Ravishekar Kannan - Academia.edu (original) (raw)

Papers by Ravishekar Kannan

43rd AIAA Aerospace Sciences Meeting and Exhibit, Jan 10, 2005

The use of overset grids in CFD started more than two decades ago, and has achieved tremendous su... more The use of overset grids in CFD started more than two decades ago, and has achieved tremendous success in handling complex geometries. In particular, overset grids have the advantage of avoiding grid remeshing when dealing with moving boundary flow problems. Traditionally overset grids were mainly used with structured grids to simplify the grid generation process, because a complex computational domain can be more easily meshed after it is partitioned into sub-domains with overset interfaces than with patched interfaces. More recently, unstructured grids are also used in the overset grid system to further simplify grid generation for each sub-domain. In this paper, two particular unstructured grids are advocated for moving boundary flow simulation, i.e., the use of overset adaptive Cartesian/prism grids. Semi-structured prism grids are generated around solid walls. These prism grids then overlap a single adaptive Cartesian background grid. With the adaptive Cartesian grid, the mesh resolution of the prism grid near the outer boundary can easily match that of the oversetting Cartesian grid cells. In addition, the tree-based data structure of the Cartesian grid can be used efficiently in hole-cutting and donor cell identification. The overset adaptive Cartesian/prism grid method is tested for both steady and unsteady flow computation. It is demonstrated that moving boundary flow computations can be automated with minimum user interferences. c = Position vector of the centroid of triangle j Re = Reynolds number v g = Grid velocity v gn = Surface normal grid velocity component V i = Volume of control volume i

International Journal for Numerical Methods in Biomedical Engineering, 2010

A high-order spectral volume (SV) method was implemented for solving the steady-state moment mode... more A high-order spectral volume (SV) method was implemented for solving the steady-state moment models, such as the hydrodynamic (HD) models and the energy transport (ET) models for semiconductor device simulations (in 1D). The first derivative inviscid/convective fluxes are handled using an approximate Riemann flux and the second derivative diffusive fluxes are discretized using the local discontinuous Galerkin formulation (LDG). The LDG method is also used for discretizing the potential equation. An implicit pre-conditioned LU-SGS p-multigrid method developed for the SV Navier-Stokes (NS) solver by Kannan and Wang is adopted here for time marching. The entire formulation is compact and hence can be easily parallelized. A n +-n-n + diode was assumed for simulation purposes and the results are compared with the existing discontinuous Galerkin simulation results. In general, the numerical results are very promising and indicate that the approach has a great potential for higher-dimensional device problems.

Computers & Fluids, Sep 1, 2011

A high-order spectral volume (SV) method is implemented for solving the steady-state, isothermal,... more A high-order spectral volume (SV) method is implemented for solving the steady-state, isothermal, line-contact elastohydrodynamic lubrication problem. The standard procedure involves solving the Reynolds equation, the film thickness equation and the force balance equation. The nonlinear, integral and implicit nature of the film thickness equation creates a complex pressure updating procedure. In this paper, the Reynolds equation is modified so as to automatically result in a pressure update during time marching. The convective flux is handled using an approximate Riemann flux. Two viscous flux formulations were attempted: a. The classical local discontinuous Galerkin (LDG) formulation and b. The recently formulated LDG2 formulation. An implicit pre-conditioned Lower Upper Symmetrical Gauss Seidel (LU-SGS) p-multigrid method developed for the spectral volume Navier Stokes (NS) solver by Kannan and Wang is adopted here for time marching. A convergence speedup of more than two orders is obtained with the implicit p-multigrid method, when compared to the traditional Runge-Kutta explicit method. The entire formulation is compact and hence can be easily parallelized. A comparison between results obtained using the SV formulation and those from an existing low order finite difference simulation confirmed that the SV delivers extremely accurate solutions in spite of using relatively small number of degrees of freedom (DOF). The LDG2 formulation delivers slightly more accurate results than the traditional LDG approach. In general, the numerical results are very promising and indicate that the approach has a great potential for more complicated elastohydrodynamic problems such as the transient line contact problem and the point contact problem.

International Journal for Numerical Methods in Biomedical Engineering, Oct 8, 2010

International Journal for Numerical Methods in Fluids, 2013

In this paper, the first in a series, the authors have developed and implemented new computationa... more In this paper, the first in a series, the authors have developed and implemented new computational algorithms for improving the scalability of CFD simulations on emerging architectures such as multicore high performance computing (HPC) platforms. These algorithmic developments and implementations are classified into three categories: (i) improved partition for multicore platforms, (ii) improved and optimized communication for HPC and (iii) enhancing scalability using computer science based methods. In the first category, the multilevel partitioning strategy was modified to reduce the number of out-of-core communications. This resulted in noticeable speedup even for small cases. In the second category, the authors came up with a next generation communication procedure optimized for the architecture and the partitioning. This next generation communication resulted in noticeable speedups. In the third category, improvements with respect to better management of memory were implemented. This again resulted in a speedup of nearly 10%. The overall scalability, as a result of the three algorithmic implementations, yielded ideal and at times superlinear scalability until 3000 processors. In general, the scalability results are very promising and indicate that the approach has a great potential for more complicated multidisciplinary problems such as fluid-structure interaction and aeroelastic simulations.

