Engines buried in the wing root have minimum parasite drag and probably minimum weight. Their inboard location minimizes the yawing moment due to asymmetric thrust after engine failure. However, they pose a threat to the basic wing structure in the event of a blade or turbine disk failure, make it very difficult to maximize inlet efficiency, and make accessibility for maintenance more difficult. If a larger diameter engine is desired in a later version of the airplane, the entire wing may have to be redesigned. Such installations also eliminate the flap in the region of the engine exhaust, thereby reducing CLmax.
For all of these reasons, this approach is no longer used, although the first commercial jet, the deHavilland Comet, had wing-root mounted engines. The figure shows Comet 4C ST-AAW of Sudan Airways.
The following figure, from the May 1950 issue of Popular Science, shows the inlet of one of the Comet's engines. "Four turbine engines are placed so close of centerline to plane that even if two on one side cut out, pilot has little trouble maintaining straight, level flight."
Wing-mounted nacelles can be placed so that the gas generator is forward of the front spar to minimize wing structural damage in the event of a disk or blade failure. Engine installations that do not permit this, such as the original 737 arrangement may require additional protection such as armoring of the nacelle, to prevent catastrophic results following turbine blade failure. This puts the inlet well ahead of the wing leading edge and away from the high upwash flow near the leading edge. It is relatively simple to obtain high ram recovery in the inlet since the angle of attack at the inlet is minimized and no wakes are ingested.
In the days of low bypass ratio turbofans, it was considered reasonable to leave a gap of about 1/2 the engine diameter between the wing and nacelle, as shown in the sketch of the DC-8 installation below.
As engine bypass ratios have increased to about 6 - 8, this large gap is not acceptable. Substantial work has been undertaken to minimize the required gap to permit large diameter engines without very long gear.
Current CFD-based design approaches have made it possible to install the engine very close to the wing as shown in the figure below. The 737 benefited especially from the closely mounted engines, permitting this older aircraft design to be fitted with high bypass ratio engines, despite its short gear.
Laterally nacelles must be placed to avoid superposition of induced velocities from the fuselage and nacelle, or from adjoining nacelles. This problem is even greater with respect to wing-pylon-nacelle interference and requires nacelle locations to be sufficiently forward and low to avoid drag increases from high local velocities and especially premature occurrence of local supersonic velocities. The figure below from Boeing shows some of the difficulty in placing the engines too close to the fuselage.
Influence of lateral nacelle position on interference drag
Structurally, outboard nacelle locations are desirable to reduce wing bending moments in flight but flutter requirements are complex and may show more inboard locations to be more favorable. The latter also favors directional control after engine failure. Finally, the lateral position of the engines affects ground clearance, an issue of special importance for large, four-engine aircraft.
Another influence of wing-mounted nacelles is the effect on flaps. The high temperature, high 'q' exhaust impinging on the flap increases flap loads and weight, and may require titanium (more expensive) structure. The impingement also increases drag, a significant factor in take-off climb performance after engine failure. Eliminating the flap behind the engine reduces CLmax. A compromise on the DC-8 was to place the engines low enough so that the exhaust did not hit the flap at the take-off angle (25 deg. or less) and to design a flap 'gate' behind the inboard engine which remained at 25 deg. when the remainder of the flap extended to angles greater than 25 deg. The outboard engines were placed just outboard of the flap to avoid any impingement. On the 707, 747, and the DC-10, the flap behind the inboard engine is eliminated and this area is used for inboard all-speed ailerons. Such thrust gates have been all but eliminated on more recent designs such as the 757 and 777.
Pylon wing interference can and does cause serious adverse effects on local velocities near the wing leading edge. Drag increases and CLmax losses result. A pylon which goes over the top of the leading edge is much more harmful in this regard than a pylon whose leading edge intersects the wing lower surface at 5% chord or more from the leading edge.
The original DC-8 pylon wrapped over the leading edge for structural reasons. Substantial improvements in CLmax and drag rise were achieved by the "cut-back pylon" shown in previous figures. The figures below show the effect of this small geometry change on wing pressures at high speeds.
