How Airplanes Work

I happen to fly a lot on business. For me, personally, airplanes are one of the most amazing things that I see on a daily basis. When I get on a 747, I am boarding a gigantic vehicle capable of carrying 500 or 00 people. ! 747 "eighs up to #70,000 pounds at ta$eoff. %et it rolls do"n the run"ay and, as though by magic, lifts itself into the air and can fly up to 7,000 nautical miles "ithout stopping. It is truly incredible "hen you thin$ about it& If you have ever "ondered "hat allo"s a 747 '' or any airplane for that matter '' to fly, then read on& In this edition of HowStuffWorks, "e "ill "al$ through the theory of flight and tal$ about the different parts of a standard airplane, and then you can e(plore tons of lin$s to learn even more&

!erodynamic Forces )efore "e dive into ho" "ings $eep airplanes up
in the air, it is important that "e ta$e a *uic$ loo$ at four basic aerodynamic forces+ lift, "eight, thrust and drag.

,traight and -evel Flight

In order for an airplane to fly straight and level, the follo"ing relationships must be true+

Thrust = Drag Lift = Weight

If, for any reason, the amount of drag becomes larger than the amount of thrust, the plane "ill slo" do"n. If the thrust is increased so that it is greater than the drag, the plane "ill speed up. ,imilarly, if the amount of lift drops belo" the "eight of the airplane, the plane "ill descend. )y increasing the lift, the pilot can ma$e the airplane climb. Thrust .hrust is an aerodynamic force that must be created by an airplane in order to

overcome the drag /notice that thrust and drag act in opposite directions in the figure above0. !irplanes create thrust using propellers, 1et engines or roc$ets. In the figure above, the thrust is being created "ith a propeller, "hich acts li$e a very po"erful version of a household fan, pulling air past the blades. Drag 2rag is an aerodynamic force that resists the motion of an ob1ect moving through a fluid /air and "ater are both fluids0. If you stic$ your hand out of a car "indo" "hile moving, you "ill e(perience a very simple demonstration of this effect. .he amount of drag that your hand creates depends on a fe" factors, such as the size of your hand, the speed of the car and the density of the air. If you "ere to slo" do"n, you "ould notice that the drag on your hand "ould decrease. We see another e(ample of drag reduction "hen "e "atch do"nhill s$iers in the 3lympics. %ou4ll notice that, "henever they get the chance, they "ill s*ueeze do"n into a tight crouch. )y ma$ing themselves 5smaller,5 they decrease the drag they create, "hich allo"s them to move faster do"n the hill. If you4ve ever "ondered "hy, after ta$eoff, a passenger 1et al"ays retracts its landing gear /"heels0 into the body of the airplane, the ans"er /as you may have already guessed0 is to reduce drag. 6ust li$e the do"nhill s$ier, the pilot "ants to ma$e the aircraft as small as possible to reduce drag. .he amount of drag produced by the landing gear of a 1et is so great that, at cruising speeds, the gear "ould be ripped right off of the plane. Weight .his one is the easiest. 7very ob1ect on earth has "eight /including air0. ! 747 can "eigh up to #70,000 pounds /that4s 485 tons&0 and still manage to get off the run"ay. /,ee the table belo" for more 747 specs&0 Lift -ift is the aerodynamic force that holds an airplane in the air, and is probably the tric$iest of the four aerodynamic forces to e(plain "ithout using a lot of math. 3n airplanes, most of the lift re*uired to $eep the plane aloft is created by the "ings /although some is created by other parts of the structure0.

747-400 Facts
Length: 232 feet (~ 71 meters) Height: 63 feet (~ 19 meters) Wingspan: 211 feet (~ 64 meters) Wing area: 5,650 square feet (~ 525 square meters) Empty weight: 538,000 pounds (~ 244,033 ki o!rams) Max. takeoff weight: 870,000 pounds (~ 394,625 ki o!rams) Max. lan ing weight: 630,000 pounds (~ 285,763 ki o!rams) (e"p ains #$% p anes ma% need to dump fue for emer!en&% andin!s) Engines: four tur'ofan en!ines, 57,000 pounds of t$rust ea&$ !eople capacity: up to 660 peop e and t$eir u!!a!e F"el capacity: up to 57,000 !a ons (~ 215,768 iters) Max. range: 7,200 nauti&a mi es #r"ising spee : 490 knots $akeoff istance: 10,500 feet (~ 3,200 meters

