Searcharxiv⌕ Search

arXiv subjects

M. Ramakrishna

Publications and source records attributed to M. Ramakrishna.

8 recordsLinked to original sources
arXiv2025

A study of troubled-cell indicators applied to finite volume methods using a novel monotonicity parameter

We adapt a troubled-cell indicator developed for discontinuous Galerkin (DG) methods to the finite volume method (FVM) framework for solving hyperbolic conservation laws. This indicator depends solely on the cell-average data of the target cell and its immediate neighbours. Once the troubled-cells are identified, we apply the limiter only in these cells instead of applying in all computational cells. We introduce a novel technique to quantify the quality of the solution in the neighbourhood of the shock by defining a monotonicity parameter $\mu$. Numerical results from various two-dimensional simulations on the hyperbolic systems of Euler equations using a finite volume solver employing MUSCL reconstruction validate the performance of the troubled-cell indicator and the approach of limiting only in the troubled-cells. These results show that limiting only in the troubled-cells is preferable to limiting everywhere as it improves convergence without compromising on the solution accuracy.

math.NA↗
arXiv2020

Fast and Accurate Proper Orthogonal Decomposition using Efficient Sampling and Iterative Techniques for Singular Value Decomposition

In this paper, we propose a computationally efficient iterative algorithm for proper orthogonal decomposition (POD) using random sampling based techniques. In this algorithm, additional rows and columns are sampled and a merging technique is used to update the dominant POD modes in each iteration. We derive bounds for the spectral norm of the error introduced by a series of merging operations. We use an existing theorem to get an approximate measure of the quality of subspaces obtained on convergence of the iteration. Results on various datasets indicate that the POD modes and/or the subspaces are approximated with excellent accuracy with a significant runtime improvement over computing the truncated SVD. We also propose a method to compute the POD modes of large matrices that do not fit in the RAM using this iterative sampling and merging algorithms.

math.NA↗
arXiv2020

Addressing the issue of mass conservation error and the connected problem of Carbuncle formation

We study mass conservation errors (momentum density spike) and the related phenomenon of post shock oscillations in numerical solutions of compressible Euler equations. These phenomena and their causes have been reported in literature [34, 1]. In this paper, first, we compare the mass conservation and post shock oscillation errors obtained using combinations of different numerical methods (Finite Volume, Finite Difference with WENO and DG with simple WENO limiter) and upwind flux functions (ROE, AUSM+-up, and others) for moving shocks, modelled using one-dimensional Euler equations. Next, the mass conservation error is quantified for stationary shocks modelled using one-dimensional, quasi-one-dimensional and two-dimensional Euler equations. It is shown that using a fine mesh or refining mesh near shocks using multiple over set meshes lead to mitigation of the mass conservation error. This is demonstrated using the problem of flow through a variable area duct modelled using quasi-one-dimensional Euler equations. The link between mass conservation error and carbuncle formation is shown and preliminary results indicating that the carbuncle can be cured using multiple overset meshes are also shown.

math.NA↗
arXiv2019

Modified symmetry technique for mitigation of flow leak near corners for compressible inviscid fluid flow

Using the standard symmetry technique for applying boundary conditions for free slip and flat walls with corners will lead to flow leak through the wall near corners (violation of no penetration condition) and a corresponding error in prediction of pressure. Also, prescribing a state at the corner as a boundary condition is not possible. In this paper, a method for tackling the `corner point state' problem is given and modifications to the standard symmetry technique are proposed to mitigate flow leak near the corner. Using this modified symmetry technique, numerical solutions to the Euler equations for flows over forward facing and backward facing step are computed employing the Shu-Osher conservative finite difference scheme with WENO-NP3 reconstruction (with a formal order of accuracy in space of 3), Lax-Freidrichs flux splitting, and TVD-RK3 time discretisation. It is shown that using this modified symmetry technique leads to mitigation of flow leak near the corner and a better prediction of shock structure even on coarse meshes.

