Organizational Unit:

Daniel Guggenheim School of Aerospace Engineering

Permanent Link

https://hdl.handle.net/1853/70742

Parent Organization

Organizational Unit

College of Engineering

Includes Organization(s)

Organizational Unit

Aerospace Design Group

Organizational Unit

Aerospace Systems Design Laboratory (ASDL)

Organizational Unit

Air Transportation Laboratory

Organizational Unit

Cognitive Engineering Center

Organizational Unit

Space Systems Design Laboratory (SSDL)

Show 1 more

ArchiveSpace Name Record

https://finding-aids.library.gatech.edu/agents/corporate_entities/106

Full item page

Publication Search Results

Now showing 1 - 10 of 21

Evaluation of Convolutional Neural Networks for Modeling Blast Propagation in Multi-room Bunkers

(Georgia Institute of Technology, 2023-12-15) Luo, Felix

The rapid evaluation of blasts in enclosed geometrically complex spaces has long eluded the design of safer blast-resistant structures. Traditional methods of determining blast responses in enclosed geometrically complex spaces oftentimes rely on the use of traditional computational fluid dynamics (CFD) solvers to compute the entire flow field of the structure. This method has an enormous computational burden, especially considering that blasts are highly transient in nature and require the transient pressure fluctuations to be determined to formulate an accurate blast response prediction. However, more efficient methods of blast evaluation are desired such that parametric sweeps or optimization processes can be performed at low cost to provide a tool for iterative design. To rectify this gap in capabilities, a convolutional neural network based (CNN) model was developed to provide rapid blast predictions for 2D structures to establish this capability to aid in the design of more blast resistant structures. This approach leverages the inherent spatial awareness of CNNs to provide predictions for peak pressures since blasts in enclosed spaces are highly dependent on the spatial relationships between blast locations and wall location. This approach provides a nearly 5,000 times speed up against CFD simulations used within this study with good convergence of errors, correlation coefficients, predicted and truth values and distributions in all situational evaluations. These computational advantages, in part, comes from using the CNN based model to directly predict peak pressures whereas traditional CFD solvers require iterations to propagate fluid flows over time. However, some limitations do exist with respect to higher errors, such as model training costs, and the capability to predict 3D structures. Nonetheless, the results provide a characterization of the capabilities CNN based models in predicting peak pressures from blasts in enclosed spaces. From these evaluations and studies, a model which can provide significant computational savings while maintaining a similar accuracy can be obtained, which enables the rapid iterative design of more blast resistant structures.
Reshock Gas Curtain Mixing Study

(Georgia Institute of Technology, 2022-06-25) Risley, Karl Robert

The current work investigates the behavior of gas curtain instabilities. A gas curtain can be visualized as an A − B − A domain, where A and B are light and heavy fluids respectively, creating a ”curtain” of heavy fluid B that is surrounded by a light fluid A. Specifically, the behavior of gas curtains following an initial shock passage and the passage of a reflected shock (reshock) through the entirety of the curtain are investigated. A gas curtain instability commonly occurs physically in a wide range of applications such as during afterburning of an explosion, inertial confinement fusion, and even supernovae explosions. Previous studies have emphasized that the physics occurring during the reshock of a gas curtain are far more complex than the behavior of a re-shock Richtmyer-Meshkov Instability, due to the interactions between the two interfaces and wave reverberations occurring. The current work attempts to understand the relationship between a gas curtain’s initial conditions and its behavior to reshock through two-dimensional numerical simulations that utilize the viscous Navier-Stokes equations. More specifically, the current work isolates the effects of the curtain’s initial thickness and shape on the post reshock mixing layer growth rate and molecular mixing of the curtain. The results for all cases indicate that the post-reshock growth rate of the curtain’s width is a function of initial thickness. The sensitivity of the curtain’s post-reshock growth rate to the initial thickness, however, depends on the curtain’s initial perturbation shape. As the initial thickness of the curtain is decreased, the interactions between the curtain’s interfaces grow in strength and impede perturbation growth, thus reducing the post reshock growth rate of the curtain’s structure width. Similarly, the results strongly suggest that a reduction in initial curtain thickness increases the late-time asymptotic molecular mixing fraction value. This result is significant, especially for reacting flows, because it indicates that faster combustion (or afterburning in an explosion) could be reached with the thinning of the gas curtain in flow systems.
UNCERTAINTY QUANTIFICATION OF DIMP PYROLYSIS KINETICS

