Aerodynamics · 1:40pm Feb 29th, 2016
The magic stuff that helps planes fly.
Obligatory XKCD.
Think of a thing that you consider to be very aerodynamic. An airplane, yeah? How about the world’s fastest airplane?
Obligatroy “Do you even read my Christmas list?”
An SR-71 is very pointy. Of course, a commercial passenger jet isn’t going to be as pointy as a super expensive military jet.
But wait, this mundane 737 seems to be more pointy on the back than it is on the front. What’s going on here? To answer this question, we look to nature.
A raindrop, being water, changes shape easily. Following the path of least resistance as it falls, what shape does it take? A raindrop, shape, of course. Raindrops fall with the blunt shape facing into the wind and the sharper tail trailing behind. This is the most aerodynamic shape for subsonic speeds, and why passenger planes are blunt at the front and sharp at the back. Long story short, how you put the air back together matters more than how you split it.
The sphere and the raindrop have similar front ends, but the tail on the raindrop makes all the difference.
If the air doesn’t come back together smoothly, it creates turbulence which adds extra drag.
You can think of it as a vacuum sucking on the back of the vehicle.
Even space capsules benefit from the shape. If you’ve ever played Kerbal Space Program, you’ve seen how even a tumbling capsule will face blunt side down once it encounters the drag of the atmosphere.
But wait, then why does the SR-71 have pointy edges? Short answer, because it flies a lot faster than raindrops. Longer answer, as speed increases, particularly at multiple Mach, it gets more difficult for the raindrop shape to put the air back together. It has to get sharper at the front to better guide the air as it forms a Mach cone.
Above a Mach, things get more complicated. The shockwave an object creates is like a boundary line where the air is highly compressed. The shockwave coming off the front of the plane can do weird things to other parts. The SR-71 had pointy cones on the engines to reduce the speed of the air going into them, because the shockwave could cause them to flameout.
Anything passing through the air will have drag. Correct shapes will help minimize it, as will smoothing out the surface. However, the easiest way to reduce drag is to make your object smaller.
And don’t fly cows.
Now that we’ve talked about drag, let’s talk about lift. If you have more lift than your airplane has mass, the plane can fly.
The center of lift should usually be behind the center of mass. I’ll talk more about that later.
To create lift, a plane has wings. You may have seen a diagram like the following:
The air moving over the top of the wing has to travel a longer distance because the top surface of the wing is curved. Therefore, to meet back up with the air on the bottom, the air on top moves faster. This creates a low pressure area on the top of the wing (the Bernoulli Effect). Compared to the higher pressure on the bottom, this lifts the wing up.
That’s not the whole story, though (see the XKCD comic at the top). While there are certainly some planes with wings of this profile, some others simply rely on forcing the air downwards. For example, on this B-52 by looking at how the wings are connected to the fuselage, you can see how much they are tipped to force the air down:
Here’s another wing diagram. The chord line is the wing’s baseline. If you had a flat plate mounted at the same angle of the chord line, you’d push the same air down and get the same lift. The rest of the wing is just a raindrop-like shape to smooth out the air around it.
As an aside, the B-52’s wings are tipped at – if I recall correctly – about a ten degree angle. This creates a lot of lift, but when the plane doesn’t need that much, you sometimes see pictures of B-52’s flying at a constant altitude, but the fuselage is tipped slightly nose down.
Wings are the hardest part of the airplane to design. Second hardest is balance. When you’re designing an airplane, you want it to be balanced for stability. That’s why the center of lift has to be behind the center of mass.
The plane rotates about its center of mass. Imagine the nose going up. The center of lift will pull the tail back up to level. But if the center of lift is ahead of the center of mass, it will only make the nose go up further until the plane goes out of control.
But what if we make a computer that can respond quickly enough to keep a plane flying straight even if we’ve purposely designed the plane to be unstable? Congratulations, you’ve just invented the F-16. A specially designed airplane with instability built in has the potential to be really great at maneuvering.
You can build some pretty strange airplanes as long as lift, mass, drag, and thrust are balanced.
If anyone liked weird planes, it was the Nazis.
NASA AD-1.
An Israeli F-15 that landed safely with only one wing after a midair collision.
An F-14 conducting an asymmetric test.