! Fe" Words !bout Fluid

! principal concept in aerodynamics is the idea that air is a fluid. -i$e all gases, air flo"s and behaves in a similar manner to "ater and other li*uids. 7ven though air, "ater and panca$e syrup may seem li$e very different substances, they all conform to the same set of mathematical relationships. In fact, basic aerodynamic tests are sometimes performed under"ater. !nother important concept is the fact that lift can exist only in the presence of a moving fluid. .his is also true for drag. It doesn4t matter if the ob1ect is stationary and the fluid is moving, or if the fluid is still and the ob1ect is moving through it. What really matters is the relative difference in speeds etween the o !ect and the fluid. 9onse*uently, neither lift nor drag can be created in space /"here there is no fluid0. .his e(plains "hy spacecraft don4t have "ings unless the spaceship spends at least some of its time in air. .he space shuttle is a good e(ample of a spacecraft that spends most of its time in space, "here there is no air that can be used to create lift. :o"ever, "hen the shuttle re'enters the earth4s atmosphere, its stubby "ings produce enough lift to allo" the shuttle to glide to a graceful landing.

;opular /and Imperfect0 7(planations of -ift 9reation

If you read any college'level aerodynamics te(tboo$, you "ill find plenty of mathematical methods for calculating lift. <nfortunately, none of these e(planations are particularly satisfying unless you have a ;h.2. in mathematics. /9lic$ here for a demo of an online aerodynamics te(tboo$ from ,tanford <niversity.0 .here are many simplified e(planations of lift that appear on the Internet and in some te(tboo$s. ."o of the most popular e(planations today are the Longer "ath explanation /also $no"n as the #ernoulli or e$ual transit time e(planation0 and the %ewtonian explanation /also $no"n as the momentum transfer or air deflection e(planation0. While many versions of these e(planations are fundamentally fla"ed, they can still contribute to an intuitive understanding of ho" lift is created.

.he -onger ;ath 7(planation

What is it? .he -onger ;ath e(planation holds that the top surface of a "ing is more curved than the bottom surface. !ir particles that approach the leading edge of the "ing must travel either over or under the "ing. -et4s assume that t"o nearby particles split up at the leading edge, and then come bac$ together at the trailing edge of the "ing. ,ince the particle traveling over the top goes a longer distance in the same amount of time, it must be traveling faster. )ernoulli4s e*uation, a fundamental of fluid dynamics, states that as the speed of a fluid flo" increases, its pressure decreases. .he -onger ;ath e(planation deduces that this faster moving air develops a lo"er pressure on the top surface, "hile the slo"er moving air maintains a higher pressure on the bottom surface. .his pressure difference essentially 5suc$s5 the "ing up"ard /or pushes the "ing up"ard, depending on your point of vie"0. Why is it not entirely correct? .here are several fla"s in this theory, although this is a very common e(planation found in high school te(tboo$s and even encyclopedias+ =. .he assumption that the t"o air particles described above re1oin each other at the trailing edge of the "ing is groundless. In fact, these t"o air particles have no 5$no"ledge5 of each other4s presence at all, and there is no logical reason "hy these particles should end up at the rear of the "ing at the same moment in time. >. For many types of "ings, the top surface is longer than the bottom. :o"ever, many "ings are symmetric /shaped identically on the top and bottom surfaces0. .his e(planation also predicts that planes should not be able to fly upside do"n, although "e $no" that many planes have this ability. Why is it not entirely wrong? .he -onger ;ath e(planation is correct in more than one "ay. First, the air on the top surface of the "ing actually does move faster than the air on the bottom '' in fact, it is moving faster than the speed re*uired for the top and bottom air particles to reunite, as many people suggest. ,econd, the overall pressure on the top of a lift'producing "ing is lo"er than that on the bottom of the "ing, and it is this net pressure difference that creates the lifting force.