(Georgia Institute of Technology, 2022-05-13) Patel, Pavan

To develop effective explosives and strategies for the rapid destruction of sarin stockpiles, a reliable understanding of sarin’s chemical kinetics is needed. Kinetic mechanisms of sarin simulants such as di-isopropyl methyl phosphonate (DIMP) are developed instead because they have a similar chemical structure as sarin and are less toxic. A detailed DIMP kinetics mechanism has been developed in the past; however, there is a considerable amount of uncertainty surrounding it. This uncertainty manifests through the choice of pathways, and their respective reaction rates, leading to large variations in outcomes predicted through simulations. Out of the many reaction pathways involved in the decomposition of DIMP, the initiating steps are the most crucial. Out of the two possible initiating pathways in the destruction of DIMP, the lower activation energy pathway is dominant for all temperatures. The purpose of this study is to investigate the uncertainties associated with the dominant initiating pathways of the DIMP kinetics mechanism. Propagating rate parameter uncertainties of the dominant pathway through computational models yields large uncertainties in predicting DIMP survivability at different temperatures. The prediction uncertainties are larger at lower temperatures than at high temperatures. This can significantly impact the ability to precisely predict collateral damage caused by partially destroyed DIMP in the far-field of an explosion. After reducing these rate parameter uncertainties, using Bayesian inference, the prediction uncertainties were within reasonable limits. The results here provide a reduced subspace for uncertainties associated with the first and most important step in the breakdown of DIMP, which shall enable more reliable predictions.
Investigating Lean Blowout of an Alternative Jet Fuel in a Gas Turbine Combustor

(Georgia Institute of Technology, 2022-01-05) Narayanan, Vijay

In the global effort to reduce the climate impact of combustion emissions, sustainable aviation fuels offer the ease and reliability of conventional petroleum-derived jet fuels without the significant pollutant effects. Ongoing research efforts include experimental testing of alternative jet fuels to identify fuel candidates that produce less pollutant combustion products and are cheaper and environmentally cleaner to source than conventional jet fuels. Fuel lean combustion already reduces the emissions of jet engines and increases fuel efficiency, but lean blowout (LBO) can occur at reduced throttle and minimum power scenarios such as descent. Lean blowout (LBO) has been identified as a critical figure of merit to ensure the stability of alternative jet fuels in the place of conventional fuels. This work aimed to further understand the LBO phenomenon, leveraging computational studies of the alternative fuel designated C-5 by the National Jet Fuel Combustion Program (NJFCP). The fuel sensitivity of LBO has been established by the NJFCP’s participants recently. In this thesis, the chemical kinetics for C-5 is first verified using zerodimensional (0-D) and one-dimensional (1-D) studies and then this is followed by three dimensional (3D) large-eddy simulations (LES). In LES to observe LBO, a direct-step and gradual equivalence ratio reduction were separately employed to assess fuel sensitivity of LBO against available experimental data. The time histories of pressure, temperature, and composition were analyzed for precursor signatures of LBO both inside and outside the flame. Localized extinction, a reduction in the vortex breakdown bubble size and magnitude, and a reduction in the exhaust velocity were all observed to occur during the LBO event.
LES of Turbulent Premixed Flame Kernel Formation and Development

(Georgia Institute of Technology, 2020-12-17) Lambert, Alexander

Spark ignition of flammable mixtures is highly sensitive to early and local conditions. Kernel formation and subsequent flame development are largely governed by turbulent conditions and interactions with igniter geometry. In order to investigate this phenomenon, the use of Large Eddy Simulation (LES) is examined for (1) modelling spherical turbulent flame development, and (2) simulating spark ignition in a channel with either laminar or turbulent inflow. A comparison between LES spherical flame simulation is made to FSD-LES results as well as experimental measurements from previous studies. For spark ignition experiments, we characterize the temporal evolution of the ignition process, and demonstrate the dependence on early velocity fluctuations and local conditions.
Investigation of ODE-based non-equilibrium wall shear stress models for large eddy simulation

