Technical Articles and Product Descriptions

Reference Materials

Additional Products

Race Engine Technology
The Premier Engine Magazine

There are several contemporary approaches to PSRU implementation. They include toothed-belt drives, link-chain drives, planetary reductions, helical and spur gear reductions. Each approach has its advantages and disadvantages.

BELT DRIVES

Belt drives are light, simple, provide a limited amount of torsional cushioning, and seem to be relatively reliable. However, the life of a belt is typically limited to two or three hundred hours, which can affect reliability considerations. The belt-PSRU’s are limited in power transmission capability, some do not handle hydraulic constant-speed propellers, and the major supplier of high capacity toothed belts threatens Very Bad Things to anyone caught using their products on anything connected with aircraft.

The bigger issue has to do with the maintenance of belt preload in the dramatically-varying temperature environments encoutered by aircraft. That is very difficult to accomplish in a fixed-center drive having no idler. In order for ANY belt drive to operate correctly, a certain amount of preload is required (covered in detail in our BELT DRIVES section). The high capacity toothbelts are made from fibers which cause the belts to shrink with increasing temperature, and conversely, expand as they get cooler. It was this very issue that caused an aftermarket belt drive in a popular experimental helicopter to snap a large number of transmission shafts in flight, leading to unanticipated flight termination (and in many cases, aircraft termination as well).

LINK-CHAIN DRIVES

Transmission systems using high capacity link-chains have been around for many years. They are commonly used in various different industrial applications, are currently used in several 4-wheel-drive automotive transfer cases, and were used for a short time, on early front-wheel-drive automobiles.

Link-chain PSRU’s have the features of simplicity, the potential to use a hydraulic constant speed propeller, and a fair amount of in-service experience. However, unless the drive is properly designed, the chains elongate in the highly-loaded, torsionally-active aircraft piston-engine environment, and therefore can exhibit relatively short life in aircraft applications.

Fred Geschwender’s link-chain reduction, using a 2” wide Morse Hy-Vo™ chain, has been used successfully in several applications, including ag-planes and the early version of the LEGEND aircraft. However, the torque capacity for a 2" Hy-Vo (from Morse engineering data) is about 440 lb.-ft. at 4000 RPM, decreasing to 425 lb.-ft. at 4800 RPM, and decreases fast as RPM goes higher.

There are other link-chain reductions as well, but none we have seen adequately address the significant design issues of

1. isolation from engine excitation and
2. reliability suitable for an aircraft application.

Further, we think that existing link-chain units are unnecessarily heavy, and the ratio selection is very limited.

Often, the claim is heard that link-chain drives are "self-damping". That claim is contrary to engineering evidence. These engineering issues are explained in CHAIN DRIVE DESIGN ISSUES.

ALL THAT BEING SAID, however, EPI has recently engineered a Morse Hy-Vo™ drive for an industrial application. Based on that design work, we think that a proper understanding of the load capacity calculations for this type of chain, along with proper design and engineering, could lead to the implementation of a reliable, low-cost, but HEAVY Propeller Reduction Unit using Hy-Vo™ chain, providing a wide selection of ratios from 2.0 to 2.78. EPI did a detail design on such a product at a customer's request. (See EPI Mark-11 PSRU)

PLANETARY GEAR DRIVES

The planetary gear approach offers some interesting possibilities including compactness, light weight, reliability derived from load-sharing, and simple support for a hydraulic constant speed propeller.

Properly designed for a specific application, the planetary approach can be very successful. Many of the big radials used planetary reductions (P&W 1830, 2000, 2800, 4360; Wright 1820, etc) and one of the geared Lycoming 480 engines uses a planetary reduction. However, with a planetary approach, the propeller centerline is coincident with the crankshaft centerline. That is desirable for a radial, a rotary, or an opposed engine, but not so fine for a "V" configuration.

In a retrofit, to maintain the existing thrust line placement, a "V" engine with the prop on the crankshaft centerline must be raised (unless you are willing to invert the engine). Alternatively the thrust line could be lowered, but reducing the distance between the thrustline and the aircraft CG is destabilizing (ref-4:8, Perkins & Hage, Airplane Performance, Stability and Control), so it’s typically not a good approach for a "V" engine retrofit.

The automotive origins of most of the current planetary PSRU’s limit them to relatively low-power aircraft applications. However, certain vendors are building their own planet carriers which provide more planet gears than the automotive configuration. That looks like a good step toward higher power applications. And there is some interesting planetary hardware in certain truck transmissions which might be applicable to moderate power aircraft applications.

