VSPAERO will analyze almost any geometry as long as the components are closed. Whether you are able to obtain meaningful, or reliable, information is another story. There are limits to VSPAERO's abilities to converge toward what you may think is the “right” answer for a given set of flow conditions. Keep in mind that this is a linear solver. VSPAERO will not, for example, model stall characteristics or, in fact, separation of any kind. This is a tool in your toolbox that will help find much of the aerodynamic traits of a model based on a given set of conditions that are within “normal” flight conditions (e.g. cruise, small alpha, small beta) which you can later modify according to whatever physical limits you feel are necessary.
This page will describe modeling practices that should increase your chances of success using VSPAERO. By no means is this list comprehensive. The goal is to prevent the usual pitfalls of the learning process while expanding on “good practice” methods for analyzing models using numerical codes. An introduction to numerical analysis is given through MIT Open Courseware at this link: "Introduction to Numerical Analysis".
A highly detailed model may look nice, but the chances of needing such fine resolution to solve for aerodynamic characteristics is quite low. Usually, a much lower resolution will converge to a solution around the expected value (see figure below). This is true for all numerical methods and holds here as well.
To compare the two models pictured above, the low resolution model has around 4100 cells and will run to completion in a matter of 5-10 minutes while the high resolution model has 19,200 cells and would take hours. The difference between the results of these two would be less than 5% which is much less than the error typically created in model geometry assumptions. Thankfully, there is a happy medium to be found.
Optimizing meshes for numerical solvers is an area of active research and generally follows the following logic: Areas with large gradients get a finer mesh and ignore sections that do not significantly contribute. For example, a fuselage will not significantly contribute to lift or induced drag, so for initial design steps it can generally be left out of most of the analyses. Also, the wing sections can be finely defined to capture effects of downwash or propeller slipstream while leaving most of the wing “coarsely” defined. This is particularly useful when observing low speed conditions where the local Cl is much higher than at cruise. An example of this practice is shown below.
This page was created and edited by: — Brandon Litherland 2015/07/02 10:47