.he ?e"tonian 7(planation

What is it? Isaac ?e"ton stated that for every action there is an e*ual, and opposite,

reaction /?e"ton4s .hird -a"0. %ou can see a good e(ample of this by "atching t"o s$aters at an ice rin$. If one pushes on the other, both move '' one due to the action force and the other due to the reaction force. In the late = 00s, Isaac ?e"ton theorized that air molecules behave li$e individual particles, and that the air hitting the bottom surface of a "ing behaves li$e shotgun pellets bouncing off a metal plate. 7ach individual particle bounces off the bottom surface of the "ing and is deflected do"n"ard. !s the particles stri$e the bottom surface of the "ing, they impart some of their momentum to the "ing, thus incrementally nudging the "ing up"ard "ith every molecular impact. ?ote+ !ctually, ?e"ton4s theories on fluids "ere developed for naval "arfare, in order to help decrease the resistance that ships encounter in the "ater '' the goal "as to build a faster boat, not a better airplane. ,till, the theories are applicable, since "ater and air are both fluids. Why is it not entirely correct? .he ?e"tonian e(planation provides a pretty intuitive picture of ho" the "ing turns the air flo"ing past it, "ith a couple of e(ceptions+ =. .he top surface of the "ing is left completely out of the picture. .he top surface of a "ing contributes greatly to turning the fluid flo". When only the bottom surface of the "ing is considered, the resulting lift calculations are very inaccurate. >. !lmost a hundred years after ?e"ton4s theory of ship hulls, a man named -eonhard 7uler noticed that fluid moving to"ard an ob1ect "ill actually deflect before it even hits the surface, so it doesn4t get a chance to bounce off the surface at all. It seemed that air did not behave li$e individual shotgun pellets after all. Instead, air molecules interact and influence each other in a "ay that is difficult to predict using simplified methods. .his influence also e(tends far beyond the air immediately surrounding the "ing. Why is it not entirely wrong? While a pure ?e"tonian e(planation does not produce accurate estimates of lift values in normal flight conditions /for e(ample, a passenger 1et4s flight0, it predicts lift for certain flight regimes very "ell. For hypersonic flight conditions /speeds e(ceeding five times the speed of sound0, the ?e"tonian theory holds true. !t high speeds and very lo" air densities, air molecules behave much more li$e the pellets that ?e"ton spo$e of. .he space shuttle operates under these conditions during its re'entry phase. <nli$e the -onger ;ath e(planation, the ?e"tonian approach predicts that the air is deflected do"n"ard as it passes the "ing. While this may not be due to molecules bouncing off the bottom of the "ing, the air is certainly deflected do"n"ard, resulting in a phenomenon called downwash. /9lic$ here for more on do"n"ash.0

:o" -ift is 9reated

;ressure @ariations 9aused )y .urning a Aoving Fluid
-ift is a force on a "ing /or any other solid ob1ect0 immersed in a moving fluid,

and it acts perpendicular to the flo" of the fluid. /2rag is the same thing, but acts parallel to the direction of the fluid flo"0. .he net force is created by pressure differences brought about by variations in speed of the air at all points around the "ing. .hese velocity variations are caused by the disruption and turning of the air flo"ing past the "ing. .he measured pressure distribution on a typical "ing loo$s li$e the follo"ing diagram+

A& !ir approaching the top surface of the "ing is compressed into the air above it as it moves up"ard. .hen, as the top surface curves do"n"ard and a"ay from the airstream, a lo"'pressure area is developed and the air above is pulled do"n"ard to"ard the bac$ of the "ing. #& !ir approaching the bottom surface of the "ing is slo"ed, compressed and redirected in a do"n"ard path. !s the air nears the rear of the "ing, its speed and pressure gradually match that of the air coming over the top. .he overall pressure effects encountered on the bottom of the "ing are generally less pronounced than those on the top of the "ing. '& -ift component D& ?et force (& 2rag component When you sum up all the pressures acting on the "ing /all the "ay around0, you end up "ith a net force on the "ing. ! portion of this lift goes into lifting the "ing /lift component0, and the rest goes into slo"ing the "ing do"n / drag component0. !s the amount of airflo" turned by a given "ing is increased, the speed and pressure differences bet"een the top and bottom surfaces become more pronounced, and this increases the lift. .here are many "ays to increase the lift of a "ing, such as increasing the angle of attac$ or increasing the speed of the airflo". .hese methods and others are discussed in more detail later in this article. It is important to realize that, unli$e in the t"o popular e(planations described earlier, lift depends on significant contributions from both the top and bottom "ing

