Publications

Ridzal, Denis R.

Abstract not provided.

Bochev, Pavel B.; Moe, Scott M.; Peterson, Kara J.; Ridzal, Denis R.

Abstract not provided.

COUPLED PROBLEMS 2015 - Proceedings of the 6th International Conference on Coupled Problems in Science and Engineering

Bochev, Pavel B.; Moe, Scott A.; Peterson, Kara J.; Ridzal, Denis R.

We present a new optimization-based, conservative, and quasi-monotone method for passive tracer transport. The scheme combines high-order spectral element discretization in space with semi-Lagrangian time stepping. Solution of a singly linearly constrained quadratic program with simple bounds enforces conservation and physically motivated solution bounds. The scheme can handle efficiently a large number of passive tracers because the semi-Lagrangian time stepping only needs to evolve the grid points where the primitive variables are stored and allows for larger time steps than a conventional explicit spectral element method. Numerical examples show that the use of optimization to enforce physical properties does not affect significantly the spectral accuracy for smooth solutions. Performance studies reveal the benefits of high-order approximations, including for discontinuous solutions.

van Bloemen Waanders, Bart G.; Kouri, Drew P.; Ridzal, Denis R.; Uryasef, Stan U.

Abstract not provided.

Ridzal, Denis R.; Kouri, Drew P.; van Bloemen Waanders, Bart G.

Abstract not provided.

Peterson, Kara J.; Ridzal, Denis R.; Bochev, Pavel B.

Abstract not provided.

Ridzal, Denis R.; Kouri, Drew P.

Abstract not provided.

Ridzal, Denis R.

Abstract not provided.

Perego, Mauro P.; Salinger, Andrew G.; Phipps, Eric T.; Ridzal, Denis R.; Kouri, Drew P.; Kalashnikova, Irina; Price, S.P.; Stadler, G.S.

Abstract not provided.

Walsh, Timothy W.; Aquino, Wilkins A.; Ridzal, Denis R.; Kouri, Drew P.; van Bloemen Waanders, Bart G.; Urbina, Angel U.

In this report we derive frequency-domain methods for inverse characterization of the constitutive parameters of viscoelastic materials. The inverse problem is cast in a PDE-constrained optimization framework with efficient computation of gradients and Hessian vector products through matrix free operations. The abstract optimization operators for first and second derivatives are derived from first principles. Various methods from the Rapid Optimization Library (ROL) are tested on the viscoelastic inversion problem. The methods described herein are applied to compute the viscoelastic bulk and shear moduli of a foam block model, which was recently used in experimental testing for viscoelastic property characterization.

Walsh, Timothy W.; Aquino, Wilkins A.; Ridzal, Denis R.; Kouri, Drew P.; van Bloemen Waanders, Bart G.

Abstract not provided.

Peterson, Kara J.; Bochev, Pavel B.; Ridzal, Denis R.

Abstract not provided.

Ridzal, Denis R.

Abstract not provided.

Peterson, Kara J.; Bochev, Pavel B.; Ridzal, Denis R.

Abstract not provided.

Aguilo Valentin, Miguel A.; Ridzal, Denis R.

A theoretical framework for the numerical solution of partial di erential equation (PDE) constrained optimization problems is presented in this report. This theoretical framework embodies the fundamental infrastructure required to e ciently implement and solve this class of problems. Detail derivations of the optimality conditions required to accurately solve several parameter identi cation and optimal control problems are also provided in this report. This will allow the reader to further understand how the theoretical abstraction presented in this report translates to the application.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Peterson, Kara J.; Bochev, Pavel B.; Ridzal, Denis R.

Transport algorithms are highly important for dynamical modeling of the atmosphere, where it is critical that scalar tracer species are conserved and satisfy physical bounds. We present an optimization-based algorithm for the conservative transport of scalar quantities (i.e. mass) on the cubed sphere grid, which preserves local solution bounds without the use of flux limiters. The optimization variables are the net mass updates to the cell, the objective is to minimize the discrepancy between these variables and suitable high-order cell mass update (the "target"), and the constraints are derived from the local solution bounds and the conservation of the total mass. The resulting robust and efficient algorithm for conservative and local bound-preserving transport on the sphere further demonstrates the flexibility and scope of the recently developed optimization-based modeling approach [1, 2]. © 2014 Springer-Verlag.

Love, Edward L.; Robinson, Allen C.; Ridzal, Denis R.

We review the edge element formulation for describing the kinematics of hyperelastic solids. This approach is used to frame the problem of remapping the inverse deformation gradient for Arbitrary Lagrangian-Eulerian (ALE) simulations of solid dynamics. For hyperelastic materials, the stress state is completely determined by the deformation gradient, so remapping this quantity effectively updates the stress state of the material. A method, inspired by the constrained transport remap in electromagnetics, is reviewed, according to which the zero-curl constraint on the inverse deformation gradient is implicitly satisfied. Open issues related to the accuracy of this approach are identified. An optimization-based approach is implemented to enforce positivity of the determinant of the deformation gradient. The efficacy of this approach is illustrated with numerical examples.

SIAM Journal on Scientific Computing

Ridzal, Denis R.; Kouri, Drew P.; van Bloemen Waanders, Bart G.

Abstract not provided.

Bochev, Pavel B.; Ridzal, Denis R.; Peterson, Kara J.

Abstract not provided.

Ridzal, Denis R.; Bochev, Pavel B.; Peterson, Kara J.

Abstract not provided.

Peterson, Kara J.; Bochev, Pavel B.; Ridzal, Denis R.

Abstract not provided.

Walsh, Timothy W.; Ridzal, Denis R.; Aguilo Valentin, Miguel A.

Abstract not provided.

Ridzal, Denis R.

Abstract not provided.

Peterson, Kara J.; Bochev, Pavel B.; Ridzal, Denis R.

Abstract not provided.

Young, Joseph G.; Ridzal, Denis R.

In the following paper, we discuss how to design an ensemble of experiments through the use of compressed sensing. Specifically, we show how to conduct a small number of physical experiments and then use compressed sensing to reconstruct a larger set of data. In order to accomplish this, we organize our results into four sections. We begin by extending the theory of compressed sensing to a finite product of Hilbert spaces. Then, we show how these results apply to experiment design. Next, we develop an efficient reconstruction algorithm that allows us to reconstruct experimental data projected onto a finite element basis. Finally, we verify our approach with two computational experiments.