Reactive Flow Characteristic Equations
The characteristic equations are derived for the reactive flow PDEs in 1-dimension with 1 irreversible reaction. They differ from the non-reacting case only by source terms proportional to the reaction rate.
Engineering topics
Publications and source records attributed to Menikoff, Ralph.
The characteristic equations are derived for the reactive flow PDEs in 1-dimension with 1 irreversible reaction. They differ from the non-reacting case only by source terms proportional to the reaction rate.
Abstract not provided.
The Zel’dovich-von Neumann-Doering (ZND) profile of a detonation wave is derived. Two basic material properties are required: i. An equation of state (EOS) for a partly burned explosive; P(V, e, λ). ii. A burn rate for the reaction progress variable; d/dtλ = R(V, e, λ). For a steady planar detonation wave the reactive flow PDEs can be reduced to ODEs. The detonation wave profile can be determined from an ODE plus algebraic equations for points on the partly burned detonation loci with a specified wave speed. Furthermore, for the CJ detonation speed, the end of the reaction zone is sonic. A solution to the reactive flow equations can be constructed with a rarefaction wave following the detonation wave profile. This corresponds to an underdriven detonation wave, and the rarefaction is know as a Taylor wave.
The SURFplus reactive burn model is intended for heterogeneous solid HE that produce an excess amount of carbon. It utilizes a fast hotspot reaction (SURF model) plus a second slow reaction for the energy release from carbon clustering. The fast reaction dominates shock initiation, though the fast rate parameters are slightly affected by the late energy release from the slow reaction. The slow reaction has a significant affect on propagating detonation waves. Propagating detonation waves are characterized by the curvature effect; i.e., detonation speed $D_n$ as function of the local front curvature κ. The $D_n(κ)$ curves have a qualitative different shape. This results from the sensitivity of the slope of the $D_n(κ)$ curve to the reaction-zone width up to the sonic point, and the sonic point moving as κ increases from the end of the slow reaction to the neighborhood of the end of the fast reaction. For PBX 9502, the steep slope for small curvature is determined by the slow reaction, while the lower slope for larger curvature is determined by the fast reaction rate in the high detonation pressure regime. The remainder of this paper is organized as follows. In section 2 we review the assumptions of the Shaw and Johnson carbon cluster model used by the SURFplus model for the slow reaction. Section 3 describes in detail the mixture model for a partly burned HE. The form of the slow rate used by the SURFplus model is described in section 4. We conclude with a brief discussion of applying a burn model with 2 rates to either a metalized or a composite PBX.
Embedded velocity gauges are standardly used at LANL on gas gun experiments of a shock to-detonation transition. The experiments are intended to be one-dimensional with a planar shock driven by the projectile from a gas gun propagating along the axis of the HE cylinder. As disussed by Gustavsen et al. [1999], the accuracy of the gauges depends on the alignment of the directions of shock propagation, magnetic field and the active gauge element. A source of inaccuracies is from tilt or misalignment angle at impact between the projectile surface and the HE surface; typically a few milliradians. This leads to the shock propagating at a slight angle to the axis of the HE cylinder and can affect the shock rise time of the velocity gauges, the time offset between the tracker gauges for the trajectory of the lead shock, and the gauge positions in the direction of propagation. It is important to note that tilt is characterized by 2 parameters; the magnitude of the angle relative to the HE axis and the direction in the plane transverse to the HE axis. Here we derive the uncertainties from tilt to leading order in the tilt angle.
It has previously been shown that the rate parameters of the SURF reactive burn model can be adjusted to fit the gap-stick data for PBX 9501 [Johnson et al., 2018, fig 4]. A recent recalibration of the SURF model for PBX 9501 (lot 730-010 at ρ = 1.837 g/cc 3 ) used different reactants and products EOS and an updated fitting form for the burn rate [Menikoff, 2021]. The new calibration fit gap-stick data [Hill et al., 2021] together with data from SDT experiments and curvature effect data. Here the results of gap-stick simulations using the xRage code are shown. They provide insight into the reactive flow in the gap-stick experiment.