surfaces. While neither of these e(planations is perfect, they both hold some nuggets of validity. 3ther e(planations hold that the une*ual pressure distributions cause the flo" deflection, and still others state that the e(act opposite is true. In either case, it is clear that this is not a sub1ect that can be e(plained easily using simplified theories. -i$e"ise, predicting the amount of lift created by "ings has been an e*ually challenging tas$ for engineers and designers in the past. In fact, for years, "e have relied heavily on e(perimental data collected 70 to #0 years ago to aid in our initial designs of "ings. 'alculating Lift #ased on (xperimental Test )esults In =B=5, the <.,. 9ongress created the ?ational !dvisory 9ommittee on !eronautics /?!9! '' a precursor of ?!,!0. 2uring the =B>0s and =B80s, ?!9! conducted e(tensive "ind tunnel tests on hundreds of airfoil shapes /"ing cross' sectional shapes0. .he data collected allo"s engineers to predictably calculate the amount of lift and drag that airfoils can develop in various flight conditions. .he lift coefficient of an airfoil is a number that relates its lift'producing capability to air speed, air density, "ing area and angle of attack '' the angle at "hich the airfoil is oriented "ith respect to the oncoming air flo" /"e4ll discuss this in greater detail later in the article0. .he lift coefficient of a given airfoil depends upon the angle of attac$.

Interesting .hings about Wings

.hese interesting facts about "ings are useful in developing a more detailed understanding of ho" they "or$.

Wing ,hape
.he 5standard5 airfoil shape that "e e(amined above is not the only shape for a "ing. For e(ample, both stunt planes /the $ind that fly upside do"n for e(tended periods of time at air sho"s0 and supersonic aircraft have "ing profiles that are some"hat different than you "ould e(pect+

.he upper airfoil is typical for a stunt plane, and the lo"er airfoil is typical for supersonic fighters. ?ote that both are symmetric on the top and bottom. ,tunt planes and supersonic 1ets get their lift totally from the angle of attac$ of the "ing.

!ngle of !ttac$
.he angle of attac$ is the angle that the "ing presents to oncoming air, and it controls the thic$ness of the slice of air the "ing is cutting off. )ecause it controls the slice, the angle of attac$ also controls the amount of lift that the "ing generates /although it is not the only factor0.

%ero angle of attack

&hallow angle of attack

&teep angle of attack

Flaps
In general, the "ings on most planes are designed to provide an appropriate amount of lift /along "ith minimal drag0 "hile the plane is operating in its cruising mode /about 5 0 miles per hour, or B0= $m per hour, for the )oeing 747'4000. :o"ever, "hen these airplanes are ta$ing off or landing, their speeds can be reduced to less than >00 miles per hour /8>> $ph0. .his dramatic change in the "ing4s "or$ing conditions means that a different airfoil shape "ould probably better serve the aircraft. .o accommodate both flight regimes /fast and high as "ell as slo" and lo"0, airplane "ings have moveable sections called flaps. 2uring ta$eoff and landing, the flaps are e(tended rear"ard and do"n"ard from the trailing edge of the "ings. .his effectively alters the shape of the "ing, allo"ing the "ing to turn more air, and thus create more lift. .he do"nside of this alteration is that the drag on the "ings also increases, so the flaps are put a"ay for the rest of the flight.

,lats
Slats perform the same function as flaps /that is, they temporarily alter the shape of the "ing to increase lift0, but they are attached to the front of the "ing instead of the rear. .hey are also deployed on ta$eoff and landing.

