Gordon Erlebacher
Florida State University, Scientific Computing, Department Member
Numerical modeling of physical phenomena frequently involves processes across a wide range of spatial and temporal scales. In the last two decades, the advancements in wavelet-based numerical methodologies to solve partial differential... more
Numerical modeling of physical phenomena frequently involves processes across a wide range of spatial and temporal scales. In the last two decades, the advancements in wavelet-based numerical methodologies to solve partial differential equations, combined with the unique properties of wavelet analysis to resolve localized structures of the solution on dynamically adaptive computational meshes, make it feasible to perform large-scale numerical simulations of a variety of physical systems on a dynamically adaptive computational mesh that changes both in space and time. Volumetric visualization of the solution is an essential part of scientific computing, yet the existing volumetric visualization techniques do not take full advantage of multi-resolution wavelet analysis and are not fully tailored for visualization of a compressed solution on the wavelet-based adaptive computational mesh. Our objective is to explore the alternatives for the visualization of time-dependent data on space-...
Research Interests:
This paper (the first in a series) focuses on using active-control methods to maintain laminar flow in a region of the flow in which the natural instabilities, if left unattended, lead to turbulent flow. The authors review previous... more
This paper (the first in a series) focuses on using active-control methods to maintain laminar flow in a region of the flow in which the natural instabilities, if left unattended, lead to turbulent flow. The authors review previous studies that examine wave cancellation (currently the most prominent method) and solve the unsteady, nonlinear Navier-Stokes equations to evaluate this method of controlling instabilities. It is definitively shown that instabilities are controlled by the linear summation of waves (i.e., wave cancellation). Although a mathematically complete method for controlling arbitrary instabilities has been developed, the review, duplication, and physical explanation of previous studies are important steps for providing an independent verification of those studies, for establishing a framework for the work which will involve automated transition control, and for detailing the phenomena by-which the automated studies can be used to expand knowledge of flow control.
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
A hardware-based approach for visualizing unsteady flow fields by means of Lagrangian-Eulerian advection is presented. Both noise-based and dyebased texture advection is supported. The implementation allows the advection of each texture... more
A hardware-based approach for visualizing unsteady flow fields by means of Lagrangian-Eulerian advection is presented. Both noise-based and dyebased texture advection is supported. The implementation allows the advection of each texture to be performed completely on the graphics hardware in a single-pass rendering. We discuss experiences with the interactive visualization of unsteady flow fields that become possible due to the high visualization speed of the hardware-based approach. 1
Research Interests:
Service-oriented application development is a time consuming task that changes little between projects. In general, development is a two step process where developers first create the core functionality of the service and then routinely... more
Service-oriented application development is a time consuming task that changes little between projects. In general, development is a two step process where developers first create the core functionality of the service and then routinely integrate details for distribution, transport protocols and user interfaces. Addressing a need to port standalone applications to web services, we have developed a compiler to automate the generation of Web Services. This compiler is called the Web Automation and Translation Toolkit, or WATT. Originally designed as a utility for porting Tcl scripts for the Visualization Toolkit (VTK) to equivalent but more efficient C++, WATT has been extended to seamlessly integrate the SOAP transport protocol to create working visualization web services. Within this document, we present details of the WATT compiler, including motivation, goals and example applications.
Research Interests:
We present a detailed description of the implementation of a library of Grid tag libraries and Grid beans for Grid Web portal development. Grid tags provide Java Server Faces (JSF) custom components for Grid services. They enable the... more
We present a detailed description of the implementation of a library of Grid tag libraries and Grid beans for Grid Web portal development. Grid tags provide Java Server Faces (JSF) custom components for Grid services. They enable the definition of attributes to the Grid service parameters in a dynamic way embedded into JSF view pages. In addition, Grid beans provide client proxies to the Grid services. Grid tags and beans together provide a platform to develop Grid portlets easily. In addition to standard Grid job submission and remote file operation tags, we also provide management and monitoring capabilities for Grid tasks. This system can persistently store bean features and job parameters, which results in a permanent storage for archiving and reference.