PBX 9501 is a plastic bonded explosive composed of 95 wt % HMX and a binder; see [Gibbs and Popolato, 1980, pp. 109–119]. SURF is a reactive burn model for shock initiation and propagation of detonation waves. It has previously been calibrated for PBX 9501. Here the SURF model is recalibrated for PBX 9501; specifically, lot 730-010 at ρ = 1.837 g/cc 3 . The new calibration uses the Davis reactants and products EOS calibrated for the AWSD model [Aslam et al., 2020]. The burn rate in the shock initiation regime is fit to the Pop plot from 5 embedded gauge shock-to-detonation transition (SDT) experiments from [Gustavsen et al., 1999, see fig 12 and table 5]. In the propagation regime, the burn rate is fit to curvature effect data (detonation speed as function of front curvature); see [Aslam, 2007]. Also the burn parameters are adjusted to fit the gap-stick experiment [Hill et al., 2018]. Simulating the detonation wave speed in this experiment requires a model that is accurate for initiation with complex shock loading; in particular, a pressure decreasing gradient behind a curved lead shock. This is more difficult than calibrating to the standard SDT experiments which are 1-D and driven by a sustained shock. Simulations of the gap-stick experiment for PBX 9501 with the SURF model will be discussed in a subsequent report.
Five shock-to-detonation transition (SDT) gas gun experiments for PBX 9012 were performed by Burns and Chiquete [2020]. The data from the experiments is available on the small-scale database. Here we focus on the tracker gauge data for the lead shock trajectory, x(t). In order to determine the shock speed U s (t) and the detonation transition point, a least squared fit to the shock trajectory is used based on a pair of ODEs with a fitting form for the shock acceleration developed by Hill and Gustavsen [2002].
Gas gun shock-to-detonation experiments provide data for calibrating HE burn models. At LANL the principal diagnostic is a magnetic gauge package that provides data on the lead shock trajectory, which is used to determine the run-distance to detonation, and the lead shock speed with run distance, which together with jump-off velocity of the velocity gauges determines Hugoniot locus points for the reactants.
Shock initiation gas gun experiments of PBX 9012 have been performed by Burns and Chiquete [2020] using magnetic velocity gauges as the principal diagnostic. In private communications, the experimenters noted that compatibility issues required a different glue to bond the gauges to PBX 9012 then previously used for experiments with PBX 9501 and PBX 9502. The silicon glue used is more viscous and can result in glue layers up to 50 microns, which is about the gauge thickness. This raises the question on what affect the thicker glue layers have on the gauge response. In particular, the width of the lead shock can affect the accuracy of inferring shock locus points used for the shock pressure of Pop plot data points and for Hugoniot data to calibrate the reactants EOS.
Reactive burn models represent a significant leap in high explosive (HE) modeling capability. The first generation of engineering models of HE detonation are called programmed burn models and they are largely based on the distance between a prescribed detonation point and each zone in a simulation. There have been many advancements to programmed burn models over the years and when the assumptions upon which they are based are met, a properly tuned programmed burn model can be highly accurate but if any of their assumptions is not met, as is the case for corner turning or weakly initiated HE burn, they will give the wrong answer. Reactive burn models represent an entirely new way of modeling HE burn. They use the local conditions of a zone – e.g. temperature, pressure or density – as calculated by a hydrocode to determine if and when the zone is going to detonate and if so, how rapidly. This difference opens up an entirely new set of capabilities for HE modeling. It makes it possible to accurately and predictively model phenomena like the effect of confinement and the formation of dead zones. Reactive burn models have seen sustained development effort at LANL for at least the last decade but several recent developments make it timely to transition reactive burn models from a research topic to a production tool. The main goal of this milestone is to facilitate and accelerate the adoption of reactive burn as a commonly available modeling option, with recommendations on the resolution that will be required and uncertainties associated with their modeling choices. To achieve this, we have performed verification, validation, and uncertainty quantification (UQ) assessments of AWSD and SURF/SURFplus in xRage and FLAG on a variety of different problems.
Advanced reactive burn models enable phenomena to be modeled that aren’t possible with programmed burn. Programmed burn propagates a burn front at a prescribed speed from a prescribed initiation point. Reactive burn uses the hydro variables (pressure, temperature, etc.) to determine when, how rapidly, and to what extent a zone detonates. This enables a completely new set of problems to be simulated. Our goal in this milestone has been to improve the usability of reactive burn modes. Thus, the closure criteria are: 1. Perform code and solution verification of the reactive burn models using appropriate solutions and simplified HE setups. 2. Perform validation assessments using small-scale experiments. 3. Evaluate sensitivities and quantify uncertainties due to model form variations and mesh dependencies.