International journal for numerical methods in biomedical engineering, 2012

Near infrared spectroscopy (NIRS) is a technique used to detect and measure changes in the concen... more Near infrared spectroscopy (NIRS) is a technique used to detect and measure changes in the concentrations of oxygenated hemoglobin, deoxygenated hemoglobin, and water in tissues based on the differential absorption, scattering, and refraction of the near infrared light. In this imaging technique, the optical properties of tissues are reconstructed from the measurements obtained from the sensors located on the boundary. A computational method for the rapid noninvasive detection ∕ quantification of cerebral hemorrhage is described using the above procedure. CFD Research Corporation's finite volume computational biology code was used to numerically mimic the NIRS procedure by (i) noninvasively 'numerically penetrating' the brain tissues and (ii) reconstructing the optical properties the presence of water, oxygenated, and deoxygenated blood. These numerical noninvasive measurements are then used to predict the extent and severity of the brain hemorrhage. The paper also discu...

Journal of pulmonary medicine & respiratory care, 2022

The efficacy of pediatric oral drug delivery using dry powder inhalers, such as Turbuhaler®, is d... more The efficacy of pediatric oral drug delivery using dry powder inhalers, such as Turbuhaler®, is dependent on the age and health of the test subjects. The available clinical data for these studies is scant and rarely provide correlations between the health condition and the regional lung deposition. In particular, the data and the correlations for pre-school children are minimal. Deposition simulations were performed using the newly developed Quasi-3D whole lung model to analyze the effect of health conditions on the regional lung deposition from the Turbuhaler® in 3-year-old children. The healthy lung model was created from CT scan data. Cystic-fibrosis models were created by uniformly constricting the airways to various degrees. The simulated drug deposition outcomes were validated against the available experimental data. The results show that, while the dose deposited in the lungs exhibits minor variations, the Peripheral:Central (P/C) ratio is strongly affected by both the health...

International journal for numerical methods in biomedical engineering, 2020

Hydroxychloroquine (HCQ) is commonly used in the treatment of malaria and rheumatic diseases. Rec... more Hydroxychloroquine (HCQ) is commonly used in the treatment of malaria and rheumatic diseases. Recently it has also been identified as possible therapeutic option in combating COVID-19. However, the use of HCQ is known to induce cytotoxicity. In 2020, we developed a Multiscale Absorption and Transit (MAT) toolkit to simulate the dissolution, transport, absorption, distribution, metabolism, and elimination of orally administered drugs in the human GIT at multiple levels. MAT was constructed by integrating the spatially accurate first-principles driven high-fidelity drug transport, dissolution, and absorption model in the human stomach and GIT using the recently published Quasi-3D framework. The computational results showed that MAT was able to match the experimental concentration results better than the traditional compartmental models. In this study, we adapted MAT, to predict the pharmacokinetics of orally delivered HCQ in healthy subjects. The computational results matched the expe...

Existing computational models used for simulating the flow and species transport in the human air... more Existing computational models used for simulating the flow and species transport in the human airways are zero-dimensional (0D) compartmental, three-dimensional (3D) computational fluid dynamics (CFD), or the recently developed quasi-3D (Q3D) models. Unlike compartmental models, the full CFD and Q3D models are physiologically and anatomically consistent in the mouth and the upper airways, since the starting point of these models is the mouth–lung surface geometry, typically created from computed tomography (CT) scans. However, the current resolution of CT scans limits the airway detection between the 3rd–4th and 7th–9th generations. Consequently, CFD and the Q3D models developed using these scans are generally limited to these generations. In this study, we developed a method to extend the conducting airways from the end of the truncated Q3D lung to the tracheobronchial (TB) limit. We grew the lung generations within the closed lung lobes using the modified constrained constructive ...