Cotating ,urfaces
Diven "hat "e $no" so far about "ings and lift, it seems logical that a simple cylinder "ould not produce any lift "hen immersed in a moving fluid /imagine a plane "ith "ings shaped li$e cardboard paper'to"el tubes0. In a simplified "orld, the air "ould 1ust flo" around the cylinder evenly on both sides, and $eep right on going. In reality, the do"nstream air "ould be a little turbulent and chaotic, but there still "ould be no lift created. :o"ever, if "e "ere to begin rotating the cylinder, as in the figure belo", the surface of the cylinder "ould actually drag the surrounding layer of air around "ith it. .he net result "ould be a pressure difference bet"een the top and bottom surfaces, "hich deflects the airflo" do"n"ard. ?e"ton4s .hird -a" states that if the air is being redirected do"n"ard, the cylinder must be deflected up"ard /sounds li$e lift to me&0. .his is an e(ample of the *agnus (ffect /also $no"n as the )o ins (ffect0, "hich holds true for rotating spheres as "ell as cylinders /see any similarities to curveballs hereE0

)elieve it or not, in =B> , !nton Flettner built a ship named the )ruc$au that used huge spinning cylinders instead of sails to po"er itself across the ocean. 9lic$ here to learn about Flettner4s Cotorship.

)lo"n ,urfaces
-et4s ta$e our cylindrical "ing from the above e(amples and find another "ay to create lift "ith it. If you4ve ever held the bac$ of your hand vertically under the faucet, you may have noticed that the "ater did not simply run do"n to the bottom of your hand and then drip off. Instead, the "ater actually runs bac$ up and around the side of your hand /for a fe" millimeters0 before falling into the sin$. .his is $no"n as the 'oanda (ffect /after :enri 9oanda0, "hich states that a fluid will tend to follow the contour of a curved surface that it contacts .

In our cylinder e(ample, if air is forced out of a long slot 1ust behind the top of the cylinder, it "ill "rap around the bac$side and pull some surrounding air "ith it. .his is a very similar situation to the Aagnus 7ffect, e(cept that the cylinder doesn4t have to spin. .he 9oanda 7ffect is used in specialized applications to increase the amount of additional lift provided by the flaps. Instead of 1ust altering the shape of the "ing, compressed air can be forced through long slots on the top of the "ing or the flaps to produce e(tra lift.

)elieve it or not, in =BB0, Ac2onnell 2ouglas :elicopter 9o. /no" $no"n as A2 :elicopters, Inc.0 removed the tail rotors from some of its helicopters and replaced them "ith cylinders& Instead of using a conventional tail rotor to steer the aircraft, the tail boom is pressurized and air is blo"n out through long slots e(actly li$e the figure above.

(ma!e &ourtes% )*+*

$he lift-c"r'e slope of a ()#) airfoil

:ere is the standard e*uation for calculating lift using a lift coefficient+

L , ift #l , ift &oeffi&ient

*rho+ , air densit% , , air -e o&it% ) , #in! area

!s an e(ample, let4s calculate the lift of an airplane "ith a "ingspan of 40 feet and a chord length of 4 feet /"ing area F = 0 s*. ft.0, moving at a speed of =00 mph /= = $ph0 at sea level /that4s =47 feet, or 45 meters, per second&0. -et4s assume that the "ing has a constant cross'section using an ?!9! =40# airfoil shape, and that the plane is flying so that the angle of attac$ of the "ing is 4 degrees. We $no" that+

A F = 0 s*uare feet +rho, F 0.00>87 B slugs G cubic foot /at sea level on a standard day0 - F =47 feet per second 'l F 0.55 /lift coefficient for ?!9! =40# airfoil at 4 degrees !3!0

,o let4s calculate the lift+ Lift F 0.55 ( .5 ( .00>87 B ( =47 ( =47 ( = 0 Lift = ./.01 l s .ry your hand at airfoil design on ?!,!4s Web site using a virtual "ind tunnel. 'alculating Lift 2sing 'omputer Simulations In the years since ?!9!4s e(perimental data "as collected, engineers have used this information to calculate the lift /and other aerodynamic forces0 produced by "ings and other ob1ects in fluid flo"s. In recent years, ho"ever, computing po"er has increased such that "ind tunnel e(periments can no" be simulated on an average personal computer. ,oft"are pac$ages, such as F-<7?., have been developed to create simulated fluid flo"s in "hich solid ob1ects can be virtually immersed. .he applications of this type of soft"are range from simulating the air flo"ing over a "ing, to mapping the airflo" through a computer case to ensure that there is enough cool air passing over the 9;< to prevent the computer from overheating

Aore !irplane ;arts

.he "ing is obviously the most important part of an airplane '' it4s "hat gets the airplane in the air& )ut airplanes have a lot of other characteristic parts designed to control the plane or get it moving. -et4s e(amine the parts you find in a typical airplane by loo$ing at a 'essna 34..

