71

Military drones, such as the Predator drone or the Global Hawk, tend to have a weird shape with a whale-like head, and engines concentrated at the back. The wings tend to be of very high aspect ratio, and v-shaped tails seem to be popular:

enter image description here enter image description here enter image description here

Why aren't there, say, Cessna-shaped drones? If Cessna 172s work fine for human pilots, why not for computer pilots? The Cessna 172's low speed and stability also seems like a good asset for military recon drones.

ithisa
  • 1,702
  • 2
  • 15
  • 18
  • 29
    Don't read too much into it, the designers are just overcompensating for something... – IQAndreas Oct 09 '15 at 06:10
  • 13
    I'd guess that a pusher configuration helps in reconnaissance: there is no engine or propeller blades obstructing the field of view of the sensors and cameras. – vsz Oct 09 '15 at 06:12
  • 12
    Are military drones shaped differently from civilian drones? – gerrit Oct 09 '15 at 09:58
  • 28
    You're assuming that a Cessna is the natural shape for a plane, but a lot of the design of a light aircraft is predicated on the thing having to carry one or more human beings, who need to see, be protected and so on. – David Richerby Oct 09 '15 at 17:01

8 Answers8

120

The whale-shaped forward fuselage covers a parabolic antenna for a high-bandwidth datalink. The operators want to receive the reconnaissance data in real time, and by giving the antenna the best place in the aircraft, it will be able to connect to communications satellites even when they are just above the horizon.

Global Hawk cut-away drawing

Global Hawk cut-away drawing (picture source)

The V-tail of the Predator is an inheritance from its origins as a Navy drone which could be folded to fit into a torpedo tube. Read all about it in this answer. If you want more info on the benefits of an inverted V-tail, make sure you read KeithS's answer, too.

In the Global Hawk the V-tail was chosen to mount the engine above the fuselage for better IR shielding from below. The two tail surfaces also shield the exhaust from the side without affecting aerodynamics negatively.

The Aerosonde's configuration was chosen after Tad McGeer had experienced in 1990 from the 1/5th scale RC model of the Perseus UAV of Aurora that a tail-mounted propeller would stabilize the aircraft to the point that it became practically uncontrollable. A pusher propeller is more efficient than a tractor propeller, which also explains the propeller location on the Predator.

If it strikes you how many of the UAVs have V-tails: This allows to use fewer components (surfaces, connectors, actuators), and since all are moved by computer control, the issue of a mechanical mixer does not exist. The lower damping of V-tails can also be easily tolerated, since the FCS can react much more quickly and commensurately to disturbances than any human pilot. The lower maneuverability possible with V-tails is also not an issue - reconnaissance UAVs fly mostly straight and level.

The high aspect ratio wings are helping in extending the range and especially the flying time of observation drones. Similar to glider wings, the long and slender wing will offer the lowest drag for a given lift at subsonic speed.

In the end, each detail has been chosen deliberately and can be well explained. There is no conspiracy to make UAVs look different from Cessnas.

Peter Kämpf
  • 231,832
  • 17
  • 588
  • 929
  • 54
    There is no conspiracy to make UAVs look different from Cessnas. Whew! :) – FreeMan Oct 08 '15 at 19:31
  • Look at the F22 or the F35 they both have V shaped rear surfaces too. – JamesRyan Oct 08 '15 at 20:48
  • 2
    @JamesRyan: In their case in order to reflect radar away from the horizontal plane. Both Global Hawk and Predator are not shaped for stealth, so radar reflection characteristics were not the reason for the V tails. – Peter Kämpf Oct 09 '15 at 04:43
  • 4
    @JamesRyan, PeterKämpf - It might be more complicated than both of your stances. The angle of the "vertical" stabilizers on the F/A-18 is to get them out of the direct path of leading edge vorticies (a problem for the F-14 and F15). It also appears that some designs use inward deflection of both rudder surfaces to add to upward pitch control. And yes, avoiding right angles and aligning edges contributes to reduced radar return. As to what the primary reason(s) is(are), only the designers could say for sure, and some details of design and motivation are likely classified. – Todd Wilcox Oct 09 '15 at 18:12
  • Please you add a parenthesis t explain what you mean by "The Aerosonde configuration". Did you mean "The Aerosonde's configuration" (so the subject is the Aerosonde in particular), or did you mean somethng generic about the Aerosonde (e.g. placing the pusher behind he wing but in front of the tail)? – ChrisW Mar 06 '19 at 23:45
  • @ChrisW: I meant that particular implementation of this configuration (which was used earlier in other designs). I hope adding the genitive-s improves the answer. – Peter Kämpf Mar 07 '19 at 00:10
  • Yes that's unambiguous now, thank you. The illustration in the answer shows a jet engine behind the wings and in front of the tail, without the genitive-s it wasn't clear to me whether perhaps that's what you meant -- something generic about the configuration, generalised from the Aerosonde as an early prototype. – ChrisW Mar 07 '19 at 00:20
  • +1 Enjoy the gold medal :D let me know if I need to ask this separately, but why was the RC model of the Perseus very stable? Is it simply because the prop placed a suction peak right aft of the tail-plane, increasing the effectiveness? –  Mar 07 '19 at 00:35
  • 1
    @ymb1: No, the prop created considerable lift and side forces, depending on angle of sideslip and attack. Since it was sized for flight at 24 km, it was huge and acted like another fixed tail. More on that here. – Peter Kämpf Mar 07 '19 at 06:29
56

