Computational fluid dynamics simulations (CFD) can serve many useful purposes. Just a few that come to mind are:
- Aircraft and UAV design and performance enhancements
- Heat transfer analysis for heating and cooling systems and electronics
- Sizing ductwork and blower motors for HVAC systems
- Wind loads on various structures
- Motor vehicle design
This is just to name a few. There are thousands of potential applications. As with any analysis, the real trick is to gain confidence that your simulations can accurately reflect the design under analysis. Having experience in both analytical predictions and real-world testing can be of great benefit. Over my several decades in the engineering business I have been asked to predict fluid dynamic effects on a wide range of objects and vehicles. Participating in many wind tunnel tests, reducing the data, and writing the reports has provided a very sound basis for making the jump from analytical predictions to reality.
Along the way, I have reviewed literally hundreds of test reports, whether they be wind tunnel, flight tests, aero-ballistic gun range, aircraft munition drop tests, parachute experiments, hot-wire anemometer tests, motor vehicle drag, both small and large caliber projectile tests, destructive testing, and so on.
Sanity Checks
By “sanity checks” I mean to be able to run a quick hand calculation on a design or design change using classic aerodynamic analysis methods that is being analyzed via finite element analysis, providing the ability to confirm or reject the validity of the analysis. Computational fluid dynamics, or CFD, has quite a large number of settings that need to be correct in order for the analyst to have confidence in the accuracy of his computerized design study. All the color – contour plots, visible streamlines, etc. – resulting from FEA looks very impressive, but should never be taken at face value.
A More Simplified Approach
Sometimes, a gradual approach, where the model is first analyzed in a more simplified state, with design complexity gradually added on can be very beneficial. Most CFD analyses are nonlinear, and as such can be tricky to get solutions in a timely manner. A stepped approach can reveal where the program is having difficulty finding a solution. Small changes in flow conditions, mesh density, turbulence models, etc. (this is a long list!) should have predictable results.
Interpolation and Extrapolation
There are times FEA can be used to interpolate or extrapolate for the final result. A short review here.
- Interpolation, simply stated, is when you have test or analysis results that can be plotted and your task is to estimate the design change that will place the new data between two of those data points.
- Extrapolation is where on a similar graph the data you wish to acquire forces you to extend the graph at either end.
The task at hand is then one of estimating a trend as opposed to absolute values. In this manner, previously acquired data can work very well combined with CFD analysis methods.
Balancing sophisticated computer methods for computational fluid dynamic analysis with test results and some hand calculations will help control costs in a big way.
Norman Neher
Analytical Engineering Services, inc
Elko New Market, MN
www.aesmn.org