Reactive burn models for a high explosive (HE) have three components: an equation of state (EOS) for the products, an EOS for the reactants and a reaction rate. For the products, we discuss the uncertainty in data for the planar detonation state and its release isentrope. For the reactants, the uncertainty in the principal shock locus data is discussed. For the reaction rate, we discuss uncertainties in shock initiation data (distance-of-run to detonation as a function of shock pressure), and curvature effect data (detonation speed as a function of front curvature). An important source of uncertainty for calibrating the reaction rate for a plastic-bonded explosive (PBX) results from variations in heterogeneities; that is the mesoscale structure of HE grains, binder and porosity. This shows up in the scatter of both shock initiation data and curvature effect data with small changes in initial density and initial temperature, and systematic changes with PBX lot. As illustrative examples, we use data for the conventional HMX based PBX 9501 and the insensitive TATB based PBX 9502 since these explosives have been extensively studied and high quality data are readily available in the literature.
Reactive burn models for propagating detonation waves in an explosive are based on the ZND (Zel'dovich-Von Neumann-Doring) theory. A key property of the ZND theory, which leads to self-sustaining detonation waves, is a sonic point relative to the front in the reaction zone. Moreover, burn models assume the explosive is a homogeneous material. This assumption is used for plastic-bonded explosives, even though they are heterogeneous materials. The heterogeneities are accounted for only with an empirical burn rate. Furthermore, the reaction-zone width can be less than the length scale of heterogeneities. This raises the question of the effect of heterogeneities on the reaction zone. In particular, whether the reaction zone can be steady and have a well defined sonic point. To examine this question we focus on PBX 9501 since there is data on the grain-binder heterogeneities and on the reaction-zone velocity time history of a steady planar detonation wave. Using pseudo-mesoscale 1-D simulations, we examine the effect of a detonation wave transversing explosive and binder segments of random length. The simulations show that the reaction zone is quasi-steady with only short wavelength small amplitude variations in the detonation wave pressure and speed. The main effect of the heterogeneities is to introduce noise in the pressure and velocity fields that propagates into the Taylor wave behind the detonation. This is consistent with the scatter in the reaction-zone measurements for PBX 9501.
A key issue for detonation wave simulations with a reactive burn model is the resolution needed for the reaction zone. The reaction-zone width is a physical length scale. For plastic-bonded explosives, it is small (on the order of 100 microns) but affects important detonation properties such as the detonation speed. Even with adaptive mesh refinement, a well resolved reaction zone is computationally expensive. Here we consider the trade-off between computational expense and accuracy when the resolution of the reaction zone is varied for simulations with the xRage code of PBX 9501 with the SURF burn model and of PBX 9502 with the SURFplus burn model. For assessing shock initiation, the numerical Pop plot is simulated, and for assessing propagating detonation waves, the cylinder test is used as a test problem. The reaction-zone resolution is varied from a cell size of 6.25 to 200 microns, and key properties of the solutions are compared. With 6 micron resolution, the numerical solution is near mesh converged and can be used to judge the accuracy of the coarser resolution simulations. For the propagating detonation wave in the cylinder test, at 200 microns the detonation speed is low by a couple of per cent. In terms of cpu-hours, the finest resolution is about a factor of 100 more expensive than the coarsest resolution.
Generating a planar steady detonation wave in a high explosive is important for accurately measuring key parameters such as the CJ detonation state and the reaction-zone width. A detonation wave can be promptly shock initiated in the TATB based explosive PBX 9502 due to a fast hotspot reaction. However, a slow carbon clustering reaction causes the detonation wave to have a slow approach to steady state. Here we study the approach of the reaction zone of a planar detonation to the steady ZND profile of an underdriven detonation wave. For PBX 9502 simulations were run with the SURFplus reactive burn model. For comparison, simulations were also run with the HMX based PBX 9501. The SURF reactive model is used for PBX 9501 since it has only a fast hotspot reaction.
Abstract not provided.
Previous verification studies of the SURF reactive burn model in the xRage code used test problems for underdriven or self-supporting detonation waves. Here we consider overdriven detonation waves. As a test problem we use a cylindrical converging detonation wave. Since there is no exact solution, the numerical solutions at different resolutions are compared for 1-D cylindrical geometry and 2-D Cartesian geometry.