The biggest reason UAVs look like they do is because they don't need to carry a human, or the life support and avionics displays that human would need. The designs, therefore, are built around what the aircraft does need, mainly communications and viewing/recon equipment in addition to the standard stuff any combat aircraft needs (fuel, weapons, propulsion, lift/control surfaces). The use of a V-tail in practically all of them is for drag reduction; the aircraft doesn't have to be as maneuverable as a manned fighter, so a twin tail with separate all-moving elevators is superfluous, and the lower drag increases flight time for more range or loitering capability.

A key design consideration in any aircraft is weight balance. The 172, a fairly conventional aircraft design, is designed for a high degree of flexibility in its payload and distribution; the high wing keeps the center of gravity below the center of lift for stability, and by the same token the cockpit and cargo area are under the wing so changes in CG don't cause quite as drastic a change in handling.

In an unmanned aircraft, weight distribution is fairly static; any ordnance is right under the wings and thus the center of lift (fuel is also placed similarly), while the contents of the fuselage besides fuel load hardly change at all between flights. That allows the aircraft designer to allow form to follow function; the front of the aircraft has most of the avionics and communications, counterbalancing the propulsion system at the rear, with the fuel in the middle. The relative size and shape of these three elements is at the designer's discretion, and then the wings are simply placed at or slightly behind the balance point.

The RQ-1/MQ-1 Predator is probably the weirdest-looking, because in addition to having a V-tail for the weight and drag reasons, the tail is inverted:

enter image description here

The inverted V-tail has several advantages in a plane the Predator's size and with the complexity of the total flight system:

  • A rudder surface placed beneath the CG, as on the Predator, will roll the plane toward the direction of yaw instead of away from it as a top-mounted rudder would, which allows for coordinated turns using only the rudder.
  • This "rolling rudder" behavior is also pretty much all the RQ-1/MQ-1 needs in terms of roll control, so the rudder surfaces can be the only control surfaces on the plane, dramatically simplifying the control layout of the craft. Theoretically the lack of ailerons allows the wings to be removed or folded easily, but the MQ-1 has weapons pylons with firing circuits that complicate this. Larger drones with longer wings have higher MOI in the roll axis, more than the rudders could overcome by themselves, so these designs have ailerons (either wing-warping or hinged), which allows the V-tail to be upright for better ground clearance on landing.
  • A crosswind on an inverted V-tail will cause the plane to pitch up and roll into the wind instead of pitching down and rolling away from it with an upright V-tail. Given the high comms latency between Nevada and Kandahar, this is an advantage, as the plane usually has plenty of room above it to fly while the pilot notices and reacts to the disturbance, but can only fly downward for so long before it runs out of air. Again, larger planes have higher MOIs and so disturbances are more minor.
  • The inverted tail masks a little of the plane's engine noise in flight from specific forward angles. Not much, but every bit helps when you're trying to stay hidden.
  • Lastly, the inverted tail protects the more expensive propeller if the pilot scrapes the back end on the ground. The larger Reaper has more ground clearance to allow for the weapons pylons, and so it can flare more on landing without involving the prop.

The RQ-2 Global Hawk actually doesn't look that farfetched; it bears some resemblance to the A-10 Warthog:

enter image description here enter image description here

The two aircraft have vastly different mission profiles (the Global Hawk is intended as the U-2's replacement for high-altitude surveillance and reconnaissance while the A-10 is a venerable armor-killer close air support craft), but there's some common design elements, such as the high engine masked by the tail surfaces to reduce IR signature, and the big low wings producing a lot of lift (for a high ceiling in the GH's case, for payload and survivability in the A-10's).

What's not similar between the two is program costs; the original A-10As had a unit cost of just \$450k in the '70s, and with new wings and a glass cockpit upgrade the A-10Cs are still just \$11 million a frame. Unit costs of the Global Hawk, including R&D, make the program the most expensive small aircraft to date at \$222 million per unit, eclipsing the "cost is no object" F-22 (\$182m).

