You mean simulate something myself? There's a whole field of study called computational fluid mechanics (CFD), it gets pretty complicated. At the moment there is no "theoretical" way to calculate fluid flows 100% accurately. You just use approximations based on how accurate you need the results. The simplest approximations let you do calculations by hand, and will probably get you within about 10 to 20% of what's actually happening (depending on how complex the shape is, how much of the flow is turbulent, etc, probably for a complex shaped car, you'd expect less than 30% accuracy). The next stage of complexity requires a computer, and would take a modern PC a few seconds to solve, you'd get probably within 5-10% of what's happening, assuming a simple shape, but a complex object like a plane or car you'd still be pretty far off. Then you can do full CFD, assuming you do it right, this can get you within a few percent, and for a whole car would take a modern computer a long time to solve (anything from days to weeks). When we want to do CFD like this we usually use our "supercomputers", which are often just 20+ PCs in a cluster, it still takes several hours to solve. But after all that, its still not accurate, because no theory exists to determine transitions, especially if things are vibrating, it becomes very hard to predict what's actually happening, unfortunately this is usually the case with cars and even planes
That's why wind tunnels still exist, we haven't gotten to a point where we can actually determine 100% what's going on with computers. Even though real, rational answers exist in aerodynamics, the maths is beyond comprehension, so we usually just use probability to get an estimate. You even have people doing phds on how wind tunnels aren't accurate enough.
I didn't know what an orifice plate was until I started studying fluid dynamics. 
Sadly, it's about as tedious as pressing buttons..!
