How much less efficient is a push pull arrangement a la Savoia-Marchetti SM-55, LeO H-24, Dornier Wal, ... than two tractors?

A reasonable approximative guesstimate to get into the right ballpark is what I would be interested in. And what can you do to make it the beast of it? Counter rotating props (long live varioprop!), higher pitch/revving pusher, ...?

From : Don Stackhouse

Simple question, but as usual no simple answer. For the types of disk loadings we see in models, it's likely to be less efficient than two independently mounted tractors.

A propeller imparts a swirl to the air passing through it. It takes energy to accelerate the air into this rotation, and therefore represents an efficiency loss. The idea behind a contra-rotating propeller system (i.e.: two props rotating in opposite directions on the same axis; "counter-rotating" means two props rotating in opposite directions on different axes, such as on the Lockheed P-38) is to have the swirl from one prop cancel the swirl from the other, eliminating the rotation in the slipstream.

Supposedly this can provide efficiency improvements of as much as 15%. The problem with this concept is that there must be that much efficiency loss already in the basic design, that is available for recovery. This is only true in props with extremely high disc loadings (i.e.: massive amounts of horsepower being forced into a relatively small diameter prop), such as the "propfans" that NASA was experimenting with back in the 1980's. Those were trying to absorb 12,000 horsepower or more in a prop only about 10 to 12 feet in diameter, less than the diameter of the 4500 horsepower props on a C-130 Hercules! When you have that much power going into such a relatively small prop, there is lots of swirl, and therefore a lot of energy that can be recovered by the second prop. That's also why those props use so many blades; more disc loading requires more blades to absorb the power.

At the disk loadings typical of our models, there is very little swirl by comparison, and I doubt that you could expect more than a percent or so of recovery from it at best, IF you did everything exactly right (a virtual impossibility in the real world).

About the only case where swirl in models is a significant factor worth doing something about is in ducted fans. In that particular case we have a lot of power going into a very small prop, so there is lots of swirl. To combat this, we put stator vanes in the duct behind the prop to straighten out that swirl. Those stator vanes are nothing more than the special case of a contra-rotating propeller system, with the second prop in the system designed to run at an RPM of exactly zero. Even so, those stators must be designed and optimized very carefully, or the energy losses due to their own drag will be greater than the energy revered from the swirl, resulting in a net loss. This is the exact same problem most energy recovery devices (such as winglets) face, that of delivering a benefit that exceeds their cost.

The other benefit of a contra-rotating system is that it can cancel out the torque and P-factor effects of a large engine. This is one of the main reasons for its use in planes such as the later Rolls Royce "Griffon" engined versions of the Spitfire, the Bugatti racer, or the Fairey Gannet. Unfortunately, that also requires the use of a fairly complex gearbox, and gearbox-driven props of any kind have a long history of nothing but trouble. The British seem to have had the best luck with them (the gear drive that combined the two crankshafts into the single propshaft on the Napier "Sabre" 3000 to 5000 horsepower H-24 engine was particularly ingenious, and very successful), but other than those successes, the propeller gearbox has historically been the ruin of many airplanes. Gearbox problems were on of the biggest factors that kept the Northrop XB-35 flying wing bomber from being ready before the end of WW II.

In any case, torque and P-factor are generally not significant issues on models. However, the asymmetric thrust in the case of a failed engine on a twin (yes, even electrics can have those), can be a significant issue. The "centerline thrust" of a contra-rotating twin arrangement can solve this. This was one of the biggest reasons for this layout on the Cessna 337 and the Rutan "Defiant".

For a typical un-ducted contra-rotating propeller system, one of those two props is a pusher prop, and therefore you have all the problems and efficiency losses inherent in a pusher prop, which can be considerable on any size airplane. The myth of pusher efficiency assumes that by putting the prop at the back end of the airplane so the rest of the airframe in line with the prop does not feel the increased speed of its slipstream, you save on airframe drag. In actual practice this may be true, although in the vast majority of cases the savings from this are microscopic. If we convert that airframe drag savings into its equivalent in terms of propeller efficiency, we're looking at typical differences on the order of a small fraction of one percent. Recent wind tunnel studies by NASA even show that the majority of the flow behind a propeller tends to be laminar, not turbulent.

Meanwhile, putting the pusher prop at the back, so the airplane does not have to fly through that prop's slipstream, means that the prop now has to fly through all the disturbed airflow coming off the airframe. The efficiency losses from that are typically at least 2-5%, and can be well in excess of 15% in some cases, not to mention the increase in vibrational stresses and noise, the added FOD ("Foreign Object Damage") of stuff coming off the airframe, rocks kicked up by the wheels on takeoff, etc.. Keep in mind that on a propeller driven aircraft, only a very small percentage of the airframe is actually immersed in the propeller slipstream, and therefore only a small percentage of the total airframe drag is affected by the slipstream. Meanwhile, essentially all of the thrust comes from the propeller, so anything you do that hurts the propeller's ability to do its job will have big effects on thrust and efficiency.

In addition, pusher props are usually restricted in diameter because of ground clearance problems. This tends to force additional efficiency losses. Diameter is probably the single most important factor in the efficiency of most propellers, and even a small restriction on it can have big effects. This is especially true at high power and low speed, such as takeoff and climb, although less so at high speeds. This is one of the major reasons the Prescott Pusher (among others) was such a disaster. Try comparing its takeoff performance with conventional tractor aircraft in the same power and payload class and you'll see what I mean.

A pylon-mounted arrangement like you're considering doesn't have ground clearance issues, but has restrictions due to the height of the pylon. All that thrust way above the C/G and the hull tends to shove the nose down, especially on takeoff. I know of at least one amphibian with a pylon-mounted engine that has been unable to accept larger engines, because any significant power increase beyond the plane's current engine tends to make the plane want to become a submarine when you open the throttle for takeoff.

With a pylon-mounted arrangement, the forward prop sees some disturbed inflow due to the flow next to the fuselage and wing, but the aft engine also sees the disturbances from the forward engine and nacelle, as well as the pylon and any external bracing.

The net result of all of this is usually little or no measurable benefits from reducing airframe drag, but quite significant losses due to these other factors, for an overall net loss. Even the possibility of recovering swirl energy, as in the case of a contra-rotating propeller system, usually does not start to see measurable benefits until you get into the sorts of horsepowers typical of turboprops and very large piston engines. I used to be an engineer for a propeller company that happened to have more experience with pusher installations than probably anyone else in the business (Voyager was one of those). Our usual first reaction when someone approached us with a new pusher application was to try to talk them out of it.

There are a number of aircraft designers (including Rutan) who have at some time in their careers been a big proponent of pusher designs. In general, they are airplane designers, not propeller designers, and tend to overestimate the benefits to the airframe of a pusher arrangement while badly underestimating the detrimental effects on the prop. There is a tendency to think of props as these mystical devices that you just bolt to the engine to make thrust, with little thought given to the prop's own needs and idiosyncrasies. To really get a decent working relationship between a pusher prop and the airframe usually takes an incredible amount of work. Piaggio came up with one of the better pusher designs (from an aerodynamic standpoint) in their P-180 "Avante", but it took a huge amount of engineering effort including over 2006 very expensive hours in Boeing's wind tunnel to achieve it.

Don Stackhouse
DJ Aerotech