In engineering, there are exceptions to nearly every rule. In aviation, with so many control surfaces on every aircraft, it becomes nearly impossible to eyeball a potential plane design and determine if it will fly. That’s why we have software for this stuff. In the old days, it was either math or trial and error.
For more on the raw how and why of airplanes, I recommend a book called Stick and Rudder.
Certain little tricks emerge as one studies more aero. Many sports cars, like the model of the Ford GT below, have heat exchangers in the nose and the air actually flows through them instead of going around the car.
That’s why many sports cars have reverse-facing scoops – to let the air out. They can also let high pressure air out of the engine bay, wheel wells, and from under the car.
In general, you don’t want cars to fly. However, due to the Bernoulli Effect and other factors, your average car wants to lift as it goes faster. Sports cars have spoilers and wings to prevent this. The wings on sports cars are like airplane wings, but turned over so they create downforce instead of lift.
Streamlining was practiced beginning with some of the earliest cars, but not widely. It was the 1930’s before civilian cars began to pick up aero designs, but due to cost and practical reasons it took a while for the technology to be fully implemented.
Since no discussion of aero and cars would be complete without mentioning the Dodge Daytona, here you are.
This particular number 71 Daytona driven by Bobby Issac won the 1970 NASCAR championship and set 28 records on the Bonneville Salt Flats, some of which still stand today. In the early 1970’s, it was arguably the world’s fastest car. It was just a simple modification that added a nosecone and a wing to an otherwise regular car, but it made such a huge difference. I’ve written about them before.
On the other end of the scale are cars singlemindedly purpose built for racing.
Formula 1 car. More wings than a Paul McCartney cloning machine.
Downforce stops the car from lifting off the road, but if you add even more downforce, the car will stick to the road even better. A Formula 1 car actually makes more downforce than it weighs, so it could theoretically drive upside down.
But here’s the downside of downforce – drag. At top speed, that Formula 1 car slows down at about 1G force when you let off the gas pedal. That’s as hard as some cars slow down with their brakes. There are things that can be done to generate downforce while minimizing drag, but generally it’s a tradeoff to figure out how you want the air to exert force on the vehicle.
While aerodynamics work with air, air is a fluid. A lot of aerodynamics also apply to hydrodynamics. It’s why large ships tend to have bulbous bows, because as we know it’s more efficient to lead with a rounded edge.
More importantly, and not as obviously, it reduces the effect of the normal bow wave, reducing the water that hits the hull and therefore reducing drag. As I stated before, flows are difficult to eyeball and guess. This is why we’re still making improvements to aerodynamics today – because people are still coming up with stuff to try that’s never been thought of.
The boundary layer is like a thin film of air that is right up against the surface of the object. It acts a little differently than the rest of the air. Even the smoothest surface will still be a little “sticky” to air of the boundary layer. By doing things like injecting special chemicals or plasma or whatnot, scientists are trying to come up with ways to modify how the air reacts.
Related, I once read an article about an experiment a racecar team did to run the engine fan backwards and blow air out the grille. When it met the high speed air the car was driving through, the forces canceled out and it acted as if the front of the car was smoothed over. This reduced the drag because air wasn’t coming into the grille anymore.
So yeah, we know a lot of the fundamentals of fluid dynamics, but we keep discovering more exceptions and loopholes. There’s no way a simple guide like this could go into the really detailed stuff, so I’ll just post one more funny picture and call it quits.
As always, an enchanting read.
The use of "magic" when answering the first question only applies to ponies
... now I want to watch a Formula 1 race on a track full of loops and shit.
Dat GT, tho'.
roa.h-cdn.co/assets/15/24/980x490/landscape-1434115309-ford-le-mans-17.jpg
Just a few things I'd like to point out because I'm bored. I went to college for this stuff.
This has less to do with the shape of the capsule and more to do with where the center of mass is vs. where most of the drag is. Aerodynamic stability depends on the CoM being in front of the aerodynamic center of the vehicle. This is why we put the feathers at the back of an arrow shaft, and the horizontal and vertical stabilizers are at the back on most airplanes.
