Regarding props that have a flatter pitch inboard and coarser pitch outboard: ...Funny thing is that if greater pitch at the tip is better, it sorta begs the question why does Zinger not make all of their props like that?

From : Don Stackhouse

What we're seeing here may be the difference between qualitative and quantitative design. Yes, in many cases it improves efficiency to unload the inboard portions of the blade so that more of the load is carried on the outboard portions. However, this must be done with some care, as the differences involved can be very subtle. On full-scale props, a difference in twist distribution (the variation in pitch along the blade) of as little as 1/4 degree can have significant effects on prop efficiency. Other than perhaps diameter, twist is one of the most important and sensitive parameters in the design of a prop. Personally I'm skeptical; to this former full-scale prop engineer, arbitrarily mixing 6" pitch with 10" pitch sounds to me a little like trying to do brain surgery with a meat axe.

The optimum twist distribution will be different for different airspeed, altitude, RPM and power settings. In general practice, the optimum pitch distribution for a low-speed prop will tend to be more non-linear than for a high speed prop, with usually a little MORE pitch inboard except for those sections really close to the hub that are so thick (for structural reasons) that they can't produce much lift efficiently. While it may make sense for the most inboard 20% or so of the blade to be unloaded a bit, the angle required to do this is extremely small.

To understand why, let's look at what goes into determining propeller pitch along a blade. The biggest component is the basic pitch angle determined by the local speed due to rotation (or "tangential velocity") and the velocity in the direction of flight (the "inflow velocity). For the moment assume that the plane of the disk is perpendicular to the direction of flight. The inflow velocity and the tangential velocity form the legs of a right triangle. The hypotenuse of that triangle is the helical velocity, and the angle between the helical velocity and the tangential velocity is the basic pitch angle.

Don't start carving your prop yet, we've barely started. A prop makes thrust by accellerating air aft. In order to do this, the air velocity in the slipstream behind the prop has to be higher than in the free stream air way out ahead of the prop. Half of this speed increase (the "induced velocity") occurs in front of the prop and half behind it, so half of the airspeed increase must be added to the inflow velocity in our little right triangle. Now is where it starts to get tricky. The outer portions of the blade are more efficient than the inboard portion. Most of the thrust is made between about 30% and 90% of the blade radius. Closer to the root the tangential velocity is too low, and the required thickness of the blade for structural requirements is too high for that portion of the blade to do very much. Out near the tip the lift drops off, reaching exactly zero at the tip for the same reason the lift at the tip of a wing is zero. As a result, the amount of local induced velocity we need to add to the local inflow velocity varies along the blade. That's the first nail in the coffin of that old idea about having "true helical pitch".

In addition to the induced velocity, our determination of local pitch along the blade must also account for the angle of attack required at each blade station to produce the necessary lift coefficient, which in turn is dependent on the local airfoil characteristics (primarily thickness and camber), and the local blade chord. We must also include the effects of the zero lift angle of each local airfoil, since the wide range of camber and thickness along the blade also results in a wide range of zero lift angles as well. The angle of attack is usually small compared to the other components of the total pitch angle, but it is extremely sensitive to small changes. Not only must it be determined very precisely, but all the other components must also be determined equally precisely if you want the result to be anywhere close to your predictions. Like I said above, as little as 1/4 of a degree is significant, and that's in the TOTAL of all the accumulated errors in the individual predictions of velocity components.

The really nasty part of all this is that unless we know how much thrust the prop is making AND the distribution of that thrust along the blade, we can't precisely determine the local induced velocities. The thrust distribution along the blade for any reasonably efficient blade will be roughly elliptical in shape, but with the peak at around 75% radius, the aerodynamic center of a propeller blade. However, for a really efficient prop we need to know the exact induced velocities, inflow velocities (which vary as well because of skin friction effects between the airflo and the fuselage or nacelle, effects which can be sigificant out to about 60-80% radius), blade local chords, local airfoils, etc., in order to determine the required angle of attack and the resulting local pitch angle. It's a classic "Catch 22", you can't determine the answer without knowing the answer already. This is where the computer comes in. We input the horsepower, rpm, airspeed, initial blade geometry and diameter, etc., then iterate on everything else until we find the pitch angle (as measured at the aerodynamic center, the 75% radius) that results in the correct horsepower being absorbed. Note, this is how it's done for variable pitch, constant speed props.The method for fixed pitch props is slightly different but follows the same principles.