Research Interests:
A hardware-based approach for visualizing unsteady flow fields by means of Lagrangian-Eulerian advection is presented. The implementation allows texture advection to be performed completely on the graphics hardware in a single-pass... more
A hardware-based approach for visualizing unsteady flow fields by means of Lagrangian-Eulerian advection is presented. The implementation allows texture advection to be performed completely on the graphics hardware in a single-pass rendering. We discuss experiences with the interactive visualization of unsteady flow fields that become possible due to the high visualization speed of the hardware-based approach.
Pruning neural network parameters is often viewed as a means to compress models, but pruning has also been motivated by the desire to prevent overfitting. This motivation is particularly relevant given the perhaps surprising observation... more
Pruning neural network parameters is often viewed as a means to compress models, but pruning has also been motivated by the desire to prevent overfitting. This motivation is particularly relevant given the perhaps surprising observation that a wide variety of pruning approaches increase test accuracy despite sometimes massive reductions in parameter counts. To better understand this phenomenon, we analyze the behavior of pruning over the course of training, finding that pruning's benefit to generalization increases with pruning's instability (defined as the drop in test accuracy immediately following pruning). We demonstrate that this "generalization-stability tradeoff" is present across a wide variety of pruning settings and propose a mechanism for its cause: pruning regularizes similarly to noise injection. Supporting this, we find less pruning stability leads to more model flatness and the benefits of pruning do not depend on permanent parameter removal. These r...
A hardware-based approach for visualizing unsteady flow fields by means of Lagrangian-Eulerian advection is presented. Both noise-based and dyebased texture advection is supported. The implementation allows the advection of each texture... more
A hardware-based approach for visualizing unsteady flow fields by means of Lagrangian-Eulerian advection is presented. Both noise-based and dyebased texture advection is supported. The implementation allows the advection of each texture to be performed completely on the graphics hardware in a single-pass rendering. We discuss experiences with the interactive visualization of unsteady flow fields that become possible due to the high visualization speed of the hardware-based approach.
Research Interests:
... A map u : N × I −→ TM is a time-dependent vector field provided that u(x, t) ∈ TxM . ... For a non-tangential vector field, a related tangential vector field can be computed by projection fromTxM to TxN, ie, by removing the normal... more
... A map u : N × I −→ TM is a time-dependent vector field provided that u(x, t) ∈ TxM . ... For a non-tangential vector field, a related tangential vector field can be computed by projection fromTxM to TxN, ie, by removing the normal parts from the tangent vectors. ...
... We can plot these points with Amira as small cubes for visualization. ... of Energy. References 1. Garbow, ZA, Olson, NR, Yuen, DA and JM Boggs, Interactive Web-Based Map: Applications to Large Data Sets in the Geosciences, Electronic... more
... We can plot these points with Amira as small cubes for visualization. ... of Energy. References 1. Garbow, ZA, Olson, NR, Yuen, DA and JM Boggs, Interactive Web-Based Map: Applications to Large Data Sets in the Geosciences, Electronic Geosciences, Vol. 6, 2001. ...
... A map u : N × I −→ TM is a time-dependent vector field provided that u(x, t) ∈ TxM . ... For a non-tangential vector field, a related tangential vector field can be computed by projection fromTxM to TxN, ie, by removing the normal... more
... A map u : N × I −→ TM is a time-dependent vector field provided that u(x, t) ∈ TxM . ... For a non-tangential vector field, a related tangential vector field can be computed by projection fromTxM to TxN, ie, by removing the normal parts from the tangent vectors. ...
Steady and unsteady vector fields are integral to many areas of scientific en-deavor. They are generated by increasingly complex numerical simulations and measured by highly resolved experimental techniques. Datasets have grown in size... more
Steady and unsteady vector fields are integral to many areas of scientific en-deavor. They are generated by increasingly complex numerical simulations and measured by highly resolved experimental techniques. Datasets have grown in size and complexity motivating the ...