Pointy edges at the front of supersonic aircraft has nothing to do with putting the air back together at the back. It's all about the shockwaves at the front. Those shockwaves mark a huge jump in pressure, and a blunt leading edge has a lot of surface area, not an ideal configuration if you want to go fast. Space capsules, on the other hand, benefit from that because they want to slow down enough to deploy their parachutes.
Another issue with shockwaves is the amount of heat they produce. It's why re-entry is such a big deal in spaceflight and why the SR-71 needed such expensive exotic materials for its leading edges. Rainbow Dash traveling at Mach 3 would incinerate herself as temperatures behind her bow shock reach over 1000 degrees Fahrenheit. SR-71 engines handle this by creating a series of smaller shocks to slow the air down gradually, because if you do it in one shot, a lot of important parts inside the engine will melt.
THIS IS WRONG. I THOUGHT YOU WERE BETTER THAN THIS!!!
Ahem...
The air on top of the wing has no idea what the air on bottom is doing and is under no obligation to "meet back up" where it was split off. What actually happens doesn't really lend itself to easy simplification, so here's a good place to read up if you're curious.
This is also not true, especially for cambered airfoils like the one in your diagram. The chord line is just a reference to determine the angle of attack. A cambered airfoil can generate lift at zero AoA and even at small negative AoA.
As a parting shot, there's a reason the F-15 was once described as a triumph of thrust over aerodynamics.
Fake Edit: I lied. Here's a video of some fun aerodynamic voodoo in action. Look up the Coanda Effect for more info.
[youtube=lb1W3AK7hRY]
I thought I'd add to this with my area of expertise:
Railways
Now, during the 20s and 30s they made lots of experiments with streamlining steam locomotives. some designs worked
stanlaundon.com/steam/duchess1.jpg
others not so much
c1.staticflickr.com/3/2782/4416925836_24ff9a1346_b.jpg
but, perhaps the most well known of these designs was that of the LNER A4s
suttonparkvideos.co.uk/wp-content/gallery/steam-trains/mallard.jpg
one of these holds the world speed record for steam, at 126mph
looking at it, you'll notice a few things.
1. the Wheel Valances are Aerofoil shaped
2. the front is practically slab-sided, instead it slopes from top to bottom, rather than from the center to the sides or in a cone shape, not only was this a more efficient design, but it directs air over the locomotive, thus lifting the smoke and giving the driver a clear view of the road ahead. But the design as a whole can be traced back to Bugatti racing cars.
of note, the trains they pulled followed the teardrop principle at the other end
lowres-picturecabinet.com.s3-eu-west-1.amazonaws.com/43/main/48/127362.jpg
3784304 I did not go to college for this stuff and I'm trying to simplify a degree I don't have into a short blog post.
Aerodynamic stability is easy to guess at but difficult to master. I'm playing SimplePlanes and they tell you where center of mass, thrust, and lift are, but not drag.
The whole air over the top/air over the bottom thing is what everyone is taught in middle school, so factual or not I figured I had to at least speak to it. In the next paragraph, I point out how it' s not the end all be all answer. Stick and Rudder discusses this. There's enough controversy surrounding the issue that entire websites exist to debate it.
3784304 Well, I'm glad to see that you saved me some time. I was going to point out the issues myself, but you hit most of them nicely. All I really have to add is a specific refusal of the RD mach cone image he posted because that is blatantly wrong to anyone who knows aerodynamics. To start, I'll post two raw images which give a better view of what she is doing along with a shadowgraph of an actual mach cone:
vignette4.wikia.nocookie.net/mlp/images/b/b7/Rainbow_Dash_charges_S01E16.png/revision/latest?cb=20130317111723
vignette3.wikia.nocookie.net/mlp/images/e/ef/Mach_cone_forms_during_first_attempt_S1E16.png/revision/latest?cb=20121111235503
images.slideplayer.com/26/8736489/slides/slide_3.jpg
Looking at those pictures, it is immediately obvious that the effect that is forming around Rainbow is very different from the actual mach cone. The real mach cone forms at the point where the air meets the body and travels back from there in a perfectly straight shockwave, while the effect in front of Rainbow forms a significant distance in front of her with a pronounced curve that is much larger than the leading edge. This may not seem significant to someone who is unfamiliar with aerodynamics like whoever put that picture together, but it is a clear sign that there is no way that picture could possibly be showing an actual mach cone. After all, the reason the air behaves the way it does is that disturbances from the aircraft cannot affect the air it is not directly impacting because those disturbances would have to travel at the speed of sound and are therefore slower than the aircraft. Furthermore, we can very clearly see the actual mach cone because the mach cone at the speed of sound is a flat plane which exactly matches the shown profile of the rainboom, something that is further accentuated by the condensation that is frequently seen around aircraft as they break the sound barrier and is perfectly capable of refracting light into a rainbow if you are standing in the right place.