Other drones are much more cost-efficient; the Predator is just \$4 million each while the larger and more sophisticated Reaper is \$16 million. Losing one still hurts, and the USAF has lost plenty, mainly to operator error (the communications latency inherent in piloting one of these remotely from a ground station half a world away is considerable), but the total costs of these programs is still a pittance compared to any manned airframe in service today, with dramatically reduced maintenance to flight hour ratios and total costs per flight hour. The A-10 is again the cheapest manned combat aircraft in service, costing about \$18,000 per flight hour to operate (no word on what all that price includes; likely some combination of fuel, pilot pay, parts, maintenance and ordnance). The Predator costs just \$3,600 per flight hour, and the Reaper about \$4,800. Again, the Global Hawk program is fairly expensive to operate (about \$49,000 a flight hour) and that's a big reason the U-2 Dragon Ladies that Global Hawk was supposed to replace are still flying (the U-2 is only about \$30,000 a flight hour).

Community
  • 1
KeithS
  • 22,891
  • 4
  • 82
  • 122
  • 1
    I believe some other answer here mentions that the take-offs and landings are actually controlled by local operator on the airport exactly to avoid problems with latency. – Jan Hudec Oct 09 '15 at 08:43
  • How much work has been done on automated systems to allow UAV self-recovery/stabilization in those very cases where the operator connection is lost? It seems like something that would be worth the investment, or does that already exist and the main issue is operator control being lost during maneuvers that the UAV cannot self-recover from in time? – JAB Oct 09 '15 at 15:07
  • To elaborate on "don't need to carry a human," a big difference in the "weird" look (if not the shape) is the lack of a windshield for humans to see out through. I suppose that's obvious but it hadn't been explicitly stated yet. To me it makes the drone feel kind of "faceless" (or eyeless). – LarsH Oct 09 '15 at 15:51
  • @Jan Hudec - I mentioned that in a previous revision of this answer, but that bit is less about looks which was the crux of the question. Yes, takeoffs and landings are handled by a local ground station in Kandahar while the bulk of the mission is flown from some cargo crates in Nevada. – KeithS Oct 09 '15 at 17:31
  • 1
    @LarsH As far as lack of cockpit, the "cockpit" is basically the bubble cam you see on the bottom of the Predator pic (there's one under the "chin" of the GH too, but that plane is deceptively big, another consequence of having no cockpit is that there's no sense of scale to these aircraft in flight). – KeithS Oct 09 '15 at 17:34
  • @JAB - The ability to tell the drone "go here" in very non-flying terms is actually a big part of the Global Hawk's capabilities (and its cost). The plane is normally controlled by just clicking on a map to draw waypoints, and the GH will figure out how to get where it's told to go, and can follow the flight path in the absence of further input. We used to do the same thing with cruise missiles, but a retrievable aircraft is less expensive to operate (though the breakeven on the GH program over a Tomahawk-based recon platform is on the order of 300 missions for each of the 50 GH aircraft). – KeithS Oct 09 '15 at 17:38
  • The downsides of this control scheme so far are that the GH isn't smart enough to avoid weather, and doesn't have weather radar or even de-icing equipment, so the mission is scrubbed if bad weather's even likely (over the Pacific, that's close to a guarantee, leading to the Navy's version the MQ-4C which is more hardy) – KeithS Oct 09 '15 at 17:40
  • @KeithS Oh, that's cool. But can it handle sharing airspace? – JAB Oct 09 '15 at 17:40
  • That's another big downside; the GH can't interface with civilian ATC by itself, so when one is being launched or retrieved from Beale AFB near Sacramento and the Bay Area (or anytime it's cruising in FAA airspace), either the plane has to be flown "manually" from a ground station patched into ATC, or NOTAMs issued and ATC instructed to vector civilian flights around the GH. – KeithS Oct 09 '15 at 17:49
  • 1
    To forestall further discussion in comments, here's a really good article on the RQ-4 Global Hawk, what it can do, what it can't, what it costs, what is may save, etc: http://foxtrotalpha.jalopnik.com/why-the-usafs-massive-10-billion-global-hawk-uav-was-w-1629932000 – KeithS Oct 09 '15 at 17:53
  • Directly piloting UAVs is just a blatant make-work job. Human involvement should be limited to a high level "go to X, do Y". The computer knows better and easily monitor every sensor on the airframe and respond within a fraction of a second to whatever condition. – Nick T Oct 09 '15 at 19:24
  • @JAB They're supposed to fly in circles or return to base when the connection is lost, but it's not unusual for them to crash instead. – Matt Nordhoff Oct 10 '15 at 08:35
  • @NickT - Your opinionated statements depend on a level of flight automation we have not yet achieved. For an O-1's pay grade, you can put a human in more or less direct control of an aircraft and get capabilities like image recognition, voice recognition, verbal command comprehension and independent decision-making that far exceed what $10 billion has given us with the Global Hawk (which is the only known program to have a command interface anywhere close to what you think all drones should have). – KeithS Oct 13 '15 at 00:01
  • @KeithS true, I'm being a bit inflammatory, but at any pay grade and level of skill, a computer's understanding of aerodynamics and physics is superior. Not that I'm not advocating for total autonomy like you mention (image recognition, etc.), as in that realm computers are still fairly lacking. – Nick T Oct 13 '15 at 00:22
26

Who says they are weird shaped? Maybe the 172 is weird...