Research Interests:
Research Interests: Statistical Analysis, Decomposition, Direct Numerical Simulation, Density, Simulation, and 15 moreFlow, Structures, Mathematical Models, Shear Flow, Eigenvalues, Functions, Eigenvalue distribution, Kinetic Energy, Energy Transfer, Mean, Compressible Flow, Probability Density Function, Eigenvectors, Pressure Gradient, and Probability Density Functions
This paper describes a self-contained, automated methodology for flow control along with a validation of the methodology for the problem of boundary layer instability suppression. The objective of control is to match the stress vector... more
This paper describes a self-contained, automated methodology for flow control along with a validation of the methodology for the problem of boundary layer instability suppression. The objective of control is to match the stress vector along a portion of the boundary to a given vector; instability suppression is achieved by choosing the given vector to be that of a steady
Research Interests:
The practical implementation of 4D-Var data assimilation for atmospheric and oceanographic models is hampered by the large dimensionality of the discrete model initial conditions, typically in the range 106- 10^7. Order reduction... more
The practical implementation of 4D-Var data assimilation for atmospheric and oceanographic models is hampered by the large dimensionality of the discrete model initial conditions, typically in the range 106- 10^7. Order reduction strategies aim to alleviate the computational burden of the 4D-Var procedure by formulating the optimal control problem in a low-order control space. In this study a proper orthogonal decomposition method (POD) is used to identify a low dimensional space that captures most of the energy and the main directions of variability of the model. Data assimilation experiments are setup with a 2D global shallow-water model using the Lin-Rood flux-form semi-Lagrangian discretization with initial conditions specified from the ECMWF 500mb ERA-40 dataset. Qualitative and quantitative aspects of the reduced-order control strategy are analyzed in a twin experiments 4D-Var framework by comparison with results in the full model space. Numerical results show that with an appropriate selection of the basis vectors the optimization in the low-order POD space is able to significantly reduce the computational cost while preserving the quality of the solution. Issues related to the generation of the ensemble of snapshots and optimal selection of the basis vectors in the context of optimal control are addressed. A second order adjoint model implemented in the reduced space is used to provide a Hessian condition number analysis and to assess the efficiency of the POD-based optimization. Further applications in numerical weather prediction to estimation of information content of data and to identification of observational data of most benefit to the analysis and data assimilation procedure are presented.
Research Interests:
With the proliferation of GPU ( graphics accelerator board) the computing landscape has changed enormously in the last 3 years. The new additional capabilities of the GPU , such as larger shared memories and load-store operations , allow... more
With the proliferation of GPU ( graphics accelerator board) the computing landscape has changed enormously in the last 3 years. The new additional capabilities of the GPU , such as larger shared memories and load-store operations , allow it to be considered as a viable stand-alone computational and visualization engine. Today the massive threading and computing capability of GPU can
Research Interests: Radial Basis Function, Stokes flow, Shared memory, Nearest Neighbor, Mantle Convection, and 9 moreMessage Passing, Wave propagation, Acoustic Diffusion Equation Model, Finite Difference, Seismic Wave Propagation, Shallow Water Equation (SWE), Vortex Method, Numerical Solution, and Spectral Element Method
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Abstract Amira (www. amiravis. com) is a powerful 3-D visualization package and has been employed recently by the science and engineering communities to gain insight into their data. We present a new web-based interface to Amira, packaged... more
Abstract Amira (www. amiravis. com) is a powerful 3-D visualization package and has been employed recently by the science and engineering communities to gain insight into their data. We present a new web-based interface to Amira, packaged in a Java applet. We ...
Research Interests:
Research Interests:
Research Interests:
ABSTRACT It is shown that the dilatational terms that need to be modeled in compressible turbulence include not only the pressure-dilatation term but also another term - the compressible dissipation. The nature of these dilatational terms... more
ABSTRACT It is shown that the dilatational terms that need to be modeled in compressible turbulence include not only the pressure-dilatation term but also another term - the compressible dissipation. The nature of these dilatational terms in homogeneous turbulence is explored by asymptotic analysis of the compressible Navier-Stokes equations. A non-dimensional parameter which characterizes some compressible effects in moderate Mach number, homogeneous turbulence is identified. Direct numerical simulations (DNS) of isotropic, compressible turbulence are performed, and their results are found to be in agreement with the theoretical analysis. A model for the compressible dissipation is proposed; the model is based on the asymptotic analysis and the direct numerical simulations. This model is calibrated with reference to the DNS results regarding the influence of compressibility on the decay rate of isotropic turbulence. An application of the proposed model to the compressible mixing layer has shown that the model is able to predict the dramatically reduced growth rate of the compressible mixing layer.