;robably the most important parts of an airplane, after the "ing, are the propeller and engine. .he propeller /or, on 1et aircraft, the 1ets0 provides the thrust that moves the plane for"ard. /9hec$ out :o" Das .urbine 7ngines Wor$ to learn about 1et engines.0

! propeller is really 1ust a special, spinning "ing. If you loo$ed at the cross section of a propeller, you4d find that a propeller has an airfoil shape and an angle of attac$. 6ust by loo$ing at the propeller pictured above, you can see that the angle of attac$ changes along the length of the propeller '' the angle is greater to"ard the center because the speed of the propeller through the air is

slo"er close to the hub. Aany larger propeller aircraft have more elaborate three' blade or four'blade props "ith ad!usta le pitch mechanisms. .hese mechanisms let the pilot ad1ust the propeller4s angle of attac$ depending on air speed and altitude. .he landing gear is also essential during ta$e'off and landing.

Front lan ing gear

-ear lan ing gear

.he 9essna =5> has fi(ed landing gear, but most planes have retractable landing gear to reduce drag "hile in flight.

9ontrolling the 2irection

.he tail of the airplane has t"o small "ings, called the horizontal and vertical sta ili5ers, that the pilot uses to control the direction of the plane. )oth are symmetrical airfoils, and both have large flaps on them that the pilot controls "ith the control stic$ to change their lift characteristics.

Hori.ontal tail wing

,ertical tail wing

With the hori5ontal tail wing, the pilot can change the plane4s angle of attac$, and therefore control "hether the plane goes up or do"n. With the vertical tail wing, the pilot can turn the plane left or right. .he plane4s main "ing is 40 feet /H => m0 long from end to end, and about 4 feet /H =.> m0 "ide. 3n the inner portion of the "ing, there are flaps used during ta$eoff, landing and other lo"'speed situations. 3n the outer ends, there are ailerons used to turn the plane and $eep it level.

Main wing

Flaps

.he flaps are actuated by electric motors in the "ing. !lso enclosed in the "ings are t"o fuel tan$s, each of "hich holds about >0 gallons of gas.

!irplane ,ensors
From this description you can see that a plane has four different moveable control surfaces, as sho"n here+

.he plane also has t"o different sensors mounted on the "ing+

.he -'shaped tube is called a pitot tu e. !ir that rams into this tube during flight creates pressure, and that pressure moves the needle on the air'speed indicator in the coc$pit. .he small opening on the right is a "histle that sounds as the "ing nears a stall. .he larger opening visible near the coc$pit is used for ventilation. For more information on airplanes, flight dynamics and other other related topics, chec$ out the lin$s on the ne(t page&

-ots Aore Information&

Celated :o",tuffWor$s !rticles

:o" Das .urbine 7ngines Wor$ :o" !ir .raffic 9ontrol Wor$s :o" )ecoming an !irline ;ilot Wor$s :o" )lac$ )o(es Wor$ :o" a <.,. ,py ;lane Wor$s :o" 9oncordes Wor$ :o" :elicopters Wor$

Aore Dreat -in$s&

Aerodynamics !irfoil -ifting Force Aisconception in I' .e(tboo$s ,talling Wing 9ross ,ections What is Dround 7ffectE ;hysics 2emonstrations+ !erodynamic concepts "aper Airplanes ! classic paper airplane ;aper !irplane !erodynamics #uilding and6or 7lying 8our 9wn Airplane 9arbon 2ragon .echnical Web site )oeing 747 ,9! Flight Aanual !ircraft 7ngine 2irectory Cotorcraft, !irship, and !ircraft :omebuilt Iits !irplay+ !ll !bout ;aragliding )adio 'ontrolled Airplanes CG9 !ircraft Info and Files CG9 !irplane'3riented Web sites CG9 ,ailplaneGFreeflight'3riented Web ,ites 9ther )esources .hose Aagnificent Aen and .heir Flying Aachines 7(tensive list of aviation museums .he First Flight 9entennial Foundation ' 9ommemorating the =00th anniversary of the Wright brothers4 first flight