the UEC is to set standards/conventions which affect the users and provide some basic structure for software installation and maintenance; it is not the intent that all 1500 LANs be made identical at an operating system and/or hardware level. The specific areas of concern include: (1) definition of a non-OS file structure; (2) definition of an interface for remote mounted file systems; (3) definition of a user interface for public files; (4) definition of a basic user level environment; and (5) definition of documentation requirements for public files (shared software). Each of these areas is addressed in this paper.
The purpose of this paper is to develop an analytical model to convert ballistic limit curves obtained from flat projectile experiments to ballistic limit curves based on equivalent diameter spheres. Results from a test program involving flat plat projectiles conducted at Sandia National Laboratories are compared against the predicted performance of equivalent spherical projectiles as determined from the Wilkinson and Cour-Palais penetration equations. The developed method demonstrates good correlation of the ballistic limit of the shield concept for the flat plate projectiles to the theoretical ballistic limit for equivalent spheres as predicted by the penetration equations. 3 refs.
The design of the potential Yucca Mountain repository is subject to many thermal goals related to the compliance of the site with federal regulations. This report summarizes a series of sensitivity studies that determined the expected temperatures near the potential repository. These sensitivity studies were used to establish an efficient loading scheme for the spent fuel canisters and a maximum areal power density based strictly on thermal goals. Given the current knowledge of the site, a design-basis areal power density of 80 kW/acre can be justified based on thermal goals only. Further analyses to investigate the impacts of this design-basis APD on mechanical and operational aspects of the potential repository must be undertaken before a final decision is made.
Calculations were performed with CTH (a finite difference hydrodynamics code) to evaluate computational capabilities for predicting residual projectile length and position in high velocity penetration events. The calculations simulated selected tests in a set of armor penetration experiments conducted and reported by Los Alamos National Laboratory. The tests and simulations involved penetration of armor ranging from 6.4 to 50.8 mm in thickness by long rod projectiles over a range of velocities from 1.0 to 1.29 km/sec. Comparisons are made between the calculated and measured final projectile lengths and positions, and the sensitivity of the predicted results to target and projectile property variations is investigated. 8 refs., 11 figs., 8 tabs.
CTH is a software system under development at Sandia National Laboratories (SNL) to model multidimensional, multi-material, large deformation, strong shock wave physics. One-dimensional rectilinear, cylindrical, and spherical meshes; two-dimensional rectangular, and cylindrical meshes; and three-dimensional rectangular meshes are currently available. A two-step Eulerian solution scheme is used with these meshes. The first step is a Lagrangian step in which the cells distort to follow the material motion. The second step is a remesh step where the distorted cells are mapped back to the original Eulerian mesh. CTH has several thermodynamic models that are used for simulating strong shock, large deformation events. Both tabular and analytic equations of state are available. CTH can model material strength, high explosive detonation, fracture, and motion of fragments smaller than a computational cell. The material strength model is elastic perfectly plastic with thermal softening. A programmed burn model is available for computing the thermodynamic properties of explosive detonation. The Jones-Wilkins-Lee equation of state is available for modelling high explosive reaction products. Fracture can be initiated based on pressure or principle stress. A special model is available for moving fragments smaller than a computational cell with the correct statistical velocity. 7 refs.