There is a generic issue with single-stage planetary gearsets of automotive origin (not the compound-reduction Ravigneau-style such as used in Powerglide transmissions). That is the fact that ratios between approximately 1.6 and 2.7 (non-reversing) are very difficult to implement because of the very small planet size required and the corresponding very high planet RPM that results. In order to produce those ratios, a two-stage planetary gearset is required (or a Ravigneau gearset, which has severe planet-bearing issues).

On that subject, a big problem we have seen when trying to use automotive planetary hardware in a PSRU is the fact that the needle bearings inside the planet gears often do not have an acceptable life under the applied loads. We require an L-10 life of at least 2000 hours for rolling element bearings. We do not consider 200 hours (or even 500 hours) acceptable.

Here is an example. Consider the use of a 6-planet E4OD front planetary gearset. The available ratios are 1.538 (Ring-driving, Sun-fixed, same direction), 1.857 (Sun-driving, Carrier-fixed, reverse direction) and 2.857 (Sun-driving, Ring-fixed, same direction). At 6000 input RPM, 300 HP applies 263 lb-ft of torque to the gearbox. Using the sun-driving, carrier fixed (1.857 ratio) arrangement, the planets spin at 14000 rpm and the load applied to each set of planet-gear needle bearings is 884 pounds. Under those conditions, the expected life of the planet-gear needle bearings is 195 hours, assuming ideal conditions and fully-equal load distribution. In the Sun-driving, Ring-fixed configuration (2.857 ratio), the planet speed drops to 9100 RPM but the bearing load is still 884 pounds, giving an expected life of 296 hours. However, using the Ring-driving, Sun-fixed arrangement (1.538 ratio), the planet-gear speed is also 9100 RPM but the bearing load drops to 475 pounds (at the same 300 HP, 6000 RPM input). In that configuration, the expected life of the planet bearings is 2344 hours. Quite a difference depending on the gearbox configuration, yes?

When adapting automotive hardware to aircraft use, however, you must keep in mind the automotive load model vs. the aircraft load model. (For more information on Load Models, CLICK HERE).

The planetary applications we have observed appear to be following a design philosophy which implements a very high first and second mode resonant frequency. As you will see in the section on torsional vibration, that approach is sub-optimal for a light, long-life gearbox, and can be devastating to a propeller. (More information on potential prop problems appears in the PROPELLER VIBRATION ISSUES section.) Further, the absence of the design and manufacturing details necessary to achieve actual (as opposed to theoretical) load sharing can cause much higher loads to applied to certain parts than were anticipated in the design.

OFFSET GEAR DRIVES

The next configuration to consider is the offset (helical and straight) gear reduction (such as used on the Continental GTSIO-520, the Rolls-Royce Merlin and the Allison V-1710 engines) in which the centerline of the propshaft is offset (usually upward) from the centerline of the engine crankshaft.

If the design includes an internally-toothed driven gear, it immediately suffers from the engineering deficiency of heavily loaded shafts in an overhung configuration. (Allison tried that approach in their early V-12’s and abandoned it.)

For the decision whether to use helical or spur gears, the issues include:

1. noise,
2. the additional cost for helical gears having the necessary American Gear Manufacturers Association (AGMA) quality level,
3. asymmetric tooth loading (edge-loading) on helical tooth pairs which are in partial contact (contact across only a part of their face width), and
4. the necessity for a helical design to include suitable bearings (ideally not washers) to absorb the considerable thrust loads generated by helical gears.

The issue of thrust absorption is non-trivial. The output shaft provides for propeller thrust absorption anyhow, but the design must accommodate significant thrust loads on the input shaft, as well as any intermediate (idler) shafting. Solidly attaching the input shaft to the back of the crankshaft in order to use the engine thrust bearings is generally a poor solution, for two reasons:

1. the crankshaft does NOT operate concentric to the bearing centers, so the rigid connection must be designed so as not to constrain the radial movement of the crankshaft, and
2. it can be difficult to make the rigid connection have the torsional rate necessary to achieve isolation from the engine torsional excitation (covered later in great detail).

On the positive side, helical gears do have a significantly greater contact ratio than spur gears of similar diameter and tooth pitch. Helical gears are also much quieter than spur gears, but in an aircraft application, who can tell the difference?

For reasons of reliability, simplicity and cost, EPI chose externally-toothed spur gears for it's geared PSRU. See the GEAR DESIGN section for a full presentation on the superiority of spur gears.