(Georgia Institute of Technology, 2019-07-30) Dzanic, Tarik

For high Reynolds number flows, wall modeling is essential for performing large eddy simulation at a reasonable computational cost. In this work, a novel low-cost ODE-based non-equilibrium wall model is introduced for wall shear stress modeling in LES. Using polynomial approximations of the pressure gradient and convective terms obtained from interpolation of the LES solution, as opposed to direct evaluation of these gradients within the wall model, the governing wall model equations reduce from coupled PDEs to uncoupled ODEs that do not require an embedded wall model grid within the LES grid. Additionally, the steady form of the wall model equations was utilized, feasible due to the spatial decoupling of the wall model equations, and the effects of the temporal evolution on the wall shear stress were modeled. The effects of polynomial degree on the accuracy of the wall shear stress predictions were explored, and an empirical lag model was built to model the unsteady effects without requiring the solution of a time-stepping problem. Wall resolved large eddy simulations of separated flow around the NASA wall mounted hump and an iced NACA 63A213 airfoil were performed and used as a reference for the comparison of the non-equilibrium wall model to a commonly used equilibrium wall model. The proposed non-equilibrium wall model was able to predict separated flow and laminar flow regions in much better agreement with the wall resolved results than the equilibrium wall model. Underpredictions in the skin friction coefficient in non-equilibrium flow regimes were reduced from 20-50% to less than 10% between the equilibrium and the non-equilibrium wall modeled approaches. Minor improvements in the pressure coefficient predictions were observed with the non-equilibrium model in the separated flow region of the iced airfoil. The results suggest that the proposed wall model can offer better predictions of separated and/or laminar flows compared to equilibrium wall models with negligible computational cost increase.
Towards multi-scale reacting fluid-structure interaction: micro-scale structural modeling

(Georgia Institute of Technology, 2015-04-15) Gallagher, Timothy

The fluid-structure interaction of reacting materials requires computational models capable of resolving the wide range of scales present in both the condensed phase energetic materials and the turbulent reacting gas phase. This effort is focused on the development of a micro-scale structural model designed to simulate heterogeneous energetic materials used for solid propellants and explosives. These two applications require a model that can track moving surfaces as the material burns, handle spontaneous formation of discontinuities such as cracks, model viscoelastic and viscoplastic materials, include finite-rate kinetics, and resolve both micro-scale features and macro-scale trends. Although a large set of computational models is applied to energetic materials, none meet all of these criteria. The Micro-Scale Dynamical Model serves as the basis for this work. The model is extended to add the capabilities required for energetic materials. Heterogeneous solid propellant burning simulations match experimental burn rate data and descriptions of material surface. Simulations of realistic heterogeneous plastic-bound explosives undergoing impact predict the formation of regions of localized heating called hotspots which may lead to detonation in the material. The location and intensity of these hotspots is found to vary with the material properties of the energetic crystal and binder and with the impact velocity. A statistical model of the hotspot peak temperatures for two frequently used energetic crystals indicates a linear relationship between the hotspot intensity and the impact velocity. This statistical model may be used to generate hotspot fields in macro-scale simulations incapable of resolving the micro-scale heating that occurs in heterogeneous explosives.
Large eddy simulation of syngas-air diffusion flames with artificial neural networks based chemical kinetics

(Georgia Institute of Technology, 2011-09-07) Sanyal, Anuradha

In the present study syngas-air diffusion flames are simulated using LES with artificial neural network (ANN) based chemical kinetics modeling and the results are compared with previous direct numerical simulation (DNS) study, which exhibits significant extinction-reignition and forms a challenging problem for ANN. The objective is to obtain speed-up in chemistry computation while still having the accuracy of stiff ODE solver. The ANN methodology is used in two ways: 1) to compute the instantaneous source term in the linear eddy mixing (LEM) subgrid combustion model used within LES framework, i.e., laminar-ANN used within LEMLES framework (LANN-LEMLES), and 2) to compute the filtered source terms directly within the LES framework, i.e., turbulent-ANN used within LES (TANN-LES), which further dicreases the computational speed. A thermo-chemical database is generated from a standalone one-dimensional LEM simulation and used to train the LANN for species source terms on grid-size of Kolmogorov scale. To train the TANN coefficients the thermo-chemical database from the standalone LEM simulation is filtered over the LES grid-size and then used for training. To evaluate the performance of the TANN methodology, the low Re test case is simulated with direct integration for chemical kinetics modeling in LEM subgrid combustion model within the LES framework (DI-LEMLES), LANN-LEMLES andTANN-LES. The TANN is generated for a low range of Ret in order to simulate the specific test case. The conditional statistics and pdfs of key scalars and the temporal evolution of the temperature and scalar dissipation rates are compared with the data extracted from DNS. Results show that the TANN-LES methodology can capture the extinction-reignition physics with reasonable accuracy compared to the DNS. Another TANN is generated for a high range of Ret expected to simulate test cases with different Re and a range of grid resolutions. The flame structure and the scalar dissipation rate statistics are analyzed to investigate success of the same TANN in simulating a range of test cases. Results show that the TANN-LES using TANN generated fora large range of Ret is capable of capturing the extinction-reignition physics with a very little loss of accuracy compared to the TANN-LES using TANN generated for the specific test case. The speed-up obtained by TANN-LES is significant compared to DI-LEMLES and LANN-LEMLES.
Large eddy simulation of heated pulsed jets in high speed turbulent crossflow