Until today the modeling of human body biomechanics poses many great challenges because of the co... more Until today the modeling of human body biomechanics poses many great challenges because of the complex geometry and the substantial heterogeneity of human body. We developed a detailed human body finite element model in which the human body is represented realistically in both the geometry and the material properties. The model includes the detailed head (face, skull, brain, and spinal cord), the skeleton, and air cavities (including the lung). Hence it can be used to accurately acquire the stress wave propagation in the human body under various loading conditions. The blast loading on the human surface was generated from the simulated C4 blast explosions, via a novel combination of 1-D and 3-D numerical formulations. We used the explicit finite element solver in the multi-physics code CoBi for the human body biomechanics. This is capable of solving the resulting large system containing millions of unknowns in an extremely scalable fashion. The meshes generated for these simulations are of good quality. This enables us to employ relatively large time step sizes, without resorting to the artificial time scaling treatment. In order to study the human body dynamic response under the blast loading, we also developed an interface to apply the blast pressure loading on the external human body surface. These newly developed models were used to conduct parametric simulations to find out the brain biomechanical response when the blasts impact the human body. Under the same blast loading we also show the differences of brain response when having different material properties for the skeleton, the existence of other body parts such as torso.

Volume 2: Biomedical and Biotechnology, 2012

Previously we had developed an articulated human body model to simulate the kinematic response to... more Previously we had developed an articulated human body model to simulate the kinematic response to the external loadings, using CFDRC’s CoBi implicit multi-body solver. The anatomy-based human body model can accurately account for the surface loadings and surface interactions with the environment. A study is conducted to calibrate the joint properties (for instance, the joint rotational damping) of the articulated human body by comparing its response with those obtained from the PMHS test under moderate loading conditions. Additional adjustments in the input parameters also include the contact spring constants for joint stops at different joint locations. By comparing the computational results with the real scenarios, we fine tune these input parameters and further improve the accuracy of the articulated human body model. In order to simulate the effect of a C4 explosion on a human body in the open field, we employ a CFD model with a good resolution and the appropriate boundary treat...

ABSTRACT The use of overset grids in CFD started more than two decades ago, and has achieved trem... more ABSTRACT The use of overset grids in CFD started more than two decades ago, and has achieved tremendous success in handling complex geometries[6]. In particular, overset grids have the advantage of avoiding grid remeshing when dealing with moving boundary flow problems. In this paper, two particular unstructured grids are advocated for moving boundary flow simulation, i.e., the use of overset adaptive Cartesian/prism grids. Semi-structured prism grids are generated around solid walls. These prism grids then overlap a single adaptive Cartesian background grid. With the adaptive Cartesian grid, the mesh resolution of the prism grid near the outer boundary can easily match that of the oversetting Cartesian grid cells. In addition, the tree-based data structure of the Cartesian grid can be used efficiently in hole-cutting and donor cell identification. In order to expedite the simulations, the entire solver was made capable of running in parallel. The overset adaptive Cartesian/prism grid method is tested for both steady and unsteady flow computation. It is demonstrated that moving boundary flow computations can be automated with minimum user interferences.

ABSTRACT Near-infrared (NIR) tomography is an imaging technique, in which the optical properties ... more ABSTRACT Near-infrared (NIR) tomography is an imaging technique, in which the optical properties of tissues are reconstructed from the measurements obtained from the sensors located on the boundary. In this paper, we describe a computational method for rapid non-invasive detection/quantification of blast lung injury using the above. The Near-infrared spectroscopy (NIRS) technique is simulated using computational models and thus in principle mimics an actual NIRS procedure which penetrates non-invasively through the thoracic wall, pleural region and the lungs to detect (and subsequently reconstruct their properties) the presence of water, oxygenated and de-oxygenated blood. These ‘pseudo’ measurements can then be used to predict the extent and severity of the lung injury. The paper also discusses ideas to obtain the location, radius and the severity of a localized injury. Simulations are performed as a proof of concept for NIRS being feasible for the above-mentioned detection/quantification. Copyright © 2010 John Wiley & Sons, Ltd.

The convergence of high-order methods, such as recently developed spectral difference (SD) method... more The convergence of high-order methods, such as recently developed spectral difference (SD) method, can be accelerated using both implicit temporal advancement and a p-multigrid (p = polynomial degree) approach. A p-multigrid method is investigated in this paper for solving SD formulations of the scalar wave and Euler equations on unstructured grids. A fast preconditioned lower-upper symmetric Gauss-Seidel (LU-SGS) relaxation method is implemented as an iterative smoother. Meanwhile, a Runge-Kutta explicit method is employed for comparison. The multigrid method considered here is nonlinear and utilizes full approximation storage (FAS) [Ta'asan S. Multigrid one-shot methods and design strategy, Von Karman Institute Lecture Note, 1997 [28]] scheme. For some p-multigrid calculations, blending implicit and explicit smoothers for different p-levels is also studied. The p-multigrid method is firstly validated by solving both linear and nonlinear 2D wave equations. Then the same idea is extended to 2D nonlinear Euler equations. Generally speaking, we are able to achieve speedups of up to two orders using the p-multigrid method with the implicit smoother.

