Just kidding, excelletn tutorial!

@@xiaojiang2610 SI units really make things a lot easier!

It would not be expected to get completely different results from changing the computational domain (unless you are making it much too small, smaller than shown in the video here). Most likely what is happening if you are using the "Automatic" mesh settings is that you are also inadvertently changing the mesh density when adjusting the size of the computational domain. Lift and drag calculations are very mesh-sensitive so that would explain large differences. For an apples-to-apples comparison between computational domain sizes you could use "Manual" mesh mode on the global mesh and enter number of basic mesh cells in X,Y,Z direction to maintain a consistent density (such as a 2" mesh size) for the various sizes of the computational domain that you attempt. Divide the length, width, and height of the computational domain by your desired basic mesh cell size to figure out what number to enter. Generally sizing the computational domain can be thought of with two metrics that would be similar to conducting a physical wind tunnel test. First would be percent area of the flow that is obstructed which is sometimes called "blockage ratio". This is the ratio of the frontal area of the vehicle out of the total cross sectional area of the domain. The other would be be the length of the computational domain upstream and downstream from the vehicle, which is often characterized as some multiple of vehicle length (or chord length for an airfoil). Generally additional length is required downstream to resolve the wake. So you can look to literature for these values for general rules of thumb for a given application or geometry (e.g. 5% blockage ratio, 3 lengths upstream, 10 lengths downstream). It will be difficult to predict in advance how much is required for your geometry without performing a sensitivity analysis. But if you keep the mesh density consistent, and vary the two parameters separately (computational domain cross sectional area and upstream/downstream length) I think you should quickly find that the values converge and also identify a point of diminishing returns. Another thing to consider is whether you are trying to replicate a wind tunnel test or the real world behavior. In this example we used free conditions on all the sides of the computational domain which should be closer to the real world behavior, whereas to replicate a physical wind tunnel test it may be better to model a box shape representing wind tunnel walls and perform an internal analysis. Additional considerations to represent "real world" behavior for a ground vehicle would be to incorporate wall motion of the ground using a real wall condition.

@@hawkridgesystems Oh thank you so so much. To be honest, I wasn't even expecting you to reply to me. This is a very detailed reply too. I will read it right after I finish work and will try to implement it tonight on a small project. Thank you so much for your time. Have a great day!

Make sure the CAD files are stored somewhere locally on your computer where you have write permissions. You could try copying them to a folder on your Desktop and try meshing again.

Yes, if the total lift force predicted exceeds the weight of the vehicle then it would be capable of flight. The amount of lift can be adjusted by varying the position of the control surfaces and the angle of attack. A cruise condition would correspond to when the lift force is equal to the weight.

@@hawkridgesystems thanks for the answer!!

The most likely causes are differences in the mesh and/or stopping or "convergence" criteria If you refine the mesh to the point the solution is mesh or "grid independent" and then also solve for a sufficiently long number of iterations, it would be expected that the differences between runs would be very minute

@@hawkridgesystems thank you a lot:D

The forces displayed are only for the portion of the geometry within the computational domain, so the goals are displaying "half" forces in this case and need to be doubled for total value.

Thanks!@@hawkridgesystems

If you are a student you may be able to access the student edition Otherwise there is a hobbyist version called SOLIDWORKS for Makers that starts at $9.99/mo or $99/yr www.solidworks.com/solution/3dexperience-solidworks-makers

Yes, the setup process is interchangeable whether it is a part or assembly.