(Georgia Institute of Technology, 2010-08-12) Pasumarti, Venkata-Ramya

The jet-in-crossflow problem has been extensively studied, mainly because of its applications in film cooling and injector designs. It has been established that in low-speed flows, pulsing the jet significantly enhances mixing and jet penetration. This work investigates the effects of pulsing on mixing and jet trajectory in high speed (compressible) flow, using Large Eddy Simulation. Jets with different density ratios, velocity ratios and momentum ratios are pulsed from an injector into a crossflow. Density ratios used are 0.55 (CH4/air), 1.0 (air/air) and 1.5 (CO2/air). Results are compared with the low speed cases studied in the past and then analyzed for high speed scaling. The simulations show that the lower density jet develops faster than a higher density jet. This results in more jet spread for the lower density jet. Scaling for jet spread and the decay of centerline jet concentration for these cases are established, and variable density scaling law is developed and used to predict jet penetration in the far field. In most non-premixed combustor systems, the fuel and air being mixed are at different initial temperatures and densities. To account for these effects, heated jets at temperatures equal to 540K and 3000K have been run. It has been observed that, in addition to the lower density of heated jets, the higher kinematic viscosity effects the jet penetration. This effect has been included and validated in the scaling law for the heated jet trajectory.
Richtmyer-Meshkov instability with reshock and particle interactions

(Georgia Institute of Technology, 2010-07-08) Ukai, Satoshi

Richtmyer-Meshkov instability (RMI) occurs when an interface of two fluids with different densities is impulsively accelerated. The main interest in RMI is to understand the growth of perturbations, and numerous theoretical models have been developed and validated against experimental/numerical studies. However, most of the studies assume very simple initial conditions. Recently, more complex RMI has been studied, and this study focuses on two cases: reshocked RMI and multiphase RMI. It is well known that reshock to the species interface causes rapid growth of interface perturbation amplitude. However, the growth rates after reshock are not well understood, and there are no practical theoretical models yet due to its complex interface conditions at reshock. A couple of empirical expressions have been derived from experimental and numerical studies, but these models are limited to certain interface conditions. This study performs parametric numerical studies on various interface conditions, and the empirical models on the reshocked RMI are derived for each case. It is shown that the empirical models can be applied to a wide range of initial conditions by choosing appropriate values of the coefficient. The second part of the study analyzes the flow physics of multiphase RMI. The linear growth model for multiphase RMI is derived, and it is shown that the growth rates depend on two nondimensional parameters: the mass loading of the particles and the Stokes number. The model is compared to the numerical predictions under two types of conditions: a shock wave hitting (1) a perturbed species interface surrounded by particles, and (2) a perturbed particle cloud. In the first type of the problem, the growth rates obtained by the numerical simulations are in agreement with the multiphase RMI growth model when Stokes number is small. However, when the Stokes number is very large, the RMI motion follows the single-phase RMI growth model since the particle do not rapidly respond while the RMI instability grows. The second type of study also shows that the multiphase RMI model is applicable if Stokes number is small. Since the particles themselves characterize the interface, the range of applicable Stokes number is smaller than the first study. If the Stokes number is in the order of one or larger, the interface experiences continuous acceleration and shows the growth profile similar to a Rayleigh-Taylor instability.