physics.comp-ph↗
arXiv2019

Fast and Accurate Proper Orthogonal Decomposition using Efficient Sampling and Iterative Techniques for Singular Value Decomposition

In this paper, we propose a computationally efficient iterative algorithm for proper orthogonal decomposition (POD) using random sampling based techniques. In this algorithm, additional rows and columns are sampled and a merging technique is used to update the dominant POD modes in each iteration. We derive bounds for the spectral norm of the error introduced by a series of merging operations. We use an existing theorem to get an approximate measure of the quality of subspaces obtained on convergence of the iteration. Results on various datasets indicate that the POD modes and/or the subspaces are approximated with excellent accuracy with a significant runtime improvement over computing the truncated SVD. We also propose a method to compute the POD modes of large matrices that do not fit in the RAM using this iterative sampling and merging algorithms.

math.NA↗
arXiv2017

A Hierarchical Singular Value Decomposition Algorithm for Low Rank Matrices

Singular value decomposition (SVD) is a widely used technique for dimensionality reduction and computation of basis vectors. In many applications, especially in fluid mechanics and image processing the matrices are dense, but low-rank matrices. In these cases, a truncated SVD corresponding to the most significant singular values is sufficient. In this paper, we propose a tree based merge-and-truncate algorithm to obtain an approximate truncated SVD of the matrix. Unlike previous methods, our technique is not limited to "tall and skinny" or "short and fat" matrices and it can be used for matrices of arbitrary size. The matrix is partitioned into blocks and the truncated SVDs of blocks are merged to obtain the final SVD. If the matrices are low rank, this algorithm gives significant speedup over finding the truncated SVD, even when run on a single core. The error is typically less than 3\%.

math.NA↗
arXiv2014

Multi-scale turbulence modeling and maximum information principle. Part 2

We consider two-dimensional homogeneous shear turbulence within the context of optimal control, a multi-scale turbulence model containing the fluctuation velocity and pressure correlations up to the fourth order; The model is formulated on the basis of the Navier-Stokes equations, Reynolds average, the constraints of inequality from both physical and mathematical considerations, the turbulent energy density as the objective to be maximized, and the fourth order correlations as the control variables. Without imposing the maximization and the constraints, the resultant equations of motion in the Fourier wave number space are formally solved to obtain the transient state solutions, the asymptotic state solutions and the evolution of a transient toward an asymptotic under certain conditions. The asymptotic state solutions are characterized by the dimensionless exponential time rate of growth $2σ$ which has an upper bound of $2σ_{\max}$ = 0; The asymptotic solutions can be obtained from a linear objective convex programming. For the asymptotic state solutions of the reduced model containing the correlations up to the third order, the optimal control problem reduces to linear programming with the primary component of the third order correlations or a related integral quantity as the control variable; the supports of the second and third order correlations are estimated for the sake of numerical simulation; the existence of feasible solutions is demonstrated when the related quantity is the control variable. The relevance of the formulation to flow stability analysis is suggested.

physics.flu-dyn↗
arXiv2010

Multi-scale turbulence modeling and maximum information principle. Part 1

We discuss averaged turbulence modeling of multi-scales of length for an incompressible Newtonian fluid, with the help of the maximum information principle. We suppose that there exists a function basis to decompose the turbulent fluctuations in a flow of our concern into the components associated with various spatial scales and that there is a probability density function $\pdf$ of these fluctuation components. The unbiased form for $\pdf$ is determined and the turbulence model is closed, with the multi-scale correlations up to the fourth order, through maximizing the information under the constraints of equality and inequality for that flow. Due to the computational difficulty to maximize the information, a closely related but simple alternative objective is sought, like the determinant or the trace of the second order correlations of the turbulent flow. Some preliminary results and implications from the application to homogeneous turbulence are presented. Some issues yet to be resolved are indicated.

physics.flu-dyn↗