The next step is to iterate on the iterations while varying all the other parameters one at a time until we find the most efficient blade geometry. That's for ONE flight condition. It's really only possible to do in depth with a computer, plus some very good software and a lot of experience in interpreting its results. We also usually have to do the same for other flight conditions of importance, then find an overall blade design that does the best job of providing the highest possible efficiency at the highest priority flight conditions while trading off as little efficiency as possible at the other flight conditions.

We have to make sure the final design's vibrational characteristics and structural strength are compatible with the application. If any natural frequencies fall within the range of frequencies it will be subjected to by the motor and by airflow distortions (this is an especially bad problem for most pushers), the blade could easily be shaken to pieces in as little as a few seconds. BTW, in this regard wood is much more forgiving than most other materials, which is why wood is such a good choice for props on piston-engined models. Structural strength is also a big factor. I don't have the numbers for typical model propellers (although I've long thought that a study of this would be an excellent subject for a magazine article someday), on a typical full scale propeller blade the numbers are astonishing.

For example, on a typical aluminum propeller blade (doesn't matter if it's a Piper Cub or a Beech King Air, because their tip speeds are similar, their centrifugal forces are similar), the centrifugal force ("CF") trying to pull the blade out of its socket in the hub is usually around 40,000 to 50,000 pounds. Kevlar blades are much lighter, so their typical CF range is around half of that. For the really large props on things like Lockheed C-130's, the forces get really astronomical. These props actually need lots of extra CF to keep the blades properly seated in their retention bearings against the enormous bending moments created by all that thrust and horsepower. They also need lots of extra bending strength in the blade roots. CF for the older all-metal blades in these types of applications is in the neighborhood of 130,000 pounds!

BTW, think about that the next time you're debating whether to keep using that old prop with the little scratch in one blade. While you're at it, also think about how that little scratch can cause a stress concentration that can easily much more than double or triple or even quadruple the local blade stress. Also consider that plastics can have some very deep but virtually invisible cracks initiating and growing very rapidly from little scratches like that. I heard about a case a number of years ago where a man was leaning over the prop disk while tuning the needle on a .60 control-line model with a nylon prop. It tossed a blade right through his heart, killing him instantly.

A book I have on full-scale props talks about propeller "incidents" and propeller "accidents". An "incident" is normally the result of a prop that someone was expecting to move, and generally results "only" in severe maiming. An "accident" is typically the result of a prop that someone was NOT expecting to move, and usually results in death.

One more major factor to consider before you start carving your prop. We now know the desired blade geometry and structure, and have a pretty good prediction of the blade stresses in flight. Now we have to figure out how much the blade will distort because of those stresses, then distort the blade that same amount in the opposite direction when we manufacture it, so that in flight it distorts INTO the desired shape. In particular, the blade will tend to untwist because of CF-related stress, which will tend to increase pitch at the tip. OTOH, the chordwise component of CF will tend to flatten out the pitch at the tip. Which effect dominates and by how much depends on the individual blade design and its application.

That's how it's done on full-scale props. On models we usually don't have the equipment, software, or even the raw data of sufficient accuracy to attempt this approach. About all we can do is make a rough estimate of what general pitch and diameter is likely to work, then try a bunch of different props till we find one that seems to work the best. All the prop makers can do is try to estimate typical horsepowers, rpm's, airspeeds, vibrational environments, etc., then try to design a range of props that should fit those situations. Or they can just bow to the altar of "TLAR" and guess at it. The "true helical pitch" approach is one example of this. They can also make guesses like "the inboard parts of a blade are less efficient, so we'll arbitrarily reduce them by some amount to unload them". This is a very dangerous approach, since just a tiny bit too much could result in portions of the blade making negative thrust, or at least messing up the thrust distribution enough to brutally murder the overall efficiency.

One final caution. We haven't heard any input from the engineers at Zinger on any of this. We have nothing (other than that company's good reputation for making props that seem to perform well) to indicate how much the engineers knew or how well they applied it. Advertising copy is usually written by advertising folks, and as such is generally NOT a reliable indicator of a product's engineering quality. The designation of a multiple pitch on this prop could indicate anything from a crude guess at the blade geometry, to a subtly refined design based on a thorough engineering study with hours of analysis and testing behind it. Be VERY careful to avoid reading too much into the advertising hype, or giving it too much credibility. In addition, even a carefully refined design will be only as good as the match between the design's assumed flight characteristics and the characteristics of your model. If they're not a VERY close match to each other, then that prop is not likely to perform well on your model no matter how well it was designed and manufactured. The bottom line is, if your model flies best with a particular prop, then that's the one you should use.

Don Stackhouse
DJ Aerotech