A curvature and entropy based wall boundary condition is implemented in the high order spectral v... more A curvature and entropy based wall boundary condition is implemented in the high order spectral volume (SV) context. This method borrows ideas from the ''curvature-corrected symmetry technique'' developed by (Dadone A, Grossman B. Surface Boundary Conditions for Compressible Flows. AIAA J 1994; 32(2): 285-93), for a low order structured grid Euler solver. After numerically obtaining the curvature, the right state (by convention, the left state is inside the computational domain and the right state lies outside of the computational domain) face pressure values are obtained by solving a linearised system of equations. This is unlike that of the lower order finite volume and difference simulations, wherein the right state face values are trivial to obtain. The right state face density values are then obtained by enforcing entropy conservation. Accuracy studies show that simulations performed by employing the new boundary conditions deliver much more accurate results than the ones which employ traditional boundary conditions, while at the same time asymptotically reaching the desired order of accuracy. Numerical results for two-dimensional inviscid flows around the NACA0012 airfoil and over a bump with the new boundary condition showed dramatic improvements over those with the conventional approach. In all cases and orders, spurious entropy productions with the new boundary treatment are significantly reduced. In general, the numerical results are very promising and indicate that the approach has a great potential for 3D high order simulations.

In this paper, we obtain inherently unsteady solutions to the Navier-Stokes equations involving m... more In this paper, we obtain inherently unsteady solutions to the Navier-Stokes equations involving moving boundaries. We employ a mapping function to map the grid and the flow features between a fixed reference frame and a moving reference frame. The actual equations of conservation (applicable on the moving reference frame) are then rewritten so as to form an altered set of equations, which are valid in the fixed reference frame. These altered set of equations are discretized and then solved using the high order spectral volume method (SV). The time advancement is carried out using the three stage Runge Kutta method. Simulations are performed to demonstrate the proof of the above concept and the ability of this method to handle more complicated motions.

The direct discontinuous Galerkin (DDG) method was developed by Liu and Yan to discretize the dif... more The direct discontinuous Galerkin (DDG) method was developed by Liu and Yan to discretize the diffusion flux. It was implemented for the discontinuous Galerkin (DG) formulation. In this paper, we perform four tasks: (i) implement the direct discontinuous Galerkin (DDG) scheme for the spectral volume method (SV) method, (ii) design and implement two variants of DDG (called DDG2 and DDG3) for the SV method, (iii) perform a Fourier type analysis on both methods when solving the 1D diffusion equation and combine the above with a non-linear global optimizer, to obtain modified constants that give significantly smaller errors (in 1D), (iv) use the above coefficients as starting points in 2D. The dissipation properties of the above schemes were then compared with existing flux formulations (local discontinuous Galerkin, Penalty and BR2). The DDG, DDG2 and DDG3 formulations were found to be much more accurate than the above three existing flux formulations. The accuracy of the DDG scheme is heavily dependent on the penalizing coefficient for the odd ordered schemes. Hence a loss of accuracy was observed even for mildly non-uniform grids for odd ordered schemes. On the other hand, the DDG2 and DDG3 schemes were mildly dependent on the penalizing coefficient for both odd and even orders and retain their accuracy even on highly irregular grids. Temporal analysis was also performed and this yielded some interesting results. The DDG and its variants were implemented in 2D (on triangular meshes) for Navier-Stokes equations. Even the non-optimized versions of the DDG displayed lower errors than the existing schemes (in 2D). In general, the DDG and its variants show promising properties and it indicates that these approaches have a great potential for higher dimension flow problems.

In this paper, we improve the Navier-Stokes flow solver developed by Sun et al based on the spect... more In this paper, we improve the Navier-Stokes flow solver developed by Sun et al based on the spectral volume method in the following two aspects: the development of a more efficient implicit/p-multigrid solution approach, and the use of a new viscous flux formula. An implicit preconditioned LU-SGS p-multigrid method developed for the spectral difference (SD) Euler solver by Liang et al is adopted here. In the original SV solver, the viscous flux was computed with a local discontinuous Galerkin-type approach. In this study, a penalty approach based on the first method of Bassi and Rebay is suggested and tested for both the Laplace and Navier-Stokes equations. The second method of Bassi and Rebay is also tested for the Laplace equation. Their convergence properties are studied in the context of the BLU-SGS approach. Fourier analysis revealed some interesting advantages for the penalty method over the LDG method. A convergence speedup of up to 2 orders is obtained with the implicit method. The convergence was further enhanced by employing a p-multigrid algorithm. The numerical results are very promising and indicate that the approach has a great potential for 3D flow problems.