A Crash Course in Aerodynamics and an Introduction to Turbulence

[Gone4]

If you want a good explanation of secondary flow there is a cool video on youtube. It cuts off the more interesting parts that occur later, that you can find on MIT's OCW

http://youtube.com/watch?v=HKG9aPRldPs

It's possible to used vanned diffusers to reorganize the flow without introducing turbulence, which the video delves into later if you go beyond the clip on youtube. For cars the vortex generators are about closing the flow on the pressure differential down, not actually assisting the flow. That is the turbulent flow interrupts the pressure differences caused by the sudden termination of the body. This is done by adding a swirl component; this is abstracted by the Euler-N equation and others. The swirl causes a theta differential in the flow direction that disrupts laminar flow that causes the wake.

My graduate studies are focused on fluids, and more specifically turbomachinery right now. I fully intend to concentrate on heat transfer in the future, so this isn't quite my first love but I should be able to help.

[trebuchet03]

Originally Posted by Daox

Great post. I especially like the explination of VGs and why they work.

Careful - just slapping them isn't any qualification for if they work... Or just looking at the car and placing them isn't really a qualifying factor either unless you put them on, test, adjust, test etc. No doubt, Mitsubishi went through several iterations to find their location...

Originally Posted by Fuzzy

No pun intended?

100 imaginary cookies for you

[KWillets]

Aerodynamics is fun. I once had a job doing wind tunnel tests, and it's a tricky business. One thing I learned is that most of the terms like Cd are approximations made up to fool amateurs. Classical fluid dynamics is all steady-state stuff where a body is assumed to reach an equilibrium state with a steady flow around it. "Turbulence" (another loosely defined term) is treated as an untidy exception to that.

One note about your example: that's a supersonic test. The lines are shock waves; the angle suggests it's moving at about Mach 2. You might want to find a subsonic example.

Reynolds number is difficult to explain, but it's basically a scaling factor between the body and the intramolecular distance of the gas. At small scales, turbulence is disproportionately less likely to occur, while large bodies have many more eddies to deal with.

Most of aerodynamics is about flow visualization. There are many tricks to getting smoke trails, pressure sensors, etc., around the model to get useful measurements. The good news is that empirical tests work -- theory is sort of a starter course on where problems occur, but the state of the art is a lot of testing and fiddling.

[trebuchet03]

Originally Posted by KWillets

One note about your example: that's a supersonic test. The lines are shock waves; the angle suggests it's moving at about Mach 2. You might want to find a subsonic example.

Yeah... I was hoping no one would notice that If I find a better picture, I'll replace it (I'm looking for something that points to each regime like that one ).

[aerohead]

Originally Posted by MetroMPG

AWESOME writeup. Thanks for posting that.

---

Here's a question: Is there a simple answer as to why introducing a small amount of turbulence in the boundary layer (ahead of the area where flow would normally separate) moves the separation point further downstream & reduces the size of the wake?

Or does that fall into the category of thinking up questions to ask god? :P

The question is a good one and before I try to answer,I'll go home and look it up.It has to do with ideal fluids,invisid flow Daniel Bernoulli's velocity/pressure relationships,momentum interchange,friction,pressure,density,etc..The reason I haven't committed
it to memory,is that from my understanding,our cars live in a world of turbulent boundary layer for their size and velocities whether we like it or not,and there's nothing we can do about it.

Above 20-mph we have the dimpled golfball scenario working for us,and drag coefficients are stable up to transonic flow(about 250-mph).

Submerged in the troposhere,as our cars move and displace larger and larger volumes of atmosphere,air at rest is accelerated until it reaches the point of frontal area for each vehicle.Remember,up to this point,it is in a positive pressure gradient,ramming the air.As the air reaches the frontal area,or area of greatest cross-section,it no longer resides in a region of positive pressure.If air was an ideal fluid,it would have no friction,and its velocity-pressure(Bernouli) would simply be converted back to static-pressure(Bernoulli).

Because of energy loss and heating due to viscous shearing within the laminations of air close to the skin of the car,thermodynamics dictates that entropy will scuttle our attempts to get the air back as it was before we came along.Subsequently,the flow will "stall" and separate from the car,creating a turbulent wake of high drag.

Slick as it is,the smooth polished paint of our cars has enough surface roughness to "trip" the air into tiny burbles,which as members have noted,feed kinetic energy into the layer of air adjacent to the skin of the car which is otherwise at a standstill,and will postpone the point of separation,say for the golfball,well behind the point of maximum diameter,thereby reducing its wake,cutting drag by half,and extending range from 150-yards,to 300-yards with a clubhead exit velocity of 110-mph.

If someone doesn't beat me to it,I'll dig out the details for next time.Main point is, that we couldn't have a laminar boundary layer if we tried.