The short answer is that it's the best design for the mission and the specific application. The 172 is designed to carry people and train young pilots so its characteristics reflect being a good fit of that mission. Drones are required to carry computers and surveillance equipment and often a weapons payload and very specifically, not people. Drones can be of a different shape than a GA plane since pilot ergonomics don't really matter. For what it's worth there are planes that are shaped like the drone you mention.

The Bonanza had a long run with the V tail design: photo of Bonanza (source)

The new Cirrus Jet has a V tail and a bubble airframe much like the drones you pictured: photo of Cirrus Jet (source)

Bottom line is that planes are designed for specific applications and some airframe aspects lend themselves to a given mission better than others. Planes will reflect these constraints over looks 100% of the time.

Side note: low speed is by no means necessary for good recon. The SR-71, which was arguably one of the most successfully spy planes, did plenty of recon from 80000ft at Mach 3 with out any problems.

Community
  • 1
Dave
  • 101,073
  • 5
  • 220
  • 364
  • 6
    The SR-71 recon mission is just a little bit different from the drones' mission. I'd say low speed is an advantage when you want it to loiter over a target area. – TomMcW Oct 08 '15 at 19:55
  • Very true, i was just providing a counter example. – Dave Oct 08 '15 at 20:04
  • 2
    +1 for the fact that they don't have to carry people or deal with the weight where people normally sit. The center of gravity can be adjusted as needed, and components can be moved to where they work best with no restrictions. – JPhi1618 Oct 08 '15 at 21:12
11

Another factor: Cessnas, like virtually all manned aircraft are optimized to get to the destination.

Military drones, however, are more interested in dwell time over the area of interest than in how fast they get there. This results in a very different optimization than for manned aircraft.

They also have a very different view of safety.

You have an aircraft that costs \$500k. There's a safety measure that has a 5% chance of saving the airframe over it's expected life but which costs \$50k. On a manned plane that's almost certainly going to be done. On a drone it wouldn't make sense in most cases.

Loren Pechtel
  • 509
  • 2
  • 6
6

For single engine pushers with a canopy, when a pilot bails out of a manned aircraft they might hit the prop or intake at the rear of the aircraft. This requires either special provisions for ejecting, like falling out the bottom or side of the aircraft, or the installation of an ejection seat powerful enough to guarantee the pilot clears the propeller. This all adds weight and cost. A lot of pilots don't like this idea.

An unmanned drone has no such problem, aircraft designers are free to choose designs without considering how a human will get out of it.

Schwern
  • 998
  • 5
  • 12
2

A big part of the reason for the bubble shaped head is reduced drag. The area does hide a parabolic antenna but the could be done with many shapes. This shape is designed to have the smallest flat surface area possible to allow the air to flow smoothly across the front of the aircraft instead of becoming disrupted and turbulent by striking a flat surface. This leads to reduced drag and a longer flight time per gallon of fuel.

Cessna's and similar aircraft have a windscreen that is basically a large flat plate for the wind to slam against, and while they work great for seeing through, they actually add quite a bit of drag to the aircraft. The cessna's designed mission, however, makes this acceptable.

damccull
  • 121
  • 1
  • 1
    Welcome to Aviation. While the idea to decrease drag seems a good one, the spherical shape is not the best at it. More elongated profile (like wings) have proved to be more aerodynamic. – mins Oct 11 '15 at 15:27
1

If you look at the front bottom side of a typical single engine Cessna you will find that it is black and oily from exhaust fumes and oil residues.

enter image description here

On reconnaissance UAVs, placing an engine at the nose (although there are a few examples - IAI Hunter for one) cause engine residues to accumulate on optical payload lenses during operation and degrade payload performance.

This is the main consideration for placing (piston) engines at the rear of reconnaissance UAVs.

Chenro
  • 11
  • 1
  • 1
    Welcome to Av.SE! – Ralph J Mar 06 '19 at 19:26
  • 1
    So why not put the engine at the front with the exhaust pointing upwards? You can't do that on the Cessna because it would dump the exhaust residue onto the windshield, but that's not an issue on a drone. – David Richerby Mar 06 '19 at 20:56
0

high aspect ratios are provided for recon activities like lockheed martin U3 which had aspect ratio of 15 but this types of manned aircraft, pilot should be very high skilled

kcihtrak
  • 431
  • 4
  • 7