Until recently, many users of CAE software viewed computational fluid dynamics (CFD) as a luxury item. The gist of frequent comments has been: “It would be nice to study the effects of fluid flow on our product designs, but we’ve heard CFD is too specialized and tough to learn. Plus, analyses take a long time to run and are demanding of computer resources. We can’t justify the additional expense for CFD software. Besides, we can rely on structural analysis and get by with overdesigning a bit to account for fluid-flow effects.”

Who, What, Where

Authored by Bob Williams
Product Manager
Algor Inc.
Pittsburgh, Pa.

Edited by Paul Dvorak

Key points
• Recent CFD software is easier to use than previous versions.
• Hardware requirements are also reduced.

Algor Inc., algor.com

The good news is that the old objections don’t hold true anymore because today’s CFD software is easier to learn and use than ever. And thanks to more reliable solvers and more powerful computer hardware, CFD analyses run much faster. Most importantly, for many products, understanding — and adequately accounting for — fluid flow is essential to ensuring performance and safety.

More jobs for CFD
Traditionally, CFD analysis has simulated the flow of a fluid around obstacles and through hollow areas to control temperatures, reduce resistance to flow, or optimize phenomena such as lift. Today, CFD analyses are often used, for example, to study flow-driven temperature changes in applications such as heat exchangers, circuit boards, and cooling and heating systems.

To show the relative ease-of-use of recent CFD software, let’s examine a steady state fluid flow coupled with the thermal analysis of a closed-loop, liquid-filled heat exchanger. The model demonstrates convective flow due to fluid buoyancy at varying temperatures and the effect of the resulting flow on the temperature distribution.

Recent developments in CFD software have expanded the types of applications that even casual users can undertake. For example, some CFD packages provide tools that automate model building and include the ability to import models from CAD packages so CFD analysis can then be performed on the exact geometry. After importing a CAD model, a user specifies boundaries for the fluid region using a built-in dialog and the software creates new parts or geometry where flow will be analyzed. For a valve assembly (not shown), users would specify surfaces for modeling the fluid in the CFD software. It would then generate a new part in which to perform the flow analysis.

Another CFD-modeling capability is boundary-layer meshing. Velocity changes the most where the fluid interacts with boundary surfaces, such as the inside a tube. Hence, a fine mesh is needed in these areas to accurately simulate flow behavior. By simply activating a dialog option, users instruct the software to generate a finer mesh near the surface while keeping a coarser mesh throughout the rest of the fluid domain. Thus, it is easy to create a finite-element mesh that simulates the behavior of fluid flow along its boundary.

In addition to modeling features, today’s CFD software provides tools that let users more easily simulate applications such as:

Flow through a porous media. Instead of modeling flow obstacles such as ground rock, filters, and catalyst beds with the detail necessary to find pressure and velocity gradients, users can model larger-scale parts with specified distributed resistance. This simplifies 3D modeling and reduces model size. For example, properly simulating underground flow through porous rock can predict the movement of contaminated fluid from a solid-waste landfill into a drinking-water supply. In industrial applications, harmful particles can be filtered from a stream by passing it through a porous solid with pores too small to permit passage of the particles. Additionally, porous media may provide sites for chemical catalysis or absorption of fluid components.

Mass-transfer simulation, also called species transport, calculates a mass transfer due to diffusion (i.e., determines mass in transit due to gradients in the concentration of species within a mixture under random molecular motion). Typical applications include mixing, particle tracing, and drug delivery (by passing chemical species through a membrane).

Open channel flow simulates a free surface between a flowing fluid and a gas above it. Typical applications include marine systems, drainage systems, and liquid- column gages. In such flows, liquids and gases are clearly separated, not mixing or interpenetrating. The density ratio between the fluids is quite large. Flow is generally governed by gravity and inertia. Due to low density and negligible viscosity, the inertia and viscous force of the gas are also negligible. So the only impact of the gas is its pressure acting on the interface. Hence, the gas region need not be analyzed. The free surface is therefore calculated as a boundary with constant pressure (for example, zero pressure by totally ignoring the air effect).

Coupled multiphysics considers two or more physical effects that influence each other (such as fluid flow and heat transfer), and couples them into one analysis for more accurate results than those from performing separate manual analyses. This manual method would first run the fluid-flow analysis, then input the nodal velocity results as a loading to a subsequent heat-transfer analysis. Using coupled analysis, heat transfer due to fluid flow will be accounted for, as well as natural convection (buoyancy) in the fluid due to temperatures. This helps in situations in which steady-state solutions may not exist due to instabilities created by buoyancy effects.