Computational Fluid Dynamics (CFD) permits the virtual simulation of fluid flow and thermal applications. Clients that leverage our CFD services do so because CFD is typically faster, less expensive and provides more detailed insight than physical testing.
But what are all the things that it can do? Based on over a decade of consulting on hundreds of projects, I will try to break it down in a concise summary here. Along the way, I will also expand on what aspects are easy, which ones are more difficult (expensive) and those that we encounter most often.
Let’s start with adiabatic fluid flow. Laminar flow is the easiest, in terms of computation complexity, to simulate. Once the velocity of the fluid exceeds a certain threshold, defined by the Reynold’s number, it becomes turbulent and requires different solver settings to simulate. FYI, for a garden hose, this threshold occurs at a rate of about 6.7 gpm.
It is usually helpful to understand basic parameters like these before setting up simulation or to validate CFD results. Since we do this quite often, we have setup up spreadsheets and scripts to expedite the process.
The vast majority of applications we simulate use air and water, which are both classified as Newtonian fluids. However, we are also called on to simulate nonNewtonian fluids, where the viscosity is dependent on the local shear state (blood, ketchup, milkshake).
Air is technically a compressible fluid (density is influenced by pressure). However, for many projects, air is assumed to be incompressible since the velocities and pressure differentials are low enough so that any impact on density is nominal. This assumption is not valid for high velocity/pressure applications and requires special attention to the physics and even the mesh quality to accurately capture shock waves.
Figure 1: Compressible flow, such as this barrel shock, is a more challenging CFD application
Now, let’s introduce heat transfer into the equation. The technical term for the combination of fluid flow and thermal analysis is conjugate heat transfer. In certain situations, the temperatures will also influence the fluid flow characteristics, as detailed in the blog, Convection, Convection, which Convection? . Conduction and convection are standard, but when higher temperatures are involved, radiation also comes into play and adds another layer of complexity. CFD can even account for radiation from the sun, if needed.
Figure 2: Example of conjugate heat transfer with forced convection.
CFD has even more complex capabilities than what was outlined here, but we will table that for later discussions. So what is our most common application? That would be conjugate heat transfer with an incompressible Newtonian fluid. This is because of its wide range of popular applications, from electronics cooling all the way up to entire building simulations.