Pressure Coefficient in vicinity of outboard pylons of DC-8.
In addition, wing pylons are sometimes cambered and oriented carefully to reduce interference. This was tested in the mid 1950's, although the gain was small and many aircraft use uncambered pylons today.
One disadvantage of pylon mounted nacelles on low wing aircraft is that the engines, mounted close to the ground, tend to suck dirt, pebbles, rocks, etc. into the inlet. Serious damage to the engine blades can result. It is known as foreign object damage. In about 1957 Harold Klein of Douglas Aircraft Co. conducted research into the physics of foreign object ingestion. He found that the existing vorticity in the air surrounding the engine inlet was concentrated as the air was drawn into the inlet. Sometimes a true vortex was formed and if this vortex, with one end in the inlet, touched the ground, it became stable and sucked up large objects on the ground. Klein developed a cure for this phenomenon. A small high pressure jet on the lower, forward portion of the cowl spreads a sheet of high velocity air on the ground and breaks up the end of the vortex in contact with the ground. The vortex, which has to be continuous or terminate in a surface, then breaks up completely. This device, called the blowaway jet, is used on the DC-8 and the DC-10. Even with the blowaway jet, an adequate nacelle-ground clearance is necessary.
The stiffness of the pylon a for wing mounted engines is an important input into the flutter characteristics. Very often the design problem is to develop a sufficiently strong pylon which is relatively flexible so that its natural frequency is far from that of the wing.
When aircraft become smaller, it is difficult to place engines under a wing and still maintain adequate wing nacelle and nacelle-ground clearances. This is one reason for the aft-engine arrangements. Other advantages are:
An aft fuselage mounted nacelle has many special problems. The pylons should be as short as possible to minimize drag but long enough to avoid aerodynamic interference between fuselage, pylon and nacelle. To minimize this interference without excessive pylon length, the nacelle cowl should be designed to minimize local velocities on the inboard size of the nacelle. On a DC-9 a wind tunnel study compared cambered and symmetrical, long and short cowls, and found the short cambered cowl to be best and lightest in weight. The nacelles are cambered in both the plan and elevation views to compensate for the angle of attack at the nacelle.
With an aft engine installations, the nacelles must be placed to be free of interference from wing wakes. The DC-9 was investigated thoroughly for wing and spoiler wakes and the effects of yaw angles, which might cause fuselage boundary layer to be ingested. Here efficiency is not the concern because little flight time is spent yawed, with spoilers deflected or at high angle of attack. However, the engine cannot tolerate excessive distortion.
A center engine is always a difficult problem. Early DC-10 studies examined 2 engines on one wing and one on the other, and 2 engines on one side of the aft fuselage and one on the other, in an effort to avoid a center engine. Neither of these proved desirable. The center engine possibilities are shown below.
Each possibility entails compromises of weight, inlet loss, inlet distortion, drag, reverser effectiveness, and maintenance accessibility. The two usually used are the S-bend which has a lower engine location and uses the engine exhaust to replace part of the fuselage boattail (saves drag) but has more inlet loss, a distortion risk, a drag from fairing out the inlet, and cuts a huge hole in the upper fuselage structure, and the straight through inlet with the engine mounted on the fin which has an ideal aerodynamic inlet free of distortion, but does have a small inlet loss due to the length of the inlet and an increase in fin structural weight to support the engine.
Such engines are mounted very far aft so a ruptured turbine disc will not impact on the basic tail structure. Furthermore, reverser development is extensive to obtain high reverse thrust without interfering with control surface effectiveness. This is achieved by shaping and tilting the cascades used to reverse the flow.
Solutions to the DC-10 tail engine maintenance problems include built-in work platforms and provisions for a bootstrap winch system utilizing beams that are attached to fittings built into the pylon structure. Although currently companies are developing virtual reality systems to evaluate accessibility and maintenance approaches, designers considered these issues before the advent of VRML. The figure below is an artist's concept of a DC-10 engine replacement from a 1969 paper entitled "Douglas Design for Powerplant Reliability and Maintainability".