Also, since I'm thinking about fluid dynamics now, I just want to say one thing. Fuck turbulence. Seriously, that shit is insane, and not in an entertaining quantum mechanics way either.
3784597 Honestly, you really should have seen this coming. After all, while you may not have gone to college for this, it was basically guaranteed that someone who did is going to read this and nitpick it like this.
3784943
No doubt. Turbulence modeling is an active research area for one of my professors, and he basically told us that the best we can do is approximate its effects in CFD. The models used in commercial solvers don't reflect the actual physics of what's going on because we still don't have the computational power to do it properly.
RE: that Rainbow Dash picture
I just write it off as artistic license on the part of the show creators, so I get a kick out of watching the slapfights that people get into about it.
Depending on the shape of the body, condensation cones can form around aircraft going as slow as Mach 0.7 (low transsonic regime) and only under certain ambient conditions (very high relative humidity). It's caused by sudden changes in density, pressure, and temperature as flow around the aircraft goes locally supersonic and then passes through a weak shock to slow back down. It's similar to what happens when commercial jets have their flaps out for takeoff and landing, but that's subsonic the whole time. High AoA creates an area of low pressure above and behind the wing, and you get pictures like this.
s-media-cache-ak0.pinimg.com/564x/78/1f/75/781f75be23b008ab3bbb8d1b0e599697.jpg
3784597
I understand. I just couldn't pass up the chance to nerd out about something I put years of serious study into
This is very interesting, and the comments too. Thanks.
3785544 Since I posted about airplanes and somehow didn't manage to work in a Hornet, and since you mentioned high AoA-
-doesn't get much higher than HARV.
2.bp.blogspot.com/-U14cFRK24yc/UE1Qb0H_4XI/AAAAAAAABis/SBmgoMHcW6U/s1600/FA18_LEX+HARV.jpg
A step up from that, a pilot once told me that a lightly loaded Super Hornet effectively doesn't have a stall speed because if you keep pulling on the stick it will keep nosing up until it stands on its tail.
3785544 Yep, turbulence is a pain as you said and condensation can definitely happen slower, it is just more common and visible at close to mach one. I just didn't bother going into more detail because that gets very complicated very fast and really isn't particularly relevant here just like how the fact that it is possible to move the incidence point of the mach cone away from an object using plasma physics (something that is being investigated as an alternative to a traditional spacecraft heat shield) isn't relevant to my earlier discussion of the rainboom. Also, while I'm on the subject of the picture, I mostly wanted to point it out earlier because I would not have expected him to use a picture with that mistake in a blog post if he realized how wrong it was so I thought he might genuinely believe it.
3785688 Honestly, the Hornet is not the aircraft I would have used for that because its thrust to weight ratio is worse than any other fighter currently in US service. While it is possible to get it up over one to one so it will stand on its tail with a light load, other jets will do that with much larger loads thanks to their higher thrust to weight ratios. The F-16 is the classic example of a very high thrust to weight ratio, although the F-22 is up there as well. Of course, the V-22 has them all beat for obvious reasons, although that's really not a fair comparison for similarly obvious reasons.
3787075 That's not what I said. I was talking about AoA.
3787301 I was specifically referring to the vertical climb because that really only requires more than a one to one thrust ratio to pull off, but if you want to get picky on angle of attack then you are really looking for the supermaneuverable F-22 which can do all kinds of stupid shit because it retains control after it enters an aerodynamic stall.
If you jump to about 1:10, you will see it completely flip end over end without actually reversing direction like it would in a loop.
Oh God, I realize I'm about a year late but I just had a flashback to bumbling through an explanation of the Magnus effect on my CFI checkride. Very well researched, sir.