Research Interests:
ABSTRACT Today there is an explosion in data from high-resolution computations of nonlinear phenomena in many fields, including the geo- and environmental sciences. The efficient storage and subsequent visualization of these large data... more
ABSTRACT Today there is an explosion in data from high-resolution computations of nonlinear phenomena in many fields, including the geo- and environmental sciences. The efficient storage and subsequent visualization of these large data sets is a trade off in storage costs versus data quality. New dynamically adaptive simulation methodologies promise significant computational cost savings and have the added benefit of producing results on adapted grids that significantly reduce storage and data manipulation costs. Yet, with these adaptive simulation methodologies come new challenges in the visualization of temporally adaptive data sets. In this work turbulence data sets from Stochastic Coherent Adaptive Large Eddy Simulations (SCALES) are visualized with the open source tool ParaView, as a challenging case study. SCALES simulations use a temporally adaptive collocation grid defined by wavelet threshold filtering to resolve the most energetic coherent structures in a turbulence field. A subgrid scale model is used to account for the effect of unresolved subgrid scale modes. The results from the SCALES simulations are saved on a thresholded dyadic wavelet collocation grid, which by its nature does not include cell information. Paraview is an open source visualization package developed by KitWare(tm) that is based on the widely used VTK graphics toolkit. The efficient generation of cell information, required with current ParaView data formats, is explored using custom algorithms and VTK toolkit routines. Adaptive 3d visualizations using isosurfaces and volume visualizations are compared with non-adaptive visualizations. To explore the localized multiscale structures in the turbulent data sets the wavelet coefficients are also visualized allowing visualization of energy contained in local physical regions as well as in local wave number space.
Research Interests:
Research Interests:
Research Interests:
ABSTRACT The main objective of this paper is to initiate a systematic investigation of the small spatial structures present in compressible homogeneous turbulence. This is accomplished through an analysis of the rate of strain tensor... more
ABSTRACT The main objective of this paper is to initiate a systematic investigation of the small spatial structures present in compressible homogeneous turbulence. This is accomplished through an analysis of the rate of strain tensor calculated both from the solenoidal and from the irrotational components of the velocity field. Statistical properties of the tensors' eigenvalues and eigenvectors are presented. We also look into the relationships between the pressure gradient, the vorticity and the eigenvectors of the rate of strain tensors, and how these are affected by the mean shear rate. At low shear rates, we expect only slight departure from incompressible statistics. However, as the shear rate is increased, both the solenoidal and the irrotational velocity components will vary in an O(S(exp -1)) time, determined by the mean shear. This leads to possible interaction mechanisms between the two components.
Research Interests:
Research Interests:
Research Interests: Economics, Data Mining, Data Analysis, Middleware, Numerical Simulation, and 15 moreInteractive Visualization, Cluster Analysis, Visual, Real Time, Client Server, Data Processing, Ease of Use, Large Scale, Software Tool, Fault Tolerant, Space Missions, User Interaction, Low Resolution, Computational Method, and Programming language
An academic directory and search engine.
Research Interests: Distributed Computing, Materials Science, Computational Chemistry, Data Management, User Interface, and 12 moreMiddleware, Computer Software, Concurrency, Data storage, Computational material science, Publish/subscribe, Web Service, Distributed Collaboration, Portal, Instruments and Techniques, Equation of State, and Collaborative Computing
Research Interests:
Research Interests:
Line Integral Convolution (LIC) is a promising method for visualizing 2D dense flow fields. Direct extensions of the LIC method to 3D have not been considered very effective, because optical integration in viewing directions tends to... more
Line Integral Convolution (LIC) is a promising method for visualizing 2D dense flow fields. Direct extensions of the LIC method to 3D have not been considered very effective, because optical integration in viewing directions tends to spoil the coherent structures along 3D local streamlines. In our previous reports, we have proposed a selective approach to volume rendering of LIC solid