Enzo A. Dari

Anisotropic mesh optimization and its application in adaptivity

The construction of solution-adapted meshes is addressed within an optimization framework. An approximation of the second spatial derivative of the solution is used to get a suitable metric in the computational domain. A mesh quality is proposed and optimized under this metric, accounting for both the shape and the size of the elements. For this purpose, a topological and geometrical mesh improvement method of high generality is introduced. It is shown that the adaptive algorithm that results recovers optimal convergence rates in singular problems, and that it captures boundary and internal layers in convection-dominated problems. Several important implementation issues are discussed.

OPTIMIZATION STRATEGIES IN UNSTRUCTURED MESH GENERATION

SUMMARY We propose a new optimization strategy for unstructured meshes that, when coupled with existing automatic generators, produces meshes of high quality for arbitrary domains in 3-D. Our optimizer is based upon a non-differentiable definition of the quality of the mesh which is natural for finite element or finite volume users: the quality of the worst element in the mesh. The dimension of the optimization space is made tractable by restricting, at each iteration, to a suitable neighbourhood of the worst element. Both geometrical (node repositioning) and topological (reconnection) operations are performed. It turns out that the repositioning method is advantageous with respect to both the usual node-by-node techniques and the more recent differentiable optimization methods. Several examples are included that illustrate the efficiency of the optimizer.

Blood flow distribution in an anatomically detailed arterial network model: criteria and algorithms

Development of blood flow distribution criteria is a mandatory step toward developing computational models and numerical simulations of the systemic circulation. In the present work, we (i) present a systematic approach based on anatomical and physiological considerations to distribute the blood flow in a 1D anatomically detailed model of the arterial network and (ii) develop a numerical procedure to calibrate resistive parameters in terminal models in order to effectively satisfy such flow distribution. For the first goal, we merge data collected from the specialized medical literature with anatomical concepts such as vascular territories to determine blood flow supply to specific (encephalon, kidneys, etc.) and distributed (muscles, skin, etc.) organs. Overall, 28 entities representing the main specific organs are accounted for in the detailed description of the arterial topology that we use as model substrate. In turn, 116 vascular territories are considered as the basic blocks that compose the distributed organs throughout the whole body. For the second goal, Windkessel models are used to represent the peripheral beds, and the values of the resistive parameters are computed applying a Newton method to a parameter identification problem to guarantee the supply of the correct flow fraction to each terminal location according to the given criteria. Finally, it is shown that, by means of the criteria developed, and for a rather standard set of model parameters, the model predicts physiologically realistic pressure and flow waveforms.

Simulation of highly elastic fluid flows without additional numerical diffusivity

Abstract A transient decoupled method to treat viscoelastic flows is presented, using Lagrange-Galerkin techniques for velocities and the Lesaint-Raviart method for extra-stresses. An important point of our scheme is the linear interpolation employed for all fields: this allows for high Weissenberg numbers to be reached with no need for streamline upwinding (SU). Numerical results for the 4:1 plane contraction are presented, including comparisons with other methods and an analysis of the effect of mesh refinement and SU addition.

NUMERICAL INVESTIGATION OF FLOW THROUGH A CAVITY WITH INTERNAL HEAT GENERATION

Laminar flow through a two-dimensional square cavity in which internal (volumetric) heat generation takes place is analyzed numerically. The inflow is a horizontal jet at the bottom of the cavity, and the outlet is at the top, in a centered position. As the internal heat generation rate is increased, several flow patterns appear, in particular periodic and quasi-periodic in time flows. Two distinct regimes are identified, one dominated by inertia in which significant hot-fluid trapping occurs, and another, stratified regime, in which buoyancy effects are dominant. The effects of Reynolds number, Prandtl number, horizontal heat generation inhomogeneities, and wall boundary conditions are assessed. Of technological importance is the inertia-dominated regime, as temperatures much higher than the mean outflow temperature appear.

Numerical Experiments in Complex Hæmodynamic Flows. Non-Newtonian Effects

Numerical experiments for non-trivial flows, close to realistic situations in hæmodynamics, are described and interpreted. Two geometries have been selected: an axisymmetric corrugated tube (with periodic boundary conditions) and a 3D bifurcation with an obstructed end (anastomosis). Results concern sensitivity of errors associated to the time-step size and mesh refinement, but essentially consist of the quantitative estimation of non-Newtonian effects based on Cassons rheological model, treated in retarded form. The time-step lag of such effects is the main reason for evaluating the sensitivity of errors. Due to the high computational cost characterizing the problems to be faced, we expect that the present results will be useful when real geometries should be modeled. The main conclusions are that non-Newtonian effects may be relevant (especially for secondary flows) and that, in most cases, for the same level of errors the use of Cassons law does not generate excessive additional computational costs. Thus, within this strategy, the user can accurately solve the problem using this rheological model without having to worry if the non-Newtonian effects are important or not.

Finite Element Modeling of Liquid Deuterium Flow and Heat Transfer in a Cold-neutron Source

A finite element simulation of flow and heat transfer in the moderator cell of a cold-neutron source (CNS), in which liquid deuterium subject to internal heat generation is flowing, is reported. The numerical scheme consists of a stabilized equal-order method. A time-accurate approach is adopted to resolve the large-scale eddies of the flow, with a Smagorinskys model for the subgrid-scale effects. The thermal coupling follows a staggered strategy, with SUPG-type upwinding. A specific wall-law is developed that accounts for the correct partition of the heat deposited at the wall by radiation between the liquid deuterium and the helium gas flowing at the outer side of the wall. The average flow and thermal structure are presented. The turbulent fluctuations are both illustrated in physical space and decomposed into spectral components. The wavenumber spectrum suggests that adequate resolution of the large-scale eddies has been attained with just 200,000 nodes, while a DNS analysis would have required at least 1010 nodes. Usefulness of the approach in the design process of the CNS is highlighted.

Transversally enriched pipe element method (TEPEM): An effective numerical approach for blood flow modeling.

In this work, we present a novel approach tailored to approximate the Navier-Stokes equations to simulate fluid flow in three-dimensional tubular domains of arbitrary cross-sectional shape. The proposed methodology is aimed at filling the gap between (cheap) one-dimensional and (expensive) three-dimensional models, featuring descriptive capabilities comparable with the full and accurate 3D description of the problem at a low computational cost. In addition, this methodology can easily be tuned or even adapted to address local features demanding more accuracy. The numerical strategy employs finite (pipe-type) elements that take advantage of the pipe structure of the spatial domain under analysis. While low order approximation is used for the longitudinal description of the physical fields, transverse approximation is enriched using high order polynomials. Although our application of interest is computational hemodynamics and its relevance to pathological dynamics like atherosclerosis, the approach is quite general and can be applied in any internal fluid dynamics problem in pipe-like domains. Numerical examples covering academic cases as well as patient-specific coronary arterial geometries demonstrate the potentialities of the developed methodology and its performance when compared against traditional finite element methods. Copyright